Hafler 200 input strangeness

[pixpop]

A friend just donated a nice Hafler 200 to me. It works fine, except that I'm seeing DC offsets. I understand that this can be adjusted with a pot, but before I do that, there's something I'm puzzled by.

With the inputs open, I get a different offset than with the inputs grounded. On one channel, it's about 4 volts with the input open, but about 10mV with the input grounded. On the other channel, it's about 50mV with the input open, and about -700mV with the input grounded.

Since there are input capacitors, I think the offset should not change when I ground the input.

This indicates to me that there must be current flowing through the input capacitors. Is my reasoning correct on this? Should I replace these caps before I adjust the offset?

Note, my DVM indicates 0V AC on the outputs, regardless of whether the inputs are grounded. With a speaker connected, I don't hear any hum in either case.

[SY]

Yeah, you've got some leakers. Replace them. And do the same with the elytics in the shunt part of the feedback network.

[Dick West]

One of the more frequent problems found with these vintage amps is DC offset and there often is no way to correct it other than installing new matched pairs of transistors. The first 4 transistors in the circuit must be matched on hFe for both the NPN and PNP pairs, else DC offset results. Time and often overdriving the inputs result in these devices drifting out of spec over the years. You need to get a population of around 30 each of the NPN and PNP transistors and measure each for hFe and then install pairs that match within 10%.

This circuit is somewhat like an op-amp and its feedback network will compensate for some degree of DC offset, but not all of what you have. There is a simple add-on circuit to null DC offset but it will only compensate for around a volt maximum. I believe that changes in DC offset between the input open and shorted reflect the ability of the feedback network to compensate but not cure the problem, which is probably in the first four transistors.

You could also check the feeback electrolytic for proper functioning.

What color is the circuit card? If it is brown phenolic this board could not tolerate a solder bath temp sufficiently high to properly solder all connections. Some connections on these cards develop microscopic "opens" and a circuit sometimes can be rejuvinated by re-flowing all the solder connections.

Keep us informed of your progress and others here may also have suggestions.

[pixpop]

OK, cool. I'll replace the caps, and then see how the offset looks. Am I correct in my belief that the pot on the PCB is for adjusting the offset?

BTW, I think the boards are the old phenolic ones. A pale brown color.

Neil

[Dick West]

There is no pot to adjust DC offset on the DH-200. P1, the pot you see, is for adjusting bias or idling current. Download the manual from Hafler.com (tech archives) and read about the amp, its layout, schematic, and bias adjustment. I believe page 11 is important for this.

What is the serial number of your amp? Are there those little brown disk ceramic caps on the circuit cards?

The feedback cap (C5?) is non-polarized, 470 uF 6.3 VDC.

The DH-200 came in 3 versions.

1. phenolic circuit card with ceramic caps
2. phenolic circuit card with mylar film caps (red colored)
3. gray-green fiberglass circuit card with through plated holes and film caps.

The first version is the worst. Versions 2 and 3 can sound about the same. The last version is the best. Which do you have?

[pixpop]

I'm not sure which version. I'll check when I get home tonight.

Neil

[pixpop]

apparently, the previous owner was not content with the original design, and has implemented a bunch of tweaks.

The input caps I was mentioning do not exist.. they have been replaced with jumpers. I open circuited the jumpers, and now one channel has negligible offset, while the other has in excess of 4 volts. So, I'm going to say the one with the low offset is good, and the other is bad. (As I said, both seem to make music OK).

I'm going to try to compare the good one to the bad one, and see where the difference is.

The amp is serial #3027057. it seems to be the version with the mylar caps, not the disc ones, even though the parts list specifies disc. The only disc cap is the one across the bridge rectifier.

[djk]

With the back of the amplifier facing you, the one with the 4V will be the left side.

The input jack is isolated above ground. There is a 2R2 ground resistor on this jack. When this resistor is not grounded properly the input off-set current has no where to go anf the DC will float up quite high.

When connecting to another unit sometimes this resistor burns open.

I recommend you change this part to a 10R 1/2W and put a pair of diodes across it and a 0.1µF film cap.

That way when you get a ground loop it will not burn open, but will buzz like the dickens.

The DH220 and DH500 changed the location of the input jacks, making them side by side. They still have a 0R5 ground resistor on the circuit card that will burn open on a bad ground loop. Fix it with the same diode cap thing as above.

[Dick West]

pixpop,

Your amp's serial number tells me you have a phenolic board with mylar film caps. Be careful when soldering on this phenolic circuit card as a too hot iron will cause traces to become unglued and lift.

Shortly after the DH-200 came out various electronics pundits began to be critical of its circuit. Dick Marsh and Walt Jung published POOGE-2 in the Audio Amateur that gave a long list of mods and improvements to this amp. One of the more discussed items was the grounding scheme of this amp and Jung/Marsh showed how it could be improved into more of a "star" grounding scheme. The amp's designer, Erno Borbely, defended his design by stating he tried different grounding schemes and then stayed with the one that produced the least hum. Your amp's grounding scheme is OK as long as all its parts are intact.

The two RCA phono inputs should have been side by side, as in the DH-220, else a ground loop can occur between the widely separated left and right channel inputs of the DH-200. The solution to fixing a possible ground loop for the DH-200 was to lift the chassis ground of the left input and connect it to the chassis only through a 2.2 ohm resistor. This is R45 on the parts list so you can see it was almost an afterthought. Check R45 and make sure it is intact and that it properly connects RCA jack ground side to the chassis.

Many amateur "experts" at that time claimed the sound improved by removing C1, a non-polarized electrolytic cap. Removing it is NOT recommended for various reasons, one of which djk alluded to. I routinely replace them with a quality 5 to 8 uF polypropylene film cap such as those from Solen or Axon. The trick is to hang this much larger cap from the board's two closely positioned mounting holes.

Study the schematic and parts list and get your amp back to its original construction as much as possible then you can better tackle the offset problem. If Q1,2 and Q5,6 and Q3,4 are OK, matched, and with correct EBC readings, then you should check for problems with the driver and pre-driver transistors. In fact, a bona fide 4 volt offset might be caused by them and careful EBC voltage measurements, comparing the good to bad channel's readings, my help isolate the problem. Yesterday eve I got one channel of a DH-200 here down from 110 mV DC offset to 80 mV by installing a matched set of NPN transistors (Q5 , Q6). Today I will install a matched set of PNP transistors (Q1, Q2) and hope this will get DC offset below 30 mV. In the real world offsets below 125 mV are supposedly acceptable for this legacy amp, though I like to keep mine below 50 mV when possible.

BTW, what are the values of the fuses in the fuse blocks on the rear channel? A DC offset of 4 volts will blow these speaker fuses when the amp is first turned on, if a load such as speakers is attached, and your speakers must really do a "thump!" These fuses should be 2 -3 amps in value, never more than 5 amps (read the manual). Some amateurs put in oversized fuses there to stop them from blowing out without realizing the damage something like a 4 volt offset can cause to their speakers.

Keep us informed and good luck.

[pixpop]

With the back of the amplifier facing you, the one with the 4V will be the left side.

Correct!

The input jack is isolated above ground.

Correct!

There is a 2R2 ground resistor on this jack.

This resistor (R45 on the schematic) is nowhere to be found.

On closer inspection, it seems that the input jacks have been replaced with 'better' ones, and there is no solder lug for the chassis end of the resistor. Also, a large solder lug has been installed on the base of the chassis (near the other input jack), and a wire runs from this to the junction of the two PSU filter caps. It looks as though this point has been established as a star ground, and then grounded to the chassis through a wire, about 6" long.

Regardless of how my amp has been modified, why does the original amp ground the two channels differently? Why a resistor on one side, but a short on the other?