Amplifier Distortion, DC-Offset, and You!
- Thread starter EchoWars
- Start date
sp58
Active Member
Nakamichi Receiver 1, Nakamichi Receiver 2
Before reading this thread: From an auction i got this Receiver 1 - now 26 years old - in good condition, this will be my everyday player - so the main things will be fixed: speaker relay, capacitors and power output transistors - original Sanken.
A very nice player this Nakamichi, especially on FM-Radio. DC-Offset not nice +56mV right, +45mV left. Found out, when lifting the bridges from pre to end-amp the offset reduces !
Found out that most of amps have a Cap -coupling between pre and end. Nakamichi Receiver 1+ 2 dont have this -means DC coupled. Other Nakamichi Receivers build in this Years have Cap - coupling.
So what - i added 10uF/25V orange LowNoise Elco-Caps briged with 0,1uF Micas to the endstage inputs.
Now the Offset-reading shows 35 right 22 left. Good, not good enough.
After reading this thread: Checked all transistors of the diff-amp ( 2sc2240 bl ( Balanced ) 3pieces per unit - one is a constant current source i think. All o.k. but thermical sensitive.
The right channel transistors got a circle change and been changed over to one position. Now the values are 28 right 22 left.
There is also an issue with the rail-voltage for the endstages, not symmetric, it shows +53V vs - 51V.
Changed the main brigde-rectifier from KBU6D to new KBU8D. The values: +54V vs -54V really symmetric !
What about the offset ?
Decreased to 22 right 15 left.
( If not otherwise noticed, all values are in millivolts )
Now a very, very nice and musical player this Nakamichi.
Greetings from Germany
Juergen
sp58
Active Member
New Article, not an answer
Nakamichi Receiver 1, Nakamichi Receiver 2
Before reading this thread: From an auction i got this Receiver 1 - now 26 years old - in good condition, this will be my everyday player - so the main things will be fixed: speaker relay, capacitors and power output transistors - original Sanken.
A very nice player this Nakamichi, especially on FM-Radio. DC-Offset not nice +56mV right, +45mV left. Found out, when lifting the bridges from pre to end-amp the offset reduces !
Found out that most of amps have a Cap -coupling between pre and end. Nakamichi Receiver 1+ 2 dont have this -means DC coupled. Other Nakamichi Receivers build in this Years have Cap - coupling.
So what - i added 10uF/25V orange LowNoise Elco-Caps briged with 0,1uF Micas to the endstage inputs.
Now the Offset-reading shows 35 right 22 left. Good, not good enough.
After reading this thread: Checked all transistors of the diff-amp ( 2sc2240 bl ( Balanced ) 3pieces per unit - one is a constant current source i think. All o.k. but thermical sensitive.
The right channel transistors got a circle change and been changed over to one position. Now the values are 28 right 22 left.
There is also an issue with the rail-voltage for the endstages, not symmetric, it shows +53V vs - 51V.
Changed the main brigde-rectifier from KBU6D to new KBU8D. The values: +54V vs -54V really symmetric !
What about the offset ?
Decreased to 22 right 15 left.
( If not otherwise noticed, all values are in millivolts )
Now a very, very nice and musical player this Nakamichi.
Greetings from Germany
Juergen
Nakamichi Receiver 1, Nakamichi Receiver 2
Before reading this thread: From an auction i got this Receiver 1 - now 26 years old - in good condition, this will be my everyday player - so the main things will be fixed: speaker relay, capacitors and power output transistors - original Sanken.
A very nice player this Nakamichi, especially on FM-Radio. DC-Offset not nice +56mV right, +45mV left. Found out, when lifting the bridges from pre to end-amp the offset reduces !
Found out that most of amps have a Cap -coupling between pre and end. Nakamichi Receiver 1+ 2 dont have this -means DC coupled. Other Nakamichi Receivers build in this Years have Cap - coupling.
So what - i added 10uF/25V orange LowNoise Elco-Caps briged with 0,1uF Micas to the endstage inputs.
Now the Offset-reading shows 35 right 22 left. Good, not good enough.
After reading this thread: Checked all transistors of the diff-amp ( 2sc2240 bl ( Balanced ) 3pieces per unit - one is a constant current source i think. All o.k. but thermical sensitive.
The right channel transistors got a circle change and been changed over to one position. Now the values are 28 right 22 left.
There is also an issue with the rail-voltage for the endstages, not symmetric, it shows +53V vs - 51V.
Changed the main brigde-rectifier from KBU6D to new KBU8D. The values: +54V vs -54V really symmetric !
What about the offset ?
Decreased to 22 right 15 left.
( If not otherwise noticed, all values are in millivolts )
Now a very, very nice and musical player this Nakamichi.
Greetings from Germany
Juergen
50mV is about the absolute maximum for a unit that been in operation untouched for 10 or 15 years. Thats what the OP is talking about. When adjusting my amps I hold to within 5mV. If I start seeing double digits offsets I know something isn't right. I don't think anybody adjusting an amp would call the job done with 50mV on the posts. That would just be sloppy, lazy work.
One of my better sounding amps may hold adjusted 50mV drift range after warmup, a MOSFET amp without dc servo offset correction.
I wish it could be narrower, but it is what it is.
My F5 build pretty reliably holds inside +/- 5mV once it's warm. I suppose you might see 10mV drift using multiple MOSFETs, but if I saw 50mV I'd probably up the source resistors a little bit. Going up .1 ohm would probably tame that.
I'm at 0.22 per on them for 8 per channel. The drop varies a fair amount between, so I assume the devices aren't matched. Possibly intentional for overlap or somesuch, but that would seem to stress some more than others.
Rinascita
Active Member
This is such a fun thread. This was my most recent DC offset related adventure. Even though I am a tube person, I have been playing with mosfets a lot lately.
Recently acquired.
B&K EX-442
Toshiba mosfets, but non Sonata version.
Initial readings
18.5mv right Channel. 15mv left channel.
After a little love, got both channels down to 12mv. Ran out of adjustment on the right channel (doh).
Good enough for now. I will do more work on this amp at a later time.
Recently acquired.
B&K EX-442
Toshiba mosfets, but non Sonata version.
Initial readings
18.5mv right Channel. 15mv left channel.
After a little love, got both channels down to 12mv. Ran out of adjustment on the right channel (doh).
Good enough for now. I will do more work on this amp at a later time.
8 .22's parallel? That's practically nothing! But if it works you're definitely getting the most out of it! I've thought about dropping my sources a little, but the thing already runs hot as a toaster. Not sure I need to introduce more volatility to the mix.
My present am
Probably should start a new thread in solid state on this interesting diversion for further discussion.
Last edited:
Fuzzbuster
Addicted Member
After watching this thread with much interest I checked my beloved Onkyo TX-2500Mkll:
Left channel 1.2mv, right channel 8.3mv. Is 7mv difference enough to be concerned with?
Left channel 1.2mv, right channel 8.3mv. Is 7mv difference enough to be concerned with?
petehall347
the brandy coffee man
you are good to go ..
RJolly87
Well-Known Member
It's not so much a comparison of the channels with each other, it's more a concern of the actual value.
For example, my Pioneer SX-780 is of concern. I don't recall the exact figures, but I believe it was something to the effect of 140mv on one side and 158mv on the other. The difference of 18mv in this case is not a concern. The problem is that both channels are quite high, and in need of attention.
A couple more for the list.
Both allowed about 30 minutes to warm-up
Technics SU-V460
R -2.5mv
L +1.9mv
Realistic STA-25. This amp has been in storage and today is the first time it's been powered-up in over 20 years. No smoke - yet
Both channels drifting and checked with 2 multi-meters.
R +14mv to +16mv
L +.05 to -2mv
Both allowed about 30 minutes to warm-up
Technics SU-V460
R -2.5mv
L +1.9mv
Realistic STA-25. This amp has been in storage and today is the first time it's been powered-up in over 20 years. No smoke - yet
Both channels drifting and checked with 2 multi-meters.
R +14mv to +16mv
L +.05 to -2mv
RocknRum
Active Member
I don't disagree. A person should adjust it as low as possible seeing as how easy the procedure is. The first Marantz receiver I bought (a 2250b) years ago, has a dc offset reading of less than one millivolt on both channels. I check it every 2 or three years and it may climb to 1 or 2 millivolts at most. One of the reasons I kept it and sold the rest of the Marantz receivers.
Fuzzbuster
Addicted Member
Ok. So I just brought home a Realistic STA-220...what a beast. After letting it warm up for 10 minutes or so the readings are at around 0.3mv but they don't stay steady. They bounce around between there and -0.2mv. What does that indicate, if anything?
