Improving Signal to Noise Ratio On The Dynaco Mark III Vacuum Tube Amplifier
ฝัง
- เผยแพร่เมื่อ 29 ก.ย. 2024
- A simple fix that may help you reduce the noise in your vacuum tube amplifiers. After studying the original Dynaco manual, Dynaco does not show the tab of the multi-section electrolytic capacitor soldered to ground. Dynaco shows a short black wire to the nearest 6550 tube socket for ground. In the original build manual they do instruct the builder to twist each of the four tabs on the can. Over years of time this will loosen and oxidize and will not provide a good electrical ground. Soldering one of the tabs to the chassis with a large soldering iron does the trick and gives a S/N ratio of 90 dB as the Dynaco specs state.
- บันเทิง
Now to make it even quieter, change ALL the 1/4 and 1/2 watt carbons to 1watt metal films. It will be worth every penny.
Great troubleshooting and demo of grounding issues with nice use of instrumentation in the setup!
Is the error signal output equivalent to common mode noise for grounding issues? Could it be separated with a differential probe setup to measure common mode noise? Eliminating the need for the Tek AA501 SA? It would be interesting to compare results.
Most don't do a SN +N measurement to verify specs. It probably is the best overall metric for success in an amp. I'll have to add this to my diagnostic tools. Keep em glowing.
I enjoy all your videos immensely and I post them on my favorite audio forum..I always increase capacitance in the power supply and the reason being is under heavy transient dynamics from the new 95 and 100db listening material as opposed to when the amp was built,it can collapse the power supply at certain frequencies especially on lower efficiency speakers.
This is a factory built unit and the riveted sockets and terminal strip is the telltale sign.
Lousy video quality. Cell phones are NOT quality cameras. Get a decent camera and a TRIPOD!
It may be factory built.... I see several areas that are riveted and not screwed in, such as the tube sockets. While it is possible the original owner used riverts as part of a kit ..... its doubtfull
+Jordan Rubin I agree and have noticed the rivets also. The octal tube sockets are becoming pretty lose and I have considered taking the rivets out and putting screws in. I didn't build the amp and got it 30+ years ago at a garage sale for $5 if I remember correctly. That long ago, (which seems like yesterday...) most all tube stuff was considered "old" and junk and non-repairable and I am sure much was just thrown in the trash. I have always horded the old tube equipment and still do.
hi, i have a dynaco mark 3 to 110v, you have an idea how to convert the 220v to 110v,
Im a dummy'''' But''' ' Couldnt you just Ground The Thing you wiggled to the Chassis'''?Solder on a ground' from & to..
(silver capped) ~ I Know Thats lame but,,,? Providing the cap is Ok ? Would this Somehow' stablize the ground properly,,,?
Chrisandra VOX That is exactly right. In this model amplifier the silver can needs to actually have its tab(s) soldered to its metal mounting bracket or directly to the chassis if it is steel - can't solder to an aluminum chassis. In my case, the can was not making a truly good electrical connection to the chassis (ground). Instead of being directly soldered to its mounting bracket, (the original Dynaco kit instructions show it grounded through a short wire to a nearby tube socket, which it was). But we can see that this did not provide the same noise canceling ground connection that the soldered tab does right at the base of the can. This is certainly not going to be the case in every amplifier. Possibly where the can is located next to the 5AR4, power transformer, etc. may have a unique effect on this Dynaco model amplifier design and build.
hi, what is the ripple acceptable for a power supply tube amp !
I just repaired a pair of these last month for a fellow in Boston. Very important point with these amplifiers--do NOT use a directly-heated rectifier tube! The voltage rating on the capacitors will be exceeded until the 6550's warm up. The amplifiers that were sent to my shop had 5U4 rectifiers installed. The correct rectifier is the 5AR4/GZ34, which are indirectly heated and whose warm up time to emissions track the 6550 output tubes. Another thing about using 5U4s is that the distance from cathode to plate is much greater, hence the voltage drop through that resistance path is greater. Which is why use of this rectifier causes the amplifier to fail to make specified power. The 5AR4 has lower resistance, and thus better voltage regulation. It will make rated power with this rectifier.
If the owner wants to use solid state rectifier, he can, but the capacitors must be replaced with 600V DC ratings. I did this for my customer, leaving him the option. The Mark III will make 72 watts with a SS rectifier and 55W with the tube.
the one that REALLY kiked my butt was this, a factory speaker cable, amp would oscillate n be nasty, nuthin i did worked, finally grabbed the speaker cable, n twisted, n voila..
I completely agree with you. It it sometimes the simplest problems that kick us around the hardest. We are looking for complicated things and it is something simple right in front of us. Even simply good visual inspection can allow us to fix things that otherwise defy high-end laboratory grade equipment and analysis.
@@ElPasoTubeAmps love your shop, id need a mit degree just ta run the stuff, i was an electrician on DC-8 DC-9 F-14. F-4 aircraft, the ATs n AQs could fly thru gear like that, i just use a scope, n a meter, in repair field, u can get by i guess! Appreciate your videos n work, always learn sumpin new from you! My best to you n your family n a very Merry happi blessed Christmas wish..,
Ive got a danelectro guitar amp that has a hum coming in on the first stage 12AX7, during testing I noticed accidentally, when attempting to add a 10uf cap from ground to DC that splits off to the plate as an extra DC filter, there was a small spark and hum went away, i ended up not adding the capacitor but dont know what happened to fix the issue?
That is interesting and I sure don't know why. I have also had nutty things like that happen to be and after being puzzled for a while I just shrug my shoulders and go on. If it happens again, you might try to determine if the hum is 60 Hz or 120 Hz before attempting any remedial maintenance. 60 Hz is likely a filament issue and 120 Hz is going to be a PS problem.
No need for all that equipment. You would hear the grounding issue.
I can certainly agree that careful listening along with a good visual checkout and wiggling/tapping of components can solve problems without the need for complicated equipment.
Thank you really learning a lot, I tinker with old audio and don't understand how to value the caps for power supplies it's a little confusing. how much can I raise the capacitance? also, if you can explain how to sub diodes for tubes with a (Zener or thermistor) (?) for spikes that would be great hands on. again thank you. P.S. my motion sickness wasn't too bad (Ha, Ha).
Bobbyj438 There is a common misconception propagated across the Internet that more capacitance in power supplies is always good. First, it probably isn't necessary and it may not always be good. Vintage capacitors often have a have a tolerance rating of -20% to +100%. If you think about that, this means a 100 uF capacitor may actually be 80 uf or it may be 200 uF or anywhere in between. I think this should illustrate a lot. On the other hand, 40-50 years ago, practically everything was vacuum tube, which means higher Z loads and lower currents so PS capacitors could be less, i.e. 40 uF at 500 volts, for example. It was much more difficult to make high value (as in uF) capacitors back 50 years ago than it is today. Economy always has and always will play a major role in manufacturing. So, designers made equipment that was adequate but did not oversize components as we naively thing they should have. Think about this - 50 years ago if you had told someone that you could hold a 4000 Farad capacitor in your hand, they would have laughed you out of the room. Back then a 4000 Farad capacitor would have been the size of a Greyhound bus. Nowadays, ultracapacitors (en.wikipedia.org/wiki/Supercapacitor) are common and cheap. Sorry for the digression - bottom line is, odds are, you do not need to raise capacitance in circuits. It will not improve them. But, if you have, for example, a 50 uF capacitor in your power supply of your tube amp and you put in a 100 uF capacitor (with equal or better voltage rating) it will work fine but probably not better if you are able to measure its performance objectively. As for using diodes for spike suppression, I think that might be worthy of a separate video but ultimately an unnecessary addition to most equipment. A thermistor (thermal resistor) in series with the 115 volt AC line is almost always a welcome addition as it will allow a "softer" power up. In a few videos I allowed YT to smooth out my camera motions and I got more comments than I do by just moving around as slowly as I can. Mounting my camera on a tripod just doesn't give me the ability to explain what I want. I do appreciate your comments.
ElPaso TubeAmps thank you for your reply very informative. my reference to diodes was as a replacement for the rectifier tube. the thermistor as part of that modification. Sometimes what I'm thinking and what I write isn't expressed well. Looking forward to future videos.
Also, did you match the input signal impedance to 470Kohms in your tests?
+Jordan Rubin I do see that the Mark III has a 470K resistor on the grid to ground and no blocking capacitor on the input but the output impedance of all of my audio oscillators is 600 ohms and will swamp any high Z it is connected across.
many thanks
Thanks ElPaso' I was looking for a common sense resolve to your issue''' Im a tinkerer w' Amps' No Pro''' Thanks for your answer !!! :)
great vid
Hello. I'm in the market for a dynaco mark3 set with a preamp. I am just asking for advice on how much should I try to spend on the 3 components knowing they need to be rebuilt. And how much you would charge to upgrade them in general.
Thanks for the great video. I'm going to verify my S/N ratio using my 8903 as a comparison. I'll be checking my Dynaco ST-70 and verifying my cap is soldered (can't remember right now).
Nicely informative . please work on the camera work zoom out to stop jurking back and forth very distracting . The voice over however is spot on as they say.
Nice video, David!! Thanx for sharing!! :)
As i´m always sayin´: Audio equipment needs *propper* and *good* grounding.
This is the very *first* rule when designing/restoring audio units, regadless of its kind.
If you only rewire the ground in this chassis (as the gounding looks /sorry/ just awful in here) you would end up with something very close to -100dB Overall S/N, turst me.
I´ve comment this about 10 times in various videos:
If you use any sort of metal chassis, do *not* believe that this chassis provides you a good ground plane for free, as this is always *not* the case!!
Instead, use a solid copper wire and solder it (or clamp/screw it on - as it´s even better) to the very fist filter cap and as close to the filter cap as possible.
This is why top notch filter caps has screw terminals instead of just wires. The connection resistance would be about 5 to 10 times lower by using screws instead of a solder joint.
So next is to use this solid copper wire where-ever you need a ground and use only this wire no other. Go all the way through your chassis. Even connect *every* jack/outlet/pin/terminal ground to that wire. *BUT* do *not* connect any ground twice with that wire. And do not return this wire to the filter cap, leave it open at the last ground you soldered onto.
You´ll end up with a very quiet, nearly noisefree amp.
In my very last Amp, the TinyTubeAmp, i´ve done this ( with a 1.5mm diameter solid Cwire) and even with that aluminium plate and my tiny aluminium chassis, i´ve archived about -88dB TS/N. If i would shield that open frame switch mode power supply and its cables i would gain another -12 dB TS/N and another -3dB when i shield the open back side of the chassis, i tried that with a thin copper plane but first i need to put everything else right in place before install that copper plane finally;)
I was into some really top notch old military Tube pre amplifiers for microphones and they done this like that, by using a 5mm diameter soild copper wire, and easily archived about -120dB S/N ratio even when they had a overall gain of 60dB the output was absolutely quiet. The input jack was grounded (=shorted) and max.input gain was selected..:)
Thanks David for this video i have three scopes i am going to hook one of them to my AA501 thanks for the posting.I hope you do a video on test a ampifier in bridge mode it seems that i can't find any on you tube i tried every thing but i failed hope to see one soon. Thanks
had a ampeg bass amp kik my rear end like that too!
thanks for the video, I recently stumbled upon a similar problem on one of my tube amps
Thanks for your many, many video all of which are very interesting and informative.
I would like to know what it took to make your HP8903 work with a computer. What
hardware is involved? What software program are you using, etc. ???
Thanks!!!
John Petzold Hi John - I have made several videos on the HP 8903 that may help you - When I started using the one I have, I found only one (1) video on YT and it was just showing it in use and not how to configure it. They are truly a jewel of an instrument as you can read audio voltage, THD, S/N, scan frequencies and capture it on your PC, read power directly, etc - no doubt at the leading edge of the technology in the '80's when they came out . Check out this video - th-cam.com/users/edit?o=U&video_id=8GD-FU27wpk
I am using the same configuration with high efficiency speakers and have been working with hum at 120hz and 60hz. It has been greatly reduced, but I am running home to get out the soldering iron!
You might keep in mind that the 120 Hz is coming from your power supply as it is full wave and soldering the cap to ground will hopefully help but the 60 Hz hum is possibly coming from the filaments of the tubes.Grounding one side of the filament line may reduce 60 Hz hum. You have to experiment which side of the line reduces the hum the best and even reversing the filament line on the 6AN8 may help with the 60 Hz. A hum balance adjustment does not necessarily do the best job. If you have a way to measure S/N you may be able to empirically get the hum down to close to -90 dBm. Good luck
Very good to know, thank you'
another great video.....
Ha, ha! Occam's razor !
I've dealt with the same issue over the years. Grounds loops in the chassis created by the sum of all the returns are the culprit behind this power supply noise. If all ground returns are not at the same potential, this is the result. As you demonstrated here, the system has a definite zero ground point which is not always at the center tap of the mains x-former where in theory, it should be but is shifted slightly to a higher potential due to ground loops in the chassis. Connection of a stages ground to any other point, raises that stage slightly above ground. Keven O'Connor's book The Ultimate tone, addresses this phenomenon directly. When I refurbish old tube amps that are to be used for recording, I always lift all chassis grounds and rewire as a star grounding system attaching only the c/t of mains x-former and beginning of the star grounds to the chassis so that all returns flow not through the chassis but directly back to the center tap. This eliminates the ground loops and makes an eye opening difference. It is a labor of love indeed. As you demonstrated here, in most cases, returning the ground back to it's proper point on the chassis is all that it is needed but in some cases, one can never find the culprit so lifting all grounds and rewiring is sure to get them all.
bain5872 I think your explanation is great - thank you. As a young fellow I built amplifiers and grounded everything at the most convenient chassis point and invariably ended up with an amplifier that hummed and nothing I did would reduce the hum. I did learn to do as you mentioned, and string a buss around the chassis and ground the buss as one point. Some believe at the CT of the PT and others recommend at the audio input connector. In any case, this approach does provide options and the results are much better. The last big amp I built and sold was a pair of KT90's which would do 90+ watts before clipping and I used a wire buss in it - it was quite although I do not remember the S/N. Always appreciate you input.
ElPaso TubeAmps
Yep, me too! It took a bit of head scratching, thinking I was going nuts and reading to finally figure out why my electronics theory was not matching my real world experiences. I discovered a small article about ground loops in a publication which I can't recall now but I did not fully understand chassis loops until I read the book I mentioned earlier. These loops, at high voltage, can be no fun for sure. Manufactures selected the chassis ground points for the lowest SN possible in order to reduce the cost of using a star grounding system however, it's the cheapest but not the best grounding system. If one moves a chassis ground point it usually increases SNR considerably. Thanks for sharing and by doing so, I'm sure this will help someone in the future.
Excellent!
Great Video Mate! Really enjoy your videos! Thanks for all the effort you do!
Wow Good Stuff.... Thank You Way Good