Old semiconductors as a general rule have no problems. Exceptions are known problem parts such as some Hitachi devices that suffer from 'lead rot' where the legs take on a purple/black/tarnished look that spreads up and into the actual die. Typically these are flatpak T0126 style parts.
Trimmers have a very wide tolerance and -/+20% would not be out of the way, particularly cooking grade carbon types.
If you were to replace a preset like yours that reads 806 ohms with a new one that read say 950 or 1100 ohms you might get enough range to set the bias correctly but the original part was not faulty, its just the replacement allows a slightly wider swing.
The resistor we need to know the value of is the one in series with the preset.
Trimmers have a very wide tolerance and -/+20% would not be out of the way, particularly cooking grade carbon types.
If you were to replace a preset like yours that reads 806 ohms with a new one that read say 950 or 1100 ohms you might get enough range to set the bias correctly but the original part was not faulty, its just the replacement allows a slightly wider swing.
The resistor we need to know the value of is the one in series with the preset.
R25 is a 1.3K CC resistor. L channel measured 1.394K, R channel measured 1.404K This is the one in series with the trimmer.
R23 has a 2.2K (2.095K measured) installed in the L amp, R amp had a 2K (2.1K measured) installed by the factory. I don't know whether that was a post assembly change during testing, or whoever assembled the amp grabbed from the wrong bin of parts, or...
The trimmer overall value of 0.806K doesn't have me concerned, it's the tracking from fully closed. There's not much resistance change the first 1/4-1/3 of wiper movement. It's not until the wiper is roughly half way through the turn that it appears to start tracking reasonably.
R23 has a 2.2K (2.095K measured) installed in the L amp, R amp had a 2K (2.1K measured) installed by the factory. I don't know whether that was a post assembly change during testing, or whoever assembled the amp grabbed from the wrong bin of parts, or...
The trimmer overall value of 0.806K doesn't have me concerned, it's the tracking from fully closed. There's not much resistance change the first 1/4-1/3 of wiper movement. It's not until the wiper is roughly half way through the turn that it appears to start tracking reasonably.
I wouldn't worry over the trimmer and how it seems to track tbh. You are turning to min resistance and still not seeing quite enough voltage developed across the bias generator to turn the output stage on.
1.3k is an odd value to be used but no matter 🙂
All you need to do at this point is incrementally reduce R23 until you can achieve the desired bias current, ideally with the preset at around 50 to 80% turned up.
Quick simulation of the bias generator duplicated four times.
On the left is what you have now. There is not enough voltage developed.
Next along has the trimmer at 400 ohms (so lowering bias) and the voltage is lower still.
Now we reduce the 2k2 to 1k (its not enough) so now we go a bit lower still to 680 ohm and you are getting in the right zone. The trimmer should adjust nicely now.
All those values are just examples, you must lower the value incrementally to get it in the right zone.
1.3k is an odd value to be used but no matter 🙂
All you need to do at this point is incrementally reduce R23 until you can achieve the desired bias current, ideally with the preset at around 50 to 80% turned up.
Quick simulation of the bias generator duplicated four times.
On the left is what you have now. There is not enough voltage developed.
Next along has the trimmer at 400 ohms (so lowering bias) and the voltage is lower still.
Now we reduce the 2k2 to 1k (its not enough) so now we go a bit lower still to 680 ohm and you are getting in the right zone. The trimmer should adjust nicely now.
All those values are just examples, you must lower the value incrementally to get it in the right zone.
Not sure if it matters for what you simulated, but the resistor values are reversed as compared to the circuit.
Not sure I'm understanding you. I get what you're saying, but as described earlier, before I ever had a way to measure the bias, the amp was working and sounding pretty good. So, the output stage had to have been turned on correct? With music playing and sounding pretty good to the ear, the L channel bias was <2mV (trimmer was fully closed), R channel had ~12mV with the trimmer fully closed when I took these first measurements.
The amp (99% of amps) sound the same with zero bias compared to normal bias. You would be hard pushed to tell any difference without very critical listening. Even then thanks to the action of negative feedback and the very high loop gain of amps like this the difference may not be noticeable.
Did not know that. I guess that once I get the bias responding properly to trimmer movement, I'll set it fairly low. It was suggested to set between 20-40mA of bias current, which, with the emitter resistors being 0.22 ohms, would equate to 8.8-17.6mV across the resistors. The Kenwood Basic M1 I have (also rated at 105W and with 0.22ohm emitter resistors) suggests setting the bias at 8-20mV, so I'll probably shoot for 10mV.
I know its hard to believe but it true 🙂
This is representative. Normal 100ma bias current vs absolutely zero bias current.
Normal bias generator and generator removed and replaced with wire link.
And the results. You would find it very difficult to tell the difference just by a causal listen. It might show up more listening at very very low levels and with very low impedance loads:
This is representative. Normal 100ma bias current vs absolutely zero bias current.
Normal bias generator and generator removed and replaced with wire link.
And the results. You would find it very difficult to tell the difference just by a causal listen. It might show up more listening at very very low levels and with very low impedance loads:
A little update here:
I had a cermet 1K trimmer (much smaller form factor than the OEM) which tracked better along the range of travel, so inserted on the trace side of the board for easy access. Bias adjusted, but again, not til the last half of travel. So today, I pulled the series 1.3K resistor out of the circuit and installed a 1K MF. Min bias wasn't 0 any longer (1mV) however, I was able to get the bias to rise more quickly and to a higher value than before. However, I ended up using a 1.1K Dale RN60 in the end game. Pulled the board apart again to get the trimmer and series resistor on the component side of the board so that fully CCW on both channels equated to minimum bias. I wanted to make sure that both channels adjusted in the same manner (clockwise) vs one channel CW the other CCW. Also had to make a standoff out of two pieces of socket material epoxied together so that the trimmer would sit far enough above the board to be adjustable.
I hope to get the R channel bolted back into the amp, check and set bias and DC offset, and then do some listening tests. I still need to get the correct amp rated (6A vs currently installed 2A) fuse. Unfortunately, I have them either above or well below the 6A rating.
I had a cermet 1K trimmer (much smaller form factor than the OEM) which tracked better along the range of travel, so inserted on the trace side of the board for easy access. Bias adjusted, but again, not til the last half of travel. So today, I pulled the series 1.3K resistor out of the circuit and installed a 1K MF. Min bias wasn't 0 any longer (1mV) however, I was able to get the bias to rise more quickly and to a higher value than before. However, I ended up using a 1.1K Dale RN60 in the end game. Pulled the board apart again to get the trimmer and series resistor on the component side of the board so that fully CCW on both channels equated to minimum bias. I wanted to make sure that both channels adjusted in the same manner (clockwise) vs one channel CW the other CCW. Also had to make a standoff out of two pieces of socket material epoxied together so that the trimmer would sit far enough above the board to be adjustable.
I hope to get the R channel bolted back into the amp, check and set bias and DC offset, and then do some listening tests. I still need to get the correct amp rated (6A vs currently installed 2A) fuse. Unfortunately, I have them either above or well below the 6A rating.
Both channels hooked up and screwed to the frame. Both channels biased very nicely to 8.5mV across the emitters, and then set dc offset to 0.0mV. This was all through the DBT, which as stated before, provides little voltage drop. I'll still check and reset both on line power and then do some listening tests.
During the R amp rebuild I did replace the 1.3K series CC resistor with MF, and now the L amp has MF in that location. The 2K and 2.2K R23s in each channel are still the original CCs. I MAY replace those with MF due to my dislike of CC and how they (typically) drift high with age. However, I now feel that most of the biasing issue was with the trimmer. Between poor tracking, some dead spots, etc. The above recorded voltage measurements were rock solid, not bouncing around.
During the R amp rebuild I did replace the 1.3K series CC resistor with MF, and now the L amp has MF in that location. The 2K and 2.2K R23s in each channel are still the original CCs. I MAY replace those with MF due to my dislike of CC and how they (typically) drift high with age. However, I now feel that most of the biasing issue was with the trimmer. Between poor tracking, some dead spots, etc. The above recorded voltage measurements were rock solid, not bouncing around.
Thanks for the update, I'm pleased it all seems OK 🙂 Good to see the bias is stable as well.
CC (carbon composition) aren't great but those things were obsolete many decades ago. CF (carbon film) are generally very stable.
CC (carbon composition) aren't great but those things were obsolete many decades ago. CF (carbon film) are generally very stable.
Yep, I normally replace all CC resistors when doing a restore UNLESS there's a compelling reason not to (like for high energy pulses) and even then, I think MOX resistors are an acceptable substitute. But, I was attempting to keep the amp as original as I could for this rebuild.
Metal oxides tend to be used in high dissipation applications and they are pretty rugged but I've never considered them particularly tight tolerance or stable. Often came across them in service of TV's and SMPS.
The dissipation of the resistors in your bias circuit is microscopically small and worst case the 1.3k would only see 0.28 milliwatts dissipation. A 1/8 watt carbon film would be just fine for your 1.3k
The dissipation of the resistors in your bias circuit is microscopically small and worst case the 1.3k would only see 0.28 milliwatts dissipation. A 1/8 watt carbon film would be just fine for your 1.3k
I have a large stock of the Dale resistors from an estate sale of an electronics tech, so I use them a lot.
Finally got around to doing bias and dc offset on full line power and then to give a listen. I dialed both channels to 9mV across the emitters, offset to 0.2mV or less on each channel, and then over into the test rack for a listen. No sound out of the right channel. All the wiring seems good, did a bunch of tone tracing. I also plugged the L amp molex into the R side and vice versa and sound came out of the R speaker, none out of the L. So, it appears the R amp board still has something amiss. So, I started taking voltage measurements under the DBT.... Thoughts and help in understanding what's going on appreciated.
Edit: When I swapped amp boards sockets, the volume out the R side was diminished vs it being hooked up into its correct socket. Don't know if that means anything. Likewise, when I initially checked yesterday, there was a definite hum/buzz in the L channel that had the sound (ground issue) that I didn't notice when plugged into the R side (though the volume was much lower)
Edit: When I swapped amp boards sockets, the volume out the R side was diminished vs it being hooked up into its correct socket. Don't know if that means anything. Likewise, when I initially checked yesterday, there was a definite hum/buzz in the L channel that had the sound (ground issue) that I didn't notice when plugged into the R side (though the volume was much lower)
Last edited:
With the highlighted voltages, does this indicate that Q11 on the R channel isn't turning on? Or could it be bad (pain to take it apart yet again, so don't want to if I don't have to). Seems that ~ 3mV difference is miniscule, but perhaps not.
The readings between the channels look very close and Q11 voltage looks perfect. The emitter of Q1 goes to the main output line and should be close to zero as the reading is a direct measure of any offset present. So-1.6 and -.5 millivolts is good.
The DC conditions all look OK to me, zero volts offset and a small bias current flowing.
The DC conditions all look OK to me, zero volts offset and a small bias current flowing.
Thanks. Perhaps there's something broken in one of the protection circuits....Voltage Limit, Current Limit, Overtemp.
Question: if the bulb(s) burned out in one of those indicators when the R channel went out (leading to an open circuit), could that prevent the right channel from producing sound? Time to check continuity in those indicators....
EDIT: Well, all the indicator (including the Epicure badge) lights show continuity EXCEPT the Left Current Limit, yet it's the R channel without sound.
Question: if the bulb(s) burned out in one of those indicators when the R channel went out (leading to an open circuit), could that prevent the right channel from producing sound? Time to check continuity in those indicators....
EDIT: Well, all the indicator (including the Epicure badge) lights show continuity EXCEPT the Left Current Limit, yet it's the R channel without sound.
Last edited:
You need to go back to first principles with the fault finding. If you are happy the DC conditions are OK, and those voltage readings suggest they are, then you must check there is continuity from the output line to the speaker terminals (any relays correctly engaged and any switches OK) and then you are dwon to injecting a known signal and tracing it from the input signals.
DC conditions that are correct usually mean the basic amp is OK.
DC conditions that are correct usually mean the basic amp is OK.