Replacement transistors for the 9090DB

CDFixer

CDFixer

Number 1 USC Trojan Fan
I have been working on a 9090DB that had low outputs, it would only get ot about100 watts before clipping. I set up another that was working properly and began trying the easy stuff. I swapped the driver boards, no difference, I then changed the output transistors on one channel using OnSemi MJ21193s and MJ21194s, no difference. I then decided to change the power supply board, this was a pain but the unit came up and one channel was putting out 141 watts, the other still only making it to 100 watts before clipping. The only difference was the outputs, and the side with the OnSemis was the low side. I pulled the old Sankens out of the parts bin and put them back in and the unit fired up with 141 watts out both channels.

I am going to test with another set of the onsemis tomorrow evening but my first impression is that they are not going to make the grade as replacements for the 9090DB. They may work in fine in other systems.

Anyone else had a problem like this?
 
Register to hide this ad
Wow. That's really...ummmm...wierd. The C-E saturation voltage of these devices can't be the reason, so the issue here is one of gain.

At what frequency did you do this? Honestly, even at 20KHz, it shouldn't matter, but obviously the drivers can't seem to saturate the output stage...the drivers (or a previous stage) are clipping, not the outputs. In the end, if you can easily clip with the Sanken devices, unusually low gain would seem to be the issue here. I'd :saywhat: about those devices.

I've put 21193's and '94's in a 9090db before, and it clipped at about 160WPC, which is consistant with the 4 or 5 units that have come to me for repairs.
 
I sure need to read before I post! I was looking at the MJW21193 or something. You are right, my bad.
 
Last edited:
Me? The 21193's and 4's I'm speaking if are TO-204AA's, otherwise known as TO-3's.

My point is that is the if the outputs had really piss-poor gain, the drivers and pre-drivers would have to work harder to make up the difference. Certainly there might come a point where they were doing all they could, but the gain of the outputs was so bad that these driver stages could do no more, and max output level was reached long before it should have been.
 
Was that the problem? Low hfe on the outputs? I really want to understand all phases of amplifier theory even if rehashing.
 
I do all the testing at 1Khz, that shouldn't have been a problem. I was thinking that the gain was the problem. I am going to put another set of 93's and 94's in the unit and test it again. I do both channels and see if the problem recurs, I will use a different set of transistors.

Any ideas what on the power supply board could have been causing the orginial problem? I can't find any defective components on the board.
 
I have same transistors

I have a 9090db that Is a keeper that I put new onsemi's in and rebuilt driver board with new caps zeners and such.If my calculations are right its about 156 rms on each channel .I have never had a problem with them.
 
Well I put a new set of OnSemi's in the beast and I have the same outputs as with the orginial transistors. Maybe I had a bad set or something. I am going to put that set back in and see what happens. I didn't change any settings on the scope so I don't think it was a cockpit error.

Will let you all know.
 
Were both the positive and neg wave clipping at about the same time with the old ones? If you put the questionable ones back in, check this.
 
The positive was clipping before the negitive. Maybe there is only one bad Xistor in the bunch. I will install the orginial set and test.

Thanks All
 
I recently repaired a 9090DB with low output on both channels, only in my case I got about 5 watts on one channel and 20 on the other. The problem was the flameproof resistors on the driver circuit board. So in your case if the output/driver transistors are ok and the +- rail voltage is in spec, I would suggest you check these resistors. I don't have a schematic so I can't give you the R numbers but these are the resistors that are kinda dull looking and stand off the circuit board about a quarter inch. If they are bad their resistance will way out of tolerance. By the way EW and others are correct, a 9090DB should bench at 150-160 watts at clipping.

Bill from Indy
 
I have several of these and was having trouble with only this one. I swapped driver boards with no change, I was able to correct the problem by changing the power supply board.
 
You must log in or register to reply here.
Back
Top Bottom