| djQUAN |
I have posted in another thread but no one replied. the Orion gurus must have missed it.......
I have a blown 225HCCA amp with all ten IRFZ44 fets blown (shorted gate to drain) and the output devices seem to be ok.
the resistor that connects both totem pole drivers to ground (from MPSA56 collector to ground, MPSA56 is blown) is burned. I need the value for those resistors.
Any help would be greatly appreciated! |
|
|
| Clipped |
bands are : brown, black, black, gold, brown
check the A06 ,A56 as well as all the diodes in the ps as well...
when the fets blow it usually takes a few other things with it |
|
|
| djQUAN |
that would be 10 ohms. :)
thanks very much! :D now I can move on. |
|
|
| Clipped |
| goodluck, i learned how to fix that amp on this very board :D |
|
|
| djQUAN |
| the A56 are cracked so I know the magic smoke is out. going to check the A06 and other parts soon as time permits. |
|
|
| Clipped |
when you check the diodes, if possible, lift the end of the leg that isnt connected directly to the ground plane, the ground plane will draw the heat away and will be harder to lift the diode/component.
found this out the hard way... |
|
|
| djQUAN |
| finally had the time to crack this thing open and see what else was damaged, I replaced both sets of drivers and I can get only a 7V peak square waves at the mosfet gates (mosfets not yet installed) is this normal? seems a little low for me. :xeye: |
|
|
| djQUAN |
ok, I decided to install the fets and run it off a regulated supply.
it is running off 13V (current limited to 1.5A) and runs fine but the transformer runs hot (around 50C) but there seems to be no unusually high input current draw or anything that tells something is wrong. the amp section seems fine and does amplify signals.
the only thing wrong in the picture is the transformer getting too hot. also, I noticed that the drains on one side of fets have a little higher spike and ringing compared to the other phase. |
|
|
| Perry Babin |
Are the gate signals the same for both banks of FETs?
Did you confirm that all gate resistors were within tolerance?
How high is the spike (from ground to the top of the spike)? |
|
|
| djQUAN |
I checked the signal from all individual gates of the fets and they are all the same.
as for the spikes, on one side, (supply at 13V) the spike measures around 16V at the highest peak and the other side is at around 18V.
the ringing dies out pretty quickly and from the several car amps I have built, I have seen worse but still runs much cooler. this thing is beyond me. :xeye: |
|
|
| Perry Babin |
What's the operating frequency of the supply.
Starting with the transformer at room temperature, if you power up the amp and let it idle (no speakers/load), how long does it take to get to 50C?
Is the spike 18v above the top of the square wave? That's pretty bad but shouldn't cause any problems. If there are no snubbers on the power transformer, you could add them. If it has snubbers, check the value of the resistors and capacitors. |
|
|
| djQUAN |
operating frequency is 20kHz (40kHz clock)
it takes 2mins to reach 50C and gets even hotter as time goes by. after 7 mins the transformer is hot enough that I can't leave my finger on it for more than a sec.
I misread the oscilloscope measurements regarding the spikes. the top of the square waves are about 25V this time (push pull! I forgot about that!) and from the top of the square waves, on one side, the spike goes up by 5 volts and on the other it goes up by 12volts above the square wave.
as far as I can tell, there are no snubbers used. I wonder, what if I start messing around with dead time and increasing it a bit? :confused:
is the low gate voltage a probable cause? I measured 7volts peak which is driven into the gates. I'm not sure if this is normal in these amps.
I have a feeling the transformer heating is caused by core DC saturation..... |
|
|
| Perry Babin |
They run them at relatively low frequencies. I have one here running at 19.97kHz.
The core should not be getting that hot. A higher frequency should give lower loss but I think there are other issues. It's very strange that it's not pulling excessive current.
Have you tried to run it into a load to see if the waveforms on the drains remain normal? If they change considerably, it could lead you to the problem. If you try this, don't run it too hard. 4 ohms stereo should be enough to see if there's a problem.
Of course, you must mount it back into the sink before driving it into a load.
The low drive voltage is normal, sort of... Generally the drive voltage is ~4.5v peak at idle (depending on the B+ voltage). These amps use Pulse Amplitude Modulation for the regulation. In general, I very much dislike this design. It causes the FETs to operate in a somewhat linear mode at low current demand. This can make the FETs run hotter than they would with PWM.
Attached is the gate of the amp I have. I'll attach another in the next post.
2v/div 10us |
|
|
| Perry Babin |
Drain of FET
Scope settings 10v/div 10us. |
|
|
| djQUAN |
the picture you posted is nothing similar to mine. :dead:
well, for the gate voltage, I get a nice clean square wave right at the gates but I get a small negative going peak before the voltage rises. (gnd clip of scope connected directly to the negative power terminal) the voltage is 7 volts peak which drops as I reduce supply voltage. when the voltage goes to 13.9V, the gate drive suddenly drops to 3v peak. there is a very slight transition between the two.
for the drain terminals, the square wave is also nice with the similar but larger spike after the rising edge.
I tried loading it at 8ohms bridged (only thing I have at the moment)
and the waves at the drains don't change much. the spike and ringing reduces until it drops off due to supply current limiting.
idle current is measured to be 640mA. |
|
|
| Perry Babin |
The gate voltage you have is correct. When the input voltage is lower than 13.9v (according to your information), the secondary voltage is below the target voltage and amp is driving the FETs harder to try to reach the target voltage. When you reduce the voltage, the available voltage for the drive circuit decreases so the peak gate voltage simply follows the B+ voltage (minus losses from components feeding the drivers).
As you increase the B+ voltage above 13.9v, the target rail voltage is reached and the regulator cuts the drive voltage to prevent overshooting the rail voltage.
The regulator is working fine.
If you increase the B+ voltage to above 13.9v to cause the drive voltage to roll back to ~3v, does the transformer get hot? |
|
|
| djQUAN |
| yep. it still gets hot but stabilizes at a lower temp compared to running at full gate voltage. I could keep my finger on it and not get burned. |
|
|
| Perry Babin |
At this point, I'd suggest reassembling it completely and running it into a normal speaker load. Monitor the temp of the transformer. If it remains under ~80C, it's probably Ok.
The only other thing I could suggest is that you remove and check the large primary filter caps. Check them for both ESR and capacitance value if possible. |
|
|
| djQUAN |
| hmmm. I have the amp already assembled while doing the measurements. the main filter caps (50V 3300uF) large cans at the 12V supply seems a little bloated and the initial spark from charging when first applying power seems a little weak. will try and do measurements. |
|
|
| djQUAN |
ok, I did a simple checkand the cap aint bloated. it was the plastic covering of the top part that simply bulged but the aluminum inside is still flat.
with the supply charged to 12V and the power terminals thenshorted, the caps still push a bit of current so capacitance and ESR may still be useable.
I tried connecting the oscilloscope probe across GND and B+(directly at the transformer primary CT) and the scope set at AC couple. I see only the spikes caused by the ringing at about 1V p-p and the supply voltage modulated by the 40Hz sine wave at around 0.2Vp-p when driving an 8 ohm load bridged to a level before my regulated supply overcurrent is triggered (around1.5A)
I guess that could mean something about the state of the main supply caps...... |
|
|
| djQUAN |
update:
I tried adding a resistor across DT and 5V ref terminals to increase deadtime but still the heating persisted. I touched the transformer both on the bare core and on the copper coils and the core is noticeably warmer than the coils so I could blame the heating due to the transformer core. why it heats up, that I don't know.
I will now reassemble the amp and install it into my friend's car to further test it and see if it would last.... |
|
|
|
