I'm trying to repair an AER 40 amplifier that has a popping/crackling sound on startup but goes away after a few minutes as it warms up. This is a solid-state, transistor-based amplifier.
I haven't repaired a guitar amp before but am fairly decent with electronics. I have a lab with a bench power supply, DSO, Brymen multimeter, isolation transformer, hot air rework Station, good soldering iron etc.
My first thoughts on this fault are: cold solder joint/cracked joint, or a transistor fault.
Visual inspection didn't show any leaking or blown capacitors. I have tapped all the small transistors that I have seen, tried cold spray when it is crackling away but haven't managed to narrow the problem area down.
Could someone help me work through a logical diagnosis process or point me towards likely problem areas?
Would the mains power supply transformation circuit be at risk (likely the primary electrolytics?). Without a schematic, I don't know where to start.
I haven't repaired a guitar amp before but am fairly decent with electronics. I have a lab with a bench power supply, DSO, Brymen multimeter, isolation transformer, hot air rework Station, good soldering iron etc.
My first thoughts on this fault are: cold solder joint/cracked joint, or a transistor fault.
Visual inspection didn't show any leaking or blown capacitors. I have tapped all the small transistors that I have seen, tried cold spray when it is crackling away but haven't managed to narrow the problem area down.
Could someone help me work through a logical diagnosis process or point me towards likely problem areas?
Would the mains power supply transformation circuit be at risk (likely the primary electrolytics?). Without a schematic, I don't know where to start.
Totally sounds like dry joints to me. Freezer spray and/or hot air wands might help locate the problem. No need for a schematic to test for dry joints / faulty connections, just a chopstick prodder, or above. Forget about faulty capacitors or other bits 'cos, good or bad, the circuit isn't going to work with dodgy connections.
Thanks for your reply. I have disassembled the amp and have inspected all joints thoroughly under magnification - no dry joints but I reflowed many of them anyway. Nothing on a prod test. No obvious leaking/bulging caps.
There is a 100Hz hum on the output with intermittent high frequency noise as well. Main filter caps seem to be soldered well. I pulled them and tested them with my cheap ESR meter - said the ESR was 0.00 ohms each which is bizarre (high value caps, 2200uF 35V IIRC). Leakage current was about 20mA each.
The 100Hz hum suggests leakage from a full-wave rectifier. The amp has a set of four discreet diodes that rectify the mains after a step-down transformer. I'm suspicious that the main filter caps have gone. What do you think?
Immediately after turn-on - high frequency crackling noise
A few seconds later - 100Hz with superimposed high frequency crackling noise (ignore freq readout on screenshot)
Pure 100Hz ripple without any crackling (rare!)
There is a 100Hz hum on the output with intermittent high frequency noise as well. Main filter caps seem to be soldered well. I pulled them and tested them with my cheap ESR meter - said the ESR was 0.00 ohms each which is bizarre (high value caps, 2200uF 35V IIRC). Leakage current was about 20mA each.
The 100Hz hum suggests leakage from a full-wave rectifier. The amp has a set of four discreet diodes that rectify the mains after a step-down transformer. I'm suspicious that the main filter caps have gone. What do you think?
Immediately after turn-on - high frequency crackling noise
A few seconds later - 100Hz with superimposed high frequency crackling noise (ignore freq readout on screenshot)
Pure 100Hz ripple without any crackling (rare!)
20mA leakage is kinda high for a 35V2200 part. But those waveforms look more like a rectifier with a serious leakage problem.
Was the electrolytic leakage tested at supply voltage, or just some volt or two applied by the test apparatus? If it's 20mA at a couple of volts, they need to be replaced.
Cheers
Was the electrolytic leakage tested at supply voltage, or just some volt or two applied by the test apparatus? If it's 20mA at a couple of volts, they need to be replaced.
Cheers
Last edited:
Hi Rick,
Cheers for the reply. I was using a standalone component tester. I am not sure what voltage it is used to actually test cap charging, but it is run off a 9V battery into a logic-level ATMEGA microcontroller so cannot be that high.
Yes, that is what I thought. Leaked full-wave rectifier at 2x my local mains frequency. Apart from faulty filtering caps, is there anything else that could cause this? I'm pretty sure not, but doubting myself, so would appreciate a second opinion.
Cheers for the reply. I was using a standalone component tester. I am not sure what voltage it is used to actually test cap charging, but it is run off a 9V battery into a logic-level ATMEGA microcontroller so cannot be that high.
Yes, that is what I thought. Leaked full-wave rectifier at 2x my local mains frequency. Apart from faulty filtering caps, is there anything else that could cause this? I'm pretty sure not, but doubting myself, so would appreciate a second opinion.
Any chance you have, or could borrow, a bench supply to sub for that rail? That would clear up several possibilities in short order.
Hi Rick,
Yes, I do have a bench power supply.
These are the main filter caps after the full-wave rectifier from the main toroidal power transformer. They produce two rails, +28 and -28VDC. This then goes off and gets split into a +24VDC (zener diode), -15VDC, +15VDC and +5VDC VCC (these three via linear voltage regulators). There is 100Hz noise present on the inputs to all of these regulators which has not been filtered out by the bypass caps on these rails.
Yes, it does seem that 100Hz should be less conspicuous at the regulators.
If you set your DMM to AC volts, what value does it show at the various regulator inputs?
Maybe choose the positive or negative rail with the highest AC voltage, and carefully substitute for it.
If you set your DMM to AC volts, what value does it show at the various regulator inputs?
Maybe choose the positive or negative rail with the highest AC voltage, and carefully substitute for it.
Last edited:
20 uA would be fine, 20 mA is way too high. It's the kind of value you could get for the first few minutes when the capacitor hasn't been used in many decades.
If the Zener diode is meant as a shunt regulator rather than an overvoltage protection, then a 28 V rail voltage is unusually low for a 24 V Zener. 28 V at nominal mains voltage means approximately 25.2 V when the mains is 10 % below nominal. Subtract the peak ripple voltage and you get awfully close to 24 V.
If the Zener diode is meant as a shunt regulator rather than an overvoltage protection, then a 28 V rail voltage is unusually low for a 24 V Zener. 28 V at nominal mains voltage means approximately 25.2 V when the mains is 10 % below nominal. Subtract the peak ripple voltage and you get awfully close to 24 V.
Hi Rick,
What I think I shall do is replace the filter caps, then proceed to retest the regulator inputs as you suggested if the problem persists. At present, there is say 32V across the main filter caps with 1.5V AC measured across them with a DMM.
What I think I shall do is replace the filter caps, then proceed to retest the regulator inputs as you suggested if the problem persists. At present, there is say 32V across the main filter caps with 1.5V AC measured across them with a DMM.
Hi Marcel,
I've got hold of the schematics but can't share them. Yes, it's a shunt regulator. Nominally +29VDC input from the rectified waveform post the toroidal transformer off the mains (I read 32V, but I think that is OK variability). It is a 24V shunt zener to ground to provide +24VDC.
I would replace the bridge, too. Or at least have the DMM on AC volts across one of the new caps -- that way you can cut the power very quickly if the ripple is too high. (Several minutes / sustained high ripple current could mean early death for the new electrolytics.)
Regards
Regards
Can you share at least a block diagram?
Most brands do that, and evem print them in their user manuals, way too general to disclose any secret, but would help us suggest more specific tests.
Most brands do that, and evem print them in their user manuals, way too general to disclose any secret, but would help us suggest more specific tests.
Replacing the caps seems to have fixed the problem! I'd say this is a success...for now. Bit more testing but we shall see.
Message back from my friend: 'still crackling a bit even without a cable plugged in at one point. It's much better now so whatever you've done was great'
Aargh. So, to recap (no pun intended). Changing the main filter caps greatly improved the problem. In fact, when I did some audio tests etc at home, I heard nothing. However, perhaps as he has played it for longer, and it has warmed up, a fault had arisen.
There is only the four discrete rectifying diodes upstream of these caps in the circuit - I've tested them with a multimeter and they seem to be okay but worth replacing them anyway?
Or am I looking in the wrong section of the amplifier??
AFAIK, most DMM's don't check reverse leakage -- only forward voltage drop. And while there is often a correlation between a 'soft knee' in the forward direction and reverse leakage, the single point test that the DMM performs will not likely catch that. The cause of the indication in your 'scope pics would be substantial leakage.
But since you're describing 'crackling' and not 'hum' as the 'scope pics showed, maybe you could post a recording . . ?
Cheers
But since you're describing 'crackling' and not 'hum' as the 'scope pics showed, maybe you could post a recording . . ?
Cheers