Hafler DH-200/220 Mods

About Hafler XL-280

Reply to Dick West,

I am curious about the upgraded PA-3B from:

http://www.musicaldesign.com/mc_amp_mods.htm

Which ones have you bought? What do you recommend out of the options that there are?

PA-3B modification kit
$299 kit, $429.00 installed

PA-3B Signature
$449 kit*, $629.00 installed,

NEW!!! PA-3B Platinum
$999 kit*, $1299.00 installed

Also are the: Power Amplifier options worth it?

Thanks
 
I have tried the MC PA-3A and now its more recent version PA-3B. The PA-3B is the best. I prefer it to the sound of the DH200 or DH220, or the DH-120, Hafler 9130.

This is the extent of my experience.

As for the more expensive versions of the PA-3B some say the extra expense is worth the cost, others say that the more expensive capacitors are not. It all depends on your ears, the resolving ability of the rest of your system, and the size of your pocketbook.

The MC power supply caps make a big difference in every amp I have tried them with, if your switch and diodes can handle the extra capacitance.

Good luck.......
 

fab

Member
Joined 2004
Paid Member
pcb with mods

Dick West said:
fab,

As I recall from a post many months ago you toyed with the idea of making a PCB that encompasses your DH-200 mods onto a circuit card which others can use to enjoy the fruits of your labor and knowledge. Any further thoughts on this?

I praise MC circuit cards highly but the price of their kit is more than I want to pay, at least not pay frequently. How about a PCB from you?

Dick and Terry

Thank you for your interest. Unfortunately, I can only say the same I said on post #29 ! Shame on me...These last months I could not do much electronics and I stayed away from the threads too. Now I am back but still working on improving the output stage, doing simulations, etc.

The project is still going but I need to "clean" my lab first and that is probably the main reason why I do simulations instead of working with real circuits at this moment...

I could eventually provide a pcb but NOT populated. The pcb would also be from a local shop so it would not be of quality such as the MC with glazing and silk screen for components locations.

BTW the MC PA-3B pcb layout is in their installation manual...
This is how I saw their enhanced cascode VAS in the circuit.
Also the MC PA-3B has supply filters on the board using higher caps (4700uf or 470uf?) instead of 100uF for stock DH-200/220. So it is closer to a regulated supply.

fab
 
"BTW the MC PA-3B pcb layout is in their installation manual...
This is how I saw their enhanced cascode VAS in the circuit.
Also the MC PA-3B has supply filters on the board using higher caps (4700uf or 470uf?) instead of 100uF for stock DH-200/220. So it is closer to a regulated supply."

The circuit card filter caps on the PA-3B are 1,000 mfd/100 VDC.
 
I own 4 PA3 modded Haflers, two DH200 and two DH220. One is the original PA3 and the other 3 are the latest PA3B revision. John Hillig is preparing to raise the price on the basic kit, so if anyone wants to get one for $199 they'd better hurry. I have the PA3 cards and the upgraded 27,000uf 80v filter caps. The only other mods I have done are beefing up the power cords , binding posts and rca's to gold plated ones and replacing the front end caps with Black Gate 100uf 100v FK and C1 and C5 with Black Gate NX bipolar caps. John recommeds the Black Gate changes , though I was leary of removing perfectly fine 1000uf 100v Nichicons and substituting the 100uf 100v FK's. It seemed like I was taking a more robust cap and going back to the stock value cap. John says the Black Gate caps will make you rethink what you have always believed, that bigger equals better and it's the quality of the component not the size that makes the difference. I believe the Black Gates were worth it, though I know many would dissagree. I've only been playing with the Haflers for the last two years but am a big fan of the Hitachi lateral mosfets as well as the Musical Concepts mods. They are very good quality and easy and fun to install. I was wondering what the general consensus around here is on the merit of Black Gate caps and if anyone else uses them.
 
Ahem . . . Chris, I believe the PA-3B kit as actually $299, not $199. And, yes, we know the price will be going up so now is the time to buy, if one is interested.

Your comments are very similar to what Hillig has on his web site so it is obvious you are a fan of his, as are many of the rest of us. However, it is my understanding that the Black Gate caps are no longer available. Also, over on the Audio Asylum forums, many have commented about the BG caps sounding "different" but not necessarily "better," and that some of them require a long break in period. Now we know the brain plays tricks on us and we adjust our expectations of what we hear based on previous experience. Therefore a different sounding cap can begin to sound OK, in time.

We know that clearly defined audio measurements do not always agree with subjective impressions of listeners but, heck, a lot of listeners do not agree with each other. And so the story goes . . .
 
Dick,
You are correct , my bad. I meant $299 not $199. My first post and I start with a mistake Doh! that's what happens when I post past 3am. I actually got my last three PA3B kits at dealer cost, so I paid less than the $199 figure. Anyhow, I was intentionally paraphrasing what John says on his website regarding the Black Gate caps, I just wanted to get some feedback as to whether or not other peoples think they are worth the additional cost. I don't know if Black Gate caps are out of production or not. I've read that Rubycon may be converting the current manufacturing plant to Organic Semiconductors . I've also read that Rubycon does not own the patent for Black Gates, and that it's possible another manufacturer might take over production. In any case they are available , the last time I spoke to John he had plenty. Percy's audio is another good source. I've read the long break in time threads too. I don't think it's necessarily the case though, I think they sound good without a prolonged break in, though they may possibly improve with time. My Parasound Preamp PHP-850 came with several Black Gates factory installed and it required no protracted break in period to sound decent.
 
So, Chris and DickW... What do you think a nice first step would be for me?

Chris, I _swear_ I'm gonna put those caps you sold me in my 200 Real Soon Now. That one's my best sounding, and has the maggies in the living room. The power switches will also help. Now go leave some feedback - I need it after that fellow sent me what looked like a perfectly good P230, that when it came out of the box looked like it'd smarted off to a hungover cop, and fallen down some stairs... And then proceded to ignore me for a week...

I've got one DH-500 that seems to have a little bit of a hum when I get everything plugged in down in The Bunker. I think it started with RCAs, but now it's got quarter inchers. It also looks like the folks bypassed some fuses. I'm going to take the schematic/assembly book, and basically go through it from Point A to Point ZZ, and make it "new" again. Good thing I've got the other 500 to check it against visually...

Thus far the planned electrical load for the basement is...

DH-500 rears
DH-500 sides
DH-225 left front
DH-225 right front
P230 center
 
My DH-200 mod summary

Many thanks to several on this and other threads that listed mods to this amp. This one is especially pertinent to point out as helpful:

http://www.diyaudio.com/forums/showthread.php?postid=398116#post398116

I made several add-on mods. No circuit card traces were cut or changed. The sonic effects of each mod were auditioned and only those that improved the sound were retained. The mods fall into four categories:

1. Bypassed loop from circuit card eyelets 8 and 11 with a jumper. Then, speaker out (eyelet 12) was daisy chained through the speaker protection fuse to the red output terminals. Speaker ground or return (eyelet 6) went direct to the main ground between the two PS caps (known hereafter as G1). The black speaker terminals were also connected to G1.

2. RCA inut jacks are floating. Their signal and ground wires went to eyelets 1 and 2, as normal. A third ground wire was also connected to eyelet 2 and is included in the twisted bundle of wires. This third wire was brought down to a terminal strip mounted at the rear halfway between each side. These two extra ground wires from each channel are now connected to the chassis ground through a 10 ohm resistor. This chassis ground reference point is also connected to G1.

3. All electrolytic caps on the circuit card are bypassed with 0.1 uF mylar film caps. Caps C1 and C10 were replaced with new 220 uF caps. The drains for both PS rails are bypassed to heat sink chassis ground via a 0.47 uF polypropylene Wonder Cap mounted behind each circuit card. Each PS filter cap (C16 and C17) is bypassed with a 4700 uF electrolytic cap.

4. C1 is replaced with a 5 uF polypropylene film cap.

Summary:

Simplifying wiring from the circuit card to and from the inputs and outlets (items 1 and 2) seemed to reduce noise floor and make the sound less "cluttered". The sound just seemed to get easier to listen to. This simplification and improvement in the star grounding scheme paid off.

Items 3 and 4 made the sound smoother with better quality mids and highs, plus more body to the lower mid range. The use of film caps as bypasses are readily recognized as good mods, but the effect of the 4700 uF add-ons to the main filter caps was a revelation. The 26,000 uF caps from Musical Concepts sound very good but I was reluctant to spend $70 to put them into this amp. However, the addition of the 4,700 uF caps made them sound almost as good as the MC caps. The replacement of the 100 uF C7 and C10 caps with new 220 uF caps made a very hearable difference. These new and larger caps helped improve dynamics and body in the sound at all frequencies. FWIW on a previous amp I replaced C7 and C10 with 400 uF caps but the sound became "ponderous." I felt they were too large. The 220 uF caps seemed to be about the right size for this circuit and they quickly became "keepers."

DC offset is 41 mV in one channel and the other channel, after replacement of Q1/Q2 and Q5/Q6 with matched pairs, is now down to ~2 mV. Bias is set at 270 mA in each channel.

I have had several DH-200 amps here over the years but can state without doubt that the above mods have made this the best sounding of the bunch. I love listening to it.

Dick
 

fab

Member
Joined 2004
Paid Member
Choices for driver circuit

Hi all

In previous posts some of you seemed to indicate that they liked the MC PA-3B circuits replacement for the stock DH-200 amp.
I just came up with 3 choices at this time:

1) Finish my new pcb layout implementing all mods I have done on the stock PC-6 board of the DH-200 (plus new BIAS circuit). The DH-200 circuit is a symetrical circuit from end-to-end.
or
2) Built a completely new circuit implementing the "design principles" of the mods I have already done on my DH-200 amps except that the front end would not be symetrical. It would use the enhanced VAS cascode as found in the MC PA-3B but will also include more stuff too.
or
3) buy the MC PA-3B but it is quite expensive and I have never heard the sound of it. I am not sure if it "sounds" better or worse than my DH-200 mods. Also, I think it is much more fun for me to develop my own circuits.

Of course the 2 diy options would be a "fit form function" for the stock DH-200/220/280. See attached picture of an unfinished experimental prototype for the DH-200/220/280. You will notice that it includes a voltage regulated front end supply in the left and right side of the pcb. The holes for mounting the pcb into the DH-200-220-280 are indicated in the 3 rectangles.

What do you think?
 

Attachments

  • proto1.zip
    79.8 KB · Views: 631
I like what I see but reading the schematic and parts list is more helpful than a picture, at least for me. The use of those voltage regulators on each side is quite creative. What voltage do you use for them? Does the card run at a higher voltage than the output drivers and MOSFETs?

I am told that the MOSFETs should clip before the circuit card as that is less nasty than the other way around. This requires two voltage levels. Higher for the front end and lower for the output end. An XL-280 transformer would fit the bill here.

Personally I vote for a non-symmetrical circuit. I have the PA-3B here and compare it frequently to other amps. It has a sound that is true to the music. No overly pronounced or flabby bass or zingy highs. It is accurate. It also has a lot of detail and subtle inner nuances I miss in other circuits, and it does so without the sound becoming "etched" or overly clinical.

Tell us more. . . .
 

fab

Member
Joined 2004
Paid Member
Choices for driver circuit

Dick West said:
I like what I see but reading the schematic and parts list is more helpful than a picture, at least for me. The use of those voltage regulators on each side is quite creative. What voltage do you use for them? Does the card run at a higher voltage than the output drivers and MOSFETs?

I am told that the MOSFETs should clip before the circuit card as that is less nasty than the other way around. This requires two voltage levels. Higher for the front end and lower for the output end. An XL-280 transformer would fit the bill here.

Personally I vote for a non-symmetrical circuit. I have the PA-3B here and compare it frequently to other amps. It has a sound that is true to the music. No overly pronounced or flabby bass or zingy highs. It is accurate. It also has a lot of detail and subtle inner nuances I miss in other circuits, and it does so without the sound becoming "etched" or overly clinical.

Tell us more. . . .

Dick,

The circuit shown is not the one I want to build. For the symetric circuit (mods already implemented in my DH-200) you can read my previous posts to see parts of the circuit (input stage, VAS). Bias is improved by CFP configuration. Driver stage for mosfet is the same except the current is doubled.

For the non-symetrical circuit, it is still in the simulator at this moment and I rather have it prototyped and tested at least electrically before claiming any good value. I am still not sure to include output error correction or not, etc..

The voltage for the regulator is the choice of the diyer. It can be adjusted and compatible with a higher front end supply. Having a higher voltage for the front end is not imperative for me for mosfet output amp but is surely most efficient power wise. I do not go into clipping since it is only for in house use. I may use a higher voltage anyway.... By the way I include the mosfet driver transistors in the regulated supply too (since I have heard the difference...).

The "details and subtle nuance" are the main characteristics of the symetrical mod DH-200 (see my previous posts) and I hope to get the same or even better result with the non-symetrical topology.

So far one vote for the non-symetrical.
 

fab

Member
Joined 2004
Paid Member
More on non-symetrical circuit

The new non-symetrical circuit pcb would have the following features:

- cascode CFP diff input stage
- current mirror in diff input load with high tail current
- cascode active current source for diff input pair
- darlington enhanced CFP VAS cascode
- cascode active current source for VAS
- bias circuit part of the isolation buffer between the VAS and mosfets
- at least 20ma for mosfets current drive
- regulated supplies in front end until mosfets
- all bjt
- adjustment of local feedback of diff input stage
- adjustment of local feedback of VAS
- possible damping factor adjustment

Any comments?
 
Bogie,
I would try Digi-key , hopefuly it's just a generic single pole slide switch. The only Haflers I ever owned that had factory installed bridging switches were a DH120 and an SE240 . I regret selling both of them particullarly the SE240 . I was all about the DH200 and DH220 after discovering Musical Concepts and all the "cool" mods I could do with those amps. The SE240 just didn'y have the DIY factor I was looking for at the time . In hindsight it was a damn fine amp right out of the box and didn't need a whole lot of mods to sound and look good. I've seen factory bridging kits pop up on ebay for the DH200 and 220, which appear to just be just toggle switches with a bracket to mount it to the rear of one of the heatsinks along with some wiring and installation instructions. If your P230 is like the DH120 I had it is a slide type switch and you can probably find a replacement at Digi-key. I'm currently awaiting delivery of some IXYS FRED bridge rectifiers for my 4 Hafler amps. I've been living with the stock bridges but I'm told these new soft recovery type rectifiers will be faster switching, quieter , and will increase reliability of my amps. Hopefully it will be a worthwhile upgrade and not just an exercise in soldering. Does anyone have any experience or opinions about upgrading the stock bridge? I picked this one because it came as a finished piece . I know I could have built one cheaper but I was concerned with aesthetics as well as performance. Next mod will be to install IEC sockets in 3 of my amps, one already has one.
 
Chris,

Check the schematic on the XL-280. It shows clearly how to bridge this amp using a DPDT slide switch. It uses a small slide switch. The DH-222 bridging kit (no longer available) shows the use of a DPDT toggle switch. The slide switch I have here is about 1" in overall length with fixing screw holes 13/16 apart. This should be a simple item to find. Heck, maybe even Radio Shack has it.

The IXYS bridge is tempting but stuff I read, like from Nelson Pass, claim that these exotic diodes are of little value in a circuit as "slow" as 60 hz (or even 120 hz with a full wave bridge). These exotic diodes are used in switching power supplies that are called on to rectify 450 hz (or higher). Many people who install them in their power amplifiers, such as a Hafler, only think they hear a difference. Also, some state that snubbers (cap resistor) are better but the values of their components need to be determined by the circuit they feed.

Now, let the flames roll. When I can see hard headed evidence of their value I may be otherwise persuaded.

Is the installation of an IEC socket a sonic upgrade? Not having to wrestle with an attached cord is a convenience and, I suppose, one might find an IEC cord that is 14 gauge or even 12 gauge to replace the 16 gauge stock power cord, but will this short length of wire really improve sound? It represents another exercise in soldering, another set of metal to metal contact points, with questionable sonic value.

Now, more flames may roll. They keep me warm in the winter.

Dick