Unclejed613, PB2, mikeAtx:
did you guys ever do anything else with your sims of the APT 1 amp discussed here: https://www.diyaudio.com/community/threads/apt-1-schematic-and-simulation-from-unclejed613.347851/ ?
did you guys ever do anything else with your sims of the APT 1 amp discussed here: https://www.diyaudio.com/community/threads/apt-1-schematic-and-simulation-from-unclejed613.347851/ ?
Not me. I got my apt-1 running again after replacing the front end devices and a few other parts. I reread the thread and don't remember doing some of the sims it has been so long. The only thing I've done system wise is switch to a soft-start circuit for my amps to stop the lights from dimming.
Thanks. I miss mine; lost to a lightning strike many years ago. I occasionally think about build a clone to replace it. Every year I get a little closer to doing just that ...
🙂
🙂
I have a soft spot too. Bought mine in 82 I think. I think if I were to redo it about all I would do is make it a bit taller. It is tight to work on. Almost like they decided the box size before they started squeezing stuff in. A mere 1/2 or even 1/4 inch would have made it much easier.
Why not just buy a used one and service/upgrade it? I've done just that to our three units. New output transistors, film caps, electrolytics, trimmers, bypass caps, and other odds and ends. Certainly less expensive than designing and building a clone, which would be difficult based on the lack of availability of many of the smaller transistor types.
This photo is while setting the impedance warning light, in preparation for conducting distortion/function testing.
This photo is while setting the impedance warning light, in preparation for conducting distortion/function testing.
@mlloyd1 No, I never did anything more with this. It would be nice if someone did circuit boards for this amp,
but of course I have to ask what design provides the best performance most economically. If anyone makes a
board for this it would be nice to have compensation options for TMC, and others.
I have a couple questions about doing a DIY variation on this design:
1. If we replaced the diff pair with JFETs is there another commercial or DIY design that
has already done this, that is otherwise almost the same topology?
2. The multiple power supplies is overly complex IMO, but if the output stage is replaced
with the Bryston with say a gain of 3-5 the front end could run on either the same or
lower voltage than the output stage.
3. It would be interesting to explore other compensation schemes such as MIC, TMC or others.
I think that the combination of all of these might make for a very interesting design.
Not as a mod but as a new DIY design.
1. If we replaced the diff pair with JFETs is there another commercial or DIY design that
has already done this, that is otherwise almost the same topology?
2. The multiple power supplies is overly complex IMO, but if the output stage is replaced
with the Bryston with say a gain of 3-5 the front end could run on either the same or
lower voltage than the output stage.
3. It would be interesting to explore other compensation schemes such as MIC, TMC or others.
I think that the combination of all of these might make for a very interesting design.
Not as a mod but as a new DIY design.
PB2:
There's been 3 that I can remember off the top of my head: Marantz M510 from the mid 70s (not one to emulate, I think), Pioneer M-90 and Luxman M120A. The Luxman was interesting in that it was a lot more similar and included a cascoded JFET front end and a Hawksford (?) cascode VAS . But they didn't use the staggered rails nor the Baker Clamp, which I always thought were critical elements for the APT1 performance. In my opinon, they didn't try anything very different with compensation, either. And they actually hung resistors to ground on the VAS outputs to drop the open loop gain!
I'd love a "new" design APT1 with JFETs in front, the same thoughtfulness throughout the design as in the original and one of the more modern compensation schemes. I think it would be very interesting and I was really impressed when Tom appeared here earlier with those great test results of an original unit. I've butchered somebody's LT Spice file (was it you or Uncle Jed that made the original?) to play around with something I'll eventually build.
So far, Ostripper has been the only person I've seen around here who's done much with a new design of the topology, but he didn't include some of the previously mentioned touches either. But his results were very, very good and he really liked it!
You make an interesting point about output stages with voltage gain. I always like the old Boulder amps. Their literature always sang the praises of output stages with gain and implied that other manufactureres might not have been "smart" enough to do the math and analysis to get all the compensations right to successfully deploy output stages with gain. It says a lot that those amps were popular for studios and musical acts on the road (they sounded good and were robust & reliable). I don't know if they still employ this approach in their "modern" products. I thought the model they used of continuously perfecting 990-style discrete op amps to use as the primary voltage gain stage for everything was a reasonably smart way to be successful.
mlloyd1
There's been 3 that I can remember off the top of my head: Marantz M510 from the mid 70s (not one to emulate, I think), Pioneer M-90 and Luxman M120A. The Luxman was interesting in that it was a lot more similar and included a cascoded JFET front end and a Hawksford (?) cascode VAS . But they didn't use the staggered rails nor the Baker Clamp, which I always thought were critical elements for the APT1 performance. In my opinon, they didn't try anything very different with compensation, either. And they actually hung resistors to ground on the VAS outputs to drop the open loop gain!
I'd love a "new" design APT1 with JFETs in front, the same thoughtfulness throughout the design as in the original and one of the more modern compensation schemes. I think it would be very interesting and I was really impressed when Tom appeared here earlier with those great test results of an original unit. I've butchered somebody's LT Spice file (was it you or Uncle Jed that made the original?) to play around with something I'll eventually build.
So far, Ostripper has been the only person I've seen around here who's done much with a new design of the topology, but he didn't include some of the previously mentioned touches either. But his results were very, very good and he really liked it!
You make an interesting point about output stages with voltage gain. I always like the old Boulder amps. Their literature always sang the praises of output stages with gain and implied that other manufactureres might not have been "smart" enough to do the math and analysis to get all the compensations right to successfully deploy output stages with gain. It says a lot that those amps were popular for studios and musical acts on the road (they sounded good and were robust & reliable). I don't know if they still employ this approach in their "modern" products. I thought the model they used of continuously perfecting 990-style discrete op amps to use as the primary voltage gain stage for everything was a reasonably smart way to be successful.
mlloyd1
I've heard about the Boulder amps using the JE990 as the front end but never saw a schematic
and didn't know that they had voltage gain in the output stage - interesting.
I'd like to see a schematic if anyone has one.
Yes, now I remember the talk about the Luxman and I wonder if the lower supply voltage on the
front end avoids saturation of anything in the output stage so that the Baker clamp is not needed.
I also remember the Marantz design and interesting that it has the roots of this topology but is
rather primitive.
I'm not familiar with the Pioneer, have to take a look at it.
I don't think that I've mentioned here that what I like about this design is that it avoids a comp diff
front end and the problem of matching devices, yet provides for a push-pull Vas by way of the mirror.
The push-pull VAS can provide symmetrical slewing and stronger drive to the output stage for very
low impedance loads.
and didn't know that they had voltage gain in the output stage - interesting.
I'd like to see a schematic if anyone has one.
Yes, now I remember the talk about the Luxman and I wonder if the lower supply voltage on the
front end avoids saturation of anything in the output stage so that the Baker clamp is not needed.
I also remember the Marantz design and interesting that it has the roots of this topology but is
rather primitive.
I'm not familiar with the Pioneer, have to take a look at it.
I don't think that I've mentioned here that what I like about this design is that it avoids a comp diff
front end and the problem of matching devices, yet provides for a push-pull Vas by way of the mirror.
The push-pull VAS can provide symmetrical slewing and stronger drive to the output stage for very
low impedance loads.
That is part of my attraction to it as well. Also attractive was Bob Cordell's comments about it's non-fighting VAS / compensation, and OS' comments about lack of 7th harmonic in his version of the topology as compared to others.
Just wish I had more time to play. If only retirement would get here faster...
Just wish I had more time to play. If only retirement would get here faster...
It would be cool to use the Sanken MT200 outputs but as far as I know they are NLA.
Valery (RIP) said that he liked these for their ability to handle HF with minimal cross conduction:
https://www.diyaudio.com/community/threads/valery-zaichenko.383857/
https://www.diyaudio.com/community/threads/improved-op.251278/
https://www.diyaudio.com/community/...-for-super-complementary-output-stage.306377/
Valery wrote but I don't see why, if saturation is avoided that it would be any worse than other output stages:
"Choosing the output transistors for this topology, make sure they are fast enough - so called RET transistors are what you'd want to look at (fT = 50MHz or so).
Sankens, like 2SC2922/A1216 or the newer ones 2SC3264/A1295, or NJW3281/1302 - would be a good choice. Transistors like MJ21193/21194 (fT = 4MHz) will most likely kill themselves in this arrangement - they're too slow while closing."
Valery (RIP) said that he liked these for their ability to handle HF with minimal cross conduction:
https://www.diyaudio.com/community/threads/valery-zaichenko.383857/
https://www.diyaudio.com/community/threads/improved-op.251278/
https://www.diyaudio.com/community/...-for-super-complementary-output-stage.306377/
Valery wrote but I don't see why, if saturation is avoided that it would be any worse than other output stages:
"Choosing the output transistors for this topology, make sure they are fast enough - so called RET transistors are what you'd want to look at (fT = 50MHz or so).
Sankens, like 2SC2922/A1216 or the newer ones 2SC3264/A1295, or NJW3281/1302 - would be a good choice. Transistors like MJ21193/21194 (fT = 4MHz) will most likely kill themselves in this arrangement - they're too slow while closing."
Last edited: