Go Back   Home > Forums > Amplifiers > Class D

Class D Switching Power Amplifiers and Power D/A conversion

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 21st June 2013, 12:45 PM   #281
Banned
 
Join Date: Jan 2005
FORGET all about the Lepai garbage!!!

If you want good sound and performance from the TA2020, you should go with a AMP6 or similar that has a board laid out according to the Tripath recommendations and using quality brand components.
  Reply With Quote
Old 21st June 2013, 07:42 PM   #282
Bare is offline Bare  Canada
diyAudio Member
 
Join Date: Dec 2001
Location: vancouver
Quote:
Originally Posted by THUMP LUMP View Post
If, in fact, the tests done by TMM are valid and the problems and noise are centered around the Korean made clone TA2020, then what would be the point in buying a "properly made and laid out TA2020 amp" when they will have the same Korean made clone TA2020?
Welll.. IMO the Korean clones were likely the good ones :-)
One would be genuinely Lucky to receive one of those.
It's the far cheaper copies 'Fabrique En Chine' that masquerade as Korean Made.. that are the crux of the problem.
But then half the parts on the PCB's are copies of copies as well.
  Reply With Quote
Old 22nd June 2013, 01:27 AM   #283
diyAudio Member
 
THUMP LUMP's Avatar
 
Join Date: Aug 2007
So you're saying that even though this one says "Korea" at the bottom (you can just make it out in this photo), that in fact it is probably a cheap Chinese clone? I know it's not a true original Tripath but, I figured Korea was Korea.
Attached Images
File Type: jpg TA2020 PCB TOP.JPG (204.8 KB, 394 views)
  Reply With Quote
Old 26th June 2013, 08:51 AM   #284
diyAudio Member
 
Join Date: Jun 2013
Hello all, this is my first post but I have been reading and following the forum for a while and have learned a lot so first let me thank everyone for all I have learned.
I am in the have the 2020A+ (actually a copy but the only change I can see is 2 relays rather than 1) and have been modding it for a while largely using what I have learned here.
I have read that replacing the output inductors will improve quality a bit. Do you feel this is worth the cost/effort and if so what are the tolerances on replacements as I am unable to find ones with the same rating.
Thanks again
I have read that
  Reply With Quote
Old 23rd August 2013, 08:59 PM   #285
Ron Q is offline Ron Q  United States
diyAudio Member
 
Join Date: Aug 2013
Impressive stuff Redshift. Can you post a component list to duplicate your mods for us beginners? Or direct me to it if you have already.

Thanks
  Reply With Quote
Old 28th August 2013, 12:09 AM   #286
lsheng is offline lsheng  United States
diyAudio Member
 
Join Date: Feb 2010
Default quality post

Thanks for the measurement. This amp has BTL output, did you measure it differentially or just single end output. For the even order harmonic distortion, measure it differetially, you can get different result from measure it single-endedly.

I measured TA2024 chip amplifier before, single endedly, the result is not good, especially with dummy load. But that is before I change the output inductors, which make sounds as the music you play(source of extra nonlinearity).


Quote:
Originally Posted by TMM View Post
This thread needs less subjective opinions about capacitors and more objectivity, so i will add my 2c.

Measured my 2020A+, completely stock:
1W @ 8ohm, 1khz test tone:
Click the image to open in full size.

4.3W @ 8ohm (near max power):
Click the image to open in full size.

0.1W @ 8ohm:
Click the image to open in full size.

1W @ 4ohm (not sure what the hell is going on with the noise floor here!):
Click the image to open in full size.

Oh dear, this is pretty awful distortion, especially at lower power levels (the most important ones!) and looks exactly the same as the Muse TA2024 amp i have. Distortion is okay i guess (except 2nd order) near max power, but since music is dynamic this is pretty moot. You can't play music at 4.3W without playing 0.1W simultaneously. It will however be much better playing loudly than quietly.

Anyone who has played with and measured speaker amplifiers and opamps will know that the level of distortion above is unlikely to be caused by the simple opamp circuits. Also the similarity to the Muse TA2024 amp (which has no preamp or tone control circuitry what so ever) immediately makes me suspect that the distortion is being produced entirely by the TA2020 chip.

Even at max power, i wasn't achieving anywhere near the quoted total harmonic distortion figure in the TA2020 datasheet which is 0.03% @ 10W 4ohm. 0.03% corresponds to -70dB, and i had single harmonics that measured higher than that! On 12v the best my chip can muster without clipping is 5Vrms @ 4ohm = (5^2)/4=6.25Watt and 6.3Vrms @ 8ohm = (6.3^2)/8=5Watt. My Muse TA2024 amp is similar. So we probably need a 13.5v supply to get rated power output, but you'd think it would still get close to the quoted 0.03% distortion.

I had a brainwave that maybe BOTH outputs need to be loaded up to achieve low distortion. So lets try that. Max power with both channels loaded with 4ohm resistors.
Click the image to open in full size.

Nope. The general trend is that total harmonic distortion gets lower as power output increases - except 2nd order, which just gets worse as you crank up the juice (but also bad at very low power levels).

The first mod i tried was lowering the gain. Default value of Rf on my board is 22k, and Ri is 15k. According to the 2020 datasheet that gives us a gain of 12*(22/15)=17.6.
I soldered 20K resistors across pins 9 and 10, 12 and 13 of the 2020 chip. This lowers the gain to 12*(10.48/15)=8.4

Here are the results of that:
Click the image to open in full size.
Odd order distortion goes down significantly, but even order seems to be relatively unchanged.

The next thing i tried was to replace the output inductors. I wound some 10uH aircores with 20AWG wire:
Click the image to open in full size.
3rd order comes down, 5th, 7th and 9th go up! Why the heck did that happen? More on this later...

Then i tried some of the 'usual' mods in this thread. First i cut the trace between R18, R19 and the volume pot, therefore completely isolating the tone control circuit at both the input and output. I measured, and there was absolutely no change.

Then i removed one of the 2.2uf decoupling caps and connected my input signal directly to the input of the 2020 chip via a 560nF polypropylene cap. No change.

At this point i was getting frustrated as to why my measurements weren't repeatable. I'd go and change the gain resistor to another value, then change it back to 20k and get a worse result than when i first measured it with the 20k resistor.

Maybe heat is affecting it? First i sprayed the heatsink with freeze spray:
Click the image to open in full size.

Then heated it up:
Click the image to open in full size.

This may also go some ways as to explain why the distortion gets lower at higher power outputs.

If we look at the waveform with the 1khz tone notched out, we can see that the distortion mechanism is completely changing with temperature (output signal on top, notched signal below):
Room temp:
Click the image to open in full size.
Cold:
Click the image to open in full size.
Hot:
Click the image to open in full size.

This in my eyes proves that the TA2020 chip itself is internally producing all of the distortion at the output and there is not much you can do to fix it. Reducing the gain is a worthwhile mod, but don't expect miracles. Running the chip hotter generally improves even-order distortion and running it cool improves odd-order. The happy medium seems to be the temperature it reaches at after running it at max power for several minutes (perhaps 45C?). Music won't let it get that hot (unless you severely clip the amp) so installing a smaller heatsink may actually improve performance in practice!


I haven't tested the amp with the tone controls enabled, but certainly there are no ill affects (at least amoungst the massive distortion caused by the 2020) in leaving it connected, plus it is generally nicer to have the knobs on the front of the unit actually do something (even if you never use them), rather than pointlessly disabling/removing them for no benefit.

I'm still yet to change any caps but having isolated all of them except the powersupply and output capacitors, and already using a decent regulated powersupply to power it, i'm going to predict that replacing the caps will be about as effective as pi**ing on a bushfire.
  Reply With Quote
Old 28th August 2013, 04:29 AM   #287
lsheng is offline lsheng  United States
diyAudio Member
 
Join Date: Feb 2010
Default My test result

This is a MUSE amplifier with TA2024 chipset, with one modification is that it is powered with 12V 2A switching power supply buffered by a Capacitor multiplier(modified by myself), with an impedance under 40m Ohm from 20 to 20kHz.

It looks better, tested at level right before clipping. one channel with 8Ohm load, the other is without load. single-ended test results.
Attached Images
File Type: jpg ScreenHunter_37 Aug. 27 21.26.jpg (130.2 KB, 30 views)

Last edited by lsheng; 28th August 2013 at 04:36 AM.
  Reply With Quote
Old 23rd September 2013, 06:55 AM   #288
diyAudio Member
 
Join Date: Jan 2013
Hello everyone,

Finally i'm join here now, i got my Lepai for about 3 weeks and i have mod it as i see here, bypassing tone control, change all thru hole caps with Panasonic, Caps input with Bennic, and it really worth for it cost...

Click the image to open in full size.
Click the image to open in full size.
Click the image to open in full size.

Last edited by jarscrew; 23rd September 2013 at 07:09 AM.
  Reply With Quote
Old 19th October 2013, 11:59 AM   #289
diyAudio Member
 
Join Date: Oct 2013
Location: Florida
Hey guys new here, I just ordered the upgraded output caps, and 5600uF low esr input cap from mouser as reconmended by "XtremeRevolution" at Parts Express forums. While I was following the post he mentions that there are 3 holes that you can solder the output caps to and that we should use the outer holes as the new replacement caps are of a larger size. I looked over my board and notice that I have two relays instead of one so I don't have a 3rd hole on the two bottom output caps. Has anyone done this upgrade with a board like this? Is it OK to squeeze the legs of the capacitor together to fit in there?

Click the image to open in full size.

I just noticed I have a LEPY model, do they spell it different all the time?
  Reply With Quote
Old 28th October 2013, 04:59 AM   #290
diyAudio Member
 
Join Date: Oct 2013
is there any guides to doing any of the suggested mods? that has like a parts list also?
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
TA2024 SA-36 or LP-2020A Terrymm Class D 3 28th September 2012 04:34 PM
Help with Lepai LP-2020a+ ChrisChris Class D 5 14th September 2012 03:24 PM
Lepai LP-2020A+ Balance Problem Nightcry7 Class D 3 14th September 2012 03:16 PM
Lepai LP-2020 mod to include sub out rziebro Class D 1 12th February 2011 03:01 PM
X-LP Mod Journey pjpoes Analogue Source 2 11th March 2007 07:35 AM


New To Site? Need Help?

All times are GMT. The time now is 05:46 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2