I can see you have gone to a lot of trouble with the readings...
The 2N5320 is short as are both 40636's in the duff channel.
----------------------------------------------------------------------------------------------------------
All (of this sort anyway) silicon transistors read similar. It doesn't matter whether it's a power device or a tiny small signal one, the readings are always in the same ballpark give or take.
Taking the BF258 which we know is good should read from B to E and B to C with the red lead on the base. Are you sure your identifying the connections OK and also old devices can have oxidised leads that hinder good contact with a probe.
The 2N5320 is short as are both 40636's in the duff channel.
----------------------------------------------------------------------------------------------------------
All (of this sort anyway) silicon transistors read similar. It doesn't matter whether it's a power device or a tiny small signal one, the readings are always in the same ballpark give or take.
Taking the BF258 which we know is good should read from B to E and B to C with the red lead on the base. Are you sure your identifying the connections OK and also old devices can have oxidised leads that hinder good contact with a probe.
Well, I'm struggling to understand the readings my multimeter is giving me... but that will be down to me not understanding the electronics behind whats actually going on, I can read up on that in time. Right now I'm going to right ahead and change the components I've highlighted as damaged. You were right about the BF258 Mooly, I hadn't copied the results right from paper to computer, they are both reading out as you expected.
Well, after weeks of waiting for some free time...
I refitted all of the components, checked and double checked them to make sure they were in the right places and soldered up correctly. I tested the power transistors to make sure they weren't shorting to the chassis and from what I can gather they don't appear to be any different from the working right channel. Then I inserted the 100w tester bulb in series with the amplifier to try firing it up.
I first switched the amp on with only the right channel fuse in place. The bulb lit up initially then went dark. I tested with a signal and the right channel is still working fine with the replacement components that I changed. Anything I had to replace on the left channel I changed on the right channel as well.
I then switched the amp on with both fuses in, the bulb glowed dimly for a couple of seconds then begun to increase in brightness, simultaneously a wire wound resistor started to smoke. I turned the amp off immediately.
The resistor appears to be this one:
Both the fuses on the amplifier are still intact, I guess the bulb saved them.My guess is that the channel is shorting somewhere, any ideas welcome?!?![Smile :) :)](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
I refitted all of the components, checked and double checked them to make sure they were in the right places and soldered up correctly. I tested the power transistors to make sure they weren't shorting to the chassis and from what I can gather they don't appear to be any different from the working right channel. Then I inserted the 100w tester bulb in series with the amplifier to try firing it up.
I first switched the amp on with only the right channel fuse in place. The bulb lit up initially then went dark. I tested with a signal and the right channel is still working fine with the replacement components that I changed. Anything I had to replace on the left channel I changed on the right channel as well.
I then switched the amp on with both fuses in, the bulb glowed dimly for a couple of seconds then begun to increase in brightness, simultaneously a wire wound resistor started to smoke. I turned the amp off immediately.
The resistor appears to be this one:
An externally hosted image should be here but it was not working when we last tested it.
Both the fuses on the amplifier are still intact, I guess the bulb saved them.My guess is that the channel is shorting somewhere, any ideas welcome?!?
Firstly just measure the resistor that burnt and make sure it reads OK or thereabouts.
The fact the bulb started to increase in brightness points to possible bias problem... maybe just requiring adjusting but this will prove it.
On the bad channel apply a shorting link as shown between the bases of the two driver transistors.
Now switch on again... the bulb should remain dim.
The outcome of that determines what we do next.
The fact the bulb started to increase in brightness points to possible bias problem... maybe just requiring adjusting but this will prove it.
On the bad channel apply a shorting link as shown between the bases of the two driver transistors.
Now switch on again... the bulb should remain dim.
The outcome of that determines what we do next.
Attachments
Last edited:
I can't find a mechanical relay anywhere inside the amp, for all I don't really know what it would look like anyway.
I tested the resistor and it reads out at 46.7 so all safe there. I did the modification that you suggested Mooly, Upon power up the bulb didn't even light initially so I checked the voltage rails and I was getting a healthy 79v.
So it would appear that whatever you suggested has made the amplifier more comfortable. I tested the pot adjusting the quiescent current before and it appeared to read out ok, or is that nothing to do with it?
Thankyou!![Smile :) :)](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
I tested the resistor and it reads out at 46.7 so all safe there. I did the modification that you suggested Mooly, Upon power up the bulb didn't even light initially so I checked the voltage rails and I was getting a healthy 79v.
So it would appear that whatever you suggested has made the amplifier more comfortable. I tested the pot adjusting the quiescent current before and it appeared to read out ok, or is that nothing to do with it?
Thankyou!
That sounds promising. Does the amp play OK now ? You may have noticeable distortion at low volume but thats because of the wire link.
If all is OK then switch off and remove the link.
Now turn the bias preset to minimum resistance. In other words so that the preset is a "short" leaving just the two diodes between the transistor bases.
Switch on (with the bulb tester) and make sure the lamp is still dim.
If that is OK then the preset is adjusted as per the manual to give the recommended current.
If you don't know that then I would adjust for around 25ma or so initially and that is best done by just measureing the volt drop across one of the 0.47 ohm resistors on the outputs. Set for around 12 millivolts DC (no audio or speakers).
Keep rechecking as the value will drift with temperature. If all is OK then remove the bulb tester and again recheck the current to make sure it's still around 25ma.
If all is OK then switch off and remove the link.
Now turn the bias preset to minimum resistance. In other words so that the preset is a "short" leaving just the two diodes between the transistor bases.
Switch on (with the bulb tester) and make sure the lamp is still dim.
If that is OK then the preset is adjusted as per the manual to give the recommended current.
If you don't know that then I would adjust for around 25ma or so initially and that is best done by just measureing the volt drop across one of the 0.47 ohm resistors on the outputs. Set for around 12 millivolts DC (no audio or speakers).
Keep rechecking as the value will drift with temperature. If all is OK then remove the bulb tester and again recheck the current to make sure it's still around 25ma.
OK, I'm getting sound out of both channels, (which I must say has had me dancing for joy in itself).
To confirm before I do this, the bias preset = the quiescent current.
And I can test this value by testing the voltage drop across the 0.47 ohm resistor. According to the circuit diagram I test this by measuring the current into the collector of one of the power transistors.
To confirm before I do this, the bias preset = the quiescent current.
And I can test this value by testing the voltage drop across the 0.47 ohm resistor. According to the circuit diagram I test this by measuring the current into the collector of one of the power transistors.
Yes, it is the quiescent current. Minimum resistance of the pot (when it's a short) gives minimum current.
What current does it say to adjust for ?
It is always easier to measure a voltage than a current (which requires breaking the circuit and inserting the meter in series). The collector current flows through both outputs and the two 0.47 ohms.
So calculate the voltage expected by ohms law. V=I * R (units in base values of amps,volts and ohms)
So I said 25ma (as a safe value). Thats V = 0.025 * 0.47 which is 11.75 millivolts. It will never stay put at that... that's normal... aim for approx the correct value when it's hot.
Most meters should have no trouble with millivolts. If it's easier then solder wires to the resistor and connect the meter... then you can watch and adjust without any risk of shorting anything.
Either resistor can be used... it doesn't matter.
I would do the adjustment first with the bulb in place to get a feel for what happens, then repeat the adjustment with full mains applied. Do both channels. Make sure the value isn't to high or creeps up as the amp gets hot.
What current does it say to adjust for ?
It is always easier to measure a voltage than a current (which requires breaking the circuit and inserting the meter in series). The collector current flows through both outputs and the two 0.47 ohms.
So calculate the voltage expected by ohms law. V=I * R (units in base values of amps,volts and ohms)
So I said 25ma (as a safe value). Thats V = 0.025 * 0.47 which is 11.75 millivolts. It will never stay put at that... that's normal... aim for approx the correct value when it's hot.
Most meters should have no trouble with millivolts. If it's easier then solder wires to the resistor and connect the meter... then you can watch and adjust without any risk of shorting anything.
Either resistor can be used... it doesn't matter.
I would do the adjustment first with the bulb in place to get a feel for what happens, then repeat the adjustment with full mains applied. Do both channels. Make sure the value isn't to high or creeps up as the amp gets hot.
The schematic says adjust for 20mA, so you were close enough. I am having trouble with the channel that was damaged however. I have managed to set the Right channel bias to 9.2 on one resistor and 10mV on the other. However the moment I start to turn up the bias on the right channel the bulb lit up, and knowing what happened last time, I switched off the amp straight away. The reading before I started to turn the bias up was 0.0 on one resistor and 1.5mV.
Either way up the sound is great, back to its old self (or rather new self, what with all the new caps), all except for the Left channel being a little louder than the right.
Either way up the sound is great, back to its old self (or rather new self, what with all the new caps), all except for the Left channel being a little louder than the right.
There may not be a problem... as the bulb will begin to light as current is drawn. A filament bulb has a low cold resistance and a high hot resistance.... its very non linear so a small increase in current could well just appear to illuminate the bulb very suddenly. Maybe pull the fuse on the good channel and try again adjusting the other channel.
Without the unit in front of me though, you have to play safe and be sure by basic tests.
While monitering the voltage across the 0.47 ohm does the voltage increase slowly to around the target of 9 mv, or does it suddenly jump ? If the current can be set OK remember to recheck when on full mains voltage. That's very important.
If the voltage jumps up uncontrollably then the pot may be faulty. In fact it's worth putting your meter across it on ohms (amp off) and seeing if it adjusts OK. As its such a low value resistance I would guess it measures OK in circuit.
As to one channel louder than the other... we'll come to that later if you want too.
Without the unit in front of me though, you have to play safe and be sure by basic tests.
While monitering the voltage across the 0.47 ohm does the voltage increase slowly to around the target of 9 mv, or does it suddenly jump ? If the current can be set OK remember to recheck when on full mains voltage. That's very important.
If the voltage jumps up uncontrollably then the pot may be faulty. In fact it's worth putting your meter across it on ohms (amp off) and seeing if it adjusts OK. As its such a low value resistance I would guess it measures OK in circuit.
As to one channel louder than the other... we'll come to that later if you want too.
The voltage did not change on the meter, with the other side of the amplifier unfused the voltage varied around 0 - 1mv. I slowly turned up the bias and nothing happened, the light didn't come on. I pushed it to about where the left channel is at and still no change, so then checked the voltage going to that amp and found that the fuse had blown.
So maybe the pot is damaged but the meter didn't pick up on the spike?
So maybe the pot is damaged but the meter didn't pick up on the spike?
You need to reconfirm the faulty channel is OK by refitting the wire link as in post #48 then start again keeping the bulb in place at all times during testing.
The fact the amp is OK with the link in place suggests a problem with the network of those two series connected diodes and the pot. If you are unsure at all about testing and interpreting the results then I would replace all three components. The diodes are common small signal silicon types and you can use 1N4148 or IN914 available anywhere.
Have you tested the pot on your meter ?
The way the bias circuit works is simple. The two diodes generate a constant volt drop of around 1.4 volts (0.7 each). The pot also generates a variable volt drop that is added to the 1.4 volts of the diodes. This voltage as it is increased by turning the pot starts to turn on the output driver and output transistors.
The fact the amp is OK with the link in place suggests a problem with the network of those two series connected diodes and the pot. If you are unsure at all about testing and interpreting the results then I would replace all three components. The diodes are common small signal silicon types and you can use 1N4148 or IN914 available anywhere.
Have you tested the pot on your meter ?
The way the bias circuit works is simple. The two diodes generate a constant volt drop of around 1.4 volts (0.7 each). The pot also generates a variable volt drop that is added to the 1.4 volts of the diodes. This voltage as it is increased by turning the pot starts to turn on the output driver and output transistors.
I'll test the pot out of circuit now.
What I don't understand is the channel functions fine without the link in place now, sounds great in fact. But I can't adjust the bias without it blowing a fuse and it's set to what must be a very high Amplitude, with the voltage being so low across those resistors.
What I don't understand is the channel functions fine without the link in place now, sounds great in fact. But I can't adjust the bias without it blowing a fuse and it's set to what must be a very high Amplitude, with the voltage being so low across those resistors.
The pot, even if faulty probably does still turn down to "zero" ohms and that is why the amp would run without the link. You have just the two series diodes in circuit and the "shorted" pot.
If there is a crack or burnt area in the carbon track then as soon as that is encountered as the pot is turned, then the bias voltage goes sky high blowing the fuse.
Does that make sense ?
If there is a crack or burnt area in the carbon track then as soon as that is encountered as the pot is turned, then the bias voltage goes sky high blowing the fuse.
Does that make sense ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Solid State
- Overhaul of an Armstrong 625 gone wrong! Please help me :)