And most probably ultra unstable too, Ive dug around, this design by apex is a clone of a yamaha amp, <snip>
Do some sims, and youll find it impossible to stabilise no matter how you try to compensate it.
I use a greatly simplified version of this amp ah-la LF351 (circa 1975 so it came before the Yamahaha), a very good sounding op-amp, which is the most stable amp I have ever seen by a country mile. That power version of the LF351 will drive any load I throw at it including using the second channel as a virtual ground for the first channel through any kind of load down to the transistors SOA limit. Mine has a parts count of 35 per channel not including the power supplies. So I really do not believe the basic idea is incorrect. Just the implementation.

U can find similar design in Pioneer, InterM, Zeck, Yamaha... amplifiers. Yamaha use this design in all amplifiers in last 10 years, not just for subwoofer amps. Do some sims of newer Yamaha (AX-596), and youll find it impossible to stabilise.And most probably ultra unstable too, Ive dug around, this design by apex is a clone of a yamaha amp, only that yamaha used it as a subwoofer amp, thats why you see compensation done like that 330pf cap across the feedback resistor, there was no intention for this amp to operate to normal frequencies required for full range amp. From what I can read about the yamaha amp is that it was discontinued prematurely because of problems, I can guess what those were......
Do some sims, and youll find it impossible to stabilise no matter how you try to compensate it.
Regards
Thanks, but if I use a single transistors for VAS many others people will be more happy.What I really like about this amp is usage of Baxandall Super-pair transistors for VAS transistors...ultra fast, ultra high output impedance, ultra low output capacitance...nice!
Regards
This amplifier is completely different to the LF351.I use a greatly simplified version of this amp ah-la LF351...
Below is the LF351 schematic from National's datasheet.
Perhaps if you actually looked at a design before commenting on it, your comments would make more sense.
I thought you might have learned that after the mess you made here.
Apparently not.
Attachments
LF351 has FET however BJT may be used. This is very similar. The same general concept except much simpler.
Godfrey obviously knows the difference between the functional schematic of the design concept for the IC and the actual schematic as put on the die inside an IC, or maybe not.
The only company I ever heard of that built the schematic as found inside an IC was Analog Engineering in Kentucky or Tennessee.
Godfrey obviously knows the difference between the functional schematic of the design concept for the IC and the actual schematic as put on the die inside an IC, or maybe not.
The only company I ever heard of that built the schematic as found inside an IC was Analog Engineering in Kentucky or Tennessee.
Attachments
Amplifier is made to be similar but better than most comercial home amps, and give me expect sound of my speakers in usual home audio sistem. This is more fidelity than best amp ever.Ultimate fidelity?! 😕
Regards
I cannot say specific about ApexAudios amp but the "simplified" version more like a power LF351 has open loop: 85kHz -3dB bandwidth; gain of 95dB; THD 1%; symmetric slew rate; and output impedance of less than 1 ohm at 10kHz. That should leave plenty of room to find a compensation scheme which works. Would expect the ApexAudio amp to be very similar with possibly higher gain and possibly lower THD.
Please remember simulations mean almost nothing until the sim is tweaked to match the real world performance. At the very least this amp seems like an excellent starting point for a "ultimate" amplifier. It is intrinsically linear, has symmetric slew rates, and will operate without feedback if DC quiescent is such that the output is normal and near ground. Most of the amps seen here on DIY have none of these features.
An ultimate amplifier for audio music, yes. Not an ultimate amplifier for those who cling to the belief amplifiers in this power range need slew rates approaching 100V/µs, full power bandwidth of more than 1mHz, hash chokes in the output providing impedance coupled outputs, and to keep their "oscillators with gain" in check . There are a lot of those guys found here on DIY.
Yes strong agreement with ApexAudio on this one.
Please remember simulations mean almost nothing until the sim is tweaked to match the real world performance. At the very least this amp seems like an excellent starting point for a "ultimate" amplifier. It is intrinsically linear, has symmetric slew rates, and will operate without feedback if DC quiescent is such that the output is normal and near ground. Most of the amps seen here on DIY have none of these features.
An ultimate amplifier for audio music, yes. Not an ultimate amplifier for those who cling to the belief amplifiers in this power range need slew rates approaching 100V/µs, full power bandwidth of more than 1mHz, hash chokes in the output providing impedance coupled outputs, and to keep their "oscillators with gain" in check . There are a lot of those guys found here on DIY.
Yes strong agreement with ApexAudio on this one.
Last edited:
Bad expirience with stability of some design not mean that will be a problem with all similar circuit. I have bad expirience with stability of many amplifier which use OP amp on input stage, and I afraid to use this topology, but use it anyway in my PA amps.
Amplifier schematics is similar anyway. There is several topology, and any of them can be good starting point for " Ultimare Fidelity Amplifier " , implementation is most important for functional amp.
Amplifier schematics is similar anyway. There is several topology, and any of them can be good starting point for " Ultimare Fidelity Amplifier " , implementation is most important for functional amp.
Hi apexaudio,
Have you tried to add complementary J-fet buffer at output? Also, folded cascode for VAS would be interesting too...
Have you tried to add complementary J-fet buffer at output? Also, folded cascode for VAS would be interesting too...
Mile, please check JFET markings in your preamp - both 2N3268 and 2N3269 are small power NPN BJTs
Last edited:
Thank's Juma, that kind of mistake is usual for me.Mile, please check JFET markings in your preamp - both 2N3268 and 2N3269 are small power NPN BJTs
Fets are K170 and J74.
Regards
Last edited:
I like ultra simple circuit for started point, can be upgrade.Hi apexaudio,
Have you tried to add complementary J-fet buffer at output? Also, folded cascode for VAS would be interesting too...
Regards
With capacity multiplication in each OP PS, and constant current source with fet in differential pair SB3.1 can be upgrade in higer sound quality.Preamp circuit SB3.1 is simplified version of my preamp from 1990.
This is the stuff music is made of- not the typical recorded sound sadly so many hear. Great work!
- Home
- Amplifiers
- Solid State
- 100W Ultimate Fidelity Amplifier