It has already been published, see here:
Old-fashioned amplifiers with a new twist (2)
For good measure, here is the EZmos as well:
Easy-MOS is a simplistic, efficient and evolutive all-MOS amplifier for beginners
If you need an even lower THD combined with zero adjustment, the Circlophone has been published (somewhere!?) in a single supply version (and it is tested, it is one of the amps I actually use).
If you have enough patience to browse through the innumerable posts of the main and builders threads, you should be able to locate it.
The C is not a super-amp having sub-ppm performances, but it does quite well, both in specs and in listening tests
Old-fashioned amplifiers with a new twist (2)
For good measure, here is the EZmos as well:
Easy-MOS is a simplistic, efficient and evolutive all-MOS amplifier for beginners
If you need an even lower THD combined with zero adjustment, the Circlophone has been published (somewhere!?) in a single supply version (and it is tested, it is one of the amps I actually use).
If you have enough patience to browse through the innumerable posts of the main and builders threads, you should be able to locate it.
The C is not a super-amp having sub-ppm performances, but it does quite well, both in specs and in listening tests
Looking at this thread, I note the quasi version and I'm thinking a PNP VAS would be able to provide more current for the positive half cycle. So, I simulated it and it worked but the THD was not great, So I added a NPN driver and the THD is much better. This also shows my favorite single supply approach. Two channels could share the supply. Note I only used junk box transistors. Supply could be higher, but I think 40V would be the limit.
Attachments
Reference post 63 2n3904/3906 as drivers would be safe at about 2 watts/ch driving 5 gain 2n3055. Unless you found original TO18 metal cans versions and & used heatsinks. My dynakit ST120 had private label 2n5320/22 in TO5 cans with heatsinks on them. The 50 mhz ft was required to reproduce tinkly bells & top octave piano, by failed experiment. Original drivers sounded good on NTE60 outputs, (MJ15003?) TIP31c/32c with 6 mhz ft sounded bad. MJE15028/29 restored top octave.
Don't see the schottky diode of the "twist" either in post 31.
Don't see the schottky diode of the "twist" either in post 31.
I think he just put those transistor types in for simulation purposes, not that you’d use them to drive 2N3055’s. Driving TIP41/2 on 25 volt supply, possibly. But I’d still use MPSA06/56 - still a junk box transistor.
Yeah, adding the extra transistor cuts down on distortion added by the asymmetrical quasi. They do that on ICs all the time and get away with it (LM386). It’s no big deal to lose another six tenths on a 40 volt supply, but on 12 (or 6) it IS. I think rail to rail design these days uses CFP-with-gain, using mosfets.
It doesn’t take golden ears to hear the difference that using sustained-beta types for DRIVERS makes in typical “normal” amplifiers, even if you’re using RCA hometaxial outputs. It is the most critical position in the amp IMO, even before getting OCD with a 2pF VAS. When I see someone using TIP 31/2 as drivers for c5200’s I cringe.
Yeah, adding the extra transistor cuts down on distortion added by the asymmetrical quasi. They do that on ICs all the time and get away with it (LM386). It’s no big deal to lose another six tenths on a 40 volt supply, but on 12 (or 6) it IS. I think rail to rail design these days uses CFP-with-gain, using mosfets.
It doesn’t take golden ears to hear the difference that using sustained-beta types for DRIVERS makes in typical “normal” amplifiers, even if you’re using RCA hometaxial outputs. It is the most critical position in the amp IMO, even before getting OCD with a 2pF VAS. When I see someone using TIP 31/2 as drivers for c5200’s I cringe.
Ya, 06/56 are better suited for drivers but the 3055 data sheet saz the gain at 12V/8 Ohms=1.5 Amps is about 50, not 5. Simulation saz the peak driver current is 45mA.
I simulated a version that squeezes a bit more +V swing by partial boot strapping the VAS. But this is a small cheap amplifier, 20W if you push it to 40V. If you want power, then you need to spend money on the parts. And there are chip amps in this power range with similar performance and better protection.
The point of that post was the auto bias without using a Schottky diode and the single supply topology. It's interesting to add improvements and see if they actually help (much). Mostly, they don't. Adding more transistors and gain quickly leads to stability problems, etc. Adding more transistors does deal with supply voltage dependencies.
TIP31/32 are old and slow. I never had much use for them even 40 years ago.
The other point of that post is that LV and Minek have been quiet so I'm bored.
I simulated a version that squeezes a bit more +V swing by partial boot strapping the VAS. But this is a small cheap amplifier, 20W if you push it to 40V. If you want power, then you need to spend money on the parts. And there are chip amps in this power range with similar performance and better protection.
The point of that post was the auto bias without using a Schottky diode and the single supply topology. It's interesting to add improvements and see if they actually help (much). Mostly, they don't. Adding more transistors and gain quickly leads to stability problems, etc. Adding more transistors does deal with supply voltage dependencies.
TIP31/32 are old and slow. I never had much use for them even 40 years ago.
The other point of that post is that LV and Minek have been quiet so I'm bored.