Need assistance with diagnosis please. My trusty M120-A power amp has developed an annoying intermittent problem. Very occasionally, I get some crackling/popping out of right channel only. Further tests show right channel bias not responding to adjustments, so I suspect a faulty transistor in the power stage, but need to diagnose further. More info on problem:
-can run for days or weeks without any symptoms
-usually occurs when been switched off for several days
-right channel only
-does not vary with volume control level
When I first noticed this problem i replaced every electolytic capacity - did not solve the issue.
Next step was to re-solder all connections on main board, expecting a dry joint - still not fixed.
I checked some adjustments today and I think I'm getting closer.
Following the service manual procedure, I started with DC offset and Bias current adjustments.
-DC offset was OK, minor tweak needed
-Bias adjustment on left channel was a tad low, increased it from 65mA to 70mA by adjusting VR201a
-Bias adjustment on right channel -VR201B had no effect on bias current at all, which sits around 50mA. I suspected faulty Q201b (bias adjustment transistor), so replaced this, no change. Checked D201B (on of those triple heat coupled diodes) checks OK at 1.9V forward voltage. Should I consider replacing this?
Before I go randomly buying/replacing power transistors, any advise on how to pinpoint the problem would be much appreciated. Further test points? Correct reference voltages? Other?
Cheers
Tony
-can run for days or weeks without any symptoms
-usually occurs when been switched off for several days
-right channel only
-does not vary with volume control level
When I first noticed this problem i replaced every electolytic capacity - did not solve the issue.
Next step was to re-solder all connections on main board, expecting a dry joint - still not fixed.
I checked some adjustments today and I think I'm getting closer.
Following the service manual procedure, I started with DC offset and Bias current adjustments.
-DC offset was OK, minor tweak needed
-Bias adjustment on left channel was a tad low, increased it from 65mA to 70mA by adjusting VR201a
-Bias adjustment on right channel -VR201B had no effect on bias current at all, which sits around 50mA. I suspected faulty Q201b (bias adjustment transistor), so replaced this, no change. Checked D201B (on of those triple heat coupled diodes) checks OK at 1.9V forward voltage. Should I consider replacing this?
Before I go randomly buying/replacing power transistors, any advise on how to pinpoint the problem would be much appreciated. Further test points? Correct reference voltages? Other?
Cheers
Tony
Luxman amplifier of that era suffer from soldering issues and dried capacitors ( all of them )
You should also look for an operating trimmer and spraying it will do nothing For this model of trimmer the problem lays on the oxidation of the rotor which was silver plated once and now obviously is black .
I don't expect anything else
Kind regards
Sakis
You should also look for an operating trimmer and spraying it will do nothing For this model of trimmer the problem lays on the oxidation of the rotor which was silver plated once and now obviously is black .
I don't expect anything else
Kind regards
Sakis
Hi Tony,
I don't think you have an output transistor problem, and do not attempt to replace that varistor (diode x3). It is possible that the bias transistor has gone leaky, but that does not fix itself. What might is a bad solder connection.
If it wasn't for the fact that you say your bias doesn't respond to adjustment, I would have bet on a bad switch contact or output replay contact. I have the same amplifier too.
Can you confirm that the bias doesn't change with the bias control position? I want to make sure we aren't chasing a bad connection with your meter probes. With the probes still measuring bias current, warm the bias sense diode (D201) while watching the bias current. The bias should drop to lower levels. If it doesn't, can you test Q201 (2SC1815)? Also make sure that the low value resistors in that area are fine, and give the solder connections a good, close look.
To redo a solder joint, it helps to remove the old solder, scrape the lead and apply fresh solder. Make sure the solder wets properly and don't use too much. Your connection should look like the profile of a volcano, not like a marble.
You also have those two wires leading back to the voltage amp stage. Inspect the solder connections and make sure the wire isn't broken (in the hole and near the soldered area).
Let us know how this goes please.
-Chris
I don't think you have an output transistor problem, and do not attempt to replace that varistor (diode x3). It is possible that the bias transistor has gone leaky, but that does not fix itself. What might is a bad solder connection.
If it wasn't for the fact that you say your bias doesn't respond to adjustment, I would have bet on a bad switch contact or output replay contact. I have the same amplifier too.
Can you confirm that the bias doesn't change with the bias control position? I want to make sure we aren't chasing a bad connection with your meter probes. With the probes still measuring bias current, warm the bias sense diode (D201) while watching the bias current. The bias should drop to lower levels. If it doesn't, can you test Q201 (2SC1815)? Also make sure that the low value resistors in that area are fine, and give the solder connections a good, close look.
To redo a solder joint, it helps to remove the old solder, scrape the lead and apply fresh solder. Make sure the solder wets properly and don't use too much. Your connection should look like the profile of a volcano, not like a marble.
You also have those two wires leading back to the voltage amp stage. Inspect the solder connections and make sure the wire isn't broken (in the hole and near the soldered area).
Let us know how this goes please.
-Chris
Hi Sakis,
The leads may be scraped using a small hobby knife (ex-acto in my country), or by using small pliers rotated around a lead (standing vertically!). Either method is effective in removing coatings on the surface of the wire lead - and pad (using the blade) in some cases. Just make sure you don't actually grip the lead is using pliers!
There are times when physical methods do work better than a purely chemical approach. Doing many joints can become extremely tedious, so it helps to identify the actual problem solder joints first. Large components often will not wet properly in the solder bath during manufacture. Running at higher temperatures will also destroy a solder connection, so power resistors and transistors warrant closer inspection.
-Chris
Nope. In this case, scrape = scrape ... as in the oxide off the lead. The solder will not bond to the lead wire unless you are down to bare metal. I also use a liquid solder flux (meant for electronics use). This greatly helps the solder to bond so you can use less solder and have a much better connection.
The leads may be scraped using a small hobby knife (ex-acto in my country), or by using small pliers rotated around a lead (standing vertically!). Either method is effective in removing coatings on the surface of the wire lead - and pad (using the blade) in some cases. Just make sure you don't actually grip the lead is using pliers!
There are times when physical methods do work better than a purely chemical approach. Doing many joints can become extremely tedious, so it helps to identify the actual problem solder joints first. Large components often will not wet properly in the solder bath during manufacture. Running at higher temperatures will also destroy a solder connection, so power resistors and transistors warrant closer inspection.
-Chris
Chris,
Looks like you were right! The connections between the voltage amp and the output stage is vie wire wraps, not solder. I soldered these connections, and all seems fine now! I guess wire wrap connections are subject to very slow oxidation, causing dry joints. I'll flow all wire wraps with solder now that I've found this.
Unfortunately while poking around with my meter probes checking bias, I blew two of the output transistors! Need to replace 2 x 2SA1106 now!
...you live and learn!
Don't buy replacements with that same designation; they will be fake. You'll have to find modern equivalents. You must also replace their partners.
Hi Tony,
Welcome is right about changing the other outputs and also that current ones have a high chance of being fakes. I would use the On Semi MJL series of transistors. If they are too large, try the MJW series. They do make an MJL1302A and MJL3281A under license from Toshiba. While you are at it, replace the drivers as well. If they aren't blown, they have been operated well beyond their specifications and may fail early if they still test good. Buy from the usual suspects - Digikey, Mouser or Newark. Allied and Arrow are also really good choices. Make sure you check some of the resistors between the drivers too. They often open and high bias current that is adjustable is the result.
The On Semi parts are excellent in every way, don't consider them as lower quality than the originals.
-Chris
Welcome is right about changing the other outputs and also that current ones have a high chance of being fakes. I would use the On Semi MJL series of transistors. If they are too large, try the MJW series. They do make an MJL1302A and MJL3281A under license from Toshiba. While you are at it, replace the drivers as well. If they aren't blown, they have been operated well beyond their specifications and may fail early if they still test good. Buy from the usual suspects - Digikey, Mouser or Newark. Allied and Arrow are also really good choices. Make sure you check some of the resistors between the drivers too. They often open and high bias current that is adjustable is the result.
The On Semi parts are excellent in every way, don't consider them as lower quality than the originals.
-Chris
I finally fixed the Luxman!
It was a failed SA968 driver transistor, not the 2SA1106 pair, although I replaced the 1106 pair anyway with new items, (possibly fake) units. I made sure I matched the two drivers - of the 10 replacements I ordered, the Beta values varied greatly.
A few other points/warnings about the Luxman M120a amp, in case anybody else is looking to work on one of these.
Fuse numbers and supply rails:
The published circuit diagrams, and repair manual, have an error with the fuse numbers.
Rail Published Correct
Fuse No. Fuse No.
R Channel V+ 801 802
R Channel V- 802 801
Setting Bias Current
If you follow the instructions in the Luxman repair manual for bias adjustments, you will get it wrong! R Channel bias will be too low. Here’s why: Fuse 802 supplies the Right Channel power amp section, and also supplies the entire L+R pre-power stage. In comparison, Fuse 805 powers only the Left channel power stage. This adds about 30ma to the R Channel bias measurement! I adjusted mine by de-soldering the R Channel power supply wire, and therefore eliminating the pre-stage current from my measurement.
The previous adjustment of my amp was obviously not done correctly – I noticed the R Channel heatsinks were always much cooler than the left, due to the R channel being under-biassed.
Anyway, all good for now, thanks to all for assistance.
It was a failed SA968 driver transistor, not the 2SA1106 pair, although I replaced the 1106 pair anyway with new items, (possibly fake) units. I made sure I matched the two drivers - of the 10 replacements I ordered, the Beta values varied greatly.
A few other points/warnings about the Luxman M120a amp, in case anybody else is looking to work on one of these.
Fuse numbers and supply rails:
The published circuit diagrams, and repair manual, have an error with the fuse numbers.
Rail Published Correct
Fuse No. Fuse No.
R Channel V+ 801 802
R Channel V- 802 801
Setting Bias Current
If you follow the instructions in the Luxman repair manual for bias adjustments, you will get it wrong! R Channel bias will be too low. Here’s why: Fuse 802 supplies the Right Channel power amp section, and also supplies the entire L+R pre-power stage. In comparison, Fuse 805 powers only the Left channel power stage. This adds about 30ma to the R Channel bias measurement! I adjusted mine by de-soldering the R Channel power supply wire, and therefore eliminating the pre-stage current from my measurement.
The previous adjustment of my amp was obviously not done correctly – I noticed the R Channel heatsinks were always much cooler than the left, due to the R channel being under-biassed.
Anyway, all good for now, thanks to all for assistance.
Last message got a bit corrupted.
Correct fuse number for R Channle V+ is Fuse 802, not Fuse 801 as documented in the Luxman circuit diagram and service manual.
Correct fuse number for R Channle V+ is Fuse 802, not Fuse 801 as documented in the Luxman circuit diagram and service manual.
Hi Tony,
That's great news. I'm not terribly surprised that you found the diff pair bad. It's not always that, but when you have front end problems, that's the place to look.
Errors in service manuals is pretty common. I always measure bias current across the emitter resistors. While I'm at it, I check for current sharing between outputs on the same channel. I started doing it this way early in my career. I hate surprises.
-Chris
That's great news. I'm not terribly surprised that you found the diff pair bad. It's not always that, but when you have front end problems, that's the place to look.
Errors in service manuals is pretty common. I always measure bias current across the emitter resistors. While I'm at it, I check for current sharing between outputs on the same channel. I started doing it this way early in my career. I hate surprises.
-Chris
I have found that if an amp is making a steady static sound, look for hot running little guys in the early stages. Several times I have found this to be the noisemaker, usually
the second stage in..
the second stage in..
Checking the bias on all the outputs is a good idea, if one is biasing much higher, it will max out and distort before the others, being a "squeaky wheel" in the bunch.. it would likely be the first to let go and blow the entire stage too..
Had this on a Denon POA 6600.
Had this on a Denon POA 6600.
Hi LiquidMids,
-Chris
No, what will happen is that if one transistor is conducting more than the others in the group, it will probably overheat and fail if the outputs are being run hard. It would clip at exactly the same point as the others. At higher current levels, the emitter resistors swamp out most of the differences between the outputs and the transistors will share current more evenly. At low currents the one conducting more current will run hotter. By the time everything is too hot, it will still be hotter than the rest, and that's why it would fail first. That's barring any defects in the other transistors.
Noise normally points to any component in the voltage amplifier stages, not just the VAS transistor(s). Defects in the VAS will often cause DC offset as well. The input transistors can cause this also, so you really have to look at everything. Even the coupling capacitor to the amplifier input.
-Chris
- Status
- Not open for further replies.