Amplifier Distortion, DC-Offset, and You!

[EchoWars]

Hey B3nut!

No, there really isn't anything terribly wrong with those figures. Again, fiddling with anything that low in an attempt to get it lower is a case of diminishing returns. Plus, IIRC, the SX-450 uses an STK output module, which means your ability to tweak and fiddle are pretty much nil. Even then, most of them worked quite well, and sounded just fine.

Glad to have ya aboard!

Edit: The SX-450 uses TO-220 outputs...no STK's. Just thought I'd update that.

 

[cort]

DC offset

Echo Wars--

You're amazing, man. I'd never hear of DC offset and am not entirely certain I truly understand it now, though I think I got the concept..My Marantz 2270 w/in what you say are acceptable limits, but the Kenwood KA 8300 has a right channel reading (on average--it ping-pongs around) of 94mv. Now my multimeter is Ratshack ( and, except for an oscilliscope is about the only diagnostic device I know how to use, and I suspect I'm kidding myself about my understanding of the o-scope). Worse, just today, I picked up a 7000 series Sansui quad for $55. Rears are out. Both front channels read 100+. How/where do I adjust this ( or are the quad Sansui's an exception to what you said above)?

This is what I get for wasting years on language/literature degrees.

This is, IMHO. an exceptional post. And you're sorely mistaken if you think no one reads these. Just look at the volume of replies.

I am not worthy. . . but I aspire to be.

 

[B3Nut]

Nope....SX-450 has discrete outputs, complementary PNP-NPN pair per channel. I hear the SX-780 had power modules though...anyone know what type? Are they darlington-pair-in-an-epoxy-case type ones or full monolithic IC amp modules?

Want to get another vintage Pio Rx, but don't want a chip-amp machine. Got a Kenwood KR-A20 40wpc chip-amp (which the 15wpc SX-450 outperforms sonically) Rx as my PC amplifier, I'd like to make it go bye-bye.

TP

 

[EchoWars]

cort...

I know 5 pages of posts are a lot to read, but if you had you would see that DC coupled amplifiers (like most of ours are) use the input differential pair to cancel out variations in output DC offset (middle of page three).

Many amps have a method to zero this voltage out. I believe the 8300 does. I am totally unsure about the Sansui. On amps that do not allow adjustment, then the input differential pairs are simply replaced. Matching them before replacement is a good thing, but if the replacement transistors have good DC gain then you almost always end up with an offset of <50mV.

Again, anyone interested in fooling with this ought to track down a manual for the particular piece of gear and first determine if there is a DC offset adjust potentiometer. If not, and you feel that this is something you need to tackle (and IMHO, anything over 50mV should be addressed) then it's time to go transistor shopping.

 

[EchoWars]

oops...forgot the second part of my post...


B3nut...

My experiance with the mid and low-end Pioneer's is limited. I do know that many of them used it. If it is indeed discrete, then the capability exists to tweak it should one be so inclined. With the readings you have, however, wayyy more trouble than it is worth.

I have built a few circuits for nulling DC offset on amps that do not have provisions to do so. The attached drawing does just that. The 'Vee' is the amplifier neg high voltage supply, and the Vcc is the high voltage positive. The outputs go to the bases of the first differential pair transistor (this circuit is for two channels....stereo, ya know.. ) The two diodes (1N4148, nothing special) set up a current source (close enough) through the 5K potentiometer. Connect this through a 80k - 150K resistor to the base of the first transistor in the differential pair, and you can inject tiny amounts of current into the base of the transistor and compensate for the differences between the two transistors.

No speaker wire, Tice Clock, Pingo discs (or whatever they were), $2000 AC power cords, or other total BS. There is actual SCIENCE behind the reasons this improves the sound (imagine that?). Equalizing the differential pair on amplifiers that use them is the most startling improvement you can make in the quality of your listening experiance. Your half-deaf granny will hear the difference.

 

[B3Nut]

The Tice Clock! ROFL!!!!! I remember that thing...I was just out of high school and was a Stereophile subscriber at the time. I remember reading the article about that and just cracking up! Improve the stereo by plugging a $5 alarm clock with a really fat power cable that cost $2000....ROFL what a joke!

Thanks for the memories

TP, gettin ready to send the schemo to his trusty HP

 

[EchoWars]

If you want to actually try this circuit, be forewarned of a couple of items....

1. The leads from the 'output' (the 82K - 150K resistor, whatever you choose) to the base of the input pair must be as short as possible (under 3" and away from any transformers). The high gain of the differential pair guarantees that any hum or noise picked up by this wire will be faithfully amplified and appear at the output. Worst-case is an oscillating amplifier that toasts its output transistors or Zobel filter or both. Mild-case being unwanted noise and hum. If the left and right amplifier input pair transistors are separated by some distance, then a separate circuit will have to be built for each channel to keep the leads short. The length of the supply leads is not so important.

2. Ground the circuit to the same board that contains the differential pair.

3. Unless you use a multiturn trimpot, such as a 25-turn Bourns trimpot, this will be so touchy you'll never get it set correctly. Use good parts.

Cheers!

 

[cruisaire]

315mV on the left, 118 mV on the right

Time to buy that "Learn stereo repair in your spare time" book!

This is my "new" Heathkit 60ish wpc receiver.

 

[marantz1300dc]

Anybody have a good source for transistors? Im looking for some 2sb557/2sd427 matched pairs. The nte # is 281MCP but I would prefer anybody other than nte

Im replacing all the drivers and finals on a marantz 170dc.

 

[LARRY PRIESTMAN]

marantz1300dc
hi mate try this link mate they are replacements but they do list your trannies

good luck mate



http://nte01.nteinc.com/nte/NTExRefSemiProd.nsf/$$Search?OpenForm


rgds larry priestman

 

[EchoWars]

Now now...don't hijack the thread...start a new one if you need a hand. I'd be glad to suggest appropriate replacements.

BTW...stay away from the junk that NTE sells. They are a last resort only if all other sources fail.

 

[LARRY PRIESTMAN]

Whoop's I've stamped on some more toes he he he

1,000 PARDENS MASTER:ntwrthy: :ntwrthy: :ntwrthy:

 

[VinylHanger]

I love this thread. It isn't very often that someone can give a straight forward and simple explanation on how to actually do something useful and meaningful when it comes to repairs, thanks EW. I assume that my "new" Sansui AU505 should not be reading 310 mv? However it is in both channels so at least it is consistant

 

[EchoWars]

Your assumption is correct. But as I've said all along, it's an easy fix, even if the amp has no adjustment for zeroing the offset.

As an aside, I had a buddy drop by with a sick Dynaco ST-120. Capacitor coupled output. After we fixed the regulator, we measured 200mV+ at the output of each channel. Not supposed to happen with cap outputs...they were leaking DC to the output.

Replaced 'em both, dropped leakage to 3mV.

 

[JimiJohnB]

Echo this is a great thread, and a neat way to quickly check the 'health' of the unit. I like these technical posts, so please keep 'em coming.

Just checked a Marantz 2226B that hasn't ever received any attention as far as I know:

Right: 3.8mV
Left: -2.4mV

I'm still in the process of repairing a 170DC, but I'm anxious to see what its DC offset values read once it's functioning again.

 

[qboneus]

Awesome thread

Technics SA-500
L/1.4mv R/.8mv
Sony STR-V5
L/20.7mv R/1.2 mv

Can this be correct on the old technics?
Is this amp just wonderful or am I even
able to test this amp using this method?

This confirmed that something is awry slightly in my V-5
as I have always been slightly concerned as the power meters L/R have never been in sync.
I blundered around with the potentiometers mounted to the amp PCB and was able to get the left channel down to 18.5 mv but obviously something else is wrong.
I didn't get a chance yet to test my Pioner Sa-7100 or my modern plastic box Yammie HT but I am guessing?? that these measurements aren't useful on BPBHT equip?

Tal

 

[EchoWars]

BPBHT? Educate me...

?

The Technics is super low, so I'm wondering what topology it uses. Regardless, looks awful good.

Honestly, the Sony doesn't measure too damn bad either. If you are actually hearing a problem, then it is doubtful that the offset is the root cause. With 20mV offset, 2nd harmonics (which are not horribly objectionable...ask any tube lover) are a bit higher than the amp is designed for, but I'm sceptical that any sonic fau-pax is due to the 20mV unbalance. If there is a real problem, then it is elsewhere.

 

[qboneus]

BPBHT
=
Black Plastic Box Home Theater
Thanks Guys!
I'm sure noone recognized it 'cuz I just made it up!
Tal

 

[rca2000]

Dc offset...

I agree that one MUST be careful, when measuring dc offset, or adjusting idle bias. A few years ago, i worked for the tandy service center. I saw a guy, who had just finished repairing an amp, that was working fine, take one last offset measurment. His probe slipped off the emitter resistor, and hit ground. a quick pop through the speakers, protect kicked in, a little smoke, (and more than a few choice words from him!), and that was that!bye-bye channel!!SO, one indeed must be careful!!


By the way, some of you out there would CRINGE at some of the"state of the art" repair methods used by tandy(radio shack) on their customers repairs. I could tell horror stories. Like,some of the techs using more hot-melt glue than Macgyver, (to "rig " EVERYTHING that could be rigged, swapping out complete chassis in computer monitors,without letting the customer know, and so on. If i was an audio fan i would NEVER trust tandy, to fix my stuff!!

 

[EchoWars]

I've blown up more than my share of amps from being less than careful with a probe lead. Been a long while since that last time, but it cost me dearly then and left a permanent impression.

There's good guys and bad guys in any repair profession. Luckily, I don't repair..I restore.


/Edit again because I still can't speel.