New to this site so may be in wrong area so happy to be redirected. Any one have any opinions of Doug Self's power amp? Im thinking of building the trimodal version with a few mods.
I'm not familiar with this particular amp but try searching the solid state forum. I know the name Self shows up in many threads. 🙂
I'll move your topic over there while I'm at it.
/Hugo
I'll move your topic over there while I'm at it.
/Hugo
Some of Randy Slone's amps, particularly those with a single LTP and a BJT output appear to be derivative. His "11.4" which refers to the illustration number in his book has been built by several here, including me. The general opinion seems to be positive. On the other hand there have been some unkind words about Self's "blameless"
The Trimodal should not be confused with the "Blameless". My take on the "Blameless" is that it isn't intended to be synonymous with "perfect", "ultimate", "deal" etc but rather a fairly basic design that avoids anything really dumb. He uses it as a baseline for exploring more exotic ideas. The Trimodal is one of those. I've never read of anyone actually building it. There seems to be something of a philosophical chasm between Class A folks and Class B folks. The chasm is filled with much smoke, noise and debate about the definition of Class AB. I suspect the both faction view the Trimodal as an abomination, i.e., a crime against nature.
That said, I think you should go ahead and build. It probably will sound much like his "Load Invariant", which is pretty much what the "11.4" mentioned above is. But keep in mind Self appears to be of the school that believes all good amplifiers used within their capabilities sound alike, and if it should have a unique sound there is something wrong with it.
I think you would generate a lot of interest if you wcould report back here on the results.
The Trimodal should not be confused with the "Blameless". My take on the "Blameless" is that it isn't intended to be synonymous with "perfect", "ultimate", "deal" etc but rather a fairly basic design that avoids anything really dumb. He uses it as a baseline for exploring more exotic ideas. The Trimodal is one of those. I've never read of anyone actually building it. There seems to be something of a philosophical chasm between Class A folks and Class B folks. The chasm is filled with much smoke, noise and debate about the definition of Class AB. I suspect the both faction view the Trimodal as an abomination, i.e., a crime against nature.
That said, I think you should go ahead and build. It probably will sound much like his "Load Invariant", which is pretty much what the "11.4" mentioned above is. But keep in mind Self appears to be of the school that believes all good amplifiers used within their capabilities sound alike, and if it should have a unique sound there is something wrong with it.
I think you would generate a lot of interest if you wcould report back here on the results.
Self's book is well worth the read, some of the stuff he points out in the blameless design isn't totally obvious or dumb.
I have built the blameless and other derivatives of my own design. It's a great base to work on if you have some tricks up your sleeve; and most of all, a wonderful thing on which to learn the basic theory of multi-stage amplifier concepts.
The results as far as sound goes, in my opinion, are pretty nice with the blameless and my variants. Very clean, high damping factor, fast, and low on the distortion. There can be issues with stability in them if not built or compensated properly, but otherwise, a good layout.
As far as voltage amplifiers go, the blameless concept and its variants are some of my favorites for work-every-time simple amplifiers of high performance. For friends I almost always build something like that if they aren't looking for fancy or expensive.
I do also have a spot in my heart for the world of variable output impedance and transconductance amps, and my adaptations of the blameless design itself were successful in transconductance mode, but sometimes very hard to stabilize how I would like.
The results as far as sound goes, in my opinion, are pretty nice with the blameless and my variants. Very clean, high damping factor, fast, and low on the distortion. There can be issues with stability in them if not built or compensated properly, but otherwise, a good layout.
As far as voltage amplifiers go, the blameless concept and its variants are some of my favorites for work-every-time simple amplifiers of high performance. For friends I almost always build something like that if they aren't looking for fancy or expensive.
I do also have a spot in my heart for the world of variable output impedance and transconductance amps, and my adaptations of the blameless design itself were successful in transconductance mode, but sometimes very hard to stabilize how I would like.
The biggest worry for me in Self's designs (and Slone's) is the CFB output configuration. When it works, its great but they are prone to oscillation. The difficulty is that they are sensitive to the layout of the PCB. With multiple output devices the difficulties increase. If there is set of rules to make is stable I'm unaware. Trail and error seems to be the only approach. This isn't to say one should avoid that configuration but rather not to despair if it the first attempt doesn't quite work as you wish.
Yes I have read D Self's books and was very impressed. He does not position himself as an audiophile but as an informed analogue engineer who really understands subtle 2nd and 3rd order effects in the electonics domain. My take is that he makes no claims for the 'sound' only that what comes out is to the maximum extent an amplified version of what went in where THD+N is the measure of imperfection with suitable qualification on the audability of lower versus higher spectral components etc. Given that it looks like a good baseline to work from I will give it a go. There is a company that makes the PCBs. This has been on my 'to do' list for more than a year so may be a while until I get around to it.
As an aside: I was amused by Self's suggestion the that Class A / Class B switch be labeled Winter/ Summer. I once spent a couple of summer weeks in the Pennine region of Yorkshire. I could have used a nice Class-A amp to drape my soggy clothes over each evening while planning the next day's excursion.
On the other hand, where I now reside it's been 40+C daily for a month and will stay that way until late September so Class is unappealing.
On the other hand, where I now reside it's been 40+C daily for a month and will stay that way until late September so Class is unappealing.
Pulling your leg Sam.
Long time ago i spent 4 continuous drousy wet backpacking weeks in the Lake District btw.
Also tossed a bag of oldfashioned Brit pound coins in a laundry machine in every village crossed, a day later everything smelled again like there was an alien fungus living in the backpack.
If i ever go England camping again, a big soab Class A amp will hop along, including the generator if needed.
Long time ago i spent 4 continuous drousy wet backpacking weeks in the Lake District btw.
Also tossed a bag of oldfashioned Brit pound coins in a laundry machine in every village crossed, a day later everything smelled again like there was an alien fungus living in the backpack.
If i ever go England camping again, a big soab Class A amp will hop along, including the generator if needed.
sam9
15W /Ch. Class A is more than adequate for MOST domestic listening situations.Even then, most of the time it will be just cruising, unless you have very inefficient speakers.. A good Class A amplifier will normally have a better soundstage with vastly improved ambience. A total power dissipation of 100W isn't going to do very much warming in a room. You can always use a much higher powered amplifier for those occasions when you have a party, or if you have a very large house.
SandyK
15W /Ch. Class A is more than adequate for MOST domestic listening situations.Even then, most of the time it will be just cruising, unless you have very inefficient speakers.. A good Class A amplifier will normally have a better soundstage with vastly improved ambience. A total power dissipation of 100W isn't going to do very much warming in a room. You can always use a much higher powered amplifier for those occasions when you have a party, or if you have a very large house.
SandyK
The design I have in mind will have adjustable rails and adjustable class A bias that track the volume pot (by a simple extra pot ganged with it in the pre amp). Idea is that it runs cooler when used quietly but can be turned up (subject to overall ratings, heatsinking etc) when required. Needs some thinking about especially the tracking power supplies (I will probbaly make them extra quiet switch mode mag amp designs).
Nice concept, however you'd be swapping dissipation in the outputs for wasted heat in the regulator.
What about a commutating supply, using a series of taps on the power transformer, and switching between them?
This would cost much more for the trafo, certainly, but switching between the various taps could hopefully be done soundlessly with considerable improvement in efficiency....
The switching could be triggered from thresholds built into the volume control, as you suggested.
Cheers,
Hugh
What about a commutating supply, using a series of taps on the power transformer, and switching between them?
This would cost much more for the trafo, certainly, but switching between the various taps could hopefully be done soundlessly with considerable improvement in efficiency....
The switching could be triggered from thresholds built into the volume control, as you suggested.
Cheers,
Hugh
hm, Variable SMPS regulator? Could be a good idea. Just, you would want an adjustable 'overhead' pot as well, which adjusted the offset between volume and power rail controls. The reason is that not all music is recorded at equal volume and some things will invariably clip at all volumes...
Some ideas anyway.
Some ideas anyway.
yes does need some careful thought to avoid clipping and also the use of 'noisey' switched mode power supplies also cuts across many audio conventions however I used to design very low noise SMPSs for the military industry (a few years ago now I have to admit) so have some confidence that this can be done. It would probably be a fixed duty cycle forward converter at about 100Khz with mag amp post regulators for the voltage tracking. SMPS supplies also have other benefits such as small storage caps hence minimised fault currents and the SMPS itself can detect and limit or shutdown on excess current or detected output offset voltage. Its all still at the thinking stage. Adrian S
- Status
- Not open for further replies.