This thread has been created to discuss to retrofitting existing old Hafler and Adcom amps using Nelson Pass's circuit design referred to as the PLH, based on work by Lohn Lindsley Hood. The thread about the design is here: http://www.diyaudio.com/forums/showthread.php?s=&threadid=64081&perpage=10&pagenumber=1
We hope that we can devise a way that this can be done so easily that thousands of these old amps will get a new lease on life based on their newly aquired excellent sound quality. To facilitate this the ground rules are to use as many of the existing parts as possibly, so that the time and expense of the upgrade will be minimal, and that newbies won't be scared off.
Eventually we hope to have a group buy of boards that can be used in amps built from scratch and to accomplish the above retrofits.
A parts buy would be fun too- at least for those of us that get the parts!! Yet the number of required parts will be pretty low..
Anyway- let's get on with it!!
Response to PSz from post above:
That's still a mystery! I'm not sure what voltage the power supply on the Hafler puts out, but if it is around 40 , then it should work. The HAfler supply isn't balanced so again, it makes sense that the Plh will be OK. My biggest question is whether the voltage that the HAfler supplies will overwhelm the heatsinks.....Possiby each amp should be converted into a monoblock...
Love the circuit. I have myself convinced that I can understand how it works- a first for ignorant me!!
It looks like Nelson has more to reveal. I'm being patient- that's the least I can do. If Nelson doesn't want to sell boards, for the conversion, I will try to organize a group buy. I love the idea of all those HAflers getting a second life!
We hope that we can devise a way that this can be done so easily that thousands of these old amps will get a new lease on life based on their newly aquired excellent sound quality. To facilitate this the ground rules are to use as many of the existing parts as possibly, so that the time and expense of the upgrade will be minimal, and that newbies won't be scared off.
Eventually we hope to have a group buy of boards that can be used in amps built from scratch and to accomplish the above retrofits.
A parts buy would be fun too- at least for those of us that get the parts!! Yet the number of required parts will be pretty low..
Anyway- let's get on with it!!
Response to PSz from post above:
That's still a mystery! I'm not sure what voltage the power supply on the Hafler puts out, but if it is around 40 , then it should work. The HAfler supply isn't balanced so again, it makes sense that the Plh will be OK. My biggest question is whether the voltage that the HAfler supplies will overwhelm the heatsinks.....Possiby each amp should be converted into a monoblock...
Love the circuit. I have myself convinced that I can understand how it works- a first for ignorant me!!
It looks like Nelson has more to reveal. I'm being patient- that's the least I can do. If Nelson doesn't want to sell boards, for the conversion, I will try to organize a group buy. I love the idea of all those HAflers getting a second life!
Kari,
let's coordinate on this! It would be great to wait find out if Nelson has made it work in the Hafler. If so, it would be REALLY great to make the circuit boards a size and shape to just bolt into the amp.They would then work with HAfler upgrades and also for people that are building a complete amp. I don't think there is any disadvantage to this other than that they might be a little bigger than you might normally make them, I believe that each channel is about 110mm square with a hole in each corner. The output transistors are separately mounted on the heatsinks. I can't check, because ....I don't have the amp here....
If it turns out that the boards are a bit big, maybe you could include places for output transistors along a side that could be used by those making an amp from scratch. The HAfler peole could just run wires from those holes to the heatsinks.
If you don't wan't too, of course that's fine also...
Mark
Still waiting on the BOSOX boards - What's going on with the post office!!
let's coordinate on this! It would be great to wait find out if Nelson has made it work in the Hafler. If so, it would be REALLY great to make the circuit boards a size and shape to just bolt into the amp.They would then work with HAfler upgrades and also for people that are building a complete amp. I don't think there is any disadvantage to this other than that they might be a little bigger than you might normally make them, I believe that each channel is about 110mm square with a hole in each corner. The output transistors are separately mounted on the heatsinks. I can't check, because ....I don't have the amp here....
If it turns out that the boards are a bit big, maybe you could include places for output transistors along a side that could be used by those making an amp from scratch. The HAfler peole could just run wires from those holes to the heatsinks.
If you don't wan't too, of course that's fine also...
Mark
Still waiting on the BOSOX boards - What's going on with the post office!!
If Kari were involved I think we could pull it off.
Just to get us going though, it would be handy to know the dimensions in inches of the boards, the distance and diameters of the holes to attach them to the brackets, and the output of the power supply and/or the transformer.
Also your opinion as to whether that voltage would work, and if we can send enough bias through the thing with the existing heatsinks.
Anyone else with a Hafler can chime in here on the measurements..
Ironically, I don't have one to measure
Mark
P.S. What do you think Kari?
Just to get us going though, it would be handy to know the dimensions in inches of the boards, the distance and diameters of the holes to attach them to the brackets, and the output of the power supply and/or the transformer.
Also your opinion as to whether that voltage would work, and if we can send enough bias through the thing with the existing heatsinks.
Anyone else with a Hafler can chime in here on the measurements..
Ironically, I don't have one to measure
Mark
P.S. What do you think Kari?
If I remember corectly-DH200 rides with 2x60 to 63 volts DC-means that we need to waste prety much of voltage,even if we count on discrepancy between 1,41 and 1,2 factors for ac to dc........
on the other hand-with just 2 x 63V x 2 x 0,1A per channel-DH 200 was pretty hot in summer days ........
different xformer - some rewind job ?
slow fan on each side?
in any case-according to my (modest) inteligence and counting on that that I didn't really measured heat transfer for DH200 heatsinks, looks that heatsinks are good enough for mebbe 50W max (each) .
I don't have DH200 any more ,so someone else must grab few power resistors etc.
on the other hand-with just 2 x 63V x 2 x 0,1A per channel-DH 200 was pretty hot in summer days ........
different xformer - some rewind job ?
slow fan on each side?
in any case-according to my (modest) inteligence and counting on that that I didn't really measured heat transfer for DH200 heatsinks, looks that heatsinks are good enough for mebbe 50W max (each) .
I don't have DH200 any more ,so someone else must grab few power resistors etc.
Perhaps one could play with the idea that the Hafler transformer is center tapped.
Oh! and with all the extra pcb real estate in a Hafler it would be relatively easy to include a capacitance multiplier circuit for the psu... then we could use an xl280 chassis for the quad output device version with two to-3 spaces left for psu transistors.
I am dreaming of a PLH, Hafler chassis, without the in/out coupling caps... perhaps someday. First, I need to get my mini-a to stop humming.
regards,
PSz.
Oh! and with all the extra pcb real estate in a Hafler it would be relatively easy to include a capacitance multiplier circuit for the psu... then we could use an xl280 chassis for the quad output device version with two to-3 spaces left for psu transistors.
I am dreaming of a PLH, Hafler chassis, without the in/out coupling caps... perhaps someday. First, I need to get my mini-a to stop humming.
regards,
PSz.
My self imposed rules for the Hafler conversion was that we had to use the existing power transformer. My memory corresponds with yours, that the transformer is approx 60 volts. That's pretty hiigh, but it would be interesting to hear from Nelson if he ever had a plan to use this voltage. If we have to dump the transformer we have to, but then we are only using the box and heatsinks.... and I guess the old MOSFETS?
Psz. ,
IS the transformer center tapped? I don't remember, I just threw the amp in the mail , forgetting to look it over first..
It would make a pretty sensational magazine article Nelson!!
probably double AudioXpress sales that month!!
Psz. ,
IS the transformer center tapped? I don't remember, I just threw the amp in the mail , forgetting to look it over first..
It would make a pretty sensational magazine article Nelson!!
probably double AudioXpress sales that month!!
A funny thing happened on my way to doing a rebuild-the-Hafler circuit. I had a half-dead DH-200 on hand. Being a resourceful lad, I figured that I would buy another one-channel-dead unit, combine the two working channels into one functioning DH-200 and use the wholly dead one for a start on some ideas I've got.
Great plan, eh?
Okay, so I scored a DH-200. It came in. I took a look inside. Well, some idjit had modified the damned thing and shorted the input to the "dead" channel with a big glob of solder. I couldn't stand it...I removed the solder and Presto!, the amp sprang to life. I trimmed the bias and put the thing to work in my AV system (say what you will about old Haflers, but they're a biiiiig step up from the inboard amplification on my Pioneer Elite receiver).
All right, so there I sat with the half-dead amp. Nothing for it but to try again. Nailed another DH-200. Well, this one came in and...you guessed it...some merde-for-brains had put weirdo metal can caps in the inputs, replacing the stock 10uF NP electrolytic DC blocker. So far, so good. The problem was that these big, honkin' cans the size of my thumb were just flopping around on long braided metal leads. What was the guy thinking? The confounded things were shorting out everything in sight. You know what happened next. I removed the caps, put in some 10uF film caps that were lying on the bench, and Bingo!, the amp woke up. So that one went into the AV system, too.
Third time's the charm, right? Got another. This one had been rode hard and put to bed wet, as the expression goes. Used in a PA system. Flogged within an inch of its life. Dead, right? Wrong. That one took a whole hour to bring up, but it wasn't all that hard. Golly. It seemed a shame not to put it into the AV system with its relatives, so it went on the subs (I run 5.2--yes, the subs play the same signal, but there are two of them).
The original one-channel-dead amp? I made it the front center channel amp. Seemed the only decent thing to do.
So now I've got this big ol' stack of Haflers in the AV rack and no spare dead channel Haflers to play with. Don't worry, I've got a one-dead-channel Kinergetics amp that will serve nicely for a test bed, not to mention a well stocked junk box.
Yes, the Hafler rails run about +-60V stock. That's at stock bias of .25A per side. Nelson suggested that the rail would come down nicely when subjected to a 2A load per channel, not to mention the loss through the resistor in a CRC power supply. I haven't tried it, but I'm not sure whether the stock domestic transformer can take that. I suspect it might run a little hot. The chassis isn't ventillated all that well, either. It might be a good idea to drill out the ventillation holes to increase air flow through the interior. Better yet, just remove the top entirely. Don't forget fans on the heat sinks. This is going to be a heavy mod.
Someone try it. If it works, great. If it doesn't, great. Report back here, either way. If it's too much for the Hafler, perhaps Nelson will suggest another circuit. If he doesn't have time, I'll try to scrounge a little time and make a run at a class AB circuit to fit into a Hafler chassis.
In the meantime, I've still got an eye out for a replacement channel for the original Hafler that started the whole thing.
Grey
P.S.: I mentioned my Hafler saga to Nelson. He seemed a bit miffed that I didn't go ahead and junk the innards and build my circuit. Sooner or later I will probably do so, but for now I figure it's the will of the Audio Gods that those amps earn their keep the way they are. Saved me a whole lot of time and time is something I ain't got enough of. Besides, it was an easy, relatively cheap fix for the sound in my AV system. It's absolutely amazing what AV folks put up with in terms of audio quality. (Don't get me started. It's all part of the "oh, hey, I can hear the words so it's all good" mentality. MP3? Oh, puh-leeze! Once upon a time, people actually strove for quality...grumble, grumble, grumble...)
P.P.S.: What does three for three tell you about the actual number of so-called dead Haflers out there? In all three cases, no semiconductors were blown. No passive parts were blown. (Well, okay, I had to replace a broken fuse holder and a power switch...big deal.) Just some poorly conceived and executed mods that needed to be undone. The amps hadn't died on their own--they'd been murdered by their owners. Sad.
Great plan, eh?
Okay, so I scored a DH-200. It came in. I took a look inside. Well, some idjit had modified the damned thing and shorted the input to the "dead" channel with a big glob of solder. I couldn't stand it...I removed the solder and Presto!, the amp sprang to life. I trimmed the bias and put the thing to work in my AV system (say what you will about old Haflers, but they're a biiiiig step up from the inboard amplification on my Pioneer Elite receiver).
All right, so there I sat with the half-dead amp. Nothing for it but to try again. Nailed another DH-200. Well, this one came in and...you guessed it...some merde-for-brains had put weirdo metal can caps in the inputs, replacing the stock 10uF NP electrolytic DC blocker. So far, so good. The problem was that these big, honkin' cans the size of my thumb were just flopping around on long braided metal leads. What was the guy thinking? The confounded things were shorting out everything in sight. You know what happened next. I removed the caps, put in some 10uF film caps that were lying on the bench, and Bingo!, the amp woke up. So that one went into the AV system, too.
Third time's the charm, right? Got another. This one had been rode hard and put to bed wet, as the expression goes. Used in a PA system. Flogged within an inch of its life. Dead, right? Wrong. That one took a whole hour to bring up, but it wasn't all that hard. Golly. It seemed a shame not to put it into the AV system with its relatives, so it went on the subs (I run 5.2--yes, the subs play the same signal, but there are two of them).
The original one-channel-dead amp? I made it the front center channel amp. Seemed the only decent thing to do.
So now I've got this big ol' stack of Haflers in the AV rack and no spare dead channel Haflers to play with. Don't worry, I've got a one-dead-channel Kinergetics amp that will serve nicely for a test bed, not to mention a well stocked junk box.
Yes, the Hafler rails run about +-60V stock. That's at stock bias of .25A per side. Nelson suggested that the rail would come down nicely when subjected to a 2A load per channel, not to mention the loss through the resistor in a CRC power supply. I haven't tried it, but I'm not sure whether the stock domestic transformer can take that. I suspect it might run a little hot. The chassis isn't ventillated all that well, either. It might be a good idea to drill out the ventillation holes to increase air flow through the interior. Better yet, just remove the top entirely. Don't forget fans on the heat sinks. This is going to be a heavy mod.
Someone try it. If it works, great. If it doesn't, great. Report back here, either way. If it's too much for the Hafler, perhaps Nelson will suggest another circuit. If he doesn't have time, I'll try to scrounge a little time and make a run at a class AB circuit to fit into a Hafler chassis.
In the meantime, I've still got an eye out for a replacement channel for the original Hafler that started the whole thing.
Grey
P.S.: I mentioned my Hafler saga to Nelson. He seemed a bit miffed that I didn't go ahead and junk the innards and build my circuit. Sooner or later I will probably do so, but for now I figure it's the will of the Audio Gods that those amps earn their keep the way they are. Saved me a whole lot of time and time is something I ain't got enough of. Besides, it was an easy, relatively cheap fix for the sound in my AV system. It's absolutely amazing what AV folks put up with in terms of audio quality. (Don't get me started. It's all part of the "oh, hey, I can hear the words so it's all good" mentality. MP3? Oh, puh-leeze! Once upon a time, people actually strove for quality...grumble, grumble, grumble...)
P.P.S.: What does three for three tell you about the actual number of so-called dead Haflers out there? In all three cases, no semiconductors were blown. No passive parts were blown. (Well, okay, I had to replace a broken fuse holder and a power switch...big deal.) Just some poorly conceived and executed mods that needed to be undone. The amps hadn't died on their own--they'd been murdered by their owners. Sad.
If the transformer output is 60 volts, we could do the LC-filter-after-the-diodes-trick to get the voltage down to 54. Then it would prob sag to less than 50 , then through a resistor...LRC.... maybe we're getting there.... There was insinuation on another thread that an "L" filter like this could be unstable though.....
I need opinions on that....
If the RAILS are 60 volts, then we are talking more like 45 volts for the transformer...right?I which case things could come out fine also.
Which is it?
It seems that using one chassis per channel might be tempting- more heatsink, half the load on the transformer, but I'm not sure I want to buy another!
I need opinions on that....
If the RAILS are 60 volts, then we are talking more like 45 volts for the transformer...right?I which case things could come out fine also.
Which is it?
It seems that using one chassis per channel might be tempting- more heatsink, half the load on the transformer, but I'm not sure I want to buy another!
If I have some $ to spare and like to use the amp long, I would buy a new transformer of the right size. I see no reason to unnecessarily waste power by installing a reduction gear for long time. Buying new transformer could be more economical in long term. Yeah . . . of course the final decision depends on . . .
Regards
jH
Regards
jH
Variac said:If Kari were involved I think we could pull it off.
Just to get us going though, it would be handy to know the dimensions in inches of the boards, the distance and diameters of the holes to attach them to the brackets, and the output of the power supply and/or the transformer.
Also your opinion as to whether that voltage would work, and if we can send enough bias through the thing with the existing heatsinks.
Anyone else with a Hafler can chime in here on the measurements..
Ironically, I don't have one to measure
Mark
P.S. What do you think Kari?
Sure i'm up to it. As you said we will need the dimension of the board, either in inches or mm, doesn't matter to me. A picture of the Hafler board would also be of some help.
It should be pretty easy to define the pcb borders in Protel.
/Kari
GRollins said:
I would not be miffed, but I do have high expectations of you.
Regarding the transformer issue, no problem. We can dual 1/2
wave rectify the output to produce twice the current at the 60
volts. The transformer will load down under a 4 amp load, so
there goes some of that voltage. Thermistors and/or coils on
the primary and secondary windings will take it down further
with little dissipation. CRC will drop a few more volts. Also, the
circuit loves a regulated supply, so there's some more drop.
Remember that there are probably just as many Adcom GFA555's
out there as Haflers, and I think that's probably a good choice
also.
A follow up article for the Hafler and the Adcom would probably
be very cool. Same circuit, but with step-by-step instructions
for each of these. I may have to let loose of some TO-3 parts
so you guys can build them, unless someone has some more
sources.
There's a good chance that we will want to be using a fan - I
have a wonderful fan controller circuit that makes them very
quiet.
/pass/: Should be working on time machine instead.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.