Hi,
I'm in the process of fixing a specialty audio system and have discovered that the Grommes continuous duty 70V power amplifier in question is damaged. The amp had a blown fuse on one leg of the bias supply. Fuse blows immediately potentially suggesting a shorted output transistor. The other supply leg seems fine.
There is a monitor speaker and it was producing a repetitive chirping noise along with a sparking seen in the surge arrestor device that sits across the 70V output. The system was immediately shut down.
I have a few questions for the more seasoned experts. This amplifier is around 20 years old, and the design is fairly straightforward and robust. My question is whether I can determine if the failure is simply due to age, or if another factor (bad 70V driver, surge, lightning, or something else) could have caused it.
The desire is to repair (and future proof) the amplifier given that this design is reliable and parts are available. The output transistors can be sourced and I can replace all 6 (two sets of 3) for about $60. The bigger question is if I should be looking for other failed parts or anything else that might do well from future proofing aside the obvious (electrolytic capacitors). Even if the other set of 3 transistors appear fine, would replacement be in the best interest of future reliability?
Thanks for any advice! Schematic is attached.
I'm in the process of fixing a specialty audio system and have discovered that the Grommes continuous duty 70V power amplifier in question is damaged. The amp had a blown fuse on one leg of the bias supply. Fuse blows immediately potentially suggesting a shorted output transistor. The other supply leg seems fine.
There is a monitor speaker and it was producing a repetitive chirping noise along with a sparking seen in the surge arrestor device that sits across the 70V output. The system was immediately shut down.
I have a few questions for the more seasoned experts. This amplifier is around 20 years old, and the design is fairly straightforward and robust. My question is whether I can determine if the failure is simply due to age, or if another factor (bad 70V driver, surge, lightning, or something else) could have caused it.
The desire is to repair (and future proof) the amplifier given that this design is reliable and parts are available. The output transistors can be sourced and I can replace all 6 (two sets of 3) for about $60. The bigger question is if I should be looking for other failed parts or anything else that might do well from future proofing aside the obvious (electrolytic capacitors). Even if the other set of 3 transistors appear fine, would replacement be in the best interest of future reliability?
Thanks for any advice! Schematic is attached.
Attachments
I would first check smoothing caps for shorts.
Then if ok I would check all components with a meter.
Diode check transistors. They usually go short CE.
Then if ok I would check all components with a meter.
Diode check transistors. They usually go short CE.
Agree and add: do all testing/repairing with the OT **disconnected**; it´s a short for DC and the probable cause only one rail fuse blew.
I expect one or more shorted transistors in the negative rail, probably the driver too.
Do not search for 2N3773 (unless you buy them at the ON factory door he he) but use modern and more robust MJ1502x instead, bought at a reputable supplier, while MJ1503x would fit as drivers.
It´s worth keeping this robust workhorse running.
Why it failed? Who knows? Who cares? The design is good, sh*t happens or the Servicing industry would not exist.
Use good components, blow dust away, lubricate pots and switches, it should work another 20 years, or more.
EDIT:if available, please post the Preamp/Mixer section also, it might help another user attracted by the Grommes search label.
I expect one or more shorted transistors in the negative rail, probably the driver too.
Do not search for 2N3773 (unless you buy them at the ON factory door he he) but use modern and more robust MJ1502x instead, bought at a reputable supplier, while MJ1503x would fit as drivers.
It´s worth keeping this robust workhorse running.
Why it failed? Who knows? Who cares? The design is good, sh*t happens or the Servicing industry would not exist.
Use good components, blow dust away, lubricate pots and switches, it should work another 20 years, or more.
EDIT:if available, please post the Preamp/Mixer section also, it might help another user attracted by the Grommes search label.
Last edited:
Thanks for the suggestions - I'll definitely check everything over thoroughly and pull the OT from the circuit before powering it again. I need to get some fuses - the installed fuses had a very thin filament, so I presume they are fast blow style given their 10A rating.
I was going to buy the transistors from a reputable vendor - is there a method to choose the proper replacement models from the MJ1502x family? Would it just be whatever is cost effective, or is there a reason to use (or not use) the higher voltage rated part? I see the MJ15022 NPN is complimentary to the MJ15023 PNP and MJ15024 NPN to the MJ15025 PNP.
I presume if they are replaced they should be of identical specs on each rail and between both rails. (Both original output transistors are available as NTE parts, and ON has a RoHS compliant 2N3773 but not a 2N6609 - but given the suggestion that the modern parts would be more robust, I would think that would be ideal here).
I presume since I'm at it anyway, that replacing the electrolytics would be cheap insurance for a long lifespan.
The only reason I'm concerned for the failure is that I don't want someone to plug in the freshly repaired amplifier into the 70V speaker system and have it get killed due to a failure there. Would a failed 70V horn driver cause this risk, or more likely just a lack of audio?
There isn't a Grommes Preamp/Mixer - the audio comes from a proprietary board that generates the audio and mixes it. That mixed signal feeds via a mono RCA cable into this barebones yet reliable power amplifier to drive the 70V speakers. It has an open chassis inside the system cabinet, and nothing on the case save a power cord, circuit breaker, single input jack, and single set of output terminals. Simple and effective!
I was going to buy the transistors from a reputable vendor - is there a method to choose the proper replacement models from the MJ1502x family? Would it just be whatever is cost effective, or is there a reason to use (or not use) the higher voltage rated part? I see the MJ15022 NPN is complimentary to the MJ15023 PNP and MJ15024 NPN to the MJ15025 PNP.
I presume if they are replaced they should be of identical specs on each rail and between both rails. (Both original output transistors are available as NTE parts, and ON has a RoHS compliant 2N3773 but not a 2N6609 - but given the suggestion that the modern parts would be more robust, I would think that would be ideal here).
I presume since I'm at it anyway, that replacing the electrolytics would be cheap insurance for a long lifespan.
The only reason I'm concerned for the failure is that I don't want someone to plug in the freshly repaired amplifier into the 70V speaker system and have it get killed due to a failure there. Would a failed 70V horn driver cause this risk, or more likely just a lack of audio?
There isn't a Grommes Preamp/Mixer - the audio comes from a proprietary board that generates the audio and mixes it. That mixed signal feeds via a mono RCA cable into this barebones yet reliable power amplifier to drive the 70V speakers. It has an open chassis inside the system cabinet, and nothing on the case save a power cord, circuit breaker, single input jack, and single set of output terminals. Simple and effective!
Last edited:
OK - the amplifier lives! It turns out just one output transistor of the 6 had a short (2N3773). I replaced all 6 with a new set of the MJ1502x just to be safe since this will be in a continuous duty application, hopefully for years to come. I replaced the electrolytics for good measure as well. It turns out that the parts with the higher voltage rating were cheaper, maybe due to higher volume, so that is what I used as I didn't see a downside. The amp was run on the bench, no load attached, and then placed back into the audio system without incident. It has been powered for over a day now and the heat sinks are hardly warm.
A few questions for those with more experience than me:
1. The bias control says to adjust for 0.90 VDC, however I cannot get the voltage that low with the new parts. I presume there may be a difference here due to the part change? My Fluke meter reads 1.02 VDC at the lowest point on the pot. There is no drift, so that is good.
2. I also adjusted the DC balance pot, the unit has less hum with it turned all the way to one end and the voltages on the rails differ by 0.3 V (+45.8 V & -45.6 V)
I presume this is probably good, as it certainly seems to be operating without much issue. Granted, I just have a small speaker attached at the moment but the amp remains powered continuously.
A few questions for those with more experience than me:
1. The bias control says to adjust for 0.90 VDC, however I cannot get the voltage that low with the new parts. I presume there may be a difference here due to the part change? My Fluke meter reads 1.02 VDC at the lowest point on the pot. There is no drift, so that is good.
2. I also adjusted the DC balance pot, the unit has less hum with it turned all the way to one end and the voltages on the rails differ by 0.3 V (+45.8 V & -45.6 V)
I presume this is probably good, as it certainly seems to be operating without much issue. Granted, I just have a small speaker attached at the moment but the amp remains powered continuously.
Adjust “what” for 0.9 V? Between two test points? How much voltage is across one of the emitter resistors? That should be at most a handful of millivolts.
The service note indicates to measure the voltage between the bases of the NPN array and the bases of the PNP array. The 100 ohm trimpot (R25) is being adjusted.
There were issues with self-oscillation of the 70V transformer (distortion present) when the amps were built without bias trim adjustment, so it was re-added according to a service note.
It says to not normally perform the adjustment unless there is a problem, but I was concerned as I replaced the transistors with the more robust versions that it might be pertinent to re-set the bias for longer life.
There were issues with self-oscillation of the 70V transformer (distortion present) when the amps were built without bias trim adjustment, so it was re-added according to a service note.
It says to not normally perform the adjustment unless there is a problem, but I was concerned as I replaced the transistors with the more robust versions that it might be pertinent to re-set the bias for longer life.
Ok, that makes sense. About 0.45-0.5 volts of vbe on the output transistors, making it pretty much pure class B to minimize power consumption at idle. Some other famous amps run it as low as 0.35v.
So if I am reading it right, it sounds like running slightly higher shouldn't be too significant, aside from a slightly increased power consumption at idle. The tech note clarifies that testing with lowered bias (no bias pot) revealed issues with self oscillating tendency of the 70v transformer when hooked to certain types of matching transformers on speakers. The result was an unusual distortion.
The note says to adjust the bias if idle temp is high, distortion is noticed, or components have been replaced.
The amplifier has been powered continuously for another full day, and sounds great when in use. The one remaining question is about the lifespan of electrolytic in the preamp section. They all look physically good and I usually replace all electrolytic caps (I tend toward overkill) but I didn't have these on hand. If they are 20 years old would I be better off just changing them before it returns to service or is their use less critical than power supply filters that take more abuse and overkill?
Last edited:
With that low a bias voltage on the outputs I doubt there is any idle current in them. Current in the drivers would go up a smidge (about 10%). From what it says in the manual, you get oscillations if the bias is too *low*. This would have to do with the nonlinearity associated with it, and not so much a Nyquist instability. It would be localized to the output stage, and only happen over a small range of output voltage/current. These “parametric” oscillations happen all the time in RF (my actual professional field) and are usually non-destructive but very annoying. The old quasi-comp Phase Linears used to do this if the bias was either too low or too high. It’s less common will full complementary amps, but the output transformer itself may have something to do with it, since it can have some frequency where it’s impedance would be almost a pure reactance. A marginal stability, a high Q tuned circuit, and something to pump it at another frequency are the 3 ingredients for a parametric oscillation.
I replace all small electrolytics in old amps that get worked on. It’s too simple and easy, and they are the ones that go out. Big soup cans often work for decades with no issues. I’ll replace them if there is some sign that they are bad and only then.
Thanks so much for the follow-up!
It sounds like the bias point will likely be OK. I have only witnessed crisp, clear sound even hooking up a cheap 70v transformer with an additional speaker. The heat sinks are barely warm.
From the note's text it sounds like there was pretty bad distortion that resulted. Reading some of the additional details it sounds like that was exactly the issue - pairing a specific type of matching transformer on the other end of the line must have created just the right conditions to cause the oscillation.
I'll get the other capacitors on order so that I can have the peace of mind that I've made sure something silly won't bring the unit out of service any time soon.
It sounds like the bias point will likely be OK. I have only witnessed crisp, clear sound even hooking up a cheap 70v transformer with an additional speaker. The heat sinks are barely warm.
From the note's text it sounds like there was pretty bad distortion that resulted. Reading some of the additional details it sounds like that was exactly the issue - pairing a specific type of matching transformer on the other end of the line must have created just the right conditions to cause the oscillation.
I'll get the other capacitors on order so that I can have the peace of mind that I've made sure something silly won't bring the unit out of service any time soon.