They all check out fine Osvalsdo, thank you for your continuing help on this by way.
Hi Osvaldo,
Apologies for the tardiness of my response, I had ordered replacement electrolytic capacitors for the board to try a last ditch re-cap if nothing else worked.
when I short C109 ( which is on the defective channel) or C209 the rail voltages remain the same. I accidentally shorted R119 and R219 as the are in almost the same place on the board and the rail voltages drop to what they should be with respect to the battery ( or primary ground) but the current shoots up. In other words it has normalised the SC operating points of the stage as you predicted but with respect to R119 and R219 rather than C109 and C209.
I have it set to a 2 am limit and it goes all the way up to 2 amps.
Last edited:
Don't worry. After cap replacement, we can continue talking (writting, bah), but very rarely such a defects are caused by caps, except one of them becomes very very leaky or shorted internally.
The short confirms that the output stage is more or less, responding, so the problem comes from the input stage.
The short confirms that the output stage is more or less, responding, so the problem comes from the input stage.
I received the parts and recapped the amp.
It hasn't cured the issue with the excessive bias current on the left channel.
Following the voltages given on the schematic on page 8 of the service manual ( in first post) I found normal voltages on the test points indicated right up until Q108.
In other words the circuit up to and including Q102-104 all read the right voltages and if ?I am reading the schematic right they drive Q108.
Instead of getting the expected values of 1.1, 0.5, -1.1 ( CBE) it reads -1.9,-4,-3.
I thought this must be the faulty component and so swapped it with it's counterpart, Q208.
Now Q208, i.e. the transistor I suspected of being faulty, is behaving perfectly and Q108 shows voltages of around where they should be when stone cold but that climb steadily as soon as the amp is powered up. I noticed that the measured gains of the two transistors were quite different, 148 vs 198, would that account for it or is the fault somewhere still in the left channel amp somewhere?
It hasn't cured the issue with the excessive bias current on the left channel.
Following the voltages given on the schematic on page 8 of the service manual ( in first post) I found normal voltages on the test points indicated right up until Q108.
In other words the circuit up to and including Q102-104 all read the right voltages and if ?I am reading the schematic right they drive Q108.
Instead of getting the expected values of 1.1, 0.5, -1.1 ( CBE) it reads -1.9,-4,-3.
I thought this must be the faulty component and so swapped it with it's counterpart, Q208.
Now Q208, i.e. the transistor I suspected of being faulty, is behaving perfectly and Q108 shows voltages of around where they should be when stone cold but that climb steadily as soon as the amp is powered up. I noticed that the measured gains of the two transistors were quite different, 148 vs 198, would that account for it or is the fault somewhere still in the left channel amp somewhere?
Tubes, transistors, MOSFET's, JFET's and OpAmp have wide tolerances, in fact the humble BC548 has DC gain (hFE´s) from 100 to 900 in ranges. So, a spread in hFE not necessarily is a problem, and any good equipment should be designed to take those variations in the parameters.
So, if changing the transistor doesn't modify the scenario, the next step is to take away one terminal of each resistor and measure carefully them, because with both terminals in circuit, may conduct to errors that possibly shadow the real problem. So, I suggest to measure any diode and resistor with 1 terminal in circuit or directly away from it.
Several years ago, a simple PC AT PSU gave me a headache. The PSU worked perfectly under no load, but a small load made it to that the 5V were only 3.5V. Several hours searching and searching, but the ******* PSU couldn't with me: a 1N4148 measured 0.7V in direct and 0.9 in reverse, in the base of one of the 2SC3039's. Hurraaaa, changing the diode became it again working OK. Say, the PSU was about 10 USD, but it became a challenge. From this I learned to take 1 pin out of circuit and measure doubtful components per se.
So, if changing the transistor doesn't modify the scenario, the next step is to take away one terminal of each resistor and measure carefully them, because with both terminals in circuit, may conduct to errors that possibly shadow the real problem. So, I suggest to measure any diode and resistor with 1 terminal in circuit or directly away from it.
Several years ago, a simple PC AT PSU gave me a headache. The PSU worked perfectly under no load, but a small load made it to that the 5V were only 3.5V. Several hours searching and searching, but the ******* PSU couldn't with me: a 1N4148 measured 0.7V in direct and 0.9 in reverse, in the base of one of the 2SC3039's. Hurraaaa, changing the diode became it again working OK. Say, the PSU was about 10 USD, but it became a challenge. From this I learned to take 1 pin out of circuit and measure doubtful components per se.
‘Morning Osvaldo, Thanks for your post.
Ok, that’s reassuring to know.
I think I have isolated the problem to drive circuit of the output stage, in consideration of what you have said I should investigate the bias diodes you mentioned earlier but out of circuit.
This amp has similarly become a “challenge” for me too, it’s probably worth as much as your power supply, but it is fun trying to work out what is wrong.
Ok, that’s reassuring to know.
I think I have isolated the problem to drive circuit of the output stage, in consideration of what you have said I should investigate the bias diodes you mentioned earlier but out of circuit.
This amp has similarly become a “challenge” for me too, it’s probably worth as much as your power supply, but it is fun trying to work out what is wrong.
All diodes in the channel test as good. they were tested out of circuit.
I'm scratching my head here, it seems that the bias drifts up as soon as power is applied and increases steadily with time.
I think this has to be linked in some way to the potential difference between the primary and secondary side but can't see how.
I'm scratching my head here, it seems that the bias drifts up as soon as power is applied and increases steadily with time.
I think this has to be linked in some way to the potential difference between the primary and secondary side but can't see how.
Last edited:
Checked R110,R111 and R114,R115 out of circuit and they tested good.
A strange thing has happened now that I have unsoldered and resoldered these components, the voltages on Q108 are much closer to what they should be.
Could it have simply been dry solder joints? I think I will over the channel and refresh the solder with the solder sucker and fresh solder.
A strange thing has happened now that I have unsoldered and resoldered these components, the voltages on Q108 are much closer to what they should be.
Could it have simply been dry solder joints? I think I will over the channel and refresh the solder with the solder sucker and fresh solder.
Ok, I've worked over the whole board and checked for solder bridges.
It still has the same issue.
I have noticed that pin 4 on the 494 is at 0.8v whereas the service manual says it should be at zero, I checked the diodes and resistors ( by lifting one leg from the board) but found nothing out of spec.
any ideas?
It still has the same issue.
I have noticed that pin 4 on the 494 is at 0.8v whereas the service manual says it should be at zero, I checked the diodes and resistors ( by lifting one leg from the board) but found nothing out of spec.
any ideas?
Pin 4 is the input to the comparator from the error amp, so 0V is not possible to have, or no voltage output from it. Besides, it must affect both channels, not only one. If the output voltage from the PSU is correct, the the TL and MOSFET's are innocent.
Rarely no other participant in this thread ?
Rarely no other participant in this thread ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.