Here's a challenge....

[Meshuganah]

So I am looking at a high power Mosfet amp...400W plus and there is a couple of burnt tracks. From the output to the speaker connector is frazzled...also there is a high power Triac from speaker output to ground and this has burnt tracks to the ground point. Its almost as if this was designed to short the output to ground given a signal on the gate! What design consideration is this? Anyone seen it before? Its like a self destruct system!
Yes I checked the circuit very carefully, can't believe my eyes. I do not have a circuit diagram so had to draw it out....doesn't make sense.
Anyone?.............( Output Mosfets seem to be OK! )
BTW this is the "Soundpower" amp I wrote about before......no info.

[sregor]

search for "crow bar" circuits. I've seen them in a few consumer products (pioneer integrated amp) I believe they are more common in pro and PA products. The idea, I think, was that without a relay, the best way to protect the speakers in case of an amplifier malfunction was to make it blow the fuse. Also found in power supplies. (I think BGW used them). YMMV Good luck.

[Meshuganah]

Thanks Steve, yes I never took 'crowbar' very seriously.
When your speakers are worth hundreds of dollars, maybe that's a way of saving money....who'd a thunk?
I will rebuild it all and check it out. The fuses were all blown but the tracks were certainly fried too........must have been one hell of a party!

[CBS240]

Hence the phrase, "My amp blew up!!!". The crow bar was triggered by something, better check the sand in the output stage.

[Meshuganah]

It runs 2SJ056/2SK176 Hitachi Mosfets, ten on each channel and they are pretty indestructable I think so should be OK. I will have a close look though... Thanks guys!

[sakis]

crow bar protection based on the DC presence in the output .If excessive Dc is there means that the amplifier is toasted any way so shunt the output to round to prevent speaker damage ...

i would say that more fear exist that crowbar will be activated for no or obvious reason than real false activations ....people really afraid of this but doesn't actually happen that often

truth is that if false activated you are about to blow an amplifier with a minor problem ... for PA use logic is good enough

regards
sakis

[Meshuganah]

Hello Sakis,
I have read some of your threads here. Ijasuu (Yahsu) Sorry I don

[Meshuganah]

Hmmmm Computer was doing maintenance!

I have read some of your threads here....Yasu? (dont know how to spell Greek words in English.)

Thanks, yes, I will check for offset before I put it back together. Need a big variac now.

Good luck with your economy, Greeks have been around along time...great history. Youll be OK. Cheers

[AndrewT]

Can you check that the correctly rated fuses have been fitted to the supply rails?
Can you also check that these fuses are fitted between the main smoothing capacitor bank and the amplifier output devices?

It would seem sensible to me to design the output stage and the smoothing and the fuses and the Crowbar such that activation of the crowbar to protect the load does not destroy the amplifier.

It also seems very tempting when the crowbar has previously triggered that the operator having lost speaker output and incurred the wrath of the audience, chose to fit bigger/wrong fuses to try to prevent a re-occurrence.

If this was the case, then Crowbar did not destroy the amplifier, the operator destroyed the amplifier !!!!

[indianajo]

When testing your repairs, get everything working, test it thoroughly for AC & DC with load resistors on the speaker terminals (insensitive to DC) before putting the crowbar back. I had to replace both triacs, and one trigger diac. Output transistor faults tend to cause a lot of damage to the bias circuit. My $55 PV1.3k amp is putting out 65 VDC right now- another handful of blown diodes + who knows. The 1.3k burned circuit traces at the triacs, so I installed 25 A 32V AG3 fuses between the power supply and the O.T. subassembly. (1000 W @ 2 ohm = 22 amps). Another O.T. that passed the double diode test blew while debugging and took a circuit trace when the triac went off again. The fuses have drops on the wires but didn't melt. So the 1000W rating is a bit suspect, considering the traces melt before the fuses. Trying 20 amp fuses this time. Main breaker didn't go off this time, either, I got to the AC strip switch first.
As far as Andrew's recommended fuses between the transformer and capacitor bank, it would take a better soft start circuit than the 1.3k has to protect the fuses due to turn on surge. It surges so hard the trouble light goes dim when I turn it on. I think refrigerator motor start NTC thermistors are about $30, but they don't work at 95 VDC. The other trouble with thermistors is that when the band takes a break, they go cold and protect again at the first chord.
My most expensive amp was $60, I'm up to $600 a pair on the speakers now, used, so I like having a crowbar on an amp with the O.T.'s direct coupled to the speakers. As much trouble as this PV1.3k has been to repair, I'm selling it off if I can't figure out a better protection circuit - and then build 3 channels of single supply capacitor coupled amp which destroys itself when overheated, not the unpurchaseable soft suspension hammond speakers.