Got an amp here that had a severely blown power supply and shorted outputs in one channel. I put in two ps fets and new gate resistors for a partial power supply and clipped the shorted output transistors out of the bad channel. The amp powers up fine and plays cleanly through one channel but the one with the shorted outputs has full rail voltage on the output terminals of the amplifier. I only removed the shorted transistors for that channel (ktb688) but left the complements in (ktd718) . I know they need to be replaced before the amp is put back into service but for the time being I left them in simply because they weren't shorted. Is is possible that even though these outputs aren't shorted they are bad enough to pass full rail voltage to the output terminals?
In my experience transistors usually go short D-S or C-E.
I have had the odd one do something unexpected like a G-S short.
I have had the odd one do something unexpected like a G-S short.
It's extremely rare for two parallel outputs to fail. Did you check them after you cut them out of the circuit?
I'm pretty sure I did but I will re check them. I believe they are still laying on my bench somewhere.
As usual you are correct Perry. One of the b688's is not shorted and appears to test normal with my meter
Reinstall it and see if the amp will produce audio in the blown channel. If it draws excessive current pull the two 718s from that channel and check them for leakage. Reinstall the good one. Does the amp produce clean audio?
If not, find the defective driver or resistor. If it does, you know that it only needs outputs.
If not, find the defective driver or resistor. If it does, you know that it only needs outputs.
Well I put the b688 back in and powered the amp up. It did not draw excessive current but it still had full rail voltage on the output terminals. I pulled the d718's and checked them and they seemed ok so I put them back in. I started comparing resistance measurements from the good channel to the bad channel and I think I am close to finding the problem. One of the driver transistors is giving me some crazy resistance readings. I pulled it to check it out and between the outer two legs of the transistor I am only seeing approximately 70 ohms of resistance. The part in question is a 2SD667 .
If I pull the transistor should the amp power up if there are no other problems? I had it out of the board and tried to power up and it started to draw excessive current
I would expect it to power up without excessive current draw. Are there any solder bridges between the pads for the driver transistor?
There was no solder bridges but I will poke around some more. Perhaps the other driver for that channel is bad as well and I just haven't found it yet.
The drivers typically have their emitters connected to the base resistors of the output transistors they're driving.
Checked the other driver and it seems ok but I am going to replace it anyway. Just wanted to ask if anybody thinks that the fja4313 and fja4213 would have a good shot as replacements for the ktd718 and ktb688
Awesome. I have exactly 2 pairs left. I will give those a shot. I am hoping I can source the drivers in town.
Well I'm not really sure if it's the drivers or something before the drivers causing the issues now. I pulled the KTD718's and KTB688's and replaced them with FJA4213 and FJA4313. I put in two nte replacements for drivers and the amp was drawing excessive current on power up. Not sure if this is because these nte parts are sub par and/or a bad match or if something else is causing this problem. I took out the nte drivers and the amp powers up normally and there is no rail voltage present at the speaker outputs. I noticed while probing around that several of the other small transistors (2sa1266 and 2sc3198) in the bad channel had full rail voltage on all three legs . Probing the good channel confirmed that something is not right with this. One of the three transistors with the full rail voltage on all three legs is what I believe to be the biasing transistor(its mounted between the outputs near the heatsink). None of these read as shorted in the board. If I have time tomorrow I will pull them and check to see if they are leaky. On another note I am throwing together a parts order for some of these smaller transistors as well as the original drivers from MCM electronics. Does anyone know of a power supply fet they might sell that would work good with 47 ohm or 10 ohm gate resistors? IRFZ44N work with either of these or what gate resistor should I order to go with IRFZ44N's?
Are you sure that you got the drivers in the right locations? Sometimes, when using subs, it's easy to get them confused.
When working in the feedback loop, if any components are missing, it's VERY difficult to predict what the voltage will be on the other components in the loop. Everything between the differential amplifier and the output transistors is in the feedback loop.
I wouldn't buy FETs from MCM. I purchased some IRFZ46s from them and they were definitely not what they were supposed to be. They read more like dual diodes than FETs. I purchased them because someone was having trouble repairing an amp and after a bit of testing, it was apparent that they were no good. I want to purchase them directly so I could confirm for myself that they were not what they should be.
Buy FETs from digikey, mouser, arrow or future.
When working in the feedback loop, if any components are missing, it's VERY difficult to predict what the voltage will be on the other components in the loop. Everything between the differential amplifier and the output transistors is in the feedback loop.
I wouldn't buy FETs from MCM. I purchased some IRFZ46s from them and they were definitely not what they were supposed to be. They read more like dual diodes than FETs. I purchased them because someone was having trouble repairing an amp and after a bit of testing, it was apparent that they were no good. I want to purchase them directly so I could confirm for myself that they were not what they should be.
Buy FETs from digikey, mouser, arrow or future.
- Status
- Not open for further replies.