I was surprised to realise that the power supply cap is actually part of the output stage signal loop and thus has a major effect on sound quality other than just ripple and current supply. There is a connection called ultrapath capacitor that connects the cathode bypass capacitor to B+ instead of ground. This capacitor determines the sound quality and takes out the power supply characteristic from the signal path. The other thing is that you can use a regulated supply to reduce noise and not have it's sound character intrude on the signal.I first came across this on Bartola 's GU50 SE amp. See Ultrapath, parallel feed and Western Electric paper by Lynn Olson and his EFT presentation on signal loops and ultrapath connection for explanation.
Yes correct , ultrapath is known for decades ............and forgatten.( as well as many other interesting tricks )..........the value of the cap has to be correct ...
I agree with the cathode bypass cap peice but were talking DC current and it's only going one direction thru output transformer and output tube to ground - not back to that 1st cap in the power supply - and it's like a split rail power supply also - it would have to go back to the de-coupling caps - chokes also - Current is not going backwards- I'd bet if we watch on a scope we'd see - checking if the DC drops at all with music and check the AC ripple for any changes - of course the B+ is in the audio path but all we need is good clean power and constant current. The parallel feed is shunting the ouput with a cap and coil (choke) @@SkunkieDesignsElectronics
We listen to the power coming out of the wall. It goes through the power supply, it’s frequency/strength gets managed by the amp and goes to the speakers. Straight line from the outlet in the wall, to the speakers. Tubes/transistors manage the flow of that power. I’ve been a firm believer that everything matters. Power supplies as well.
Always loved Solen capacitors and have been using them for 20+ years in my amps for any large value caps. Even though we draw the circuits as if the current flows to ground, of course it actually flows from ground, so this makes many parts of the circuit we think of as unimportant to the sound actually very important.
It is wonderful to see someone so knowledgeable learn something new. I would love to see more videos like this where attention is given to one tiny aspect of the design and you report to us what you find
These observations were gradualy published for already decades by experienced tube /triode mania audiophiles and some tube amplifier manufactures . It is a good thing to refresh this knowledge and to share it....
The caps Film vs. Electrolytic are generally different in architecture and so in their behavior. Loading-Capacity and Release-Capacity are totally different especially in "speed" of doing that. So they will release demanding power in shorter or longer time. This has an impact on the tubes. When you compare caps of the same architecture, there will be also differences in their behavior. Depending on the chemical as well as building differences. Cheers Stefan
Yes I do like the warm n mellow sound of vintage Western Electric console wires in my 300b monoblock amp. Hope to change some cap in the future. Thanks for taking the time to talk me/us thru step by step.👍👍👍
Well, one way to look at it is that the amp modulates the DC from the supply to make a larger version of it at the output. The input signal never makes it past the 1st grid... or base or gate for that matter. Thanks so much for this video!
One of the important things that power supply caps do which is often overlooked is they provide a low AC impedance path to ground. A higher ESR cap, especially one that increases with higher frequency, effectively makes the plate load higher in value. It becomes the load resistor or transformer primary plus the impedance of the capacitor. It ends up being effectively a variable plate load. The other thing that can happen, mostly with bad caps, is you get poor decoupling between the stages so the output affects the driver. Extreme amounts mean motor-boating, but small amounts might produce some IMD which doesn't sound great.
What an excellent hobby that a person can still learn something and also make you think of endless possibilities like using caps to tune amp for speaker characteristics. Good on you. 👍
Great sharing, thank you. I recap gear regularly using a combination of DNM slit foil smoothing caps and other electrolytic caps from Elna, Nichicon and Panasonic. As regards WIMA, I use their MKP-10 polyprop’s in place of film caps and I will say they take a long time to run in. But then they sound great. It makes a lot of sense to me that amp sound is impacted by the choice of caps in the PSU. Amplifiers work by combining the signal with power to lift the output. Any artefacts in that power will be present in the output stage as noise or distortion. The cleaner the power the better in my experience.
I recently changed the WIMA MKS caps used as coupling caps in a Chinese 300B tube amp for Audio Note Copper Foil oil filled ones. Wow! The website says that these capacitors were specifically designed for this application and they sound great.
5:50 The last cap in the power supply is in the signal path (assuming this statement makes any sense at all). However, here it is the first cap that is replaced (from the looks of it, this amp sports a C-L-C filter, without knowing anything else about it) so it does seem less likely to affect the sound. The problem with these unregulated power supplies is that they only have 3-4 components so all of them play a large part in the performance, i.e. I'd venture saying that all their parasitic elements have an effect, there could be oscillations taking place (shrill top end does point to that being a possibility) and so on. I think it is possible to make a power supply get excited and thus affect the whole amp but I have never looked into it in detail. So much stuff to do, so little time... Following this train of thought, the construction of these film caps (which are made to a price point, hence why we can get these stupendous values now in film caps that we could only dream of half a century ago) really affects their performance. Maybe these Wimas are of this slit foil construction, I dunno. That might or might not mean something when used in an audio circuit but one has to remember that Solens, despite being budget caps, are made for audio whereas Wimas are not (hence the term DC-link). This is food for thought as I always wanted to stock up on Wima film-foils or metallised films for all sorts of projects. Maybe I won't now, hmmm....
Very intriguing and exciting findings of yours for sure. I can't resist the urge to immediately make use of that finding by experimenting in my amps to see if something I like can be obtained.
@@SkunkieDesignsElectronics Hi there. Into testing film cap as first cap in power supply again... That Solen you used, I guess it was the "Solen PPE Polypropylene 630Vdc Capacitors" , not the "Solen PPM Polypropylene 1000Vdc Capacitors" ? I am now planning to test a Solen 22uF as 1st cap after 5AR4 in my Skunkie-modded Reissong A10. Also finally, last week of my summer vacation, started rebuilding the front end of my "Oldchen 300" direct coupled style 300B amp, to cascoded 6SN7:s , with elevated heater centertap as per your recommendation. Very curiously awaiting the results when the ordered components arrive from Mouser.... If it turns out really well, I will buy and try the Shuguang Treasure 300B:s, since the Electro Harmonix 300B:s I have currently do not sound that interesting to my ears. My Skunkie modded A10 actually outperforms that Oldchen 300 stock amp by far as it is.
@Skunkie Wow! This was news on me too as I didn't think the power supply caps would be that critical.to the sound. While I am an avid user of Solen caps, I saved those for the signal path and just used good quality electrolytic caps in the power supply. I can't tell you what is going on though. Even though I am an electronics pro when it comes to solid state audio, tubes is still a new world for me. Yes, I know designs and topology, but how different components react as how you described here is very much new territory. Up to now, if someone told me a first cap in a pi filter would drastically change the sound of an amp, I would have thought someone would be blowing smoke up my a$$. Great work on the real world tests. You might consider yourself a DIY lay-person, but you have done more with your real world testing than many that I consider pros in the tube audio field. Keep up the great work. Your channel has become one of the most interesting ones that I have discovered in the tube audio world!
I believed the same as you, the power side had nothing to do with the signal response. I like to build Class A amps between 3-8 watts. I built one using a 12AX7 and a 6V6 into a 30 caliber ammo can for camping and travel. I included a switchable circuit for valve or diode rectifier filtering. Originally I used some cheap IC caps and was OK with the sound but switching between rectifier circuits there was a difference in tone. At first I wrote it off to the extra punch of the voltage with the diodes but it didn't make sense, the caps in the power section were the same. I'm just a hobbyist and not an engineer so I took a shot and changed the power filtering caps to F&T's, all of a sudden I noticed that the diode circuit sounded about the same but the valve circuit sounded much better, lower volume than the diodes but the tone really came alive with the F&T's. Go figure ........
Every cap has an ESR and inductance which can form a resonant circuit with the other components in the power supply. This resonance can occur at audible frequencies, or at least frequencies that otherwise impact the amp performance. Ironically, low ESR film caps can be worse as there is less series resistance to damp the ringing. It seems even the humble power supply has a lot of nuances and things that can impact sonic performance!
Interesting result. I would recommend trying some X rated safety capacitors. Even in coupling locations, these for some reason, sound drastically better than standard films.
The sound of capacitors - omg Skunkie, I went down that rabbit hole with my preamp and amp builds sometime in the mid 90s - the hours and dollars that I've spent! With regard to their use in the power supply, my experience is yes, definitely the differences are audible. In all cases except for one power amp I ended up choosing good quality electrolytic caps as my choice for the most natural sound. My experience with film caps (Solen mostly which I found to be very good all around caps versus the boxed not so good sounding Wima caps) was that a touch of high end tip up could be gained but at the expense of grainyness. With regard to bypassing, I don't recommend it. In every case i could hear the bypass cap - either tipped up highs again with some grain or an added flavour to the mid range if the value was higher. I have also tried oil type motor run caps - though they may add some smoothness to the mid range, speed / dynamics suffer. Also, be careful with the size of the film caps, I found that if the value gets too large, like the 39uF film I used on a 5AR4 I experienced a kind of pumping action to the bass, no problem though with 2 x 22uF.
The difference is the ESR. I have tested Capacitors in a lab for 23 years. I have also modded lots of less expensive equiptment with capacitors to milk all the sound possible. The power supply has a huge effect on sound. That lower ESR takes away the lag in delivery of power and increases the slew rate at the higher frequencies bringing out the detail. Another thing I have found is dont use any parts with copper clad steel leads(cheaper caps used to use that alot) Or any ferrous metals where the signal passes through. I am impressed with your channel thinking about building the KT 120 monoblocks. I have heard that womens hearing is much more acute which may be why it sounds harsh to you. Myself I love that detail.
Hi Stephe! We were using 'Wima MKS' 47nF 1kV caps & got a 10 to 30% failure rate! We swapped to blue 'Multicomp' caps & had no further problems since. This was in a high voltage power power supply, but not for an amplifier though. Good to see you got your great bass back by replacing the 'Solene'!
I should add that our problematic 'MKS' series Wima capacitors had the identical hard red plastic outer casing as your dodgy 33uF cap. Their epoxy sealant was a yellow colour, rather than red which yours has. So yours may be from a different series, possibly the 'MKP' series. Older, identical looking 47nF Wima caps, from a few years ago, gave us zero problems. We never had to replace them. So they've messed with the metallised polyester film, or p.e.t film, or whatever it is. A fault in these capacitors would cause arcover inside the eht transformers, which would destroy them, & reflect back through to the primary winding & blow the hv switching transistor & all associated driver transistors. Catastrophic destruction of our hv psu's! They'd go off with a loud "BANG!!!!", & blow the glass fuse like it had been spray painted silver & black from the inside! Sometimes even cracked the glass! I would have expected MUCH more from a German made cap! Their latest batch were CRAP!
Just a guess here, but I'm thinking Class A or not, more current is going to be drawn by the speakers at the lower frequencies than at higher frequencies. This of course causes a voltage sag on the high voltage. The power supply having more capacitance being able to supply the voltage that bass frequencies demand with less sag by the speaker is probably being supplied by the better capacitors would be my guess as to why the bass sounds a lot better with the better cap. As an aside if you're familiar with the pre-amps and amps built by Angela-Gilbert (on TH-cam). Part of the design differences that AG gets into is large capacitance arrays in external enclosures are plugged in as an accessory to the amps which supposedly increases the quality of the sound immensely. Those cap banks built by AG are somewhere in the few Farad range.
13:05 I think that's what most people involved in electronics (except very specific/niche/cutting edge areas) have anyway. 13:12 I think maybe an oscilloscope probe in the circuit would help here. One can also check caps on their own, do all sorts of X-Y curves stuff but unless one can somehow work out how the results of these measurements would translate in a specific circuit and operation, then I think it would be better to just measure the thing in the circuit and compare against alternative components this way. All this is very useful testimony especially for those of us who go into circuit building thinking that a part is a part and that's all there is to it. We must keep in mind that most of the electronics art works because it is based on simplifications: there are just far too many factors involved in any given circuit for our mind to be able to process and take into account. Measuring and experimenting is key.
I think that the output feedback has been designed for the less responsive electrolytic cap from the PS. Adding the small bypass means faster response at the output and feedback will enhance thus having overshoot.
Hey, if you happen to have the impedance analyzer adapter card for the analog discovery, you could really go deep dive on comparing the capacitors and all their parasitics. The adapter is reasonably priced and very useful and capable. I highly recommend it.
I wonder if the plastic cap is mylar or polyester, whereas the Solen is metalized polypropylene. I have noticed harshness with mylar/polyester caps when I used them as coupling caps...I would not be surprised if it made a difference with the power supply. I have always spent more time and simulations building and designing the power supply than the input and output stages. Thanks for all your work and sharing all the info you have learned...it can be a tough world in the internet.
Hi Stephe.. The PSU is IN the signal path IMHO (current loops), especially for single ended amps... A 3uF difference can still be sonically heard IME.. Has the B+ stayed the exactly the same with a 30uF Wima DC Link input cap vs the 33uF Solen? Try the Wima paralled with a plain 3uF/630V Mundorf Mcap (white).. I only tried Kemet DC Links and they are not harsh, it takes time (break-in) for the bass to subjectively heard and felt though..
I will try to make a short story: two years ago my DAC died. I write the seller. He tells me send it back to me and as soon as I see the mailing proof I will send you a new unit. I receive it and try it out. It sounded really bad. No bass at all and thin sound. I write this to the seller. He says: I don't understand because I sent you an upgrade version with all very good capacitors. Let it play for a while a couple of days. I opened the DAC and all the cheap $0.15 electrolytes were really replaced with Wima red caps. I played music for 3 days in my basement. Then tried it again on my main system. He was right. The sound was even better than the other unit. The bass was as good but the mids and highs were fantastic. Cleaned up. To resume my story: caps need breakin'in. Not only the electrolytes (reforming), the other caps to! 50 hours at least! I don't understand why. Normal for a speaker to loosen the suspension. Electrolytes often need reforming but Wima's... ???? A film or polyester cap???... Next time play music for at least 50 hours before you make your mind as I did. My short story seems quite long. Sorry my friend!
The Solen PB are great in power supplies, I have them filtering for my front end tubes. My amp is a PP AB triode strapped EL34 amp, my rectification is SS) The character you describe of the Solen's is the same as I and others describe of them when used in the PS.
This has been my experience. I typically bypass the second / B+ caps. To my ear adds detail. I built the ability to change out coupling caps in one of my basement builds. I was similarly shocked (by the sound changes). I think this is much more noticeable in se Amps, less so in pp amps, at least in my experience. Keep going, it feels like snake oil… but clearly not.
Another wonderful thread. I love your no BS approach with your honest opinion. I am in the process of choosing one of your builds. Is this project worth the cost and effort over your EL34 build which looks interesting. Thank you and keep the content coming, I am a big fan. BTW I build guitar tube amplifiers from scratch, have a EE degree and was in the audio business years ago (class D amplifiers at the beginning of the technology).
Well that was a bit of a surprise. But WIMA's are known to need some break in time...but just thought that was if used as coupling cap. But it's easy to forget that the power supply is the heart...you only modulate through the audio sections what it produces. I'm chuckling here because you're experiencing the same thing I got from using Mundorf coupling caps in my Kegger SE KT88...except I didn't know right off that they were the cause of some terrible high frequency distortion. And to make it crazier, vocals sounded awesome. I had some Cornell Dubilier 940C of the right size and replaced the MCap Supremes...immediately better. As far as bypassing electrolytics, I believe there was an overall improvement when I bypassed the cathode caps on the KT88's...but it was small. Now I'm dealing with speaker cables...they have capacitance issues also. Hang in there!
Input caps and the caps between the input tube and output tube make a difference on how it sounds. Look at BASS and Treble circuits in amps and you will see resistors and caps to change the tone output. Bass in the amp takes more current so you need heavy storage caps in the power supply for deep bass.
Capacitors are my bane. Those Wima caps are bright sounding caps IME, which works great in signal path of certain tube circuits to detail up the sound. Doesn't surprise me they brighten up in power supply. Then there are those who swear by big oil caps in the power supply..
Wow, this was an exceptionally interesting video, and that's saying quite a bit because your videos are always interesting. I recently replaced a similar WIMA cap with an Audio Note Copper Foil oil filled cap that was used as the coupling cap in my BrzHiFi 300B amps. I wasn't surprised that it made a substantial improvement, but the power supply?!? As you say, a coupling cap in the signal path is one thing but everyone almost always uses electrolytic caps in the power supply. My first tube amp was a Dynaco Stereo 70 that I rebuilt, and adding film bypass caps to the electrolytics in the power supply completely opened up the top octave. So I guess that I should be so surprised. I think that we simply don't completely understand what about capacitors affects sound quality and how to measure it. So, the impact on sound quality remains effectively magic. It can only be observed through trial and error. We can form some general rules but that's about it. Pete
Upgraded my HIVi diy 2.2A speakers crossovers. Shrillness was removed by switching out caps with $15 each film caps. Then sound opened up and sounds like it has so much more air and life, and cleaner bass with the speakers plugged into 8 ohm taps. Can’t put my finger on it, but they sound so much better in 8 ohm tap of my R8, than 4 ohm taps. Run other speakers in 4 ohm and forgot to switch at amp.
hi if you have room for solen in the supply ,put only solen its sound better cause these caps have a very low resistance and this is very fast and keep stability all the time ,electrolytics change properties with use .
So I wanted to tell you that I am really glad I found your channel. I watched you vid on the X7 Muzi. Of the ones you have seen can you recomend one 1000?
One more suggestion, to bring yr amp to another level, pls try and tell me your results! The Vol pot voltage divider has real impact to audio quality. Not sure what pot is used in yr amp? Replace it with TKD and add 10x TKD resistors to the 3 vol pot tabs.If yr pot is 100k, add resistors 10x (1 mg ) to pot in and pot out to yr driver tube. This shunt mod fr late Bob Crump is a super audio hack. Try it?
glad i watched. i was thinkin about what to do about the caps comin off the rectifier on a coupla amps im messin with--do i replace the can caps? restuff em? leave em and stick new caps under the chassis? i may try a 4uF solen film cap.
I am wondering if a comparison of the Equivalent Series Resistance (ESR) values would provide useful data in this case. Also worth noting, some designers place a smaller value cap (of different material type) in parallel with larger ones and then note sonic improvements - but I do not recall if this is only in the signal path or also in the power supply section.
Power supply can well be viewed as a positive feedback loop if that help with understanding the effect. Mullard called tubes valves as in valve regulating the flow from the power supply. Different perspectives help solve problems. The ESR of the Solen's are a bit different than the Wima's prefer ASC to both of them .
What size series asc are you using and do you parallel them?I ask because I have some 20uf 400v from all electronics as well as some 60uf that I forgot the voltage.large polypropylene.
@@johnstuchlik5828 Some series resistors give a better filtering than paralleling basic Radiotron stuff. Stay small for the first one to keep from having big peak current spikes for the diodes. Save the big ones for the later stage of the supply . Any of the ASC will work for me this is in the range of flavor to taste. Common diodes state 32 mfd as max for first cap the 20 is a good choose most of the time for the first one .
@@johnwilliamson467thanks for reply.im going to heed the tube manual first cap size limit w/o surge resistor.i am planning to use 5v4 on my low watt projects mainly because I have some and like the slow warm up.first cap will be 7.5 to 10 first cap, probably try different types,motor run ,electro,poly, what I have on hand and use larger or parallel smaller after choke/resistor.definite going to look at b+ with scope for problems after watching this video and reading comments.
Analog discovery 2 has an impedance analyzer feature you could use to measure the complex impedance characteristics over operational frequency. You probably knew that but it would be cool to see how they measure differently. I bought the little breakout board but you don’t need to if you can supply the resistance
That's a surprising finding for me, like you I thought a power supply must give the correct voltage with sufficient current, as close to smoothe DC as possible and that's it! In fact I have been very tempted to use a solid state PSU when I eventually get to building this as there are advantages with space and filtering but this has me thinking...
That's very weird being both are film caps - also being in the power supply never would have thought you'd be able to even hear a difference- I never heard of Solen caps failing before they are very good caps- but your's got physically crunched. I too would like to see what putting an electro in that 1st caps place will do also as it will store up more energy even if it is a low value 30uF -33uF but the films should be better cap at filtering the ripple. Weird
Yeah, I seriously just thought I was replacing a film cap with a film cap. The difference was NOT subtle! Like I hear was less difference swapping coupling caps around.
I agree with Stefan. The difference between the capacitor types will affect the signal. Every with a charge has a Magnetic and Electrical field offset by 90 degrees. So the fields around the aluminum plates is affected by the shape. The shape of the capacitor could affect the density of these fields and in turn affects the signal. There could be constructive and destructive interference. Remember all energy is moved through fields and not the conductors. Electrons don't really move from one end of a wire to the other. They basically vibrate. There is some electron drift but this is pretty slow. Energy is transmitted through fields as well as audio signals. So each type size and shape of capacitors will be dissimilar, because their fields will be different. Energy moves through fields, not wire. This is why overpriced audio cables are a scam.
I hate to say it, but I suspect there are other impacts in the power supply. That one wacky thing that china hifi said about large chokes choking the sound. I think there may be something to that. If you have a cap inductor, cap filter the power supply, the second cap needs to be big enough to supply power for quick demands in current on the power supply. I suspect with larger inductors, an amp needs larger caps, or the choke will not allow current to change quick enough for good dynamics. Just a hypothesis at this point. Chokes also add resistance which result in voltage drop with increased current. Not a big deal in tube amps, but the last cap needs to keep up.
Yes Steph , ...........your ears don't lie ! There are numbers and measurements ...........but in the end experience is the teacher ... The bypass hype was a rage and hype by audiophiles in the eighties, every cap had to be bypassed , often even with different and multiple bypass caps .........today we know different . Different caps with different values and materials as well as different technologies have a different tone that often don't mix well ( someting like different cone materials used on loudspeakers ) changing tone /timbre that shifts with rising frequencies and they have also different resonance points witch when combinerd " CAN" create stress , incoherence of the harmonics and musical tonality becoming out of a natural balance , ...and that's what happens and is perceived as one of the causes of rstress /listening fatigue ...when this happens it tells that something is wrong .....yes one may notice a daker background and excessive detail or hyper dynamicas on only part of the frequency spectrum but it is out of balance to the nature and natural perception od the music .............in some cases this may work after countless experiences and listening results .............. ONLY your ears will tell if you got it right ore wrong .............they don't LIE ! It remembers me of an incident that happened with a customer of an expensive Audio Note amplifier .........and oh boy these are not cheap ..The customer complained that after having a look inside he was upset beccause the filter "input capacitor " was only a regular standard quality brand in his very expensive amplifier .........he felt as being cheated by the company ! The answer of Peter Quortrup was ( and I agree ) that one buys such an amplifier for it's exceptionnal musical performance ..........explaining that the first most perfect high performance cap when choosen as the very BEST capacitor to use can be your worst enemy .............;the better this capacitors , the more current is draw
It would be interesting to see if breaking in those Wimas electrically, would change the sonic signature. I know break in is a controversial subject, and I did hear you say that the new Solen changed the sound immediately, but it would be an interesting experiment. I put some Russian paper in oil caps in my Klipsch Cornwalls, and they sounded awful until broken in. Some say 300 hours, but the first 100 does the most. I’ve also been told that when breaking anything in, not to just run current through the unit, but also to turn it off from time to time. Thanks Skunkie!
Perhaps you want to look into room, equalization wizard (REW). At the risk of spouting heresy here, I think a little digital signal processing could change any of the characteristics that you’re talking about being affected by capacitors. I have heard similar discussions involving cables. sounds awful lot like people are shooting in the dark trying to get a certain sound with caps and cables when you can zero in on the kind of frequency curve you like in your room with DSP. Personally, I have tested my hearing and know how it falls off in the higher frequencies. I created a compensation curve in Roon and I found I like it quite a bit. Of course at first it will sound a bit bright, but very quickly. It began to sound right to me. Just something he might want to consider.
Any type of EQ box I've inserted between my turntable preamp and the amp adds a veil to the sound, even if the EQ is better, it masks part of the sound in the process. That's been my experience anyway.
Admittedly DSP (digital signal processing) makes more much more sense when the sound source is digital and you can get it done prior to the digital to analog conversion. But since many of us are going to streaming as a main audio source it is very doable and Roon is my tool of choice for doing this since DSP is part of the package anyway. On the other hand Room Equalization Wizard is a free piece of software you use with a calibrated microphone to chart the flatness of the frequency spectrum coming out of your speakers. Similar to how you use you oscilliscope to measure distortion directly from an amp. So REW could chart the change in the signal created by those capacitors you were talking about. Then my thought was, at least for digital signal sources, some DSP could be used to make the same modification to the sound. I've been using it to learn what sort of frequency curve I like so that I can duplicate it across various pieces of listening gear.@@SkunkieDesignsElectronics
I know what REW is and have used it testing speakers. And in testing, this software/microphone combo isn't sensitive enough to decern some of these nuances.
Remember; the amp's circuitry through to its output stage is just a delivery system to dump power supply energy on the speakers at the rate of the input signal. What's first? Look backwards from the output into the transformer, and circuitry into that pool of energy! What do you see?? Along the way, what do you encounter in your mind's eye?
Nice job. A video on capacitors and application would be great. Bypass caps, etc. Does capacitor sound change over time? I often hear this and maybe just when out of spec.
I think majority of burn in is for metalised caps and electroyitic only, since the metalisation needs an applied voltage to self heal and form a better coating
Howdy. I would suspect the ESR of the cap.s are responsible. Equivalent Series Resistance. The ESR can be large for regular el. cap.s. Film cap.s have way smaller ESR. So, they can take current and deliver current fast. I would be careful about changing the first filter cap. , between the rectifier and the choke, to a film cap. The low ESR may cause current surges through the rectifier tube it can't handle. But yes. After the choke the filter cap may be a film cap. Regards.
I stated that the ESR both stated and measured was the same. Also measured the impedance vs frequency and the curves look the same. That's why I said I was baffled, as something other than the obvious is going on.
@@SkunkieDesignsElectronics Howdy again. Hm Ok. I was going to say there is an ESL too. Equivalent Series Inductance. But as the impedance vs. frequency is checked to be similar, it likely covers the ESL aspect. Entering the speculation zone. I think dielectrics have hysteresis as magnetics have. These produce resistive like losses. In English. AC resistance is not the same as DC resistance. One way would be to check the voltage - current phase shift. But I would think that is difficult to do with regular lab. gear. I would understand that radial lead cap.s show more ESL than PCB mounts. Regards again.
I had to come back to this video as I'm thinking of switching to Solens 47uf as the first caps in the power supply in a recent build that I really like..if I have the space. I've been told in the past to just shunt or bypass the Electro's with film's but I've never had luck with bypassing..in any position, Though I've only tried it twice now. What cap type & brand has been your go-to or good results as far as coupling the 2 tube stages?
Mundorf EVO alum/oil. Don't be tempted by their higher priced offerings, IMHO they don't sound as good. And like you, I've had zero luck bypassing 'lytic caps with film ones.
Welcome to the party :) The electricity from the wall is what arrives at the speakers, so the sound is shaped in the PSU first. The more caps you upgrade, the stronger the effect of each will be. When it comes to the brands of audio-grade ones, in general "you get what you pay for". A common style when you buy the really fancy ones is to double the voltage and half the capacitance. The higher quality electrolytics allow you to run them lower uF and actually get better sound quality. have fun
Gat a question for you I've noticed a difference in capacitors. But what about resistors? A lot of exotic snake oil out there. So what brand of resistor do you use for anode, cathode, grid ? Thanks in advance.
Curious: The 33uF Solen film cap you used , it was used as 1st solo cap after rectifier tube, ( as in original scheme ) not as bypass for a 1st electrolytic in PS , right ?
A power supply tube shouldn't impact the sound. The choke shouldn't unless it's with a cap and then in makes an lc filter. So yes a cap will have an effect. Also you have to remember caps need running in. It may sound harsh at the start but will settle down.
Is it possible that the single ended circuit is just more sensitive to these kinds of issues...like parts quality or materials? I've never had another piece of tube gear act like this before...and I've done cap work on Dyna Mk III's and IV's. Of course, the Dynaco's didn't have the sound quality either.
I get that to a point, but "in the signal path" I always was thinking more like inline with the input signal. Like if you put a scope probe at this point, you aren't going to see any signal.
Some of the best power supply caps I've ever used are motor run caps for AC compressors. NOT motor start capacitors, polypropylene motor run caps. Dirt cheap and top notch. From what I've heard, the LAST caps are the ones that matter most (the ones that actually supply voltages to audio stages) and the first ones are less important. YMMV as always.
See that's what I thought too, but this first cap made more difference than any other cap change I have ever done. There are two stages of choke/cap filters after this one and this amp has the power supply split rail after the first choke. The front end each channel has it's own RC filter, yet somehow this first cap made a HUGE change in the sound of this amp.
I have been told you can’t Mix those DC Link caps with electrolyte caps. If you use one DC Link cap in the power supply you have to use them in the entire power supply. He told me “ if you mix the two kind of Caps together weird shit happens “ . I haven’t tried it my self, but so far He seams to be correct by your outcome. 🤷♂️
Everything can make a difference. Us Audiophools learned this a long time ago. It is very harmful to the pocketbook. Thank goodness for diminishing returns. Listen and choose what is important to you. As far as the power supply, remember, the input signal itself is not what is amplified, it is remade, at a higher amplitude or otherwise changed, at each stage. Each stages output is just the power supply being modulated. The power supply IS what you are, ultimately, listening to. Dan
did we take any masurements ? every cap have 20% +- value , and the ERS is different , use smal caps parralell over the catodes in the preamp tubes to get the shrill out
As I understand it, some caps require a burn in period of up to 100hrs before that bad behavior goes away. Maybe test that theory out with one of those bad sounding caps.
IMHO there is something wrong with a part that sounds bad until it's "burned in" for 100 hours. Now maybe a part that sounds good and after some time sounds slightly better I can deal with. Or even a part on initial power up is a little cringy but after 30 minute to a hour clears up. 100 hours of music play time is absurd to expect anyone to suffer through.
@@SkunkieDesignsElectronics I actually agree with you on that. I do know that materials vary in the time it takes for them to stabilize. Speakers surrounds are a notorious example of this. I am currently building two audio note M77 clones. One has precision matched to .001 resistors ( z foil, audio note silver and other high end types, and audio note copper and kaisei caps also matched from sonicraft as well as Russian polystyrene and silver mica caps in the phono section with a personally made precision resistor attenuator, shielded tubes, shielded power supplies, silver wire everywhere, shortest wire runs everywhere, military grade soldering and device mounting, yada, yada, and the other with more of a regular build. My fear is that after all of that work and expense, it will lay a bunch of crap on me and give me massive buyer’s remorse. Imagine spending well over 1.5k in parts and having to hope it will get better because of my choice in caps. I’d be crazy if the regular build with less exotic parts ended up crushing it.
Hmm, perhaps generates the WIMA-Caps a very high frequency on the direct current signal to the secondary site of the Amplifier. I mean, the d.c. current with HF is with the WIMA-Cap not clean enought for the Tubes. You understand, what i mean? It's only a idea, but i'm not sure ... and please, you all excuse for my bad school English 😵💫!
I assume the Solen cap is a "Fast" cap - is the Wima fast ("Fast Capacitors" is trademarked by Solen so WIMA can't use that exact term)? Solen has a datasheet explaining the Fast concept. I have no idea if it makes any difference.
I need to, and then figure out where to look for this very audible change. And this was clearly not a case of expectation bias, I wasn't expecting any change and was in fact baffled that it changed lol
You should try 0,022 or 0.047uf instead lf 0.33uf. To big bypass will give that shrill sound and sign that its to big value. Alao the wima caps are kinda famous for having extreme long burn in time and i usually think its bs talk to in wimas case its not. I do not like them i signalpath that much because theyre to laid back for my taste
It's not a bypass cap, it's the main reservoir "first cap". And as I said, I have tried bypass caps across electrolytics in the past and didn't care for the results.
Honestly, that could be the case and I didn't do that. I guess I was more interested at the time to put the Solen cap back in and see if this actually did change or was it all in my head!
I was surprised to realise that the power supply cap is actually part of the output stage signal loop and thus has a major effect on sound quality other than just ripple and current supply. There is a connection called ultrapath capacitor that connects the cathode bypass capacitor to B+ instead of ground. This capacitor determines the sound quality and takes out the power supply characteristic from the signal path. The other thing is that you can use a regulated supply to reduce noise and not have it's sound character intrude on the signal.I first came across this on Bartola 's GU50 SE amp. See Ultrapath, parallel feed and Western Electric paper by Lynn Olson and his EFT presentation on signal loops and ultrapath connection for explanation.
Yes correct , ultrapath is known for decades ............and forgatten.( as well as many other interesting tricks )..........the value of the cap has to be correct ...
Interesting !!!
I agree with the cathode bypass cap peice but were talking DC current and it's only going one direction thru output transformer and output tube to ground - not back to that 1st cap in the power supply - and it's like a split rail power supply also - it would have to go back to the de-coupling caps - chokes also - Current is not going backwards- I'd bet if we watch on a scope we'd see - checking if the DC drops at all with music and check the AC ripple for any changes - of course the B+ is in the audio path but all we need is good clean power and constant current. The parallel feed is shunting the ouput with a cap and coil (choke) @@SkunkieDesignsElectronics
Thanks, never connected those dots. Never looked past the obvious.
We listen to the power coming out of the wall. It goes through the power supply, it’s frequency/strength gets managed by the amp and goes to the speakers. Straight line from the outlet in the wall, to the speakers. Tubes/transistors manage the flow of that power. I’ve been a firm believer that everything matters. Power supplies as well.
Always loved Solen capacitors and have been using them for 20+ years in my amps for any large value caps. Even though we draw the circuits as if the current flows to ground, of course it actually flows from ground, so this makes many parts of the circuit we think of as unimportant to the sound actually very important.
Thanks for sharing!
It is wonderful to see someone so knowledgeable learn something new. I would love to see more videos like this where attention is given to one tiny aspect of the design and you report to us what you find
That experience was really useful for us. A follow up video will be appreciated with such tweaks and experiments. This cannot be seen in books.
These observations were gradualy published for already decades by experienced tube /triode mania audiophiles and some tube amplifier manufactures .
It is a good thing to refresh this knowledge and to share it....
More to come for sure on this.
The caps Film vs. Electrolytic are generally different in architecture and so in their behavior. Loading-Capacity and Release-Capacity are totally different especially in "speed" of doing that.
So they will release demanding power in shorter or longer time. This has an impact on the tubes.
When you compare caps of the same architecture, there will be also differences in their behavior. Depending on the chemical as well as building differences.
Cheers
Stefan
Cool discovery. And you’ve got me hooked on that album. I bet we’ve listened to it 10 times this past week.
Nice
Yes I do like the warm n mellow sound of vintage Western Electric console wires in my 300b monoblock amp. Hope to change some cap in the future. Thanks for taking the time to talk me/us thru step by step.👍👍👍
Well, one way to look at it is that the amp modulates the DC from the supply to make a larger version of it at the output. The input signal never makes it past the 1st grid... or base or gate for that matter. Thanks so much for this video!
One of the important things that power supply caps do which is often overlooked is they provide a low AC impedance path to ground. A higher ESR cap, especially one that increases with higher frequency, effectively makes the plate load higher in value. It becomes the load resistor or transformer primary plus the impedance of the capacitor. It ends up being effectively a variable plate load. The other thing that can happen, mostly with bad caps, is you get poor decoupling between the stages so the output affects the driver. Extreme amounts mean motor-boating, but small amounts might produce some IMD which doesn't sound great.
What an excellent hobby that a person can still learn something and also make you think of endless possibilities like using caps to tune amp for speaker characteristics. Good on you. 👍
Right? It's like a giant rabbit burrow lol
Great sharing, thank you. I recap gear regularly using a combination of DNM slit foil smoothing caps and other electrolytic caps from Elna, Nichicon and Panasonic.
As regards WIMA, I use their MKP-10 polyprop’s in place of film caps and I will say they take a long time to run in. But then they sound great.
It makes a lot of sense to me that amp sound is impacted by the choice of caps in the PSU. Amplifiers work by combining the signal with power to lift the output. Any artefacts in that power will be present in the output stage as noise or distortion. The cleaner the power the better in my experience.
Loving this series of videos - you should try Mundorf tubecap, Miflex, Duelund and/or Audio Note copper. As far as I’m aware these are the top ones.
I'll look into them.
I recently changed the WIMA MKS caps used as coupling caps in a Chinese 300B tube amp for Audio Note Copper Foil oil filled ones. Wow!
The website says that these capacitors were specifically designed for this application and they sound great.
Yes it does change sound quality certainly
5:50 The last cap in the power supply is in the signal path (assuming this statement makes any sense at all). However, here it is the first cap that is replaced (from the looks of it, this amp sports a C-L-C filter, without knowing anything else about it) so it does seem less likely to affect the sound. The problem with these unregulated power supplies is that they only have 3-4 components so all of them play a large part in the performance, i.e. I'd venture saying that all their parasitic elements have an effect, there could be oscillations taking place (shrill top end does point to that being a possibility) and so on. I think it is possible to make a power supply get excited and thus affect the whole amp but I have never looked into it in detail. So much stuff to do, so little time...
Following this train of thought, the construction of these film caps (which are made to a price point, hence why we can get these stupendous values now in film caps that we could only dream of half a century ago) really affects their performance. Maybe these Wimas are of this slit foil construction, I dunno. That might or might not mean something when used in an audio circuit but one has to remember that Solens, despite being budget caps, are made for audio whereas Wimas are not (hence the term DC-link). This is food for thought as I always wanted to stock up on Wima film-foils or metallised films for all sorts of projects. Maybe I won't now, hmmm....
Very intriguing and exciting findings of yours for sure. I can't resist the urge to immediately make use of that finding by experimenting in my amps to see if something I like can be obtained.
You and me both!
@@SkunkieDesignsElectronics Hi there. Into testing film cap as first cap in power supply again... That Solen you used, I guess it was the "Solen PPE Polypropylene 630Vdc Capacitors" , not the "Solen PPM Polypropylene 1000Vdc Capacitors" ? I am now planning to test a Solen 22uF as 1st cap after 5AR4 in my Skunkie-modded Reissong A10. Also finally, last week of my summer vacation, started rebuilding the front end of my "Oldchen 300" direct coupled style 300B amp, to cascoded 6SN7:s , with elevated heater centertap as per your recommendation. Very curiously awaiting the results when the ordered components arrive from Mouser.... If it turns out really well, I will buy and try the Shuguang Treasure 300B:s, since the Electro Harmonix 300B:s I have currently do not sound that interesting to my ears. My Skunkie modded A10 actually outperforms that Oldchen 300 stock amp by far as it is.
@Skunkie Wow! This was news on me too as I didn't think the power supply caps would be that critical.to the sound. While I am an avid user of Solen caps, I saved those for the signal path and just used good quality electrolytic caps in the power supply. I can't tell you what is going on though. Even though I am an electronics pro when it comes to solid state audio, tubes is still a new world for me. Yes, I know designs and topology, but how different components react as how you described here is very much new territory. Up to now, if someone told me a first cap in a pi filter would drastically change the sound of an amp, I would have thought someone would be blowing smoke up my a$$. Great work on the real world tests. You might consider yourself a DIY lay-person, but you have done more with your real world testing than many that I consider pros in the tube audio field. Keep up the great work. Your channel has become one of the most interesting ones that I have discovered in the tube audio world!
I believed the same as you, the power side had nothing to do with the signal response. I like to build Class A amps between 3-8 watts. I built one using a 12AX7 and a 6V6 into a 30 caliber ammo can for camping and travel. I included a switchable circuit for valve or diode rectifier filtering. Originally I used some cheap IC caps and was OK with the sound but switching between rectifier circuits there was a difference in tone. At first I wrote it off to the extra punch of the voltage with the diodes but it didn't make sense, the caps in the power section were the same. I'm just a hobbyist and not an engineer so I took a shot and changed the power filtering caps to F&T's, all of a sudden I noticed that the diode circuit sounded about the same but the valve circuit sounded much better, lower volume than the diodes but the tone really came alive with the F&T's. Go figure ........
Every cap has an ESR and inductance which can form a resonant circuit with the other components in the power supply. This resonance can occur at audible frequencies, or at least frequencies that otherwise impact the amp performance. Ironically, low ESR film caps can be worse as there is less series resistance to damp the ringing. It seems even the humble power supply has a lot of nuances and things that can impact sonic performance!
That's TOTALY correct signor ...
"Humble power supply" is how I always looked at it, clearly there is more going on :)
I remember someone having to put small cap and possibly series resistance across filter cap as a snubber to control bad behavior of main filter cap.
Didn't even know about Hugo Kant but after your description i've become a fan.
Solen caps are great. I love them. Affordable and great sounding.
Interesting result. I would recommend trying some X rated safety capacitors. Even in coupling locations, these for some reason, sound drastically better than standard films.
The sound of capacitors - omg Skunkie, I went down that rabbit hole with my preamp and amp builds sometime in the mid 90s - the hours and dollars that I've spent! With regard to their use in the power supply, my experience is yes, definitely the differences are audible. In all cases except for one power amp I ended up choosing good quality electrolytic caps as my choice for the most natural sound. My experience with film caps (Solen mostly which I found to be very good all around caps versus the boxed not so good sounding Wima caps) was that a touch of high end tip up could be gained but at the expense of grainyness. With regard to bypassing, I don't recommend it. In every case i could hear the bypass cap - either tipped up highs again with some grain or an added flavour to the mid range if the value was higher. I have also tried oil type motor run caps - though they may add some smoothness to the mid range, speed / dynamics suffer. Also, be careful with the size of the film caps, I found that if the value gets too large, like the 39uF film I used on a 5AR4 I experienced a kind of pumping action to the bass, no problem though with 2 x 22uF.
Great info!
The difference is the ESR. I have tested Capacitors in a lab for 23 years. I have also modded lots of less expensive equiptment with capacitors to milk all the sound possible. The power supply has a huge effect on sound. That lower ESR takes away the lag in delivery of power and increases the slew rate at the higher frequencies bringing out the detail. Another thing I have found is dont use any parts with copper clad steel leads(cheaper caps used to use that alot) Or any ferrous metals where the signal passes through. I am impressed with your channel thinking about building the KT 120 monoblocks.
I have heard that womens hearing is much more acute which may be why it sounds harsh to you. Myself I love that detail.
These two caps have the same ESR. That would indeed been a simple explanation if that was the case.
Hi Stephe! We were using 'Wima MKS' 47nF 1kV caps & got a 10 to 30% failure rate!
We swapped to blue 'Multicomp' caps & had no further problems since.
This was in a high voltage power power supply, but not for an amplifier though.
Good to see you got your great bass back by replacing the 'Solene'!
I should add that our problematic 'MKS' series Wima capacitors had the identical hard red plastic outer casing as your dodgy 33uF cap.
Their epoxy sealant was a yellow colour, rather than red which yours has. So yours may be from a different series, possibly the 'MKP' series.
Older, identical looking 47nF Wima caps, from a few years ago, gave us zero problems. We never had to replace them. So they've messed with the metallised polyester film, or p.e.t film, or whatever it is.
A fault in these capacitors would cause arcover inside the eht transformers, which would destroy them, & reflect back through to the primary winding & blow the hv switching transistor & all associated driver transistors. Catastrophic destruction of our hv psu's! They'd go off with a loud "BANG!!!!", & blow the glass fuse like it had been spray painted silver & black from the inside! Sometimes even cracked the glass!
I would have expected MUCH more from a German made cap! Their latest batch were CRAP!
Just a guess here, but I'm thinking Class A or not, more current is going to be drawn by the speakers at the lower frequencies than at higher frequencies. This of course causes a voltage sag on the high voltage. The power supply having more capacitance being able to supply the voltage that bass frequencies demand with less sag by the speaker is probably being supplied by the better capacitors would be my guess as to why the bass sounds a lot better with the better cap.
As an aside if you're familiar with the pre-amps and amps built by Angela-Gilbert (on TH-cam). Part of the design differences that AG gets into is large capacitance arrays in external enclosures are plugged in as an accessory to the amps which supposedly increases the quality of the sound immensely. Those cap banks built by AG are somewhere in the few Farad range.
Very interesting video thank you
13:05 I think that's what most people involved in electronics (except very specific/niche/cutting edge areas) have anyway.
13:12 I think maybe an oscilloscope probe in the circuit would help here. One can also check caps on their own, do all sorts of X-Y curves stuff but unless one can somehow work out how the results of these measurements would translate in a specific circuit and operation, then I think it would be better to just measure the thing in the circuit and compare against alternative components this way.
All this is very useful testimony especially for those of us who go into circuit building thinking that a part is a part and that's all there is to it. We must keep in mind that most of the electronics art works because it is based on simplifications: there are just far too many factors involved in any given circuit for our mind to be able to process and take into account. Measuring and experimenting is key.
I think that the output feedback has been designed for the less responsive electrolytic cap from the PS. Adding the small bypass means faster response at the output and feedback will enhance thus having overshoot.
IMHO bypassing the power supply caps doesn't sound good. YMMV
I had a similar experience with my tubelab sse I built. I added a solen 47uf across my 40 uf motor run cap and it became really shrill sounding.
Thanks for sharing, this stuff is what keeps this hobby fun!
Hey, if you happen to have the impedance analyzer adapter card for the analog discovery, you could really go deep dive on comparing the capacitors and all their parasitics. The adapter is reasonably priced and very useful and capable. I highly recommend it.
Thanks for the tip! I just ordered this breakout board.
I wonder if the plastic cap is mylar or polyester, whereas the Solen is metalized polypropylene. I have noticed harshness with mylar/polyester caps when I used them as coupling caps...I would not be surprised if it made a difference with the power supply. I have always spent more time and simulations building and designing the power supply than the input and output stages. Thanks for all your work and sharing all the info you have learned...it can be a tough world in the internet.
I looked it up, that wima is a metalized polypropylene
www.mouser.com/ProductDetail/WIMA/DCP4I053007HD4KYSD?qs=gYqc6UMArTPQ5olMylyEvQ%3D%3D
Awesome topic and video. I'm so glad you share your real world experience with us. I love hearing about changes that impact the sound.
Hi Stephe.. The PSU is IN the signal path IMHO (current loops), especially for single ended amps... A 3uF difference can still be sonically heard IME.. Has the B+ stayed the exactly the same with a 30uF Wima DC Link input cap vs the 33uF Solen? Try the Wima paralled with a plain 3uF/630V Mundorf Mcap (white).. I only tried Kemet DC Links and they are not harsh, it takes time (break-in) for the bass to subjectively heard and felt though..
I will try to make a short story: two years ago my DAC died. I write the seller. He tells me send it back to me and as soon as I see the mailing proof I will send you a new unit. I receive it and try it out. It sounded really bad. No bass at all and thin sound. I write this to the seller. He says: I don't understand because I sent you an upgrade version with all very good capacitors. Let it play for a while a couple of days. I opened the DAC and all the cheap $0.15 electrolytes were really replaced with Wima red caps. I played music for 3 days in my basement. Then tried it again on my main system. He was right. The sound was even better than the other unit. The bass was as good but the mids and highs were fantastic. Cleaned up. To resume my story: caps need breakin'in. Not only the electrolytes (reforming), the other caps to! 50 hours at least! I don't understand why. Normal for a speaker to loosen the suspension. Electrolytes often need reforming but Wima's... ???? A film or polyester cap???... Next time play music for at least 50 hours before you make your mind as I did. My short story seems quite long. Sorry my friend!
Very enjoyable and interesting video, thank you.
The Solen PB are great in power supplies, I have them filtering for my front end tubes. My amp is a PP AB triode strapped EL34 amp, my rectification is SS) The character you describe of the Solen's is the same as I and others describe of them when used in the PS.
I was actually stunned at the difference!
This has been my experience. I typically bypass the second / B+ caps. To my ear adds detail. I built the ability to change out coupling caps in one of my basement builds. I was similarly shocked (by the sound changes). I think this is much more noticeable in se Amps, less so in pp amps, at least in my experience.
Keep going, it feels like snake oil… but clearly not.
I know right? And some of this will depend on the speakers if this "extra detail" sounds good or not too.
Another wonderful thread. I love your no BS approach with your honest opinion. I am in the process of choosing one of your builds. Is this project worth the cost and effort over your EL34 build which looks interesting. Thank you and keep the content coming, I am a big fan. BTW I build guitar tube amplifiers from scratch, have a EE degree and was in the audio business years ago (class D amplifiers at the beginning of the technology).
Well that was a bit of a surprise. But WIMA's are known to need some break in time...but just thought that was if used as coupling cap. But it's easy to forget that the power supply is the heart...you only modulate through the audio sections what it produces. I'm chuckling here because you're experiencing the same thing I got from using Mundorf coupling caps in my Kegger SE KT88...except I didn't know right off that they were the cause of some terrible high frequency distortion. And to make it crazier, vocals sounded awesome. I had some Cornell Dubilier 940C of the right size and replaced the MCap Supremes...immediately better. As far as bypassing electrolytics, I believe there was an overall improvement when I bypassed the cathode caps on the KT88's...but it was small. Now I'm dealing with speaker cables...they have capacitance issues also. Hang in there!
Input caps and the caps between the input tube and output tube make a difference on how it sounds. Look at BASS and Treble circuits in amps and you will see resistors and caps to change the tone output. Bass in the amp takes more current so you need heavy storage caps in the power supply for deep bass.
Capacitors are my bane. Those Wima caps are bright sounding caps IME, which works great in signal path of certain tube circuits to detail up the sound.
Doesn't surprise me they brighten up in power supply. Then there are those who swear by big oil caps in the power supply..
Yeah, I may have to try that. I used one on my 47 globe amp and it sounds great.
Wow, this was an exceptionally interesting video, and that's saying quite a bit because your videos are always interesting.
I recently replaced a similar WIMA cap with an Audio Note Copper Foil oil filled cap that was used as the coupling cap in my BrzHiFi 300B amps. I wasn't surprised that it made a substantial improvement, but the power supply?!? As you say, a coupling cap in the signal path is one thing but everyone almost always uses electrolytic caps in the power supply.
My first tube amp was a Dynaco Stereo 70 that I rebuilt, and adding film bypass caps to the electrolytics in the power supply completely opened up the top octave. So I guess that I should be so surprised.
I think that we simply don't completely understand what about capacitors affects sound quality and how to measure it. So, the impact on sound quality remains effectively magic. It can only be observed through trial and error. We can form some general rules but that's about it.
Pete
Upgraded my HIVi diy 2.2A speakers crossovers. Shrillness was removed by switching out caps with $15 each film caps. Then sound opened up and sounds like it has so much more air and life, and cleaner bass with the speakers plugged into 8 ohm taps. Can’t put my finger on it, but they sound so much better in 8 ohm tap of my R8, than 4 ohm taps. Run other speakers in 4 ohm and forgot to switch at amp.
hi if you have room for solen in the supply ,put only solen its sound better cause these caps have a very low resistance and this is very fast and keep stability all the time ,electrolytics change properties with use .
Great point!
So I wanted to tell you that I am really glad I found your channel. I watched you vid on the X7 Muzi. Of the ones you have seen can you recomend one 1000?
One more suggestion, to bring yr amp to another level, pls try and tell me your results! The Vol pot voltage divider has real impact to audio quality. Not sure what pot is used in yr amp? Replace it with TKD and add 10x TKD resistors to the 3 vol pot tabs.If yr pot is 100k, add resistors 10x (1 mg ) to pot in and pot out to yr driver tube. This shunt mod fr late Bob Crump is a super audio hack. Try it?
glad i watched. i was thinkin about what to do about the caps comin off the rectifier on a coupla amps im messin with--do i replace the can caps? restuff em? leave em and stick new caps under the chassis? i may try a 4uF solen film cap.
I am wondering if a comparison of the Equivalent Series Resistance (ESR) values would provide useful data in this case.
Also worth noting, some designers place a smaller value cap (of different material type) in parallel with larger ones and then note sonic improvements - but I do not recall if this is only in the signal path or also in the power supply section.
Or degradation. Bypassing does not always = better.
Think of any audio amplifier as nothing but a modulated power supply. Everything matters.
Power supply can well be viewed as a positive feedback loop if that help with understanding the effect. Mullard called tubes valves as in valve regulating the flow from the power supply. Different perspectives help solve problems. The ESR of the Solen's are a bit different than the Wima's prefer ASC to both of them .
What size series asc are you using and do you parallel them?I ask because I have some 20uf 400v from all electronics as well as some 60uf that I forgot the voltage.large polypropylene.
@@johnstuchlik5828 Some series resistors give a better filtering than paralleling basic Radiotron stuff. Stay small for the first one to keep from having big peak current spikes for the diodes. Save the big ones for the later stage of the supply . Any of the ASC will work for me this is in the range of flavor to taste. Common diodes state 32 mfd as max for first cap the 20 is a good choose most of the time for the first one .
@@johnwilliamson467thanks for reply.im going to heed the tube manual first cap size limit w/o surge resistor.i am planning to use 5v4 on my low watt projects mainly because I have some and like the slow warm up.first cap will be 7.5 to 10 first cap, probably try different types,motor run ,electro,poly, what I have on hand and use larger or parallel smaller after choke/resistor.definite going to look at b+ with scope for problems after watching this video and reading comments.
Analog discovery 2 has an impedance analyzer feature you could use to measure the complex impedance characteristics over operational frequency. You probably knew that but it would be cool to see how they measure differently. I bought the little breakout board but you don’t need to if you can supply the resistance
No I didn't know that, is there info online somewhere? I did just order the breakout board, thanks for this info!
That's a surprising finding for me, like you I thought a power supply must give the correct voltage with sufficient current, as close to smoothe DC as possible and that's it!
In fact I have been very tempted to use a solid state PSU when I eventually get to building this as there are advantages with space and filtering but this has me thinking...
Me too, I was really shocked.
That's very weird being both are film caps - also being in the power supply never would have thought you'd be able to even hear a difference- I never heard of Solen caps failing before they are very good caps- but your's got physically crunched. I too would like to see what putting an electro in that 1st caps place will do also as it will store up more energy even if it is a low value 30uF -33uF but the films should be better cap at filtering the ripple. Weird
Yeah, I seriously just thought I was replacing a film cap with a film cap. The difference was NOT subtle! Like I hear was less difference swapping coupling caps around.
I agree with Stefan. The difference between the capacitor types will affect the signal. Every with a charge has a Magnetic and Electrical field offset by 90 degrees. So the fields around the aluminum plates is affected by the shape. The shape of the capacitor could affect the density of these fields and in turn affects the signal. There could be constructive and destructive interference. Remember all energy is moved through fields and not the conductors. Electrons don't really move from one end of a wire to the other. They basically vibrate. There is some electron drift but this is pretty slow. Energy is transmitted through fields as well as audio signals. So each type size and shape of capacitors will be dissimilar, because their fields will be different. Energy moves through fields, not wire. This is why overpriced audio cables are a scam.
I hate to say it, but I suspect there are other impacts in the power supply. That one wacky thing that china hifi said about large chokes choking the sound. I think there may be something to that. If you have a cap inductor, cap filter the power supply, the second cap needs to be big enough to supply power for quick demands in current on the power supply. I suspect with larger inductors, an amp needs larger caps, or the choke will not allow current to change quick enough for good dynamics. Just a hypothesis at this point. Chokes also add resistance which result in voltage drop with increased current. Not a big deal in tube amps, but the last cap needs to keep up.
Hey Skunkie, what's the name of the third track of Hugo Kant that is so mind blowing to you ?
Clouds
Yes Steph , ...........your ears don't lie !
There are numbers and measurements ...........but in the end experience is the teacher ...
The bypass hype was a rage and hype by audiophiles in the eighties, every cap had to be bypassed , often even with different and multiple bypass caps .........today we know different .
Different caps with different values and materials as well as different technologies have a different tone that often don't mix well ( someting like different cone materials used on loudspeakers ) changing tone /timbre that shifts with rising frequencies and they have also different resonance points witch when combinerd " CAN" create stress , incoherence of the harmonics and musical tonality becoming out of a natural balance , ...and that's what happens and is perceived as one of the causes of rstress /listening fatigue ...when this happens it tells that something is wrong .....yes one may notice a daker background and excessive detail or hyper dynamicas on only part of the frequency spectrum but it is out of balance to the nature and natural perception od the music .............in some cases this may work after countless experiences and listening results ..............
ONLY your ears will tell if you got it right ore wrong .............they don't LIE !
It remembers me of an incident that happened with a customer of an expensive Audio Note amplifier .........and oh boy these are not cheap ..The customer complained that after having a look inside he was upset beccause the filter "input capacitor " was only a regular standard quality brand in his very expensive amplifier .........he felt as being cheated by the company !
The answer of Peter Quortrup was ( and I agree ) that one buys such an amplifier for it's exceptionnal musical performance ..........explaining that the first most perfect high performance cap when choosen as the very BEST capacitor to use can be your worst enemy .............;the better this capacitors , the more current is draw
It affects class a amps the most but yeah supply caps matter. Supply is about clean dc with current capability, esr and transients.
It would be interesting to see if breaking in those Wimas electrically, would change the sonic signature. I know break in is a controversial subject, and I did hear you say that the new Solen changed the sound immediately, but it would be an interesting experiment. I put some Russian paper in oil caps in my Klipsch Cornwalls, and they sounded awful until broken in. Some say 300 hours, but the first 100 does the most. I’ve also been told that when breaking anything in, not to just run current through the unit, but also to turn it off from time to time. Thanks Skunkie!
Could be, I guess the main point was: This changes my thoughts of power supply design and it's influence on the sound.
Very interesting!
Perhaps you want to look into room, equalization wizard (REW). At the risk of spouting heresy here, I think a little digital signal processing could change any of the characteristics that you’re talking about being affected by capacitors. I have heard similar discussions involving cables. sounds awful lot like people are shooting in the dark trying to get a certain sound with caps and cables when you can zero in on the kind of frequency curve you like in your room with DSP. Personally, I have tested my hearing and know how it falls off in the higher frequencies. I created a compensation curve in Roon and I found I like it quite a bit. Of course at first it will sound a bit bright, but very quickly. It began to sound right to me. Just something he might want to consider.
Any type of EQ box I've inserted between my turntable preamp and the amp adds a veil to the sound, even if the EQ is better, it masks part of the sound in the process. That's been my experience anyway.
Admittedly DSP (digital signal processing) makes more much more sense when the sound source is digital and you can get it done prior to the digital to analog conversion. But since many of us are going to streaming as a main audio source it is very doable and Roon is my tool of choice for doing this since DSP is part of the package anyway.
On the other hand Room Equalization Wizard is a free piece of software you use with a calibrated microphone to chart the flatness of the frequency spectrum coming out of your speakers. Similar to how you use you oscilliscope to measure distortion directly from an amp. So REW could chart the change in the signal created by those capacitors you were talking about. Then my thought was, at least for digital signal sources, some DSP could be used to make the same modification to the sound.
I've been using it to learn what sort of frequency curve I like so that I can duplicate it across various pieces of listening gear.@@SkunkieDesignsElectronics
I know what REW is and have used it testing speakers. And in testing, this software/microphone combo isn't sensitive enough to decern some of these nuances.
Remember; the amp's circuitry through to its output stage is just a delivery system to dump power supply energy on the speakers at the rate of the input signal. What's first? Look backwards from the output into the transformer, and circuitry into that pool of energy! What do you see?? Along the way, what do you encounter in your mind's eye?
Elna or Nichikon for electrolytics!
I've been using nich'ies
Nice job. A video on capacitors and application would be great. Bypass caps, etc. Does capacitor sound change over time? I often hear this and maybe just when out of spec.
I think majority of burn in is for metalised caps and electroyitic only, since the metalisation needs an applied voltage to self heal and form a better coating
Howdy.
I would suspect the ESR of the cap.s are responsible. Equivalent Series Resistance. The ESR can be large for regular el. cap.s. Film cap.s have way smaller ESR. So, they can take current and deliver current fast.
I would be careful about changing the first filter cap. , between the rectifier and the choke, to a film cap. The low ESR may cause current surges through the rectifier tube it can't handle. But yes. After the choke the filter cap may be a film cap.
Regards.
I stated that the ESR both stated and measured was the same. Also measured the impedance vs frequency and the curves look the same. That's why I said I was baffled, as something other than the obvious is going on.
@@SkunkieDesignsElectronics Howdy again.
Hm Ok. I was going to say there is an ESL too. Equivalent Series Inductance. But as the impedance vs. frequency is checked to be similar, it likely covers the ESL aspect.
Entering the speculation zone. I think dielectrics have hysteresis as magnetics have. These produce resistive like losses. In English. AC resistance is not the same as DC resistance.
One way would be to check the voltage - current phase shift. But I would think that is difficult to do with regular lab. gear.
I would understand that radial lead cap.s show more ESL than PCB mounts.
Regards again.
Clearly there is SOMETHING different, but it's not seen on basic measurements or specs.
I had to come back to this video as I'm thinking of switching to Solens 47uf as the first caps in the power supply in a recent build that I really like..if I have the space. I've been told in the past to just shunt or bypass the Electro's with film's but I've never had luck with bypassing..in any position, Though I've only tried it twice now.
What cap type & brand has been your go-to or good results as far as coupling the 2 tube stages?
Mundorf EVO alum/oil. Don't be tempted by their higher priced offerings, IMHO they don't sound as good. And like you, I've had zero luck bypassing 'lytic caps with film ones.
@@SkunkieDesignsElectronics TY Skunkie
40 yrs ago.... on cyrus and nad ss amps,,, we proved this too
Welcome to the party :) The electricity from the wall is what arrives at the speakers, so the sound is shaped in the PSU first. The more caps you upgrade, the stronger the effect of each will be. When it comes to the brands of audio-grade ones, in general "you get what you pay for". A common style when you buy the really fancy ones is to double the voltage and half the capacitance. The higher quality electrolytics allow you to run them lower uF and actually get better sound quality. have fun
That's indeed PART of the truth and one of the golden rules know by those who know........
Yes I can't give it all away because the fun is in discovering :) @@frankgeeraerts6243
Cool, thanks
Gat a question for you
I've noticed a difference in capacitors.
But what about resistors?
A lot of exotic snake oil out there.
So what brand of resistor do you use for anode, cathode, grid ?
Thanks in advance.
I honestly haven't gone down that rabbit hole. I'm using Vishay pro series.
Curious: The 33uF Solen film cap you used , it was used as 1st solo cap after rectifier tube, ( as in original scheme ) not as bypass for a 1st electrolytic in PS , right ?
Correct. 1st cap all by it's lonesome self
@@SkunkieDesignsElectronics Thx!
another great video !
I have never seen anything good or measured written about wima film caps. Polypropylene in oil motor run cap would be interesting to try.
Yes and I've used them in one very nice sounding amp.
Try bypassing the electrolytic capacitor on your cathode resister with a film capacitor.
Tried this, made the amp sound shrill and fatiguing. Some people believe this always improves the sound of an amp, I've not found this to be the case.
Awesome. Awesome.
The DHTs would be my guess. It stands to reason that you are running filtered DC directly to the cathode, no?
The cap wasn't part of the cathode/heater voltage supply.
the dynamic impedance of the power supply matters.
But the measured impedance curves of these two caps are identical. There is something else going on.
Stephe, you're transitioning into an audiophile! Yep...EVERYTHING little thing matters and affects the sound in some way.
LOL
A power supply tube shouldn't impact the sound. The choke shouldn't unless it's with a cap and then in makes an lc filter. So yes a cap will have an effect. Also you have to remember caps need running in. It may sound harsh at the start but will settle down.
Is it possible that the single ended circuit is just more sensitive to these kinds of issues...like parts quality or materials? I've never had another piece of tube gear act like this before...and I've done cap work on Dyna Mk III's and IV's. Of course, the Dynaco's didn't have the sound quality either.
For SET amps, the power supply is in the signal path
I get that to a point, but "in the signal path" I always was thinking more like inline with the input signal. Like if you put a scope probe at this point, you aren't going to see any signal.
Solen are nice musical caps, just not expensive enough for some audiophiles.
I wasn't crazy about them as coupling caps, they are very nice for PS use.
Some of the best power supply caps I've ever used are motor run caps for AC compressors. NOT motor start capacitors, polypropylene motor run caps. Dirt cheap and top notch. From what I've heard, the LAST caps are the ones that matter most (the ones that actually supply voltages to audio stages) and the first ones are less important. YMMV as always.
See that's what I thought too, but this first cap made more difference than any other cap change I have ever done. There are two stages of choke/cap filters after this one and this amp has the power supply split rail after the first choke. The front end each channel has it's own RC filter, yet somehow this first cap made a HUGE change in the sound of this amp.
@@SkunkieDesignsElectronics Definitely worth looking into.
What cap did you switch it to in the end that got the goodsound back?
Solen "Fast cap"
@@SkunkieDesignsElectronics Thank you. Have a wonderful day!
I have been told you can’t Mix those DC Link caps with electrolyte caps. If you use one DC Link cap in the power supply you have to use them in the entire power supply. He told me “ if you mix the two kind of Caps together weird shit happens “ . I haven’t tried it my self, but so far He seams to be correct by your outcome. 🤷♂️
Everything can make a difference. Us Audiophools learned this a long time ago. It is very harmful to the pocketbook. Thank goodness for diminishing returns. Listen and choose what is important to you.
As far as the power supply, remember, the input signal itself is not what is amplified, it is remade, at a higher amplitude or otherwise changed, at each stage. Each stages output is just the power supply being modulated. The power supply IS what you are, ultimately, listening to.
Dan
I was really just shocked how different this sounded, WAY more difference that any coupling cap change.
did we take any masurements ? every cap have 20% +- value , and the ERS is different , use smal caps parralell over the catodes in the preamp tubes to get the shrill out
yes measured the caps obviously, same value and ESR. And adding small caps across big caps does the exact opposite.
@@SkunkieDesignsElectronics across the anode resistor over the gainstage (pretubes)
First you say the cathode, now you are suggesting putting a cap across the plate load resistor?
@@SkunkieDesignsElectronics that anode cap was just to get rid of the shrill hights ,my mistake i meant anode
As I understand it, some caps require a burn in period of up to 100hrs before that bad behavior goes away. Maybe test that theory out with one of those bad sounding caps.
IMHO there is something wrong with a part that sounds bad until it's "burned in" for 100 hours. Now maybe a part that sounds good and after some time sounds slightly better I can deal with. Or even a part on initial power up is a little cringy but after 30 minute to a hour clears up. 100 hours of music play time is absurd to expect anyone to suffer through.
@@SkunkieDesignsElectronics I actually agree with you on that. I do know that materials vary in the time it takes for them to stabilize. Speakers surrounds are a notorious example of this. I am currently building two audio note M77 clones. One has precision matched to .001 resistors ( z foil, audio note silver and other high end types, and audio note copper and kaisei caps also matched from sonicraft as well as Russian polystyrene and silver mica caps in the phono section with a personally made precision resistor attenuator, shielded tubes, shielded power supplies, silver wire everywhere, shortest wire runs everywhere, military grade soldering and device mounting, yada, yada, and the other with more of a regular build. My fear is that after all of that work and expense, it will lay a bunch of crap on me and give me massive buyer’s remorse. Imagine spending well over 1.5k in parts and having to hope it will get better because of my choice in caps. I’d be crazy if the regular build with less exotic parts ended up crushing it.
I'm learning that some of these "exotic" parts aren't better or even as good as "normal" ones.
Hmm, perhaps generates the WIMA-Caps a very high frequency on the direct current signal to the secondary site of the Amplifier. I mean, the d.c. current with HF is with the WIMA-Cap not clean enought for the Tubes. You understand, what i mean? It's only a idea, but i'm not sure ... and please, you all excuse for my bad school English 😵💫!
It could be impedance or somehow frequency related? I bought an advanced breakout board for my analog discovery to look further into this!
I assume the Solen cap is a "Fast" cap - is the Wima fast ("Fast Capacitors" is trademarked by Solen so WIMA can't use that exact term)? Solen has a datasheet explaining the Fast concept. I have no idea if it makes any difference.
Yes it is a "Fast Cap".
parts quality is everything. KT88s can be too much going on the top end.shrill.
Just like any engine.........
Did you run some output tests to see if the sonic differences show up in objective testing?
Objective testing ................what's that ?
I need to, and then figure out where to look for this very audible change. And this was clearly not a case of expectation bias, I wasn't expecting any change and was in fact baffled that it changed lol
@@SkunkieDesignsElectronics Could it be down to the slight difference in the values between both caps? Do you have some 30uf Solen caps to test this?
That's not it, I tried but 33uf and 22uf solen caps and those sound the same.
New caps need some burn-in time to sound better. could take 20-40 hours.
Just put a 1 uf bypass
Yeah,... no. That will have zero impact on the bass response.
You should try 0,022 or 0.047uf instead lf 0.33uf. To big bypass will give that shrill sound and sign that its to big value. Alao the wima caps are kinda famous for having extreme long burn in time and i usually think its bs talk to in wimas case its not. I do not like them i signalpath that much because theyre to laid back for my taste
It's not a bypass cap, it's the main reservoir "first cap". And as I said, I have tried bypass caps across electrolytics in the past and didn't care for the results.
Novice here. Maybe the Wima cap needs several hours of burn in?
Honestly, that could be the case and I didn't do that. I guess I was more interested at the time to put the Solen cap back in and see if this actually did change or was it all in my head!
@SkunkieDesignsElectronics I wonder if they have burn in rigs... so they are burnt in so she doesn't have to wait.
🖐
👍♥️🫂
just buy a gryphon apex and be done xx