That is close to the base design (14mv when cold 11-12mv warm).
Actually , they stay at 11-12mv except on a VERY cold morning , and only
when you first turn it on.
(1 mv stable).
I run the sankens at @50-60ma , BTW.
OS
Brought the bias down to 12mV over a single Re, nice and stable there. Sounds fine on my crappy test speaker, need to put together a pair and have a real listen...
Sorry I did not see this sooner. The load is actually 8 ohm, a series-parallel arrangement of some 12 ohm 5W I had laying about sandwiched between two scrap heat sinks.
Yes , Sanken/toshiba 50-60ma ON = 80-100.
My old HK w/sankens ....OEM manual says so.
The new HK990 with the ON's , it's manual advises "hotter" bias.
PS - some of the old Sankens (1970's), advise only 35ma 😕.
(different dies - ring emitter/planar/monolithic).
Glad to hear you are thermally stable .. I "sweated" this issue in
regards to the large drivers.
Sanken 2sc6145 and a few others are the
"perfect" to-3p drivers ... only 250-300pf Cob and high gain.
The OPS would not "know" the difference between these and
the more typical TO-220 counterparts. I will use these in mine. 🙂
OS
OS,
The question will be in a sonic shootout which will sound more musical, the Gnome or the CFA-X?
The question will be in a sonic shootout which will sound more musical, the Gnome or the CFA-X?
My main purpose with that last posted amp is to (reduce) some of the
CFA "weaknesses".
If you were to run the "X" against a typical VFA (like the symasui) , they
would be in the same ballpark as far as offset and PSRR.
Some of these attributes will make for a more reliable amp ... the PSRR
might produce a better sound-stage (especially with a single PS).
In the end ..what it would come down to - does the 200V+ slew(or just the current FB), produce the "better sound" ?
Or is it just the simplicity (low group delay of just 1 stage) , that makes
the difference ?
OS
OS,
Good question about what is really important on the reproduction of best sound. I don't think we can count on sim distortion values to answer the question. I know Thimios can crank of these different front ends really fast but he doesn't seem to have the speakers he would need to make those judgement calls. Perhaps he has a friend with good to great speakers where he could more acurately tell us the differences in sound from one version to the other of these CFA amps? The other problem with these subjective answers are the use of so many different active devices that are being used and the different power supplies that could be used. If the psrr and offset voltage is improved this much would that mean you would use a simpler power supply or would a better optimized power supply still make a significant difference in the end product?
Good question about what is really important on the reproduction of best sound. I don't think we can count on sim distortion values to answer the question. I know Thimios can crank of these different front ends really fast but he doesn't seem to have the speakers he would need to make those judgement calls. Perhaps he has a friend with good to great speakers where he could more acurately tell us the differences in sound from one version to the other of these CFA amps? The other problem with these subjective answers are the use of so many different active devices that are being used and the different power supplies that could be used. If the psrr and offset voltage is improved this much would that mean you would use a simpler power supply or would a better optimized power supply still make a significant difference in the end product?
Perhaps he has a friend with good to great speakers where he could more acurately tell us the differences in sound from one version to the other of these CFA amps?
This is my plan
In this way we have 2 or more opinions about difference, but this time my friend have burned twitters.😡
This is my plan
In this way we have 2 or more opinions about difference, but this time my friend have burned twitters.😡
Thanks Thimios,
Perhaps your friend can get some replacement diaphragms without to much trouble?
Perhaps your friend can get some replacement diaphragms without to much trouble?
I have the opposite situation of Thimios.
Just spent $100's on "twitters"/caps 😀 Found a free sub (tangband wt-644)
Boxes done for full range speakers (e-waste missions) , new X-over caps
to replace 20 YO ones. Sub box done- tuned to 28hz.
Have both the sub and fullrange amp toroids ,big caps,35A bridges
2 "badger" boards and 2 "slew" boards (Thanks jason).
Need HS's and outputs. So close yet so far away .. spent my
money at parts-express 😡 .
I have the speakers to at least listen to my OEM's (amps) 😱 .
OS
Just spent $100's on "twitters"/caps 😀 Found a free sub (tangband wt-644)
Boxes done for full range speakers (e-waste missions) , new X-over caps
to replace 20 YO ones. Sub box done- tuned to 28hz.
Have both the sub and fullrange amp toroids ,big caps,35A bridges
2 "badger" boards and 2 "slew" boards (Thanks jason).
Need HS's and outputs. So close yet so far away .. spent my
money at parts-express 😡 .
I have the speakers to at least listen to my OEM's (amps) 😱 .
OS
I tried IPS almost same as CFA-X without servo (feedback resisitor = 2K ohm) and IPS like VSSA without servo (feedback resistor = 1K1 Ohm).
Sim result for VAS using Hawksford cascode is lower THD than simple VSSA. But I like much more simple VSSA in listening test. The high frequency is more transparent and no harsh sound.
May be I should try to get slew rate equal and compare them again. The simple VSSA has higher slew rate.
Bimo,
The question I would have is why would you intentionally handicap a design for comparison reasons? If the VSSA has higher slew rate by design that is what it has and should have the advantage of that in any test. It to me would be like saying you were going to compare two cars, one has a diesel and one has a gasoline engine. Would you put diesel fuel in the gasoline engine so they ran on the same fuel, that wouldn't make any sense at all, or gasoline in the diesel so they were running on the same fuel? No you would just handicap one or the other artificially in my eyes. Run each amp or input section the way it is designed and let the best device come out on top, no handicaps.
The question I would have is why would you intentionally handicap a design for comparison reasons? If the VSSA has higher slew rate by design that is what it has and should have the advantage of that in any test. It to me would be like saying you were going to compare two cars, one has a diesel and one has a gasoline engine. Would you put diesel fuel in the gasoline engine so they ran on the same fuel, that wouldn't make any sense at all, or gasoline in the diesel so they were running on the same fuel? No you would just handicap one or the other artificially in my eyes. Run each amp or input section the way it is designed and let the best device come out on top, no handicaps.
Just what I have expected from this thread.
Any cascode or current mirror has the leakage and "eat" the sound from the signal. That's what I am telling the OS and other guys here for years now. If you loose micro signals information from the input signal, you lose it and if it's not there no sim will tell you that, simply because sim is working in just an opposite way, looking what is happening with large signals and their harmonics. The only relevant test would be harmonically very complex low level input signal structure and its perfect copy on the output. Some kind of software should perform this test as a simulation or a measurement. Until then your only relevant test bench is your ears and what's between them.
The more you go complex the more you wash the sound, examples: Halcro, Soulution, etc.
Hi Lazy Cat,
Disagree with that. Current mirror is rather linear, preserving the original signal "information" detail. Can you please give a bit more detail on what "leakage" are you talking about?
Valery
All CFA that I tried is simple CFA without diamond buffer. First CFA, I use Hawksford cascode VAS. At that time I did not confident to design high slew rate amp. So, I set slew rate around 225V/uS. Second design, I use single transistor for VAS and set slew rate around 400V/uS.
Now I have feeling and some knowledge how to design CFA 😎
Lazy Cat, I will try some experiment to prove your opinion that seem not logic to me.
Your VSSA was surprising to me how an amp that has worse THD can perform better in listening test.
Buffer in front of the input pair, especially unstable and nervous one like the Diamond is, would even worsen the whole situation.
Some amplifiers are simply better! 😀
Explanation in my previous post. 😉
Any impedance disorder in the signal path, via current leakage to GND or in between the stages. The result is big difference in uV region, losing micro signals in audio signal.
Tested many amps in ABX test, the same sch only one difference between the two. At the end simpler always sounded better. Of course not too simple.