MTX TA7801 repair

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
MTX7801 folder. Read the \mtx7801notes.htm file.

The previous tech probably powered it up without having the regulator transistor clamped to the sink. Unfortunately, I did the same thing. These are some of the notes:

When working on an amplifier that you're not intimately familiar with, you need to have the transistors mounted to the heatsink. Case in point... When troubleshooting this amp I unclamped the heatsink mounted transistors so that I could remove the heatsink (to locate the traces for the thermistors). Before re-clamping the transistors, I powered up the amp for a couple of seconds. In that time (less than 5 seconds), one of the regulators overheated and the series pass transistor of the regulator (FET191) failed. When it did, it did a huge amount of damage. It blew out two transistors, one resistor and one zener diode in the regulator circuit. It blew out both op-amps in the class D driver stage. It blew two surface mount electrolytic capacitors, it destroyed both fans and the transistors that drive the fans. Because I didn't follow the rules I set for others (clamp transistors when troubleshooting), more damage was done and more time was unnecessarily spent repairing the damage.


I don't know if yours has this much damage but it's probably close to the same condition.

Q106 and 110 are MMBTA42s (marked 1D). Q107 is probably an MMBT5401 (marked 2L).
Well it was still warranty sealed when I received it so there was no previous tech inside it. I think the failure is due to bad battery connections and someone must have dropped the amp for a cap in the PWM IC to break loose and a cap in the level shifter for the high side output fets to also break loose.

I'll pull the surface mount transistors today and test them as well as the 15V Zener.

Thanks Perry, I found the files and pictures, this amp had oxidation so I could no longer read the part numbers in the regulator section, but now I can see what the values should be from your pictures.
I took a break from studying tonight and fixed the 15v+ regulator, so I was able to do some tests.

The HCPL-3120/J312's check out except pin 2 and 3, they don't read open they read 18 ohms on both chips with red on 3, and black on 2.
On diode check, im supposed to get 1.4 volts, but both chips read 0.13v with the red on 2, and black on 3 like your diagram.

On the protection circuit comparators the the numbers are off a bit, in your picture the bottom voltages are as follows:

(The 2nd set of numbers are your measurements)
Pin# / Yours
1= 5.96v / 6.71v
2=0.002v / 0.004v
3=11.12v / 13.26v **
4=4.76v / 4.60v
5=3.44v / 3.47v
6=0.56v / 0.85v
7=3.44v / 3.47v
8=3.37v / 4.02v
9=5.04v / 5.08v
10=5.04v / 5.08v
11=6.75v / 8.04v **
12=0.001v / 0.004v
13=4.77v /4.78v
14=4.77v / 4.78v

Top chip
1=0.115v / 0.140v
2=1.8v / 1.8v
3=11.12v / 13.26v **
4=1.9v / 1.9v
5=1.9v / 1.9v
6=1.7v / 1.7v
7=0.919v / 0.626v
8=2.54v / 2.54v
9=4.5v / 5.2v
10=0.08v / 0.075v
11=1.73v / 1.738v
12-0.001v / 0.004v
13=4.91v / 6.0v **
14=0.667v / 0.676v

And the 494 measurements

I still have no audio output, if the 494 measurements look ok to you, then i'll start looking at the output fets I guess?
2 and 3 have to be tied together as the amp powers up or it will pull EXCESSIVE current. After the mute delay, the jfet that's shorting them together is switched off and that will remove the low resistance between 2 and 3.

The 3120s have to be checked out of the board. Leakage in the board through other components makes it impossible to get definitive readings unless you pull them. Don't pull them yet.

The voltage on the comparators and the 494 looks OK. For each pair of inputs on the comparators, look at the relative voltage on the inputs. If x is higher than y on both yours and the reference amp, the voltage is OK. For some of the inputs, the power supply voltage will make a difference.

Is it producing rail voltage?

Do you have ~20v DC between pins 5 and 8 of each of the 3120s?
I am getting a rail voltage of +82V and -83V, and my readings between the 5 and 8 pins of the 3120's are 15.5V on the left one and 18V on the right one.

I plugged in the "prizm effects" board to see if the amp is turning on but im not getting any light, but I did test the pins in the connector and im getting about 1V on 3 of the pins, and the 4th is ground.
Yes, the 3120s.

That means the optos are not driving any of the outputs on. The outputs may have failed. It doesn't appear that they're shorted because they aren't driving full rail voltage to the speaker terminals. If they're leaking, you may have to remove them from the board to determine which one is defective. After finding the defective one, reinstall the others to see if you still have DC on the speaker terminals.

There are 5 outputs in parallel on each side of the sink. See if you can find one that has more voltage from the gate to the source (1-3) than the others. If you find one, it's probably the culprit.
I pulled the output mosfets which all 10 are IRFP240's, I tested them and im getting ~.5V across the Drain and Source on all of them, and nothing in any other combination. With this particular fet, should I be getting a connection between the Gate and Source since there is "virtually" a Zener diode between the two?

Here is the data sheet:

I tested them in the board powered up, but of course there was no voltage difference between each on the same side of the board because there in parallel.

On the negative side I was getting -rail voltage on the Gate and Source, and the drain had around ~8V if I can remember clearly, and on the positive side of the board I was getting +rail voltage on the Drain, the Gate and Source had ~8V.
You shouldn't get any reading (ohms or diode-check) between the gate and either of the other legs. The intrinsic diode is between the drain and the source. That's the 0.5v you read. You should have only read 0.5v when the red lead was on the source and the black lead on the drain.

What happens if you power up the amp without the outputs (do you still have DC on the speaker temrinals)?
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.