Fasten seat belts. TDA8932 pessimistic review.

Hi,

I've read almost the full thread and I have the following questions I haven't seen on this thread:
  • What is the performance of two modules syncronized and run in paralleled BTL mode? Is it similar to one single module in SE?
  • Has anyone ever tried to apply nfbk from the (filtered) output to the + or - inputs (depending if the output is in phase or not with the input) considering it as an OPAMP to reduce gain?
  • Has anyone ever applied differential feedback to these modules?
  • I have speakers with bi-amping inputs, in this case I would say I could just buy three single-chip modules: two running in BTL for woofers, one running in (dual) SE for tweeters?

Thank you,
Kind Regards
i use 8 pcs. BTL mono boards for a 4-way stereo speakers
i have modded many boards.
the main enhancement is, when changing to good input capacitors
did not notice any enhancement with several better output chokes
no preamp needed for speakers above 90dB
 
  • What is the performance of two modules syncronized and run in paralleled BTL mode? Is it similar to one single module in SE?

I have tried two TDA8932 paralleled after the LC filter using current sharing resistors and caps (to accommodate offset voltage) and didn't notice an improvement when I did that.

  • Has anyone ever tried to apply nfbk from the (filtered) output to the + or - inputs (depending if the output is in phase or not with the input) considering it as an OPAMP to reduce gain?

I tried adding opamps with feedback from before the output filter. This seemed to give a slight improvement in clarity, however stability wasn't great, there was some low-level 'ticking' noise.

The best improvement to the stock implementation remains low noise regulation of the signal stage power supply.
 
Thanks for your feedback ralf!
So the 470 nF caps shown on all inputs of diagrams at page 37 and 38 of the attached Philips Datasheet.


I tried adding opamps with feedback from before the output filter. This seemed to give a slight improvement in clarity, however stability wasn't great, there was some low-level 'ticking' noise.

The best improvement to the stock implementation remains low noise regulation of the signal stage power supply.
Thanks abraxalito, I would like to ask you:
  • did you filter with a low pass the signal from the output (before output filter) before reinjecting it to the input? Is it needed to add opamps?
  • you mean to regulate the voltage of VDDA (pin 8) and VSSA (pin 9)? At what voltage?
 

Attachments

  • PHILIPS TDA8932.pdf
    3.9 MB · Views: 214
I fed the output back into the opamp inverting input via a fairly high valued resistor and the opamp's configured as an integrator (with a feedback cap). So no filtering beyond that provided by the opamp's operation.

Yes, regulate VDDA at the same voltage as VDDP. I fit a diode between the two so that this supply can't go significantly above the VDDP. If you let it go higher, the chip can blow up. Lower is OK from the pov of chip safety but then you don't get full output swing.
 
Could then be that "tickling" was the feedback recoverying from full range to zero?
I've always seen a low pass on the feedback taken before the output filter, just to have again an "analog" signal to feed back to the input.

Is VDDA really needed to be as high as VDDP to get full output swing?
Looking at the block diagram on page 9, isn't just a VDDP the power supply that is switched to the speaker by the output stage?
Am I interpreting it wrong?
 
I didn't try to uncover the reason for the 'ticking' noise, it was rather low level and only showed up with my ear very near the speaker, not noticeable with music playing.

Yes indeed, reducing VDDA reduces the maximum output swing, I once tried lowering it to 10V with the output still above 20V and indeed there was clipping well before full output. Most of the power dissipation of the chip at idle comes from the signal stage quiescent current (40mA) so I experimented with lower signal supply to reduce the quiescent power. No gains could be made.
 
You can find the schematic for my version in the first post of this thread : https://www.diyaudio.com/community/threads/transformer-input-tda8932-mono-amp.357678/

I use a separately regulated (with LM317) supply for VDDA. I originally started the design with an LC filter for VDDA but an active solution sounded distinctly better. Lower noise than LM317 also sounds a bit cleaner. For input I use a trafo to get balanced drive from unbalanced and also to reduce the gain, 36dB in BTL mode is rather high.
 
Thank you again, I've receipt all the needed to perform some tests during the weekend.

TDA8932BTW.jpg
 
I took some measurements today
idle power consumption: 50mA @24V (1.2W)
No signal output (@8 Ohm load): 5mVpp approx 360kHz sinusoidal plus 20mV switching spikes containing damped 50MHz energy
Small signal frequency response (@8 Ohm load)
-3dB : 6Hz to 18kHz
-6dB : 3Hz to 42kHz

George
Hi, I'm curious, how is the board you posted image for quality? The reason I'm asking, there seem to be TDA8932T and TDA8932BT in SOIC-32 package, as OP mentions T version might be inferior to BT. To be on the safe side I ordered 8 boards in SOP-32 configuration TDA8932BTW but the seller sent 2 of these and 6 TDA8932T. On the other hand the thread started in 2015 so lots of things changed since. Appreciate your answer.
 
Hi, I'm curious, how is the board you posted image for quality? The reason I'm asking, there seem to be TDA8932T and TDA8932BT in SOIC-32 package, as OP mentions T version might be inferior to BT. To be on the safe side I ordered 8 boards in SOP-32 configuration TDA8932BTW but the seller sent 2 of these and 6 TDA8932T. On the other hand the thread started in 2015 so lots of things changed since. Appreciate your answer.
I managed to get refund for the wrong chip sent so this question^^^ is not releveant any more, thank you.
 
Hi guys,

I just bough three cheap TDA8932 mono BTL amplifiers from aliexpress and they behave nasty.
I wired one of them to test it and it makes loud audio holes (gaps) when playing beyond a certain level which is far lower then should be. I measured only 100-120mA on supply current when happens, audio output is around 2-4W
It is supplied from 26V LiFePO4 big batteries, no small power supply. Short wires, everything as per rules. Source is a phone.

I had read in datasheets before buying and it was supposed to have current limitation just to avoid exactly this phenomena.

What can be done? From where can arise the problem?
I have the exact same problem with two of these tda8932 board, ordered from different sellers. The problem appears more on higher supply voltages: while it's almost absent at 12-14V (but output power at those voltages is low on 8 ohm loads), it cuts out on bass notes with more than, say, 16V, so the board can't be used to its full capability.
Did anyone else have this problem and managed to fix it? Or could they just be fake chips?