Mystery amp schematic - assistance/comments/guesses/insights welcomed
- Thread starter bugbear
- Start date
BadgeringThe virtual ground (halved from +V by a pair of 10KΩ resistors) is given a 0.15 Hz
cut-off frequency by the 100µF capacitor.
The pre-amp +V is (in turn) derived from a 35V rail by a 230Ω resistor dropper, 24V zener, and 220µF capacitor
Virtual ground ought to be pretty clean, assuming the capacitors are still capacat-ing.
cut-off frequency by the 100µF capacitor.The pre-amp +V is (in turn) derived from a 35V rail by a 230Ω resistor dropper, 24V zener, and 220µF capacitor
Virtual ground ought to be pretty clean, assuming the capacitors are still capacat-ing.
A youtube video on guitar amp repair pointed out something that should have been obvious.
If the smoothing/filtering capacitors are failing, the result (courtesy of the rectifiers) is a 100Hz wave form.
OTOH, if the fault is mains interference on the signal path, you get a 50Hz wave.
This will be very helpful in diagnosing my fault when I next get the PCB out of the cabinet.
If the smoothing/filtering capacitors are failing, the result (courtesy of the rectifiers) is a 100Hz wave form.
OTOH, if the fault is mains interference on the signal path, you get a 50Hz wave.
This will be very helpful in diagnosing my fault when I next get the PCB out of the cabinet.
OK. The PCB is back out of the cabinet (and a less amp is being used for guitar playing).
First check (from previous post). The signal on the output of the input op-amp is 20mV, 50Hz. It's NOT the power rails, it's mains hum pickup.
So my lovely new capacitors won't be needed, and nor will the helpful tips on removing the old capacitors.
So where's the hum coming from?
badgers_earth by plybench, on Flickr
Worryingly, the mains earth only gets as far as the transformer - the front panel, and the metal frames of the pots are not earthed.
Moving to the signal path, I've found something interesting. With no jack plugs inserted, the (switched) jack inputs connect the signal leads to (signal) ground. I've checked this with a resistance meter, and - yep - shorted to ground.
And yet...
badger in/out by plybench, on Flickr
The + input of OA2 has 2mV of 50Hz on it. The "shared node" of the 2k2, 68k and 1Nf capacitor also has 2mV of 50Hz. The "normal" and "bright" leads do NOT have any signal. But the node that joins the 220pF to the 2k2 has the 2mV of 50Hz.
My small brain says the filter capacitor (all 220pF of it) is the source of the introduced mains hum.
EDIT: checking under magnification, the polystyrene cap is marked "2200 MX", so perhaps 2200pF (2.2nF) not 220pF.
Is this possible/feasible/nonsense? Advice on what to check, what to try, how to proceed welcomed.
First check (from previous post). The signal on the output of the input op-amp is 20mV, 50Hz. It's NOT the power rails, it's mains hum pickup.
So my lovely new capacitors won't be needed, and nor will the helpful tips on removing the old capacitors.
So where's the hum coming from?
badgers_earth by plybench, on FlickrWorryingly, the mains earth only gets as far as the transformer - the front panel, and the metal frames of the pots are not earthed.
Moving to the signal path, I've found something interesting. With no jack plugs inserted, the (switched) jack inputs connect the signal leads to (signal) ground. I've checked this with a resistance meter, and - yep - shorted to ground.
And yet...
badger in/out by plybench, on FlickrThe + input of OA2 has 2mV of 50Hz on it. The "shared node" of the 2k2, 68k and 1Nf capacitor also has 2mV of 50Hz. The "normal" and "bright" leads do NOT have any signal. But the node that joins the 220pF to the 2k2 has the 2mV of 50Hz.
My small brain says the filter capacitor (all 220pF of it) is the source of the introduced mains hum.
EDIT: checking under magnification, the polystyrene cap is marked "2200 MX", so perhaps 2200pF (2.2nF) not 220pF.
Is this possible/feasible/nonsense? Advice on what to check, what to try, how to proceed welcomed.
Some searching online reveals that polystyrene caps have an inner and outer foil, and that the outer foil should be connected to the lower impedance side of the circuit.
Small Signal Audio Design - Page 240
Douglas Self · 2014 · Computers
Polystyrene capacitors have two foils, and one of them will be on the outside of the component, and so vulnerable to capacitive crosstalk and hum pickup.
https://www.aikenamps.com/index.php/where-to-connect-the-outside-foil-on-capacitors
Could this be my problem - did Badger get it the wrong way round?
Small Signal Audio Design - Page 240
Douglas Self · 2014 · Computers
Polystyrene capacitors have two foils, and one of them will be on the outside of the component, and so vulnerable to capacitive crosstalk and hum pickup.
https://www.aikenamps.com/index.php/where-to-connect-the-outside-foil-on-capacitors
Could this be my problem - did Badger get it the wrong way round?
Last edited:
Putting a quick 'n' dirty short across the polystyrene cap on the rear side of the board reduced the output of OA2 from 30mV down to 7mV, which is much better. This is with nothing plugged into the input jack sockets.
So the cap is introducing a fair bit of hum signal at the input of OA2.
So the cap is introducing a fair bit of hum signal at the input of OA2.
I'm an idiot - that "short" shorted the 2K2 to ground, which would (of course) alter "everything".
I have now tried (per the information on valve amp and vintage radio forums) reversing the polystyrene cap, to get the claimed "outer foil shielding" effect. It made no difference at all - if anything the main hums output of OA2 went up from 24 to 27 mV. I don't have anything sensitive enough to measure the change on the input to OA2.
a.palfreyman
pfm Member
Yes, know the problem well. Had some random noise in a diy MM phono amp that could only be measured at the output of the first gain stage. Was a PITA to trace...
Anyway on your photo above you have yellow tap and possibly two earth wires going to board marked 26Vac. These go very close to purple and green wires marked signals. Do these two wires (purp gn) carry signal and return? If so you could be getting pick up in the open loop formed between them.
Anyway on your photo above you have yellow tap and possibly two earth wires going to board marked 26Vac. These go very close to purple and green wires marked signals. Do these two wires (purp gn) carry signal and return? If so you could be getting pick up in the open loop formed between them.
Purple and red are both "signal", and green is the ground/return.
The "2 greens and a yellow" are a centre-tapped triplet from the transformer, running 26V RMS.
With no jack inserted in either input socket, both purple and red are actually connect to green (!!).
And it hums; in order to help the conversation, I've redrawn the input buffer/gain stage schematic, and labelled nodes.
hums by plybench, on Flickr
With no jacks in the socket (see above) the voltages (measure with my toy scope, peak to peak) are as follows:
I have been having vague thoughts about making a tinfoil covered cardboard cage, to see if that helps (either fix the fault, or at least diagnose it).
But I am (very) open to suggestions as to how to proceed.
The "2 greens and a yellow" are a centre-tapped triplet from the transformer, running 26V RMS.
With no jack inserted in either input socket, both purple and red are actually connect to green (!!).
And it hums; in order to help the conversation, I've redrawn the input buffer/gain stage schematic, and labelled nodes.
hums by plybench, on FlickrWith no jacks in the socket (see above) the voltages (measure with my toy scope, peak to peak) are as follows:
- 0 V
- 2mV, 50Hz
- 2mV, 50Hz
- 0v
- 2mV, 50Hz
- 30mV, 50Hz (but very dirty waveform)
I have been having vague thoughts about making a tinfoil covered cardboard cage, to see if that helps (either fix the fault, or at least diagnose it).
But I am (very) open to suggestions as to how to proceed.
a.palfreyman
pfm Member
At one end of the 1nF cap you have 100k and the other you have 68k // 2k2 that is itself decoupled with a 2n2. This gives 2mV at nodes 2, 3 and 5 with a subsequent gain of 15 giving 30mV on oa2 output. Shunting node 2 to virtual ground pulls this down by a factor of 4 to 7mV oa2 output. This tends to indicate that the 1nF, which is floating between a 2k2 and 2n2 cap // 68k at one end and high impedance of 100k the other could be the culprit, or possibly an interaction between this cluster of caps and resistors. Is this portion of circuit near some ac?
Might be worth scoping across nodes 2&3 and 3&5.
Edit, just had another look at your image top of the page. Red and purple are an open loop that could be picking up mains passing current through 2n2, 2k2 and 68k. Can you try reducing the loop area between these wires?
Might be worth scoping across nodes 2&3 and 3&5.
Edit, just had another look at your image top of the page. Red and purple are an open loop that could be picking up mains passing current through 2n2, 2k2 and 68k. Can you try reducing the loop area between these wires?
I haven't stated (people may have guessed) that I am using my scope AC coupled.
The sad thing is, I was so convinced that the hum was caused by 40 year old electrolytics allowing mains hum onto the power rails, that I've already ordered (and am in possession of) a full set of replacements...
The sad thing is, I was so convinced that the hum was caused by 40 year old electrolytics allowing mains hum onto the power rails, that I've already ordered (and am in possession of) a full set of replacements...
a.palfreyman
pfm Member
This is all guesswork on my part. Could be totally wrong.
a.palfreyman
pfm Member
Ah, I didn't know the earth's are separate. It could be a bit of remaining ripple across the 100uF that goes to the node of the two 10k resistors, esp if that cap is on the way...so measure between the earth planes too.
The really big question is am I trying to repair a broken (or worn/aged) amplifier, or improve a badly designed one?
Given the high quality of most of the components, fancy pants tone controls, I suspect the former, but I can't be sure.
It appears that Badger were selling "serious" practice amps (perhaps "rehearsal amps"), aimed at professional musicians.
Given the high quality of most of the components, fancy pants tone controls, I suspect the former, but I can't be sure.
It appears that Badger were selling "serious" practice amps (perhaps "rehearsal amps"), aimed at professional musicians.
The class-B power amp runs off plus/minus 35V rails, with a centre tap in between. For the purposes of the overall amplifier PCB I am calling this "ground" (although it is not a true earth).
For the pre-amp/tone section, the +35V rail is dropped across a resistor/zener pair to +24V, and a mid point at +12V established by a pair of 10k resistors working as a resistor divider. From the point of view of the OP-amps it is simpler to view this not as a 0/24 power supply, with a mid point at +12V, but as a -12/+12 power supply, and to consider +12v as "virtual earth".
(this is apparently quite common, according to instructional electronics websites)
For the pre-amp/tone section, the +35V rail is dropped across a resistor/zener pair to +24V, and a mid point at +12V established by a pair of 10k resistors working as a resistor divider. From the point of view of the OP-amps it is simpler to view this not as a 0/24 power supply, with a mid point at +12V, but as a -12/+12 power supply, and to consider +12v as "virtual earth".
(this is apparently quite common, according to instructional electronics websites)
a.palfreyman
pfm Member
OK, if the 24V PSU section to the opamps is zener stabilised up front, there shouldn't be any ripple at the mid point (virtual earth). Can you scope across nodes 2&3 and 3&5, then twist these wires together and remeasure as it has to be loop across the red and purple picking up AC.
a.palfreyman
pfm Member
Lucky bar-steward.
You could just try twisting them together and see if it has any effect. Don't even have to unsolder anything, just pinch together roughly in the middle and twist them. It'll tell you if you are on the right track.
You could just try twisting them together and see if it has any effect. Don't even have to unsolder anything, just pinch together roughly in the middle and twist them. It'll tell you if you are on the right track.