Hafler DH-200/220 Mods

[Doubtingthomas51]

Well i guess this is a place for a newbie to jump in.
My son and I have bridged to Hafler 200's to monoblocks.
We have upgraded the power switch, the inputs, the binding posts and the power cord.
We are very excited by our progress as we feel the results compare favorably with my McIntosh MC 2200 which they now will replace.
Since this is our first real solid state project( we are fairly conversant with tubes) we are looking for two pieces of advice. 1)how to progressively increase our theoretical understanding, and, 2)the logical sequence of mods ahead.
We are serious DIYers to the point that we built my electric bass, modified my Oliver tube head, etc, ad nauseum and are very excited by the knowledge we can find here. Thanks.

[tvrgeek]

The 200 was a terrific amp when Erno designed it, a long time ago. It has no protection circuits, outputs are no longer available, parts quality are not up to modern standards. You could do a lot, but you could also buy newer designs for less off e-bay. It is also a pain to work on as you can't get to the back of the board easily. I know, I have been playing with a 120 to learn about amps.

On e-bay, there is a guy with replacement modern main caps and higher quality signal caps kits. I also put in a hexfred rectifier that made a big measurable difference, but some suggest correct snubbers do just as well. Bigger lower ESR caps you may want a soft start and bigger rectifier.

Anything is possible; twin die outputs, complete new drivers, protection, mute, soft start, FET inputs, anything you want.

[DarthBubba]

Originally Posted by tvrgeek

The 200 was a terrific amp when Erno designed it...

Still is, bro!

Originally Posted by tvrgeek

On e-bay, there is a guy with replacement modern main caps and higher quality signal caps kits. I also put in a hexfred rectifier that made a big measurable difference, but some suggest correct snubbers do just as well. .

Can you share your source for the HexFRED rectifier with your friends, please?

[Doubtingthomas51]

This is off to a great start. Thanks very much. I am off to learn about solid state rectifiers with the hexfred rectifier.
Snubbers is a term I've not heard before?
Can some one point me in the right direction.
Any books out there that serve as essential source material?
And a little off topic but if you haven't ever seen Mr. Hafler's chess set it's worth googling.
Definitely not DIY!

[jackinnj]

Originally Posted by Doubtingthomas51

Snubbers is a term I've not heard before?
Can some one point me in the right direction.

Snubber is the term referred to a device which damps the ringing of a resonant circuit -- in the case of the rectifier diode the junction capacitance and the leakage inductance of the transformer form a resonant circuit -- if it's bad enough the ringing can be heard or it may force other PN junctions to conduct when they shouldn't. In the case of switch mode power supplies it can cause significant interference to communications devices.

The folks who manufacture diodes (On Semi, ST Micro, IRF etc.) as well as the capacitor manufacturers put out white papers on damping the ringing, but one of the easiest to understand is this:

http://www.hagtech.com/pdf/snubber.pdf

[tvrgeek]

Originally Posted by DarthBubba

Still is, bro!



Can you share your source for the HexFRED rectifier with your friends, please?

Sure, again.
As I also put in caps 4 times as large and half the ESR that are 25 years newer. To withstand any higher inrush, I put in a much larger bridge. 35A steady state. I measured the harmonic content before and after on the amp output. The LF noise floor dropped by 5 dB and the 60 Hz and harmonics dropped between 8 and 13 dB. How much was the very expensive bridge, how much caps? Can't say. I woudl ascribe all of the LF improvements to the caps. The Hafler did not have proper R-C snubbers, just a single cap across the AC. Gentlemen like Mr. Curl say shut up and use them, where Mr. Pass says correct snubbers are just as good. By correct, I am taking that as correctly tuned, not just a few .01uF caps thrown at it. It is an RLC issue. I suspect both are correct. Search the WEB on snubbers and learn how to do it correctly. The short answer is you have to prototype and tune. The technical notes on the bridge say "reduced need" for snubbers, not none. I have not had time to play with that. The DH-120 is a royal pain to use as a test bet. We are not talking any magic cures for cancer.

I repeat "were in their day".
There are a lot of very good arguments to use FET front ends, the gate resistors were not understood yet, method of dominant pole compensation, 5% carbon film resistors, mica and electrolytic caps, and transistors out of the stone, age put them in the "were" category, without even getting to a speaker fuse as protection that won't protect anything other than maybe the speaker cable. Perhaps that they are valid at all and listenable today serves to give credit to his design expertise. They also fail my wife's critical hearing test of some distortion when trumpets blare. Exactly what is a mystery. Many modern amps fail this too. So far, of the amps I can actually pay for, only he small Rotel 8 and 9 series pass the test. As we know, for some reason MOSFETS have very weak bottom ends. So, 35 years on, you can do better. Should that be a surprise? The first transistor was only 20 years before and L-MOSFETs were brand new. No SPICE simulation yet.
You can spend a lot of money redesigning a 220, even more tweaking it once done. When done, you have the same box and heat sink, not a DH-220. I have spent half the cost of a new Halo on textbooks and parts, countless hours studying all I can including learning LTSpice and have only an improved amp that still will not pass muster and any fault will fry both the amp and speaker. As a hobby for my curiosity, this has been great. TO get better sound, well a $150 e-bay Rotel suits us far better. Jump up a notch or two: Acrus, Forte, Aragon, Parasound, Adcom. A bit more, Nak, Krell, or whatever. Far cheaper. If you know which ones, Denon, Onk, Marantz all have made super units.

[tvrgeek]

Oh yea, it came from Parts Connection.

[DarthBubba]

Originally Posted by tvrgeek

Sure, again.

That was in another thread; doesn't count here!

Originally Posted by tvrgeek

As I also put in caps 4 times as large and half the ESR that are 25 years newer. To withstand any higher inrush, I put in a much larger bridge. 35A steady state. I measured the harmonic content before and after on the amp output. The LF noise floor dropped by 5 dB and the 60 Hz and harmonics dropped between 8 and 13 dB.

That's A LOT of improvement!

Originally Posted by tvrgeek

The Hafler did not have proper R-C snubbers, just a single cap across the AC.

He/they may have been aiming to bypass the high-frequency rectification noise and demodulated FM radio signals away from the rest of the circuit rather than snubbing the LCR ringing between the transformer winding and the rectifier capacitances.

Originally Posted by tvrgeek

Search the WEB on snubbers and learn how to do it correctly. The short answer is you have to prototype and tune.

You're making me wish I'd never sold that 5MHz o'scope again...

<Balance deleted for lack of time to respond>

[gootee]

Practical snubber design procedure is at post 292, here:

paralleling film caps with electrolytic caps

[Doubtingthomas51]

Wow is this place an intellectual feast of what.
I feel like I've jumped into the deep end of the pool.
Thanks for pointing me to post 292(I think!)
Meanwhile the new inputs, fuse upgrade, new binding posts
and DIY power cord upgrade on my monoblocked Hafler 200's has already yeilded results.
Caps and bridge probably next. Gotta pause a full measure while I try to digest this stuff before plotting my next move.
Thanks again.