Twisted Sounds TS3.5KW Advice.

I need advice on a Twisted Sounds 3.5K.

The amp came in with the following damage:
  • Power Supply(PS) MOSFETs failed
  • PS gate resistors failed
  • One bank of PS drivers leaking
  • Rail Capacitors failed
  • One section of the audio output failed
  • Audio MOSFET Driver Transistors seemed to be leaky on failed section. Uses KTA1715 and KTC2814 drivers

The PS section was repaired, defective audio section MOSFETs removed, but the amp will not run without shutting down the remaining audio section. This was accomplished by pulling pin 2 on the IRS21844S low.

With the audio section disabled I was able to verify the following:
  • Rail voltages as expected, +110V and -110V. (For others, I believe this amp uses transformers in series)
  • The +15V and -15V regulators are working as expected and all opamps on the main board had +-15V
  • The 12V (above the negative rail) is produced by a transistor boosted zener circuit and is working as expected. Output is 11.6V on the emitter
  • Audio signal is reaching pin 5 (center header) on the class D PCB maximum of about 10V p-p
  • The ZNCM_HP V10 board is used
  • There are no obvious shorts on the IRS21844S drivers, when testing the two LO and HO.
  • I do see some input close to the negative rail on pin 1 of the IRS21844S drivers.

I am sure this is a generic design. Does anyone have a schematic diagram that may be close?

What is the next steps in this repair? It seems that everyone suggest just replacing the 21844 drivers, even if I cannot determine definitively that they have failed.

Thanks All for the help!
 

Attachments

  • IMG_2930.jpg
    IMG_2930.jpg
    750.7 KB · Views: 365
  • IMG_2892_C.jpg
    IMG_2892_C.jpg
    580.6 KB · Views: 305
  • IMG_2893.jpg
    IMG_2893.jpg
    1 MB · Views: 357
Disaster Strikes

Well, while trying to confirm the signal on U4, my hand slipped and the probe ground accidentally touched some components that must have been at the negative rail level.

That destroyed several smd components and no doubt other components.

Is there a source for these driver boards?
 
This amp has been challenging to repair.

I rebuilt the driver board, repaired some of the via's that was damaged, but also discovered that R18 was open. Similar to R32 in the next image. It was a 2.2 ohm resistor that seems to be a flame proof. I have replaced it with a 1/2W fusable 2.2 ohm resistor for testing. CT1 was also damaged (a 10uF 25V tantalum) and my question is, can I use an electrolytic capacitor in this location for testing only?

Also, the IRS21844S keep dying. I replace one just to have the other one fail the next time I power up. When it does, it shorts the negative rail to the output and R16 gets smoking hot.

Right now, my plan is:
  • Pull all ouputs, check which ones are still working (just that I can use them during later testing)
  • Check the driver transistors, both on diode setting and for any leakage in resistive mode on my multimeter
  • Replace both IR21844S at the same time, again! This is not done with hot air but with a soldering iron set to 650 F with a very fine tip. I do not overheat these delecate ICs during installation.
  • Check that there is no low resistance beteween the outputs of the IRS21844S and VCC, COM, VS and VB.
  • Reinstall the driver board.
  • Make sure the 12V source is still working as expected.

Any Advice?

I have included a picture of the complete board for reference.

Thanks all!!
 

Attachments

  • IMG_3072_C.jpg
    IMG_3072_C.jpg
    522.9 KB · Views: 201
  • IMG_3077_CT.jpg
    IMG_3077_CT.jpg
    316.6 KB · Views: 177
  • IMG_3077_CB.jpg
    IMG_3077_CB.jpg
    190.1 KB · Views: 77
  • IMG_3079_C.jpg
    IMG_3079_C.jpg
    513.7 KB · Views: 90
  • IMG_3075_C.jpg
    IMG_3075_C.jpg
    951.6 KB · Views: 97
Yes, I was referring to the second pin of the 21844s.

Monitoring the negative rail and 21844s pin 2 NO 21844 INSTALLED, I observe the following during startup.
  • As the negative rail builds, from about -50V pin 2 becomes more positive than the negative rail
  • Then 21844s pin 2 settle on about -30V, while the negative rail goes down to -112V
  • On the other 21844s pin 2 I observe similar behavior with slightly different voltages, starting from about -45V, settling at about -25V and a rail of -112V
  • That leaves more ~80V over the MAX (BTD1304N3) Collector and Emitter and I believe that is a 20V Vceo transistor.
  • This is using a 5A supply, so the rail builds over several seconds.

Pin 2 of the 21844S is also connected to pin 1 on the first header to the left and to pin 8 of the header to the right. With the PWM board removed from the main board, the voltage to these pins are as expected. When no remote is applied these are pulled to the negative rail, with remote applied they seem to float at the secondary ground level.

~80V also seems high for the 21844s SD pin?
 
Using differential probes, one on the negative rail and one on the pin 2 pad of the 21844s, I get between 60V and 80V during the first 12 seconds. It stabilize around 60-70V.

Can it be floating? Could the circuit rely on the 21844s to clamp it to the VCC?

I cannot find any direct current path for what I am observing.

Should I just fit two new 21844s, check driver transistors, confirm all voltages and try again?
 
Check your setup. The readings should be similar to the multimeter but would be able to catch short-lived spikes.

If you actually had 60v between the negative rail and pin 2, that could damage the ICs.

In the amps that I've seen using the 21844s, the SD terminal is held at the negative rail, then released by the muting transistor/ At that point, the internal CCS/pull-up resistor of the IC pulls the voltage up to about 5v. The limit for the voltage on the SD terminal of the IC is 10v, referenced to the negative rail. As far as I know, there shouldn't be any voltage on that pad without the ICs in the circuit or a pull-up resistor from SD to the VCC pad.
 
Current State

Hi All,

I have started to work on this Amp again. The current status is the following:
  • All Power Supplies Working +-112V Rails, +-15V Op-amp Supplies, +-5V Class D Card.
  • 12V supply for the Class D card also working. Transistor boosted zener circuit.
  • IRS21844(s) removed from Class D Driver.
  • TL072, LM211 and LM2903 Replaced.
  • The following transistors replaced. Q1 and Q3 (drivers following LM211)
  • Q2 and Q10 (IRS21844 Shut Down(SD) transistors)
  • Q5 (over current protection circuit output)
  • All output buffer emitter follower pair transistors removed
  • All output MOSFETs removed

Outstanding Questions
What does your scope show if you probe the negative rail and pad 7?
Between 12V to 15V. In the differential mode I do not have the option to use the measuring of the scope, so have to rely on the graduations on the screen and it is not precice. When I measure 12V it shows up as 15V if using the V5 scale and approximately 12V if I use the .1V scale and 10X probe, not exact.

I do have a switching signal above the negative rail on pin 1 of the IRS21844. I did not measure the amplitude of this signal.

Using the differential probe method, I still see too much voltage on the pad for pin 2 of the IRS21844. Using a Voltmeter I see up to 7V, slowly draining (possibly from my meter load).

I also expected the Driver card to have the time delay circuitry to keep the SD pin low until the rail builds to be on the card, it is not. The only function of the LM2903 is to detect over current. I verified that the over current is working as expected, using my Diode test setting on the non-inverting input pin of the Over current circuit.

If anyone has a circuit diagram of the main board that would be useful. I would like to see if the delay is implemented on the main board.

I want to make sure everything is functioning as expected, before fitting new IRS21844(s), since I have lost quite a few so far.

Any thoughts/advice welcome!
 
Last edited: