@fatmarley
Mat - you know where to find me - I've a nice working Philips 20Mhz scope with probes that's yours free, old friend.
I was given it, to release into the wild per my incentive on the other place: it's wholly yours if you'd like.
ATB.
Mat - you know where to find me - I've a nice working Philips 20Mhz scope with probes that's yours free, old friend.
I was given it, to release into the wild per my incentive on the other place: it's wholly yours if you'd like.
ATB.
I've been watching Youtube tutorials on how to use an oscilloscope, and I'm slowly learning (emphasis on slowly). If anyone wants to give me an idiots guide to amplifier repair, I'd be very grateful. Even if it's just a very basic guide, because I really wouldn't know where to start.
The fact that I have one working channel is going to help a lot for comparisons, but I'd like to know what would be the most efficient way to fault find.
Any hints or tips would be very much appreciated.
The fact that I have one working channel is going to help a lot for comparisons, but I'd like to know what would be the most efficient way to fault find.
Any hints or tips would be very much appreciated.
The best primer on using such a thing remains Tektronix 'XYZs of Oscilloscopes.'
You can download it here as a PDF:
https://www.tek.com/en/documents/primer/oscilloscope-basics
: D
You can download it here as a PDF:
https://www.tek.com/en/documents/primer/oscilloscope-basics
: D
What a kind offer regarding the scope 🙂
Once you have it in front of you it will start to make a whole lot more sense. You will need a decent signal source (sine wave) and that is easy using either MP3/WAV files played on a suitable device or a CDR that you burn yourself. It takes seconds to create your own files with Audacity which we can do here if needed.
I feel very lucky to know Martin.
I'm quite excited at the thought of playing with an oscilloscope. It's like a new toy 🙂
Blueglow Electronics has some great tutorials -
I always keep a copy of Audacity on my pc, but I have never used it to create test tones.
Can I use my laptop to play sound into the amp?
Can I use my laptop to play sound into the amp?
As long as you have an audio output on the laptop (so probably the headphone socket) you can.
I know Martin wouldn't have let me down, but I'm itching to fix this amp, so bought a scope off eBay. My son has been puking all night, and now he's finally asleep, I thought I'd have a quick look at this amp.
Playing a 1000hz sine wave though Audacity into the amp. The peaks of the sine wave are higher on the left, bad channel. There doesn't appear to be any distortion of the sine wave. Thinking about it, I may have to try at a louder level to see what happens. With the volume turned up, the distortion subjectively got a lot worse.
Playing a 1000hz sine wave though Audacity into the amp. The peaks of the sine wave are higher on the left, bad channel. There doesn't appear to be any distortion of the sine wave. Thinking about it, I may have to try at a louder level to see what happens. With the volume turned up, the distortion subjectively got a lot worse.
Last edited:
This is what it looks like with the volume cranked up. I flicked both switches to ground and made sure the horizontal lines were at exactly the same level before switching to AC.
You're doing well 👍 It looks like the amp can not deliver the negative going part of the signal under power and that may relate to the lower part of the output stage, however...
Can you put some numbers into those images.
1/ If you disconnect the load on the faulty channel then does the distortion go?
2/ What amplitude are we looking at here on the bad channel. Make sure the red knobs on the scope are in the 'Cal' position.
3/ What is the load resistor value?
As an example, in the last image and if the scope trace were centred with no signal and if the scope were on 1v/div then it would be showing approximately 3 volts positive swing (which is clean) and clipping at just over -1 volt.
Be sure to account for the probe setting (X1 X10). Use X1 to avoid confusion here.
Can you put some numbers into those images.
1/ If you disconnect the load on the faulty channel then does the distortion go?
2/ What amplitude are we looking at here on the bad channel. Make sure the red knobs on the scope are in the 'Cal' position.
3/ What is the load resistor value?
As an example, in the last image and if the scope trace were centred with no signal and if the scope were on 1v/div then it would be showing approximately 3 volts positive swing (which is clean) and clipping at just over -1 volt.
Be sure to account for the probe setting (X1 X10). Use X1 to avoid confusion here.
(1) Yes the distortion goes if I disconnect the load
(2) Red knobs are in the Cal position.
(3) Each resistor is 8ohm
I've set the probes/scope to 1x 1v/div.
Here's it is with the load:
And with No load below:
(2) Red knobs are in the Cal position.
(3) Each resistor is 8ohm
I've set the probes/scope to 1x 1v/div.
Here's it is with the load:
And with No load below:
An odd fault. I can't just get a handle on this tbh.
Is there any shift in the rail voltage when it is clipping like that? Can you compare the two rails with and without load. For example you have -/+38 volts when its OK (or with no signal) but does that split change as the fault shows, for example becoming say -5/+71. If it does then it points to a DC offset issue but because the Quad has a virtual ground it shows in peculiar ways.
The other angle on this is that it just looks like the output stage can not deliver any current in a negative direction and yet the 306 can deliver up to 1 watt into 8 ohms in Class A and it can do that without T8 and T10 even being fitted as the two 560 ohms provide the Class A current needed. Beyond that point it would clip but yours looks to be clipping at a much lower negative swing.
So lets try and see which scenario it is.
Is there any shift in the rail voltage when it is clipping like that? Can you compare the two rails with and without load. For example you have -/+38 volts when its OK (or with no signal) but does that split change as the fault shows, for example becoming say -5/+71. If it does then it points to a DC offset issue but because the Quad has a virtual ground it shows in peculiar ways.
The other angle on this is that it just looks like the output stage can not deliver any current in a negative direction and yet the 306 can deliver up to 1 watt into 8 ohms in Class A and it can do that without T8 and T10 even being fitted as the two 560 ohms provide the Class A current needed. Beyond that point it would clip but yours looks to be clipping at a much lower negative swing.
So lets try and see which scenario it is.
I think you should replace T8 and/or T10. These are the components that control the negative excursion. Up to a point T7 can do that, and it is, but when you need more current it comes from T10, phase-inverted by T8, so one or both of them is bad, or a surrounding component.
C10 measures 40v on both channels with or without the load.
C11 measures 38v on both channels with or without the load.
EDIT:
Whilst taking the voltage measurements, I kept the oscilloscope connected to make sure the clipping was still showing...
Last edited:
Thanks, in a way that's good as it shows the basic DC conditions must be correct. If the amp was putting out an offset then the load would draw a little DC current (up to the limits of the virtual ground generator) and then skew the rails. So that is not happening.
Has anything happened around T8? Check the print is OK and that you haven't accidently moved or pushed the transistor and cracked/lifted the print around one of the pads. That kind of thing can be hard to spot.
Has anything happened around T8? Check the print is OK and that you haven't accidently moved or pushed the transistor and cracked/lifted the print around one of the pads. That kind of thing can be hard to spot.
I can't see any problems around T8.
Are there any more checks I can do with the scope?
Perhaps I could swap T8 and T10 over to the other channel to see if the problem moves over?
Are there any more checks I can do with the scope?
Perhaps I could swap T8 and T10 over to the other channel to see if the problem moves over?
Yes. Using just one channel of the scope and with no load attached look at the waveform on each end of that coil in the bridge (L3). You should see the same signal at each end, nice and clean like the output.
Now connect the load and confirm the fault is still present and repeat the check. Is there a big difference in how the signal looks between each end of the coil? Is there a lot more signal on one end than the other?