Absolutely!
I am with Ric on this.
Once the distortion is sufficiently low as to be inaudible under any plausible conditions,
And the SR is sufficient as never to be a limit under any plausible conditions
Then don't waste time and money, just try to make it simple, robust and reliable.
Best wishes
David
Last edited:
That's the OPS + VAS + Cherry.
The complete amp is well within Bode/Cordell/Bonsai recommendations. see #823
You get this behaviour as a matter of course with Cherry.
Conditional but sooo.o far from (-1,0) that its moot. Several types of fancy two pole compensation eg TPC often leads to 'conditional' stability like this.
It is hard to make a law, correlating the few experience we can have with our listening impressions. One thing i'm sure of, about distortion, is anything under 10ppm don't exist. At 20KHz, it is like a dream. And trying to get better is like trying to travel faster than light speed.
About slew rate, i was very interested in this question during 3 decades , but how to be sure of something ?
I tried a lot of OPAs, professional line level. Several in serial. I was able to prefer 500V/µs to lowers. verified with different OPAs. It seems as a limit at line level. 1000V/µs were not better. We made blind test to chose CF-OPAs for our big mixing desk.
It looks crazy ? What to answer. It is what we found.
I believe we can fight for the best we can in power amps.
ANd you will have to make your own experience with your amp in real world: a fast version VS a low HD version.
Attached a sim with > 1000V/µS, but, of course, more distortion. (0.0004% at 1000Hz 50W)
I'm pretty sure it will sound better than a slower and more complicated one, it is a lot simpler. (AS long we ensure it run stable in real life).
Can be improved comp side at 20KHz ?
PS: you did not answer to my question about your slew rate measurement method .
Last edited:
"I am strangling with lateral MOSFET OPS and how to tame it. If I go for very low distortion, suffers slew rate, and vice versa. It’s quite easy to get very high SR, about 400 V/usec, but then distortion is more then 10 ppm at 20 kHz, and if I go for below 5 ppm I can’t get SR better then 150 V/usec. I tried modified TPC(with OPS included in one C branch) in combination with shunt compensation (it looks that whiteout it CFA is not very stable):
dadod, I wish everyone would stop being obsessed with single digit ppm!!!!!!!!!
We have gone from having subjectivists hijack the discussion to Doug Self and his accolytes do the same thing but in the opposite direction and now anything above 1 or 2 ppm is not good.
Your ear brain system can detect 0.5% on a good day and speakers also distort like hell.
400V/us at 10ppm distortion is a damn fine amplifier.
Whats more important here is that the harmonic structure below clipping is low order (2nd and 3rds)
dadod, I wish everyone would stop being obsessed with single digit ppm!!!!!!!!!
We have gone from having subjectivists hijack the discussion to Doug Self and his accolytes do the same thing but in the opposite direction and now anything above 1 or 2 ppm is not good.
Your ear brain system can detect 0.5% on a good day and speakers also distort like hell.
400V/us at 10ppm distortion is a damn fine amplifier.
Whats more important here is that the harmonic structure below clipping is low order (2nd and 3rds)
Yes. The OPS is inevitably where the excess phase comes in earliest and is the "weakest link" that limits the whole amplifier.
So a 13 MHz ULGF here looks ambitious compared to the usual 1.5 to 3 MHz.
What #823 shows is not the complete amp but part of the outer loop.
This is the crux. In the common MC amp the outer loop determines the feedback around the OPS and thus is an excellent indicator of stability.
In an OIC amp the outer loop splits and you must look before the split to check the total Return Ratio around the OPS.
To look at the Return Ratio only via the IPS feedback node is useful for MC but deceptive for OIC.
It is also deceptive for TMC, as JCX has repeatedly pointed out, and was the source of interminable confusion.
So the conformance of #823 to Bode et al's recommendations does not prove stability.
Yes. Probably why so many people had problems with it.
But I do think we can learn a lot from the ones that worked, about how far to push the limits of speed.
I am not too concerned about this myself. It is only just conditional, as you say.
JCX has also noted that conditional stability is not such a problem with solid state as it was with thermionic valves that s-l-o-w-l-y reached maximum.
But it does help explain your incredibly low distortion, so worth some consideration. How much conditional stability is OK?
Best wishes
David
Last edited:
"I am strangling with lateral MOSFET OPS and how to tame it. If I go for very low distortion, suffers slew rate, and vice versa. It’s quite easy to get very high SR, about 400 V/usec, but then distortion is more then 10 ppm at 20 kHz, and if I go for below 5 ppm I can’t get SR better then 150 V/usec. I tried modified TPC(with OPS included in one C branch) in combination with shunt compensation (it looks that whiteout it CFA is not very stable):
dadod, I wish everyone would stop being obsessed with single digit ppm!!!!!!!!!
We have gone from having subjectivists hijack the discussion to Doug Self and his accolytes do the same thing but in the opposite direction and now anything above 1 or 2 ppm is not good.
Your ear brain system can detect 0.5% on a good day and speakers also distort like hell.
400V/us at 10ppm distortion is a damn fine amplifier.
Whats more important here is that the harmonic structure below clipping is low order (2nd and 3rds)
I am not obsessed with single digit ppm. Just taying to get the best of one cofiguration and then deside what to make simpler afterwards. I know that simulation does not show real life distortion and even SR, as layout is critical at that level, but if we don't try get the best in simulation what purpose have all this CFA discussion. I am quite confident in VFA design but CFA is new for me and I'd like to learn some tricks here too.
It's easy to make 2nd and 3rds harmonic dominant but not at this level of GNFB, or you introduce that distortion intentionaly to mask higher harmonics and go much higher in THD20k then 100 ppm, and I don't buy that.
BR Damir
May-be they listen to music with calculators instead of their ears ?
Or it is a way to heal his anguish ?
Same thing here, as long as we will stay in simulators, we will turn in circle and be unable to lean something consistent about the best compromise HD vs Slew rate.
Dadod, you did not answered my question about your slew-rate measurement method.
Last edited:
Sorry Esperado I will, I am a kind busy now, but soon enough.
P.S.
Esperado I have this .asc file but I need all models you are using, and by the way lateral models are wrong, try to use Cordell's.
Last edited:
I was sure they came from Cordell's site. Will try to change them this night. They are not Lateral but HexFets.
Edmond, do you have, by chance, better models for lateral, I use those from Cordell?
BR Damir
At this time, it was not called lateral. But, yes, they have a negative temp behavior.
I don't find those models in your Cordell-Models.txt. If you have better models than mine (i don't have invented them ;-), can-you provide them to me ?
Simulators are really not my tool ! I have no such questions with a soldering iron and datasheets !
Last edited:
There's a good discussion in Cordell's book.
He points out that the 'usual' MOSFET models are good for stability & response but almost useless for THD. (That's what's in Cordell-Models.txt)
For THD, you need EKV models. Bob had some but seems to have misplaced them. He hopes to find & post them now that he's retired.
Lateral MOSFETs are my favourite o/p devices. Their real cost seems to have gone down since I was playing with them in the early 90's. But the last time I burnt solder on amps, 'pure Cherry' seemed allow equivalent performance with VFETs.
When we get reliable EKV models, I'll abandon dem evil BJT OPS .. at least in SPICE world.
Last edited:
Thank you Andrewbonsai said:
I beg to differ.Dave Zan said:
IMHO, the Tian probe on the IPS looks at the 'whole' amp and the probe at the VAS input within the Cherry or evil Miller loop looks at the local VAS + OPS loop (Even with simple Miller, the OPS introduces evil poles. see Cordell & others.)
This is the approach Cherry takes in his various NDFL & other feedback papers and in fact 'looking before the split' is one way to see how a zillion NDFLs amp increases the 'Return Ratio around the OPS'.
I'm just using the simplest possible NDFL .. one which my single brain cell can unnerstan en analiz
There was a long discussion (I think in Cordell's thread) on whether unstable inner loops would be visible if you looked at outer loops including the Closed Loop response. I side with Bob in thinking this is the case.
But the real 'proof of stability' is ringing the changes with different evil loads, levels, different devices, overload etc. of the complete amp. Doing this in 'real life' is as, if not more, important than emulating this as I'm doing in SPICEland.
You can use 2SK134, 2SJ49 models from his list, thy are the same as 2SK135, 2SJ50 just lower voltage.
Main advantage of laterals is their temperature behavior: no comp with their inconveniences.
With a Rdson of 0 Ohm ?
Last edited: