Hafler DH200 DC Offset

[ronzeman]

There is no adjustment on the driver board to null out the DC offset. My old DH200 sounds good...but today I measured 8.9mV DC on RIGHT channel, and ~300mV DC on the LEFT channel (measured at speaker terminals. Is this extraordinarily high? A problem? Any solution / suggestions? ( I see the DH220 has a trim pot for this...not so in the DH200) THANKS !

[eyoung]

If you do a search you will find much discussion on hafler dc2**
amps. The short answer is to match the differential transistors(q1,q2,q5,q6) for current draw. Transistor matching is an art unto itself, also a search is in order(ck out Nelson Pass DIY page for additional info).
There is a ton of info on the forum just search Hafler DH200/220 mods for the thread...Happy Discovery


Regards, Elwood

[ronzeman]

I spent much time sifting through some of the materials mentioned....However...I'm not a very technical person...and I have little test equipment beyond a DMM and soldering iron. I was hoping there would be a resistor to tweak or a trim-pot I could solder in to use to null the offset to something under 25mv.
Perhaps a new set of driver boards (musical concepts?) could help but I'd rather not invest another $400.00 into my 1981 vintage Hafler DH200. ( I see where the "newer" Haflers i.e. DH220 has a trim pot to offset DC on output but no such function is on the DH200 driver board) . In the real world how big a problem is a 300mv offset on a channel?
Thanks

[EchoWars]

Should be a lot easier (and cheaper) to match and install four new transistors than an entire board.

[eyoung]

Transistors are cheap(at least these are)buy 10 or 20 and use your dmm to match. all you need is a cheap power supply and a DMM.RadioShack has all you need to build a jig to match transistors, and MOUSER or DIGIKEY has the transistors. take a little time relax have a beverage of choice, and match a few transistors . put them in the board an check your offset, anything under 50 mv is good. goto www.Oldcrows.net/~patchell/matcher/matcher.html , for a simple circuit to build a jig to match transistors(bjt type) Go to Hafler.com for owners manuals with parts list for DC200 and DC220 to get transistor part numbers. both of these manuals have schematics and parts list.

Stay after it, You Know you want to...

Regards, Elwood(forgive the typing errors)

[john65b]

Try this - measure resistance from power ground to speaker grounds and power ground to input grounds on both channels.

You can also try to jumper the speaker output grounds and signal grounds directly to power ground.

DC offset should then drop.

[ronzeman]

Perhaps trying the Xistor change out will be the way to go (for now). I recall somewhere in the depths of the recommended readings that there are better substitute transistors to use in place of Q1,2,5,6 in the DH200. Can anyone tell me what they are ? (If hope they won't require other component changes too).

Anyone know if there are "matched sets" pre screened available out there? THANKS

[eyoung]

The 2N5401/2N5550 are ok for especially more modern versions.
Try our friends at http://tech-diy.com/Store/store.htm for matching and availability. You may also consider replacing all the electrolytic caps on the boards as well as the P/S as they are getting along in age, and doing some research on grounding schemes ...

Good luck.

Regards, Elwood

[djoffe]

To get 300 mV of offset, something has to be VERY unbalanced, or kind of leaky.

My best guess from far away is that C5, a 470 uF non polar electrolytic, has gone very leaky, or quite shorted.

That raises the DC gain from 1 to around 22. That's enough to make a LOT of offset at the output.

Don't operate the amp with the cap removed, as it probably won't be stable.

Replace the cap with a new one, and cross your fingers.

Best of luck...

Dan

[bwaslo]

I've seen several DH200s that developed those kinds of offsets, and in both cases it was the input transistors that were doing it. Rather than sending off for new ones, I fixed the amps by changing the input stage into a cascode arrangment using different transistors (unmatched 2n5210 and 2n5087) as the 4 input parts and the existing high voltage (but apparently a little leaky) 4 transistors as the upper output parts. It's a pretty easy change, using the same PC layout and only needing the transistors and a few wires.

It would be easier to replace the existing transistors with the same type (no wires to jumper, leads to twist, or circuit changes to make), though, but matching transistors is a fair bit of effort to arrange to do. The matching probably isn't needed anyway (I doubt Hafler did any when they made the amps), and you'd get a much lower offset by just putting in new (nonleaky) replacements of the same type transistors. A 300mV offset is a lot(!) and is NOT going to happen just from differences in Vbe between the transistors, more likely to be collector-base leakage.