Dynacord PM2600 wierd fault.

[Jimbo69Jones]

Hi. This is my forst post here, so go easy on me 

I've a Dynacord 2600 i'm working on. When it came in to me, it was blowing fuses, and I traced that down to 5 failed output transistors in the Channel A amp. I've replaced those, and now the amp will power up. In addition to the output transistors, 2 of the 2 way 0.33, 7w ceramic resistors were short <EDIT: Open on 1 side each, sorry I made a mistake there). I can't find those anywhere, but I've made my own x2 by putting 2 x 7w wirewound 0.33 ohms in their place, and checked they give the same values as the originals.

So here's where it gets funky. I get one of 2 scenarios when starting the amp now.

Scenario A: Amp starts, fans come on for 2 seconds, but the channel A High limiter LED is lit, as well as the Protect LED which goes out after a few seconds when a relay kicks in and the fans stop. The High limiter remains on, and the channel is innoperable. The other 3 channels (A low, B Low and B High) all work in this scenario. I've tested this with a 40hz, and 1khz signal as per the service manual. Amp is drawing anywhere from 130 - 300w in this state.

Scenario B: Sometime when switching on, the unit draws over 2KW of mains AC and the A amp high side gets hot real quick. When the amp is like this, I read 183V DC across the NPN/PNP sides of AMP A. It does not do this all the time, and most times, I get the previous scenario.

I've checked for shorts etc, and none found. I've checked all the filter Caps, they are OK, checked the relays, they are all fine and done general testing of diodes/transitors etc (admittedly, not thoroughly). With the Amps disconnected, I get the first scenario. Can't decide if the fault with the limiter is in the amp, or before it.

If anyone has come across an issue like this before, or can offer some constructive help, I would be eternally grateful.

Warmest regards from Ireland

Jimbo

[Audiorepair]

Hi. This is my forst post here, so go easy on me 

I've a Dynacord 2600 i'm working on. When it came in to me, it was blowing fuses, and I traced that down to 5 failed output transistors in the Channel A amp. I've replaced those, and now the amp will power up. In addition to the output transistors, 2 of the 2 way 0.33, 7w ceramic resistors were short.

Just a couple of clarification questions, why did you replace an odd number of output transistors? 
Does that mean you left some originals in place on the blown amp(s), i.e. you didn't replace them all?

Also, I assume you mean some of the ceramic emitter resistors were open circuit, not short?

[Jimbo69Jones]

Just a couple of clarification questions, why did you replace an odd number of output transistors? 
Does that mean you left some originals in place on the blown amp(s), i.e. you didn't replace them all?

Also, I assume you mean some of the ceramic emitter resistors were open circuit, not short?

I took all 24 out (12 PNP, 12 NPN) and tested them 1 by one and in total there were 3 MJ15023's short circuit, 1 MJ15023 reading low, and 1 MJ5022G short circuit. I did not replace all 24, because at nearly 10 euros per transistor in Ireland, It would make the repair uneconomical. I was happy enough, that the ones I put back were in full operational order.

And yes, I meant open on the ceramics, my mistake. 1 side on each,

[Audiorepair]

OK, generally if an amp has blown several transistors it is prudent to perhaps just replace all of them on just that blown amp.
(I believe there are 4 amps in this particular model)

A transistor may test good under cold static conditions, then fail in circuit or at higher voltages/currents etc, especially when it has been fed huge amounts of current during such a catastrophic failure.

A shorted transistor will make all transistors on the other side pump loads of current into it, and they then can pop one by one, blowing emitter resistors in the process.
It's possible the emitter resistor will fail before the transistor does, but that poor old transistor has had a traumatic experience and may never recover.

[Jimbo69Jones]

I 100% agree in changing them all, but cost and stock availability here doesn't allow me to do so. Even so though, with the amp 100% disconnected, the AMP A High limiter is on, so there's another fault somewhere else. That's the one i'm currently trying to track down.

[Audiorepair]

OK, to clarify, I believe this particular unit has 4 separate amplifiers of similar power.

You say you removed all 24 output transistors, so that equates to 6 transistors per amplifier, yet you only replaced 5.
It is possible those 5 transistors all came from the same blown amplifier, meaning you still have one of those 6 in circuit, which may be your problem, or at least one of them.

It is also possible that one or more of the transistors failed open circuit in any of those 4 amps, so simply stopped helping in the amp it was in.

I would check those 5 transistors you replaced, and see how many of them are shorted Collector to Emitter, that would help to confirm a possible scenario such as I am suggesting.

You would not need to replace all 24 transistors if only one of the 4 amps had failed, just those 6.

[Audiorepair]

I took all 24 out (12 PNP, 12 NPN) and tested them 1 by one and in total there were 3 MJ15023's short circuit, 1 MJ15023 reading low, and 1 MJ5022G short circuit.

Sorry, I was lazy and didn't go back and check what you said.

So, it is likely these 4 shorted transistors and the one reading low, and the blown emitter resistors, were all the consequence of a catastrophic failure of the A Low amp channel.
The one remaining transistor is back in circuit, and is likely not very well.


Edit:  And is possibly now in the A High amp channel causing your issues.

[Jimbo69Jones]

It has 2 seperate boards, each with 24 output transistors on. Each board is divided into 2 amps, so 48 output transistors in total, 12 per Amp if that makes sense.

I'll take them all out again (I at least marked the new replaced ones so I can identify those easy enough. I ordered a Peak Atlas DCA 75 Pro which arrived yesterday which should be able to test them further. Hopefully this will turn something up. I'll keep you posted.

[David_AVD]

If you ever have blown output transistors, you should check the driver and bias transistors as well. Also check all resistors in the area.

[Jimbo69Jones]

If you ever have blown output transistors, you should check the driver and bias transistors as well. Also check all resistors in the area.

Good advice, will do.

I've tested all the power transistors on this amp board, both sides (so 12 NPN, 12 PNP) and they are all testing with very similar readings, except the gain figures (hFE) vary quite a bit. On the PNP transistors (MJ15023G) The hFE figures are around 70 - 82, on the NPN's (MJ15022G) gains are from 30 - 54 (this is being measured with a Peak Atlas DCA75 Pro) I'm not convinced I need to go and order a whole batch of new transistors, but maybe i'm wrong. Would love some input on this, and if there is any way to test these transistors more thoroughly without very specialist equipment. Here in Ireland, i'm looking at around 10 euros per transisor + 23% VAT and then shipping, so hoping to avoid having to replace a huge amount of them for a 20+ year old amp.


Update on this:

Both amp boards are identical. I've referenced every capacitor, transistor, resistor etc between the 2 boards and everything is checking out identical. I've made the decision, that i'm going to swap the boards around. If the fault I am now seeing is in the transistors, then then fault will move to the other channel. I suspect when I do this, that won't happen. My gut is telling me the fault is before the amp section, but I guess i'm about to find out.

[elecdonia]

I've tested all the power transistors on this amp board, both sides (so 12 NPN, 12 PNP) and they are all testing with very similar readings, except the gain figures (hFE) vary quite a bit. On the PNP transistors (MJ15023G) The hFE figures are around 70 - 82, on the NPN's (MJ15022G) gains are from 30 - 54 (this is being measured with a Peak Atlas DCA75 Pro)

For nearly all pairs of complementary audio power transistors the PNP versions have higher hFE than their NPN complements. Contrary to some opinions it really isn’t necessary to match the hFE of audio output transistors. “Matching” is important for tube (valve) amplifiers but not nearly as important for solid-state amplifiers. The hFE figures you listed indicate that all of your output transistors are still OK to use. My experience is that nearly all audio output transistor failures result in a dead short (or totally open circuit) condition. Typically the short is from E to C. However, if the transistor isn’t shorted or open, and it measures “as a transistor” by a transistor tester, then it is still good.

Both amp boards are identical. I've referenced every capacitor, transistor, resistor etc between the 2 boards and everything is checking out identical. I've made the decision, that i'm going to swap the boards around.

This is a good plan.

Another way to test is to remove all of the output transistors and then connect 47 or 100 ohm 5 watt resistors across E-B in place of one of the (removed) parallel output transistors for each leg of the output stage. For ordinary amplifiers with balanced +/- power rails, this requires 1 resistor in place of one NPN output transistor and the other resistor in place of one PNP output transistor. For high-power amplifiers like this PM2600, 4 resistors are needed, one for each leg of the “H bridge” output stage. Do not load the amplifier’s output terminals. Leave them open circuit. The DC output offset should be < +/- 200mV. The voltage drop across each of the added resistors should be from 400 to 550mV. If not, then there is still a fault in the driver stages which must be fixed before reinstalling the output transistors.  With this test setup an audio signal applied to the amplifier input should appear (in amplified form) at the output terminals as long as the output is “unloaded.” Basically the driver circuit should be capable of functioning as a voltage amplifier. But of course it cannot drive a load without the main output transistors being present.

[Jimbo69Jones]

For nearly all pairs of complementary audio power transistors the PNP versions have higher hFE than their NPN complements. Contrary to some opinions it really isn’t necessary to match the hFE of audio output transistors. “Matching” is important for tube (valve) amplifiers but not nearly as important for solid-state amplifiers. The hFE figures you listed indicate that all of your output transistors are still OK to use. My experience is that nearly all audio output transistor failures result in a dead short (or totally open circuit) condition. Typically the short is from E to C. However, if the transistor isn’t shorted or open, and it measures “as a transistor” by a transistor tester, then it is still good.

I wasn't worried about the hFE differences between the NPN - PNP transistors, I was referring to the differences between each NPN, each PNP, sorry I should have been clearer there. You've basically confirmed exactly the way I was thinking that these differences (in tolerence from the datasheets) won't be the issue, so thanks a million for confirming that.

This is a good plan.

Another way to test is to remove all of the output transistors and then connect 47 or 100 ohm 5 watt resistors across E-B in place of one of the (removed) parallel output transistors for each leg of the output stage. For ordinary amplifiers with balanced +/- power rails, this requires 1 resistor in place of one NPN output transistor and the other resistor in place of one PNP output transistor. For high-power amplifiers like this PM2600, 4 resistors are needed, one for each leg of the “H bridge” output stage. Do not load the amplifier’s output terminals. Leave them open circuit. The DC output offset should be < +/- 200mV. The voltage drop across each of the added resistors should be from 400 to 550mV. If not, then there is still a fault in the driver stages which must be fixed before reinstalling the output transistors.  With this test setup an audio signal applied to the amplifier input should appear (in amplified form) at the output terminals as long as the output is “unloaded.” Basically the driver circuit should be capable of functioning as a voltage amplifier. But of course it cannot drive a load without the main output transistors being present.

Thats brilliant information, and I will definately be saving that for future reference/use. It's getting a bit late in the day here now, so I think I may wait until tomorrow now to swap the amp boards over completely. Thanks so much for your response, it's made me feel a bit more confident in the way i'm approaching this.

[Jimbo69Jones]

Unfortunately, life has a way of throwing obstacles in the way, and I haven't had a moment to look at this amp since.

I will have time in the coming week, and I will be back to it, and will post my findings, and hopefully, end up with a working amp.

Thanks for everyone's help/input so far.