replacing 2SC3281/2SA1302 outputs, transistor matching

[Giaime]

Hello all!

Originally posted by anatech
Hi Giaime,

This from you??? Okay, I'm surprised.

Dear Chris, I was just answering a simple question, this:

Originally posted by nick78
Any suggestions on what kind of amp to build around a nice toroidal that has 95.6-0-95.6 outputs and also 19.3v outputs?

I agree with you in the opportunity of repairing the downconverter, for sure. But he asked, I replied...

[megajocke]

The output stage should hold up even with a shorted downconverter, the outputs are 400V devices and the drivers 200V. As long as there is no signal or the output is unloaded they should be fine. The instant blowup of the output stage suggest shorted drivers or an open bias network.

It probably started with a shorted driver or open (intermittent?) diode string. Then this takes out the outputs, and the drivers. The shorted output stage then causes the downconverter fets to blow, the fuses blow after all the semiconductors as usual...

[anatech]

Hi Giaime,
I did get your attention, didn't I? Still ..........

Hi megajocke,

The output stage should hold up even with a shorted downconverter, the outputs are 400V devices and the drivers 200V. As long as there is no signal or the output is unloaded they should be fine. The instant blowup of the output stage suggest shorted drivers or an open bias network.

No way. There is minimal heat sink area and the dissipation would run the outputs into thermal runaway before long. You just need a little bias at that voltage to heat things up and start the slide to destruction.

The first failure may have toasted the downconverter. For sure, non-matched outputs would cause issues with no emitters to force a balance. The bias circuit may have possibly partially failed also. I would expect it to go short.

My guess. An older, powerful sub amp develops broken solder connections. When the amp was moved, flexing of the board caused the failure. The high fault current caused the downconverter to fail short.

Hi nick78,
Normally I check and resolder connections in sub (and car amps) for this exact reason. It's part of my normal service. It's hard to tell other people what you do without even thinking about it.

The beauty of this design, like other Carver amplifiers, is high output power with low dissipation and relative freedom from 2nd order problems. 2nd order issues are reduced, not eliminated with these designs.

-Chris

[megajocke]

Of course, if it has a standing current, it will start running away with that voltage as it is not compensated. The OP didn't say if they blew immediately, like power - bang, or after a while (which might not need to be too long). I assumed he meant they blew instantly, which should not happen even with this high voltage.

[anatech]

Hi megajocke,

I assumed he meant they blew instantly, which should not happen even with this high voltage.

True, there should be some time delay. This delay could be very short if the devices were not properly mounted, but there would be some delay.

This is why I suspect an intermittent connection. Next to instant bang if the bias network went open at this voltage. On the Lightstar, the static rest voltage is 13 VDC on the rails. If one got stuck to 125 VDC and the bias network opens briefly, I would expect some damage.

-Chris

[taj]

Originally posted by anatech

Hi Todd,
No. Not even close to a reasonable answer.

I am frankly surprised that some people would recommend replacing the amp instead of repairing the downconverter. It's easier than class D for pete's sake! You can test the amp section from a "laboratory" type dual tracking supply. In this way you can get the amp working before getting into the downconverter. What you can't do is run the supplies up so that it overheats. Make sure the bias section is working. So this is only as difficult as you want to make it.
-Chris

I wouldn't hesitate to replace the amp with something more conventional -- that I can understand, Chris. But of course, I'd ship the Carver design off to you since you can appreciate much more than I. But since it's not mine to do that with, it's a moot proposal.

..Todd

[nick78]

The first failure was instantaneous (~ 1/2 second) while moving the sub. I have spoken with NHT and the likely failure is an intermittent connection between the sub's controller and the sub itself (they are connected with a special cable). Basically the two drivers went to max excursion and 1/2 second later the fuse pops.

The 10" drivers are tough birds, I believe they have the same magnet/motor/coil structure of the infamous 1259's. They have shown no apparent damage (measured or observed) from this, and seem to still function perfect.

The second failure (when the outputs blew again after I replaced them) was thermal. After being powered on for a minute, the output heated up enough to thermal short and that popped the fuse again after about 1/2 second. I found that, after replacing the blown output once more, that the amp could be turned on for ~ 10 seconds at a time without additional failures occuring, but the outputs on the (+) side would get nice and warm, much warmer than normal

I don't know yet if anything further is damaged in the outputs. The speaker output has no significant DC offset, so I don't believe that the (+) side outputs are being driven on to any significant level. My understanding is that the downconverters are designed to keep the voltage drop across the CE junction < 6 volts, a safe bias at 6 volts might be deadly at 130V.

fingers crossed.

[anatech]

Hi nick78,

The first failure was instantaneous (~ 1/2 second) while moving the sub. I have spoken with NHT and the likely failure is an intermittent connection between the sub's controller and the sub itself (they are connected with a special cable). Basically the two drivers went to max excursion and 1/2 second later the fuse pops.

That fits. The downconverter was damaged at this time.

The second failure (when the outputs blew again after I replaced them) was thermal. After being powered on for a minute, the output heated up enough to thermal short and that popped the fuse again after about 1/2 second. I found that, after replacing the blown output once more, that the amp could be turned on for ~ 10 seconds at a time without additional failures occuring, but the outputs on the (+) side would get nice and warm, much warmer than normal

This hints that the downconverter was shorted already.

I don't know yet if anything further is damaged in the outputs. The speaker output has no significant DC offset, so I don't believe that the (+) side outputs are being driven on to any significant level. My understanding is that the downconverters are designed to keep the voltage drop across the CE junction < 6 volts, a safe bias at 6 volts might be deadly at 130V.

Without a doubt. No way will these circuits survive full voltage. They were not designed for this.

-Chris