TPA3255 300W Bench Test update - How Good is it?
ฝัง
- เผยแพร่เมื่อ 4 ก.ย. 2024
- In this video 'TPA3255 300W Bench Test update - How Good is it?', I will run a series of tests to build a specification for this amplifier. More Amps will be specified in the future. #TPA3255, #ClassD #classdamplifier #quantAslyum
Link to TPA3255 Amp board: amzn.to/3zVBt65
Link to GW Instek MDO-2204-EG: amzn.to/3Vo6UzW
Link to ICEpower 125ASX2: amzn.to/3Taksx5
Link to ICEpower 125ASX2 (lower cost): ebay.us/NPZUWK
Link to Power Supply (looks better than the one I have): amzn.to/3pgfQsp
Link to box for Class D amp project: ebay.us/ETpmJx
Link to T-Shirts and more: www.redbubble.c...
Please use links that cost nothing extra but will help support the channel:
TPA3255 Power Amplifier board: amzn.to/3LgRZTl
GW Instek 4 channel MDO Oscilloscope: amzn.to/3rvY7PK
MicSig Oscilloscope: amzn.to/3B4DcXe
MicSig Differential Scope Probe: amzn.to/3HvRG4A
Pico Oscilloscope: amzn.to/35UytvC
IR Thermal Camera: amzn.to/3J92LJu
Silicone Blue work mat: amzn.to/37a9txL
Kaiweets Store (using this link will help support this channel): kaiweets.com/c...
Other videos that might be of interest:
EEVblog Comments on Fluke Rant: • Capacitor Types in SPI...
Fluke Rant - Are they over priced? • Fluke Rant - are they ...
Fluke Rant - Better Test Leads Please; • Fluke Rant - Better Te...
Fluke TPAK Hanging meter Kit $42 or $10; magnetic hanger comparison: • Fluke TPAK Hanging met...
EEVblog is Santa Claus - Thanks Dave!!: • EEVblog is Santa Claus...
EEVblog comments;) Fluke Rant - are they Over Priced?: • EEVblog comments;) Fl...
The Iconic Fluke 87V: • The Iconic FLUKE 87 wi...
Fluke 27 Exposed and giveaway: • Fluke 27 Exposed and G...
Please support this channel that will not cost anything extra by using my link to Borbede BD-99, this meter is around $47 which is low cost for a high count meter with high accuracy:
amzn.to/3DRgUbZ
Links to support this channel that cost nothing extra;)
Fluke 87V link: amzn.to/3l4QUCq
Hioki 4256 multimeter: amzn.to/3idmexS I’ll show the enclosure that I plan to use for this project. I’ll also show the options for a pre-amplifier.
Link to TPA3255 Amp board: amzn.to/3zVBt65
Link to Power Supply (looks better than the one I have): amzn.to/3pgfQsp
Link to box for Class D amp project: ebay.us/ETpmJx
Link to T-Shirts and more: www.redbubble.c...
Please use links that cost nothing extra but will help support the channel:
TPA3255 Power Amplifier board: amzn.to/3LgRZTl
GW Instek 4 channel MDO Oscilloscope: amzn.to/3rvY7PK
MicSig Oscilloscope: amzn.to/3B4DcXe
MicSig Differential Scope Probe: amzn.to/3HvRG4A
Pico Oscilloscope: amzn.to/35UytvC
IR Thermal Camera: amzn.to/3J92LJu
Silicone Blue work mat: amzn.to/37a9txL
Kaiweets Store (using this link will help support this channel): kaiweets.com/c...
Other videos that might be of interest:
EEVblog Comments on Fluke Rant: • Capacitor Types in SPI...
Fluke Rant - Are they over priced? • Fluke Rant - are they ...
Fluke Rant - Better Test Leads Please; • Fluke Rant - Better Te...
Fluke TPAK Hanging meter Kit $42 or $10; magnetic hanger comparison: • Fluke TPAK Hanging met...
EEVblog is Santa Claus - Thanks Dave!!: • EEVblog is Santa Claus...
EEVblog comments;) Fluke Rant - are they Over Priced?: • EEVblog comments;) Fl...
The Iconic Fluke 87V: • The Iconic FLUKE 87 wi...
Fluke 27 Exposed and giveaway: • Fluke 27 Exposed and G...
Please support this channel that will not cost anything extra by using my link to Borbede BD-99, this meter is around $47 which is low cost for a high count meter with high accuracy:
amzn.to/3DRgUbZ
Links to support this channel that cost nothing extra;)
Fluke 87V link: amzn.to/3l4QUCq
Hioki 4256 multimeter: amzn.to/3idmexS
Ann Dustin Wilson " Heart " last night at Stiefel Theater with new and legend songs. She has the Voice and a great band. 🥰 Thanks for the TPA3255 update. 😎
Wow - you get some great music at the Stiefel Theater! Thanks for the feedback!
@@KissAnalog Thanks.
That holographic thumbnail is fantastic!
Thanks so much! My wife started doing my thumbnails. I really liked this one;)
@@KissAnalog I appreciate her creativity. Please ask her to continue doing so.
Thank you! You made her day;)
I'm still trying to wrap my head around all of this information but it still is interesting. Just bought a Topping D10s DAC to put in front of my Fosi TB10D. Should be interesting. Thanks for your extensive video Eddie.
Thanks for your feedback! I'll start testing my Fosi amps. People love them, so it will be interesting to test them;)
@@KissAnalog Thanks Eddie
Good work Eddie, really interesting series of videos. Reminds me of when i was trying to set my HP8903 distortion analyser to give accurate reliable readings, set up was critical, especially as i was using a PC to control it, that and the monitor were big sources of noise.
I have been looking at the QuantAsylum stuff very interesting, may look at getting one of those when i retire in a few years as i plan to do some of my own designs. 👍
Thanks - it sounds like you can feel my pain;)
I like the improvements you are making. Improving the connections will go a long way towards decreasing noise. Probably never going to be perfect, but that is good in a way. You are doing real world testing. The results got me thinking what is that amp really good for? It performed pretty poorly doing bass and highs. This could end up being an excellent series for the channel. Especially when you improve on designs. I wish I had something like a Quant Asylum. One thing I have always wanted to do was buy 2 lower cost amp kits maybe 15-20 bucks and build one as is and the other one use high quality components in the build to see if there is a noticeable improvement in sound quality.
Cool idea !
Thanks for your awesome feedback! I will do another setup improvement to see if I can get a repeatable setup - and then I'll see if I can improve on the amp. But I have another amp that has had great reviews - so I want to test this to see if it has much better results.
Actually, I'm mid way through that right now ... a pair of the Sui Yuan amps that Eddie started out with... Doing some pretty heavy mods on one of them and I will compare when done.
Thing is Eddie isn't modifying the amplifier... he's still trying to get his test setup sorted out. The connections he's changing are the connections from his QA test machine.
@@Douglas_Blake_579 Do you ha a channel please ?
@@andymouse
Nope. My friends all tell me I have the perfect face for radio... (Grin)
Seroiusly though. I thought about it and even tried to make a list of topics. What I realized is that I'd be good for about a month then I'd just run out of stuff to show...
Eddie my friend... you're getting closer.
Use a single signal lead out of your QuantAsylum analyser.. hook the center wire to the RCA center wire. Hook the lead's shield to the RCA shield ring. That will give you a clean input signal. You can check that by using a single ended cable back into your analyser and checking the inputs to the Phase Splitter chips (The op-amps)... it should read almost as clean as a loop-back test.
Now for the amplifier outputs... According to QA their inputs are balanced... so you would use two test leads... one goes to the hot side of the BTL bridge with it's shield grounded to the power input ground. The other goes to the cold side of the BTL bridge (of the same channel) and it's shield is also grounded to the power ground.
Ground noises etc. are less of a concern with a solid ground reference, because the whole thing is moving synchronously and won't affect your readings.
I use a very similar set up with my signal generator and oscilloscope and usually the readings I get are pretty much spot on the spec sheets for the various chips. Of course the QA unit gives you the advantage of charting results, which I have to do manually.
One note of caution ... IF the input signals are reading 20volts ... on a BTL bridge you have to double that, so it's actually 40 volts ... (40 x 40) / 8 == 200watts ... if you can't see the output waveforms, it is super easy to clip the hell out of an amplifier and not even know it. So I always have a voltmeter or oscilloscope handy and check the output voltage before testing.
Thanks Douglas! I did look at the output with my diff probes to be sure that I wasn’t clipping. I Enron dialed down the output of the power supply to see if I could get it to clip and I still had some headroom. I’ll work with the setup and see if I can get it any cleaner.
@@KissAnalog
Good stuff ... No amplifier will ever read better than your test setup... so getting it right is always the first step.
Hi Eddie, thank you for the information video. It is amazing.😃
Thanks so much! I appreciate you!
I have just recently purchased a QA403. whilst I was doing some tests in loopback I noticed a reasonable amount of noise. Moving the QA403 off the bench and on the floor in an electrically quiet area solved the noise issue. My bench is far too noisy for these measurements :)
Thanks for the feedback! This is a sensitive piece of equipment. I'll try isolating it;)
Nice work Eddie, did you ever get in touch with Stuart ?
Yes I sure did - thanks! You will see a video tomorrow that highlights Stuart;)
The new tpa 3255 pffb designs are even better!
Thank you - can you give me a link to one? I'd like to get it;)
The negative output of a switching supply will be connected to mains earth.
I had about 1kHz noise, not mains hum, in an audio installation at home, with computer, DACs DSP, amplifiers and such all connected up.
When I isolated the switching supply output from earth, the noise disappeared.
Safety earth was not compromised, because mains connections had the earth, but the noise path was broken on the low voltage side.
Audio signals and USB have a mains earth connection, but the PSU was one too many.
Thanks for the great feedback. Just to see if I understand; the PS input has 3 wires and one is 'Safety/Earth Ground' and it is connected to the PS chassis - is that correct so far? Then the output (which is isolated from the input) is also referenced back to the Chassis? So you opened the connection from the output (negative terminal) from the Chassis? Some applications people think it is safer to tie (reference) the output of an isolated PS back to the chassis. This can cause noise problems that you found, so disconnecting it would help. The PS input still has its safety so that if 'Hot' wire shorts to chassis it will pop the breaker. The isolated outputs, from what I understand from safety standards, can be isolated from the chassis ground as well. This is a systems question on how it is used. I have used an inductor in some audio amps between the neg and chassis to take away the noise but still provide that connection/reference. It sounds like you have experience here so you know what you are doing. Thanks for sharing!
FWIW ... the TPA3255 internal PWM switching is at 400khz.
20:32 the amplifier was clipping heavily.
Thanks Douglas, but the TI spec shows 450kHz with a +/-20kzHz tolerance. I think I measured 450kHz in a previous video, but I will double check. Splitting hairs but I want to be sure others that are reading this get the specs from TI.
That's why I had the Micsig connected again - but then removed it for the final measurments. I would swear that it was clipping because of the harmonics. I'm sure that that is what you are thinking too, but I measured it and it is actually a far ways from clipping. The waveform looks wobbled if you know what I mean. The peak is OK, but the sinewave is bent it that makes sense.
@@KissAnalog It would make sense if you are still floating all your grounds.
Also with the PWM at 400K plus... it would not show on the 100k range of your analyser.
That's better, less clutter and just looking at the left channel with a single ended input I think we now are seeing the amps true colours and the true abilities of the Quant at least much better than it was.Clearly the 50 Hz has gone and now the THD and THD+N are closer together, however the amp does fall (consistently) apart at 8 KHz but as you pointed out it looks a lot more like the TI specs. I'm a class A/B man all ways have been probably always will be so I hate this amp anyway ! to me it still looks like junk and that's nowt to do with you ED !! would like to see a simple cheap 1980's NAD or a Creek or perhaps a Cambridge Audio (I am that old and stupid) run through the mill or maybe a decent amp that you own ??..cheers! oh yeah I would prefer a Linear X Axis but that's just me.
Thank you! The noise is getting lower - but I think I've only affected the ambient noise caused by the Class D switching, 60Hz common-mode, and the SMPS. But the Amp itself still has a lot of harmonics and does have that bump in the worst part of the audio spectrum. I'll test other amps and I do have some good ClassA/B amps to test.
Ever since seeing that board type I've been really skepical about the layout; having two sizable toroid inductors right next to one and other. It can't be doing the sonic qualities any favors, that's for sure. I'm itching to put one inductor on the other side of the board, maybe turn it 90° for good measure. Making a board pretty at the expense of performance is a no-no to me.
Thanks for the great feedback! You are so right - the board layout is critical - and not done for looks:) One thing to think about though - is the beauty of the toroid is that the magnetic field is enclosed in the toroid. It doesn't spread out like it does in just any other structure. Still - Toroids do need a bit of breathing room;)
will you help to test the streo separation? I got similar module and it smoke out when I test stereo separation. left channel input at GND using 100uF E. capacitor, right channel input with 1 KHz signal generator then read the output at the left channel at 10% clipping then smoke coming out form the heatsink -- then I stopped. Luckilly, the TPA3255 still working.
Wow - that is crazy! Thanks for asking - I do have some testing that I want to do in stereo mode.
Hi Eddy, do you know any budget 2.1 amp boards with acceptable low noise? I have a TPA3116 with annnoying noise. I bought a TDA7498E (TPA?) 160+160+220. Fingers crossed its better.
Please test at 8 ohm load
Thanks I'll post those test results here very soon;)
You are making audiophile noises. IIR your rant(s) Correctly, the PSU cleans all the mains noise up.
Great question! A good PSU can do a lot I think. A switching power supply with a PFC front end should keep the 60 Hz to a minimum and supply much cleaner power. A good EMI filter will also help. The power cable is essentially invisible in any case as long as it is large enough to handle the current. A regulated output should help the amp as well.
I would like to point out that 20KHz is the upper limit of the *average* human hearing. Meaning, half of us are born perfectly capable of hearing beyond 20KHz. However, the cilia in our ears that pick up high frequencies die off as we age. Not many 70yos will be able to hear 20KHz (if any).
Thanks for this. I have understood that most people can hear up to 17kHz but the max is 20kHz. I’ll have to do some research on this.
I'm 76 and I can't hear anything beyond 12 kHz. AFAIK, only kids and (some) young women can hear 20 kHz -- 15 kHz is a more typical upper limit for adult humans
Actually almost nobody hears 20khz. The average is about 40 to 16k ...
@@Douglas_Blake_579 I read something about a dozen years ago about a high school classroom that was using frequencies above 15KHz to communicate during class while the older teacher heard nothing. It's indicative of my point.
Young children often can hear 20KHz, and I've experimented and found that about half of the pre-teens that I've tested can hear something beyond 20K. But it seems their limit is about 22 to 23K. Most research you will find use college aged test subjects because they are the most readily available for university level experiments. Besides the ethics of it, I'm not quite sure how to test babies and infants. But I'm willing to bet their range is slightly better than the pre-teens I tested.
Of course, if your test subjects are all over the age of 40, you're going to get different results. It all comes down to the definition of "average". Is "average" a 9yo, or a 22yo or 40 something?
@jdlech Thanks for the great info! I keep reading that the average is 20kHz, so what is the max or top end? Is it 22 or maybe 25kHz? Info is not easy to find. I do see that they usually say max for many is 16kHz. But then maybe we are splitting hairs as 20kHz or even 30kHz should be easy to accomplish.
🌈 Promo_SM
Thank you