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.
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.
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.
Last edited:
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!
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!
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
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
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
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
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 !!!!
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 !!!!
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.
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.
Last edited:
Andrew T, Thanks yes I suspect this but have thrown out the blown fuses...should have checked.
Indiana Joe, Wow. that 1.3k is heavy metal. I have blown tracks too and I suspect the scenario from Andrew.
Household circuit breakers may be appropriate at this voltage. At least they are easily resettable....I suspect they may be much too slow.
Perhaps the bottom line is more speakers in series and less in parallel.
This is supposed to be 600w/channel, goodness knows how much with the bridge mode on but we will be looking at huge currents in the tracks under full load.....scary stuff!
Thanks for the input..
Indiana Joe, Wow. that 1.3k is heavy metal. I have blown tracks too and I suspect the scenario from Andrew.
Household circuit breakers may be appropriate at this voltage. At least they are easily resettable....I suspect they may be much too slow.
Perhaps the bottom line is more speakers in series and less in parallel.
This is supposed to be 600w/channel, goodness knows how much with the bridge mode on but we will be looking at huge currents in the tracks under full load.....scary stuff!
Thanks for the input..
600W into what?
Into a resistor dummy test load, or into a reactive speaker load?
Anyone seen it before?
Hi, just for reference, yes i have, The British manufactuer Quad started using this method of protection years ago. If they wern't the 1st to do so, they were certainly one of the early adopters of it.
It's actually an "interesting" way of helping to protect speakers, some of which can cost $k's, or more ! Even if they don't, nobody want's theirs to go up in smoke
Also it enables the output to be relay free, for those that prefer it Of course then having soft start etc is preferable as well
Hope you get it sorted.
Hi, just for reference, yes i have, The British manufactuer Quad started using this method of protection years ago. If they wern't the 1st to do so, they were certainly one of the early adopters of it.
It's actually an "interesting" way of helping to protect speakers, some of which can cost $k's, or more ! Even if they don't, nobody want's theirs to go up in smoke
Also it enables the output to be relay free, for those that prefer it
Of course then having soft start etc is preferable as well Hope you get it sorted.
Thanks a lot for your input Guys. I have no doubt that this will get resurected although I might bounce a few more points off you. I think I will start by removing the triacs. I can fit big fuses in the output lines and dump into a room heater. Still looking for a variac. Should have got one long ago.
Cheers.
Cheers.
Hundred watt light bulbs in series with the wall plug provide some O.T. protection while debugging according to some people. Thank you government for banning them just when I decided I needed some.
I've decided, the triac crowbars on the output are going the way of the dodo, burning lands off the PCB is no protection. Fuses have an overcurrent feature for motor starting and light bulbs, whereas speakers are more sensitive to just a little overcurrent. For speaker fuses make sure you get instantaneous blow fuses (which don't come in 25 amp versions I don't believe). I'm going to use the diac circuit triac trigger to detect dc, then use the triac to pick up a relay instead of shorting the output, which I happen to have in the attic. PM17Dy, 3/8" dia copper contacts. Big enough, but we'll see how it sounds. I'll have to latch the relay up somehow, an auxillary N.O. contact would be nice to parallel the triac. I have some contactors with 117 VAC coils and aux contacts, but they are 3"x3"x7" and won't fit in the box.
I've decided, the triac crowbars on the output are going the way of the dodo, burning lands off the PCB is no protection. Fuses have an overcurrent feature for motor starting and light bulbs, whereas speakers are more sensitive to just a little overcurrent. For speaker fuses make sure you get instantaneous blow fuses (which don't come in 25 amp versions I don't believe). I'm going to use the diac circuit triac trigger to detect dc, then use the triac to pick up a relay instead of shorting the output, which I happen to have in the attic. PM17Dy, 3/8" dia copper contacts. Big enough, but we'll see how it sounds. I'll have to latch the relay up somehow, an auxillary N.O. contact would be nice to parallel the triac. I have some contactors with 117 VAC coils and aux contacts, but they are 3"x3"x7" and won't fit in the box.
Last edited:
That's interesting and may well solve the problem for you. I am thinking about sampling the clipping indicator with a full bridge rectifier and a capacitor such as is used in telephone answering systems to open the line after so many rings. If you check out the common circuits that are available, you will see that it adds pulses to the capacitor until it pulls the relay and latches, requiring a reset. A small amount of clipping would be passed and 'forgiven' but if continued for too long, sufficient voltage would be stored to toggle the relay and either reduce power substantially (connecting a series resistor) or trip it off altogether. I do not like the primitive system of self destruction and will just remove the triac altogether.
This simple circuit can have a small pot added to adjust the sensitivity of the relay to the capacitor, thereby extending the allowable clipping tolerance or reducing it to a small amount. The added benefit is that this can be used to protect the speakers in the event of high power clipping which also destroys the speaker coil with overheating. A small reset button drops power to the relay via the latching contacts, and then lets it start again through the cap.
All you need is a bridge, 1000mfd cap, 2000 ohm 12v relay, pot and an impulse (reset) switch N/O. Use a room heater element as a series load!
This simple circuit can have a small pot added to adjust the sensitivity of the relay to the capacitor, thereby extending the allowable clipping tolerance or reducing it to a small amount. The added benefit is that this can be used to protect the speakers in the event of high power clipping which also destroys the speaker coil with overheating. A small reset button drops power to the relay via the latching contacts, and then lets it start again through the cap.
All you need is a bridge, 1000mfd cap, 2000 ohm 12v relay, pot and an impulse (reset) switch N/O. Use a room heater element as a series load!
Thanks for your input. Peavey had so much trouble with guitar people overdriving square waves into the amp and burning up speaker coils that they put "DDT" into even this low priced PV1.3k, which protects against what your clipping indicator prevention circuit does. So I think I am okay on that front. See Harley peavey's company history on their website about making their own speakers and having them blown up by bands.
Shorted speaker wires that overheat the output transistors and cause O.T. avalanche is the fault I'd like to prevent. That requires SOA current measurement, and that is too hard for me. All I can do with the diac-triac-speaker relay is prevent the blown O.T.'s (that produce DC out the speaker terminal) from subsequently blowing the speakers the next time they are wired up correctly.
I've got a room heater minus the tipover switch in the junk pile, thanks for your suggestion on safe test rig.
Shorted speaker wires that overheat the output transistors and cause O.T. avalanche is the fault I'd like to prevent. That requires SOA current measurement, and that is too hard for me. All I can do with the diac-triac-speaker relay is prevent the blown O.T.'s (that produce DC out the speaker terminal) from subsequently blowing the speakers the next time they are wired up correctly.
I've got a room heater minus the tipover switch in the junk pile, thanks for your suggestion on safe test rig.
Wow all assembled and switched on with no load for a start......The fan is not working. The PCB is brown under the dropping resister and the fan is not getting any power (low voltage).....that'll be why it overheated in the first place.....at least it hasn't blown any fuses yet.......a good sign!
Is there a thermostat control to the fan?
An anecdote from long ago: A locally assembled version of large KEF Ref.105 monitors was fitted with high quality aluminium electrolytics of 11,000 uF, in-line. These protected against amplifier failure when they were really cranked up. Small amplifiers are notorious for speaker failure when there is no certain way of regulating power output and these big, Dalek monsters gulped up power. Anything under 100W/8R was likely to be clipping regularly or failing - shorted to rail, usually.
So, if it sounds good as these did with a decent Electro. inline, why not fit this as a matter of course whilst playing with and proving protection systems? I don't see that it can hurt and it could save a lot of tears.
An anecdote from long ago: A locally assembled version of large KEF Ref.105 monitors was fitted with high quality aluminium electrolytics of 11,000 uF, in-line. These protected against amplifier failure when they were really cranked up. Small amplifiers are notorious for speaker failure when there is no certain way of regulating power output and these big, Dalek monsters gulped up power. Anything under 100W/8R was likely to be clipping regularly or failing - shorted to rail, usually.
So, if it sounds good as these did with a decent Electro. inline, why not fit this as a matter of course whilst playing with and proving protection systems? I don't see that it can hurt and it could save a lot of tears.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.