Amplifier Distortion, DC-Offset, and You!

Time for one of those long, boring semi-technical posts that no one here reads...

As a few of you know, I bought a rare Kenwood 700M amplifier a few weeks ago on eBay. It arrived with a weak channel which was taken care of by replacing a bypass cap. Since then, I have gone through the entire amp and replaced all the electrolytic capacitors with the exception of the big power supply caps (not that expensive...maybe $20 in caps).

The previous owner described the amp as 'perfect' sounding, and compared with the big Mac's and Krell's and such. Before buying it, he described his current system which was quite high-end. I assumed that since he owned expensive equipment that he knew what he was listening to.
:withstpd:

Got the amp back together today (parts finally in), and fired it up with my small Dynaudio's (can't use the Heil/Dynaudio's for this, as they are bi-amped). Ummm...it sounded like ass. OK, it didn't always sound like ass, but at low volume levels it was obvious that something was wrong. I had a pretty good idea what was going on, so I grabbed my meter. Sure enough, there was 100mV of DC offset in the left channel, and almost 250mV (!!!) in the right! 250mV is almost enough for the protection circuitry to kick in!! Not good. I pulled the driver boards out and replaced the amp input differential pairs for both channels with new Zetex HG PNP's ($5 in transistors, no biggie). DC offset is now about 12mV in both channels. An input pair being as unbalanced as the Kenwood was when it arrived probably generates 10x the distortion as a properly balanced pair, especially at low volume. If you would like to read more about the benefits of a balanced differential pair, read here.

OK, sit down for another listen. NOW we're cookin'!! Amazing night and day difference. I can honestly say that it is without a doubt the nicest amp I have ever listened to, and there have been more than a few that impressed me. The bottom end on this thing is as clear as spring water, and it has an openess that is difficult to describe. As for power, my dummy loads cannot take the power output of this thing, but I can crank it for 10-15 seconds without destroying them. Left channel - 218W before clip, right channel - 220W, this into 8-ohms. Totally cool!! I just have never heard bass like this...wow...

Bottom line...if you expect to hear great sound, you just might...regardless of the reality. The guy I bought this from was well-meaning, but did not know how to listen subjectively. His new spendy amps could be performing horribly, and his expectations of what he felt he was supposed to hear would rule out anything to the contrary.

Your own subjectivity could be suffering too, so give yourself a reality-check.

As a semi-poll, I'd like to see those on this board whip out their multimeters and take a look at the DC that is being presented to the speakers. This means..

1. Speakers disconnected (or connect the meter to the 'B' speakers and set the front panel speaker control accordingly)
2. Input set to an unusued position (not Phono)
3. Volume control at minimum.
4. Balance in center
5. Tone controls either defeated or set to mid position
6. Set your meter to read DC, and set to a low scale (300mV scale is common) Connect directly to the Pos and Neg of the speaker terminals
7. Give the amp 10 minutes to settle. Report back...I'd like to see how healthy all these old amps are.

If you read:

0 - 15mV: Damn good!! If you read '0V', you may have a capacitor output, or your meter is set wrong

16mV - 50mV: An acceptable value, especially at the lower end of this range. 2nd harmonic distortion is probably twice to four times what manufacturer's spec calls for at higher frequencies. Probably not audible, as the distortion is mostly in the upper octaves. At the upper end of this range I begin to raise an eyebrow. :saywhat:

50 - 85mV: Something is certainly amiss, and while this is not enough to put your speakers or equipment in jeopardy, the amp is running nowhere near where it should. I'd venture to guess that most of the DC-coupled amps that are in use by forum members here fall into this range.

100mV to ?: A high enough voltage will cause the DC protection to kick in. This happens at a level determined by the designer, but is usually equivalent to about a diode drop (600mV)or so. Needless to say, if you are listening to an amp with 100mV or more of DC offset, you have no idea what the amp really is supposed to sound like. Indeed, some amps without a differential input are actually designed to have a bit of DC at the outputs, but this is triple-rare, and I don't think anyone here owns one. (in my book it's piss-poor design, but if you can sell it WTH..)

Soooooo...go grab a meter and tell me what you find...
Hey, I'm sort of late to the "check your DC offset" party but I did bring my multimeter. And I plan to test most of my receivers. Thanks for posting the process. But first I have a few dumb questions if this thread is alive -- and if you have the time to reply:
1. The receiver must be on during all testing, I'm assuming?
2. So if you have both A and B speaker jacks, is there any need to test both A and B -- or does one reading apply to both?
3. Speakers disconnected -- but you can leave speakers connected to "B" while testing "A" with speaker switch set to "A."
4. But what if the receiver only has one set of speaker jacks -- rare, but found on some low-wattage units? Leave speakers disconnected while testing the speaker jacks? Safe to do?
5. "Give the amp 10 minutes to settle." So, turn receiver on and wait 10 minutes before taking a reading? Or take two readings: one immediately and another after 10 minutes?

If you can reply, many thanks. If not, I understand and still appreciate all the info you've already shared.
 
So just bought a Kenwood ka-9100 and here are the readings. the left channel -129mV DC that needs adjusted? I was able to adjust the bias both right and left inside. Why woukd that have drifted so far off?
 

Attachments

  • image1.JPG
    image1.JPG
    36.6 KB · Views: 28
  • image2.JPG
    image2.JPG
    33.7 KB · Views: 26
  • image3.JPG
    image3.JPG
    35.5 KB · Views: 27
Last edited:
Measured my sansui 9090DB last night. Was positively surprised to find "just" 18.5mV (R) and 9.0mV (L). The only thing that's been done to this old (40 yrs!) girl is replacement of bulbs that had gone dark behind the front panel.
 
My old faithful standby "when all else fails" Sansui AU-5900

Left 10mV
Right 20mV

I would normally adjust but don't have the option on this unit.
 
Carver M500t -1.5mvdc right channel 64mvdc left channel.
No issues with sound quality. Inverted style amplification internal wiring.
 
Just checked my "new" Accuphase P-300L. Came to me from HifiDo in Japan, who said that it was checked through by Accuphase last year.
DC offset is 0.3 mV on the left channel, and -0.4 mV on the right channel. I suppose that's quite good :) Love that pretty little monster :)

My Sony TA-3200F measured a bit differently... I just did a full recap on it (it had some weird noise in the left channel). I am now measuring:
9 mV on the right channel
between 0 and 35 mV on the left.
The left channel is constantly fluctuating. Does that mean there is something wrong?
 
Being still new at this hobby, I was beginning to realize my recently acquired Pioneer SX-780 didn't really seem to be living up to its reputation, with one channel fading out at low volume, that seemed to improve as I fiddle with input cabling.

I checked bias voltage as mentioned, and found -130mV in the left channel, and -150mV in the right channel. I didn't let the multi-meter settle for 10 min admittedly, but the amp had been running for at least that long, and readings were stable when I stopped.

Needless to say, the amp is now getting sidelined for the time being.
 
Checked my Marantz 2245.

Left 14mV, Right 95mV.

Sounds much better for the time being, but it's obvious I have my work cut out for me.
 
Appears to be a itnteresting thread. Yes, Just noticing this thread.

Prior to reading through 225 pages. I will be attacking this read later today.

Question to any member who may dare to answer involved as :
What Experience or Relationship between Power Amplifier DC Offset / Distortion to Replacement Parts primarily Transistors used - Generic Partnumbers vs Original Manufacturer Partnumbers ???

IT WOULD BE A WONDERFUL EFFORT FOR SOME MEMBER OR ADMINISTRATOR TO SUMMARY THE HIGHLIGHTED INFO OF THIS THREAD AND POST SOMEWHERE !


cferry1
 
Last edited:
B&K ST2140 - right around 6.7 and 6.5

Just received it yesterday. So far so good, there is a bit of a pop when I turn it on. Sounds great otherwise!
 
Hi All
I have designed hundreds of audio power amplifiers during my career @ many company's like BGW, JBL and others. The DC OFFSET for an audio amplifier should be a max of 10mv. The DC OFFSET caused current to flow into the speaker and this can cause higher speaker distortion and hotter output transistors. If the offset is >10mv fix the problem.
Duke
 
Agree. I higher DC Balance Offset Voltage and higher Idle Current Adjustment will result in much higher temperature Output Final Power Transistor's. The Heatsink Fins as well.
 
Very nice tread! Havn't read all the posts here, but some of them. I remember to have a SONY TA-3140F in storage with a "little" offset problem. It has about 196mV at the right channel. Left is at 43mV DC.

Offset is not adjustable, so i have also to change some transistors. Can anyone tell me which transistors are affacted? Is it Q502/3? Usually the original types are not longer available, so i woud need recommendations for substutes as well.

Thank you in advance!


Solved the problem today. It was the transistor Q601. Changed it from the left channel an DC was at 30mV. At the left side i replaced the konw missing Q501 with a BC547 resulting in 17mV DC.
 
Hi !

My Harman Kardon HK 3490 gives ok results:

L 25
R 21

I have set trimpots as close as possible to 21.12 as manual stated. Probably I will need to look closer into hardware.
 
Back
Top Bottom