DC offset adjustment on a Hafler DH-500?

[analog addict]

After I measured the DC offset on my "new" Hafler Dh-500, I found that the left channel was ~100 mV and the right was ~500 mV. Ouch.:worried:

I did some searching and reading, and apparently according to the manual, it takes 1.8 VDC to trigger the protection relay. I also found out that the 500 had three variations of PC-19 board, and only one of them the newest, had a pot for DC offset adjustment. IIRC, this board was supposed to have two pots one for DC offset, and the other for bias....Here's the picture of my board(s)...

I only see the one black pot in the middle of the board. Can I rightly assume that this is the bias pot, and I don't want to be screwing around with it? If so, what else can I do to adjust the DC offset? Anything? Do I even need to worry about it? My intention for this amp is to run prolly the bottom end of my RSIIb's, and use another amp for the top end. Does this sound reasonable? I have tried this amp as is, and it sounds fine....

 

[jpdylon]

Yes, the one and only pot is the bias adjustment.

Sounds like the gains of the differential amp pair have drifted apart. You'll need to replace the differential amp transistors with a matched pair to get your offset down.

My .02

 

[Fred Longworth]

Also, make sure that those little heat sinks on the driver transistors don't come loose and short against anything. They've been known to. A dab of glue often helps.

Fred

 

[dshoaf]

Adding to the other excellent suggestions:

- I'd make sure the bias pot itself doesn't have dirt in it by rotating it end-to-end a number of times with the power off and even measuring resistance from the wiper to the end to make sure there's no dropouts along the range of travel.

- I'd also check the fan for ease of rotation or just disassemble it and clean the bearing races - not squeaky clean since these are scintered bearings but enough to make sure it runs smoothly. The amp doesn't look like it was used in a dusty environment unlike the P500 I purchased that was road-abused and rusty inside. Still worked, though.

Don't you just love it when all of us 'armchair bench techs' come out of the woodwork?

Cheers,

David

 

[jpdylon]

- I'd make sure the bias pot itself doesn't have dirt in it by rotating it end-to-end a number of times with the power off and even measuring resistance from the wiper to the end to make sure there's no dropouts along the range of travel.

Don't forget to note its original position and resistance before adjusting it

 

[analog addict]

Don't you just love it when all of us 'armchair bench techs' come out of the woodwork?

I couldn't do much without you guys. AK has been a godsend as far as getting me even part way down the road. I'm trying to do more to contribute too if at all possible. Anyway....

Someone wanna point out the differential amp transistors and the driver transistors to me?

Are all 4 of the heat sinked transistors all driver transistors? I'll try and ID the parts from the manual, and then see if I can cross reference a relacement set. Do the driver transistors also need to be matched, or just the differential amp transistors? Manual is here if anyone needs it.....

Thanks again......:thmbsp:

 

[Old1625]

As my high-speed service is down I cannot within reasonable time retrieve the whole manual that you have linked, but in looking at the schems you show Q3,Q4,Q5,Q6 are your diff pairs, with current sources Q1 and Q2 feeding the emitters. IIRC Trimpot P1 should be your offset null and P2 is the bias (idle current) adjustment for the ouchputs. If the amp is running cool and clean don't monkey with P2. If P2 is adjusted so that there is insufficient current at idle you will develop notch distortion. If P2 is set too far the other way then the sinks will run hot.

Adjust P1 to null output.

If the amp is sounding good to you I'd be reluctant to fool with replacement transistors; you may be opening up a can o' worms in trying to find viable subs to go in there.

 

[EchoWars]

In the pic, it apppears that P1 may have been replaced with a fixed resistor...

As far as subs, the stock 2N5550 and 2N5401 are pretty pedestrian devices. I think it would be easy to find replacements quite superior to them.

 

[Old1625]

In the pic, it apppears that P1 may have been replaced with a fixed resistor...

As far as subs, the stock 2N5550 and 2N5401 are pretty pedestrian devices. I think it would be easy to find replacements quite superior to them.

'Fraid you're right, now that I look at it again. Also in the manual through the link what is shown in the schematic as P2 is referred to as P1 in the paragraph describing the bias reference circuit.

If a hatful of the jedec #s you post can be obtained and then matched then that's probably what should be used. While I'm sure there are many faster devices around it sometimes doesn't pay to use them. In my experience with these if a sub's F-sub-t is markedly different then there can be problems with parasitic oscillations, etc.

And I try to avoid the use of subs such as SK, ECG, TCG, etc. :thumbsdn: They can be bad news.

I've got 2 of these Hafler 500s, which have been the utmost in reliable and clean sounding. I diod had to reconstruct a burnt ground land in one of them after I got it back from a rental. Hope nobody got hurt.

 

[analog addict]

Translation????

SO replace Q3/Q4 with an equivalent to 2N5550, and Q5/Q6 with an equivalent to 2N5401? Of course I'll need to do this for both sides, right? And the suggestion is to stay away from certain types of substitutions?

May I be so bold as to ask for a suggestion or two? :ntwrthy:

Digikey/Mouser a good source?

 

[Old1625]

I'd try to chase down the original #s if I were you, and also the same vendor. Now that my memory's refreshed as to what the #s are I'm feeling a little more optimistic that NOS devices of such can probably be had for a cantata. IIRC most of these were Motorola.

There is sometimes the temptation to use the more available Japanese transistors of the 2SA,B,C,D series, but my experience has largely been that while the advertised characteristics often seem to match in the aggregate it usually pans out "east is east and west is west" so to speak.

Best to stay with what is in there. All the Hafler amps I've serviced I've used the same #s as originals, but that was back in the day. The rules might have changed a bit. I'm not doing near that kind of work as I once did. Perhaps some younger--and still functioning--techs can chime in here as to current availability.

In summary, though, I'm of the old school, and while I could be a perfectionist at times when I was younger I am getting to be more pragmatic. IMHO a half volt of offset on an old standby as yours is nothing that should cause you to want to keep a fire extinguisher handy. I'd say let it run a bit that way, keep looking for deals on those old JEDEC #s and enjoy the amp. I can't see where any major damage can occur, and I haven't thus far with my 2 units. I have to admit that the second one I got I've never bench tested--just used. God knows what minor DC offset it may have on one output or the other.

The other thought I have is that perhaps you could retrofit your driver board in accordance with the schematics you show by switching out whatever fixed resistors are there and wiring in a Bourns or Allen Bradley trimmer pot to give you a bit more control over the situation. That would be very likely what I would do if it were mine.

 

[EchoWars]

If a hatful of the jedec #s you post can be obtained and then matched then that's probably what should be used. While I'm sure there are many faster devices around it sometimes doesn't pay to use them. In my experience with these if a sub's F-sub-t is markedly different then there can be problems with parasitic oscillations, etc.

My criticism of the 2N parts has nothing to do with speed, and everything to do with their middling gain figures. I'd replace the 2N parts with ZTX795A (PNP) and ZTX694B (NPN). About the same speed (or maybe slightly slower) as the 2N parts, more linear, and MUCH higher gain to reduce base current.

And I try to avoid the use of subs such as SK, ECG, TCG, etc. :thumbsdn: They can be bad news.

Yup...

IMHO a half volt of offset on an old standby as yours is nothing that should cause you to want to keep a fire extinguisher handy.

No, but it destroys sonics. On your old Yorx console, who cares? On something you want to call 'hi-fidelity', anything over 100mV needs attention.

The other thought I have is that perhaps you could retrofit your driver board in accordance with the schematics you show by switching out whatever fixed resistors are there and wiring in a Bourns or Allen Bradley trimmer pot to give you a bit more control over the situation. That would be very likely what I would do if it were mine.

A no-brainer. Absolutely.

 

[Blue Shadow]

I have a question about this Hafler thread. Can we expand it to include the DH-200 which seems to have a very similar driver board. It seems to use the same 2N5550 and 2N5401 transistors, lacks an offset adjustment and has the less than top quality bias trim pot. There is the attention and responses from the pros and I would (probably many others, too) like to 'tune up' my Haflers.

I would be glad to provide the schematic or to start a new thread. TIA.

 

[BalancedXLR]

...

 

[dshoaf]

Note that there were 2 or 3 different versions of the Hafler driver boards - I may have the original PC-19s around here from the swapout I did for my P500s. There was also a PC20 version, too, and perhaps a PC19A. This could account for some of the differences between the schematic you have and the actual board in use.

If our OP chooses to go after the diff pairs, I'd also be in there swapping out those electrolytic caps, too.

Not to take this thread off track, here's another - but more expensive alternative to rebuilding the original boards. I have these in my P500. They're about $350 the pair but well worth it, IMHO. I believe Van Alstine still sells replacement driver boards, too.

Hope that helps in some way.

Cheers,

David

 

[BalancedXLR]

...

 

[analog addict]

The Musical Concepts board upgrade...

is one possibility I've considered. Let me ask the experts....How difficult will it be to match transistors? Would you need any specialized equipment? How many of each would you need to buy to have a chance to match a set? Of course, I'd have to get a quantity of 2 different types of transistors.....

My DH-220 already has an old MC upgrade board....Maybe I'll sell it to partially fund the upgrades for the DH-500...:scratch2:

As for for the 500, I've read that there are three iterations of manufacture. I have "Five Hundred" spelled out in white. I've also seen "500" in red numbers. Anyone know the sequence????

 

[BalancedXLR]

...

 

[dshoaf]

My DH-220 already has an old MC upgrade board....Maybe I'll sell it to partially fund the upgrades for the DH-500...:scratch2:

You don't have to sell it, necessarily. Those boards can be transfered to the '500 but it does require a resistor change to accomodate the additional output devices. Problem there is that the poor DH220 will have to live its life out incomplete.

Cheers,

David

 

[analog addict]

Interested in...

I have a question about this Hafler thread. Can we expand it to include the DH-200 which seems to have a very similar driver board. It seems to use the same 2N5550 and 2N5401 transistors, lacks an offset adjustment and has the less than top quality bias trim pot. There is the attention and responses from the pros and I would (probably many others, too) like to 'tune up' my Haflers.

I would be glad to provide the schematic or to start a new thread. TIA.

going together on a transistor deal? Mouser would sell 25 of each for about $60 or so total. Also, Would one of the experts humor me and suggest a replacement trimmer pot, and which resistor it would replace?

Lastly, are the pinouts on the transistors the sames as the pieces they replace? I screwed up the last transistor replacement I attempted, so I really want to do this one right.....BTW, these are both TO-92 cans, is that OK???

Thanks again....:thmbsp: