Thrystor question in an amp that's perpetually blowing fuses

OK, returning to this problem. The new thyristor ran for awhile and the amp sounded good through headphones at least. Then the dim bulb tester I was running the unit through suddenly lit up and the sound distorted. I shut it down quickly and another thyristor had blown.

What would cause a thyristor like this to die? There's obviously something in this circuitry that's not right.
 
Q19 or Q20 leaking could trigger the triac, have you replaced them?
Try replacing C77 and C78 in the low pass filter in the protection circuit. If they have dried out and lost capacitance it could be letting bass pulses through to 'false' trigger the triac . Mouser has some 470uf bipolar electrolytics at 10v and under for about a dollar or less each. They have 5mm lead spacing.

I modeled the low pass filter in LTspice and a worst case full power 20 Hz sine wave signal into the LPF only passes a ~2.5mV p-p signal to the gate of the triac. (Unless I did it wrong, I'm new to LTspice.)
EDIT: more info
 
Last edited:
Does the schematic give some indication of voltages that you can expect around this component - perhaps taking some measurements around adjacent transistors could help?

This sounds like a frustrating problem - I hope you sort it out.
 
I tried to visualize for myself what happens and here are my thoughts:


To comment on this, one will NEED to observe this amp schematic. I have a schematic having the hifiengine marking.


The triac will short positive side of voltage only. Once triggered, it will never stop conducting, remaining current from the cap will ensure during dip of 50 or 60 Hz cycle.

first: initial triggering of the triac.
At the very moment it is triggered, it takes current from a 8000 microfarad cap, in total you have the energy of 8000 microfarad charged to about 35 Volts.
It does not matter how you look at it, the triac shorts it through the 0.25 ohm resistance (in series with triac) and voltage will go down quickly.The energy difference will be dissipated in the triac and resistors. Max initial peak current will be some 35 Volts from one cap over the 0.25 ohms resistor so excluding losses this would be over 100 amps if components were ideal. Now, because of ESR, wiring losses, and "stuff" this will probably be much less but it adds to hurting the triac on its way to destruction.
Above is ONLY for the ONE half wave maximum (8 milliseconds or until when the cap is depleted to 0.25 ohms resistor voltage whichever comes first, since the explanation below takes over.
Note: a 3 amp triac is capable of surge current way over 3 amps.



ADDITIONALLY IF NO DBT WOULD BE IN USE


The amp is fed by 120V AC
the fuse is 3 amps, which would be 360 Watts offtake primary for max operating without blowing the fuse.

A 3 amp fuse will blow quickly at approx 4 amps, which needs some 480 Watts to be taken from mains.

At the secondary side of the transformer, you need enough average current through the conducting triac, diodes and parallel power resistors to take away the 480 Watts.

Average current through the 0.25 ohms limiting resistor (consisting of 2 times 0.5 parallel in series with the triac) of some 30 to 35 amps would give you some 8 Volts over the resistors (and an additional 3 Volts over the shorting triac and the rectifier diodes). This about 11 Volts at 30 amps or more would be the minimum required to have way over 300 Watts dissipation?

This all assumes a transformer behaving properly and no losses ( unfortunately I do not know enough about transformer behavior and forgot anything more complex than very simple math...)

So, the 2 parallel resistors limit dissipation and max current but at the same time slow down the process as it is not really SHORTING the transformer (at even higher currents)

(note the educated EE's will definately find flaws in my explanation but IMO it more or less works this way)
I wonder if I made major calculation mistakes.

One thing is for sure: the triac gets way more than 3 amps during at least the time required to blow the fuse, and may go way over its safe operating area for current versus time

For whatever reason the triac triggered.... (which even can be a loose ground, a main amplifier bad transistor which in another amp would just be causing "crackling sound", oscillations, or whatever, observe dV/dt spec of the triac too versus real life situation).



Now the question remains: Is the power smoothing capacitor capable of destroying the triac, since total dissipation is limited to a safe level by the dbt lamp so the second part of my explanation does not apply....and I can not think of anything else.
 
Last edited:
A crowbar protection like this is intended to blow the supply fuse. It is not intended to crowbar the power supply indefinitely.

Check the fuse ratings; it's possible someone just kept putting bigger fuses in until they stopped blowing... So what happens now is the thyristor blows instead...

As to why the thyristor is firing in the first place; that may be a genuine fault in the amp, or a fault in the protection circuit...
 
A blown fuse and even a fried triac are cheaper (and easier to repair) than damaged speakers. I played with Ohm's law a bit and my calcs show that the surge power pulse through the triac could easily exceed 4000 watts. Pop goes the weasel :eek:. :(
A shorted sacrificial triac still performs its intended function and blows the mains fuse.
 
Last edited:
OK, put in a new thrystor. Using a dim bulb tester with 60 watt bulb. Have checked caps C77 and C78, in fact C77 was totally missing. I put in a 470uf 6.3v cap in the C77 position. Would its absence be enough to have caused this? It's in the schematic, but it wasn't there in my machine when I got it.

Turned the amp on... bulb glowed slightly and died down to nothing. OK that's good.

Testing voltages gingerly. Switching machine on and off for ten seconds at a time to test. No outrageous voltages around thrystor. Have not checked all voltages but noted that the collector of Q15 is supposed to have -20v if I'm reading the schematic right. it's got -29v ! I've pulled and checked R100, R101 and R123. All are supposed to be 2 watt resistors 62 ohms each. They all check out (out of circuit) at around 62 ohms. So why does the collector of Q15 have -29 v instead of -20 v?

Nothing's blown yet, but I'm not leaving it on to cook like the last one did. I think I remember checking the voltage before R 100, R101 and R123 and I believe it's -40v as it should be. Would Q15 even be in play for the thrystor problem blowing???

Oh, also checked the one fuse in the machine. It's supposed to be 3A. It was 3A. Also checked R110 and R111 (both 2 watt .5 ohms). Both checked out, out of circuit.
 
Have checked caps C77 and C78, in fact C77 was totally missing. I put in a 470uf 6.3v cap in the C77 position. Would its absence be enough to have caused this? It's in the schematic, but it wasn't there in my machine when I got it.
. . .
So why does the collector of Q15 have -29 v instead of -20 v?
. . .
Would Q15 even be in play for the thrystor problem blowing???

C77 missing is the first solid clue . . .
When I took C77 out of my LTspice model the signal out of the LPF was about 24 times as large as when it was present, 60mv p-p as compared to 2.5 mv p-p. It functionally blocks everything above DC. If half the filter is gone (C77 missing) I think that would be a problem. Replace Q19 and Q20, you said you tested them but a low voltage out of circuit test may not reveal a transistor that leaks under full voltage, they have 40v to 80v across them. Did the thyristor pop when there was no sound playing? Or only when sound was playing.

FWIW trying to play audio while on a DBT is not a good indicator of proper function, or so I have seen said here on AK. YMMV.

Q15 is part of a constant current regulator that feeds the 2 meter lamps and the base of the transistor driver of the 'power on' LED. Are the meters lighted, does the power on LED light? Check Q15, the 2.7v zener D27, the lamps, etc.
Note if you are still on the DBT power supply voltage measurements are not trustworthy. Once you have seen they are 'close' get off the DBT for final measurements. Fix all the power supply issues first, get off the DBT, then move back to functionality testing. I do not see how the negative current supply to the lamps could affect the thyristor.

After Q19 and 20 are replaced try powering up without the DBT and no input, with volume all the way down. The fact that it plays ok without the thyristor and has no dc offset looks very promising to me. Once everything looks good then try it with the new thyristor.
Good luck!:thumbsup:
 
The way I would troubleshoot this is to run it with the triac removed and use a scope to see what triggers the triac, then trace the source of that signal. Running it on a DBT prevents the fuse from blowing so the triac has no protection and when it crowbars the supply, it conducts current for longer than it normally would and is destroyed.
 
The way I would troubleshoot this is to run it with the triac removed and use a scope to see what triggers the triac, then trace the source of that signal. Running it on a DBT prevents the fuse from blowing so the triac has no protection and when it crowbars the supply, it conducts current for longer than it normally would and is destroyed.
I agree, and the DBT does nothing to protect the thyristor from the initial current dump from the power supply filter cap. Once triggered it stays on until power is removed, DBT or no DBT.
An old electronics saying: "The semiconductor fails in order to save the fuse." :rflmao:
 
Last edited:
Are you buying all of these Triacs from eBay? That would likely explain why they aren't lasting very long.
Is the a known sub that the Pioneer folks use that you can get from a parts house?
 
Max input voltage to the triac gate is 1.5v. Surge current should be less then 30A. The tab on the triac is non insulated. AC voltage will be present on the tab when it is on. Be careful.
 
OK, put in a new thrystor. Using a dim bulb tester with 60 watt bulb. Have checked caps C77 and C78, in fact C77 was totally missing. I put in a 470uf 6.3v cap in the C77 position. Would its absence be enough to have caused this? It's in the schematic, but it wasn't there in my machine when I got it.

Turned the amp on... bulb glowed slightly and died down to nothing. OK that's good.

Testing voltages gingerly. Switching machine on and off for ten seconds at a time to test. No outrageous voltages around thrystor. Have not checked all voltages but noted that the collector of Q15 is supposed to have -20v if I'm reading the schematic right. it's got -29v ! I've pulled and checked R100, R101 and R123. All are supposed to be 2 watt resistors 62 ohms each. They all check out (out of circuit) at around 62 ohms. So why does the collector of Q15 have -29 v instead of -20 v?

Nothing's blown yet, but I'm not leaving it on to cook like the last one did. I think I remember checking the voltage before R 100, R101 and R123 and I believe it's -40v as it should be. Would Q15 even be in play for the thrystor problem blowing???

Oh, also checked the one fuse in the machine. It's supposed to be 3A. It was 3A. Also checked R110 and R111 (both 2 watt .5 ohms). Both checked out, out of circuit.

Hello,
I'm currently going through the exact same story !!!! Bought this amp with a blown rectifier diode and blown triac. Changed all rectifiers and the triac and I have the same behavior. It plays ok under the DBT but all of a sudden it blows the triac (in may case I believe it tends to do so when powering on or off).

Any chance you can share how the story ended on your side ?

Also, I have noticed that both vu-meters shows a "spike" approx. 1 or 2 seconds after power off. Is that normal ? They go up to let's say one fifth of the full scale.

Thank you
Tibia.
 
Back
Top Bottom