Turn the bias down a bit. Plug it into the wall directly, you cannot set bias with varying AC mains voltage.
Once plugged in, set the bias on each channel, put the lid on and leave it watching for excessive heat. The bias should drop a lot. Pull the top and set the bias again, replace the top and let it sit.
You must come up to the right bias from the cold direction on these or they may run away. Patience.
Once plugged in, set the bias on each channel, put the lid on and leave it watching for excessive heat. The bias should drop a lot. Pull the top and set the bias again, replace the top and let it sit.
You must come up to the right bias from the cold direction on these or they may run away. Patience.
Dim bulb tester?
So what did you use for new o/p's and where did you get them?
Change anything else other that the o/p's?
So I looked at the schematic, it uses a bjt bias spreader, Q139. if one channel is working, with no signal, measure and compare the DC bias voltage across the drive, at Q116,Q117 bases, where the arrows are.
If using the dim bulb, it is for a quick "go, no-go" test only.
Use with bias set to minimum as a start, at power on lamp is bright but as ecaps charge, lamp will dim, once the lamp dims, the test is over, go on full AC. If you are worried, let it soak on the DBT but "DO NOT TURN UP THE BIAS"
Use with bias set to minimum as a start, at power on lamp is bright but as ecaps charge, lamp will dim, once the lamp dims, the test is over, go on full AC. If you are worried, let it soak on the DBT but "DO NOT TURN UP THE BIAS"
D
Deleted member 543346
Bias will not set. If i adjust bias on left channel from 1mw to 10mv , and do the same on right channel then the left channel go instantly down to a few mv again.
Has nothing to do with temps. Bias directly interfere with eachother somehow.
Hi RagnarB,
Okay, have you checked for oscillation with an analogue oscilloscope? Digital scopes can lie and not show a waveform if the horizontal time base isn't set right.
Now as I have been trying to point out to you, if the AC mains voltage changes, it will affect bias. Period. What you are describing is classic. So, please confirm the amp is plugged directly into the mains and it is not oscillating. There is zero electrical connection between those channels, one will not electrically affect the other, but mains voltage changes absolutely does affect both.
I need you to confirm before continuing or we are wasting time.
-Chris
Okay, have you checked for oscillation with an analogue oscilloscope? Digital scopes can lie and not show a waveform if the horizontal time base isn't set right.
Now as I have been trying to point out to you, if the AC mains voltage changes, it will affect bias. Period. What you are describing is classic. So, please confirm the amp is plugged directly into the mains and it is not oscillating. There is zero electrical connection between those channels, one will not electrically affect the other, but mains voltage changes absolutely does affect both.
I need you to confirm before continuing or we are wasting time.
-Chris
Okay, hang on here.
If you have the amp at full supply, it should be safe. If it blows outputs you have an additional fault and the amp is probably not at full supply.
Have you checked for oscillation as I described? With an analogue scope? Continue monitoring and kill the power if the current spikes or you see oscillation.
If you are not going to follow advice, I'm afraid I can't help you at all.
If you have the amp at full supply, it should be safe. If it blows outputs you have an additional fault and the amp is probably not at full supply.
Have you checked for oscillation as I described? With an analogue scope? Continue monitoring and kill the power if the current spikes or you see oscillation.
If you are not going to follow advice, I'm afraid I can't help you at all.
D
Deleted member 543346
I understand what you are saying. Please reread what i write about the imbalance when adjusting bias.
Been setting bias on amplifiers before.
D
Deleted member 543346
Hi RagnarB,
All due respect, I have been servicing for nearly 50 years now professionally, and was authorized warranty for Nakamichi and worked on these under warranty and ever since. I will not steer you wrong, but you have got to listen and do things properly. You must confirm the state of your test setup and the condition of the amplifier.
If you decide to ignore advice or fail to confirm things are as we would assume they are, you will run into trouble. So you must tell us everything as it isn't on my bench with my equipment and setup.
So, I am not trying to insult you, but now let's get a grip on the situation. We'll start with you. How much professional experience have you had on audio power amplifiers, and what types? What equipment do you have?
If you can get the amp to full supply on a variac, do so. DO NOT USE A BLOODY DIM BULB TESTER!! Are we clear? Check for oscillation throughout. Sometimes an amplifier won't break into oscillation until full supply is reached, or once a critical level of bias current is reached. So watch for this. You can fuse the mains for a lower current than the amp draws in service, but maybe 50% higher than idle current.
I'll try to help you, but if you do your own thing we are wasting everyone's time.
All due respect, I have been servicing for nearly 50 years now professionally, and was authorized warranty for Nakamichi and worked on these under warranty and ever since. I will not steer you wrong, but you have got to listen and do things properly. You must confirm the state of your test setup and the condition of the amplifier.
If you decide to ignore advice or fail to confirm things are as we would assume they are, you will run into trouble. So you must tell us everything as it isn't on my bench with my equipment and setup.
So, I am not trying to insult you, but now let's get a grip on the situation. We'll start with you. How much professional experience have you had on audio power amplifiers, and what types? What equipment do you have?
If you can get the amp to full supply on a variac, do so. DO NOT USE A BLOODY DIM BULB TESTER!! Are we clear? Check for oscillation throughout. Sometimes an amplifier won't break into oscillation until full supply is reached, or once a critical level of bias current is reached. So watch for this. You can fuse the mains for a lower current than the amp draws in service, but maybe 50% higher than idle current.
I'll try to help you, but if you do your own thing we are wasting everyone's time.
So lets determine exactly what you did,
Did you turned up the bias while on the DBT, left it that way and then went to full AC and smoked your new O/P's?
If so, a real shame.
When I power up a new OPS, I always
1) connect a scope to the o/p, DC coupled, 20V/Div
2) monitor OPS bias with a DMM
3) connect a THD analyzer to supply input signal and monitor the o/p, then use a 8-ohm load to do THD measurements
You can increase bias, a bit, while on the DBT but turn it down to zero when switching over to full AC and then re-set it.
One can even set bias to lower than spec, initially, while monitoring how the bias reacts as the heatsink/devices temperature rises to equilibrium.
What bother me about helping folks, is when they do not answer questions and then withhold pertinant information as to what they have done and then expect the experts to bale them out.
Kind makes me wonder how an OPS with 7 pairs blows in the first place.
Good Luck
P.S The cheapest option is to use the Toshiba 2SA1943N and 2SC500N 150W devices available at Mouser or Dig-Key
Did you turned up the bias while on the DBT, left it that way and then went to full AC and smoked your new O/P's?
If so, a real shame.
When I power up a new OPS, I always
1) connect a scope to the o/p, DC coupled, 20V/Div
2) monitor OPS bias with a DMM
3) connect a THD analyzer to supply input signal and monitor the o/p, then use a 8-ohm load to do THD measurements
You can increase bias, a bit, while on the DBT but turn it down to zero when switching over to full AC and then re-set it.
One can even set bias to lower than spec, initially, while monitoring how the bias reacts as the heatsink/devices temperature rises to equilibrium.
What bother me about helping folks, is when they do not answer questions and then withhold pertinant information as to what they have done and then expect the experts to bale them out.
Kind makes me wonder how an OPS with 7 pairs blows in the first place.
Good Luck
P.S The cheapest option is to use the Toshiba 2SA1943N and 2SC500N 150W devices available at Mouser or Dig-Key
D
Deleted member 543346
The problem with an amplifier that breaks into oscillation ONLY at full voltage is that once the variac gets turned up that far it is too late, and you’re replacing all 2N outputs, both drivers, predrivers and the Vbe multiplier again. Fuse or no fuse. Preventing this on new design was the original function of the dim bulb tester. It later found its way into “servicing departments”. When an amplifier design is fully vetted, you aren’t supposed to need one anymore. If an amp oscillates and explodes at full voltage on a service bench, either something is wrong with the original design or something is wrong with what the tech did to it. In either case, one could argue that “he got whats coming”.
Yeah, the DBT is the only fail safe way to bring up a Phase Linear until everything is “perfect”. One could also argue that there are serious design problems with one. They latch to rails at low voltage and oscillate at high voltage until tweaked in. Most amplifiers don’t have those problems. The Nak should fare MUCH better if it was serviced correctly.
Yeah, the DBT is the only fail safe way to bring up a Phase Linear until everything is “perfect”. One could also argue that there are serious design problems with one. They latch to rails at low voltage and oscillate at high voltage until tweaked in. Most amplifiers don’t have those problems. The Nak should fare MUCH better if it was serviced correctly.
Hi wg_ski,
Yes, Nakamichi amps are very stable - unless someone has done something with them.
I've been fixing Phase Linear amps for years on variacs without trouble. Yes, they are problematic in design for sure, changing output transistor types is a good way to create a radio station.
Hi RagnarB,
I'd like to help you, but I need to know some things so we can direct you the best way. I'm extremely familiar with these amps as I'm sure a few others are. They are extremely reliable, the design is pretty solid. I never had even one fail under warranty that didn't fail for a reason no amp could survive - and that was just one. A copper box staple fell inside and shorted the outputs. As I mentioned before, the staple did not survive the experience either.
Yes, Nakamichi amps are very stable - unless someone has done something with them.
I've been fixing Phase Linear amps for years on variacs without trouble. Yes, they are problematic in design for sure, changing output transistor types is a good way to create a radio station.
Hi RagnarB,
I'd like to help you, but I need to know some things so we can direct you the best way. I'm extremely familiar with these amps as I'm sure a few others are. They are extremely reliable, the design is pretty solid. I never had even one fail under warranty that didn't fail for a reason no amp could survive - and that was just one. A copper box staple fell inside and shorted the outputs. As I mentioned before, the staple did not survive the experience either.
Nobody mentioned here that you can't just buy 7 pairs of transistors and dump em in the amplifier and expect everything will work as it should. They must be matched first for Pete's sake. Some saying nowadays transistors don't need matching but I say they have to be matched (vbe and beta if possible), even if they're from the same batch. This means buying more transistors than you need. Expensive repair if you ask me. Been there, done that many times. Now you can throw sticks and stones at me. Besides, through mistakes you will learn. No pain, no gain ! 
Hello HighTec,
Yes, you are absolutely correct. This is standard servicing procedure.
Match on beta with the devices on the same thermal platform. They must be at identical temperatures or your readings aren't worth anything. vBE isn't a fine enough indicator. I have bought more devices in my time to service one piece. Your match yield was much worse in the 1980's, better than the 1960s and 1970s. Today your yield is far higher with good devices. Thank goodness I had a big shop and we would go through most anyway. I did throw many out, or used them as pass devices in power supplies.
I'm just trying to get him to stop using the dim bulb tester. You can't do that when setting bias currents.
-Chris
Yes, you are absolutely correct. This is standard servicing procedure.
Match on beta with the devices on the same thermal platform. They must be at identical temperatures or your readings aren't worth anything. vBE isn't a fine enough indicator. I have bought more devices in my time to service one piece. Your match yield was much worse in the 1980's, better than the 1960s and 1970s. Today your yield is far higher with good devices. Thank goodness I had a big shop and we would go through most anyway. I did throw many out, or used them as pass devices in power supplies.
I'm just trying to get him to stop using the dim bulb tester. You can't do that when setting bias currents.
-Chris
The Onsemi NJW0281/0302 are among the more well matched types right out of the tube. The chances of buying a whole tube and not getting 7 to match well enough to be used here are pretty slim.
You use the oddballs or ones that fall a little out of the group as drivers or in amps that only use a single pair.
You use the oddballs or ones that fall a little out of the group as drivers or in amps that only use a single pair.