Ok, thanks. Thats simply exellent info.
I can still at this point in time easily insert the Ls anywhere in the PSU-chain. 🙂
I can still at this point in time easily insert the Ls anywhere in the PSU-chain. 🙂
Yea, that sounds more than wise enough.
I am now running CRC, 200 000uF per channel.
Going to add 4 of the 159ZJ, one per rail. Inspired by your old 159ZJ recommendation.
Yes. I know it is absolutely fully insane: Because them twin stereo Aleph Js in parallel per channel sounds already simply beyond beautiful.
Yea well… 🎷🙂🎸
My A75 underwent CLC-surgery 4 years ago after working with standard C-only (66mF per rail)) PSU for 11 years. Ripple voltage went from 0,22V RMS to 0,011V RMS (1,7A bias) by adding 1mH(160mOhms) and 22mF per rail (+/-40VDC).
The only instantly recognisable difference was the increase in weight (app. 4 kg) after carrying the amp downstairs back into the living room.
There have been some changes in my audio system over the years - A75-setup (folded cascode, feedback), speaker position, acoustic dampening, different CD-players, new turntable, harddisk/network players, ears getting older - with by far greater sonic consequences than CLC.
Conclusion: I can only tell that my A75 has CLC by lifting it.
The only instantly recognisable difference was the increase in weight (app. 4 kg) after carrying the amp downstairs back into the living room.
There have been some changes in my audio system over the years - A75-setup (folded cascode, feedback), speaker position, acoustic dampening, different CD-players, new turntable, harddisk/network players, ears getting older - with by far greater sonic consequences than CLC.
Conclusion: I can only tell that my A75 has CLC by lifting it.
only if they are slit magnetic strip toroid cores (made for DC) there is use of them
in any other scenario (standard toroid magnetic strip, powder core), they're simply saturated
that's what logic sez; if figures you gave are actual measurements, I'm curious to read what those chokes are exactly
in any other scenario (standard toroid magnetic strip, powder core), they're simply saturated
that's what logic sez; if figures you gave are actual measurements, I'm curious to read what those chokes are exactly
In a single ended amp, the reduction in PS ripple is quite noticeable with high efficiency speakers. All my single ended amps have CLC power supplies. I have switched to my Quad ESL speakers (83 dB/W/m) for the summer (less heat) so I don't have to biamp my JBL system and the Quads are dead quiet compared to a very small amount of noise with my JBLs.
Fastron TLC/5A-102M, 1000µH, 5A DC. Datasheet says "Rated DC Current Ιsat, based on the Inductance Losses (Lo/L Load) where the Inductance decrease 30% max. (TLC/5A: 749µH). The core material is an iron-powder mixture optimized for high saturation currents"
These are measurements i wrote down in May 2021, when dealing with "crackling" output MOSFETs, made with Agilent U1253B (V) / Metrahit 18s (A):
L channel +/-39,60VDC, Bias 1,7A, Ripple before L 0,22Vrms, Ripple behind L 0,010Vrms
R channel +/-39,55VDC, Bias 1,7A, Ripple before L 0,22Vrms, Ripple behind L 0,011Vrms
My Isophon speakers are 86db/W, close to the tweeter i hear little hiss, maybe it was a bit more without CLC, i can´t tell. I forgot about this bit of noise - until yesterday -as i´m constantly enchanted by this amp 😍
Yes, Semikron P3, used for power rectifiers in my company.
then all good
nice to get (again, if there was first time and I forgot) that you did all thinking in time
This is the L:s inserted in one of the Aleph Js running in parallel. Forming CLCRC filters. Both L&R amps have got them added Ls.
I have been running them now for a couple of weeks, and my preliminary conclusion is:
A calmer, smoother and even more clear, inviting and unharch sound. The bass seems to have even more juicyness and substance. The bass grip and power sounds starts somehow further back in the silence.
I really like it so far. 🤘🙂🎸
I have been running them now for a couple of weeks, and my preliminary conclusion is:
A calmer, smoother and even more clear, inviting and unharch sound. The bass seems to have even more juicyness and substance. The bass grip and power sounds starts somehow further back in the silence.
I really like it so far. 🤘🙂🎸
Good idea, surely gonna test it. Since I loose ballpark 0,5V rail at the L and ca. 0,5V at the R.
But i am of cource in no real musical enjoyment hurry. 🙂🎸
The only downside with this arrangement is that the sound is more clear and (could be to much) chill.
Well? That easily mitigated: Simply Turn it the F up a notch or two or five! 😎❤️
But i am of cource in no real musical enjoyment hurry. 🙂🎸
The only downside with this arrangement is that the sound is more clear and (could be to much) chill.
Well? That easily mitigated: Simply Turn it the F up a notch or two or five! 😎❤️
maybe split coils prim /sec can solve this problem ? EI filter better as don't have the bw of toroids
Hey nicoch58, that’s probably a wrong reference? 🙂
It was about rectifier induced ringing, not about CLC ringing, which I was unable to get anyway as a function of dynamic power supply load or at PS startup.
However, EI transformers have higher leakage inductance than toroids and rectifiers, with EI transformers, produce enough ringing that it is a good thing to use snubbers. Split bobbin will provide less capacitive coupling between primary and secondary, so less junk will pass from the mains.
But, any decent toroidal transformer will have a grounded static shield between primary and secondary, thus being way better than EI transformer.
It was about rectifier induced ringing, not about CLC ringing, which I was unable to get anyway as a function of dynamic power supply load or at PS startup.
However, EI transformers have higher leakage inductance than toroids and rectifiers, with EI transformers, produce enough ringing that it is a good thing to use snubbers. Split bobbin will provide less capacitive coupling between primary and secondary, so less junk will pass from the mains.
But, any decent toroidal transformer will have a grounded static shield between primary and secondary, thus being way better than EI transformer.
what is the consensus value for L on 25w class A amp?
2 or 3 mH or more?
realistic value! not more the better....
2 or 3 mH or more?
realistic value! not more the better....
I like using separate CLC filtering for each channel with Hammond 159ZJ (10mH, 0.16 ohm) chokes. The CLC filter consists of 44 mF capacitance, Hammond 159ZJ, followed by another 44mF of capacitance. This gets the ripple to the millivolt range...
Hopefully , someone can check my algebra from post #134 .
Recall , with an inductor , it should be correctly dampened , otherwise , there will be over shoot , ringing and a setting time .
A CLC looks to be a 3 pole LPF .
Trying to figure out the transfer function for a 3 pole is LPF , is a algebraic hell .
However , tinkering around with microcaps software shows the roll off is - 40 dB per decade .
So that says a CLC filter is a 2 pole LPF .
Blowing the dust off my old math notes , post #134 shows the general form for a 2 pole LPF .
Grinding out the algebra for the correct value of R , to dampen L , gives
--------------------------------------------- *
R = 2 x zeta x sqrt ( L / C2 )
---------------------------------------------
plug in the values
R = 2 x 0.707 x sqrt ( 10 mH / 44 mF )
R = 0.67 ... but r = 0.16 for the 159ZJ
so connect a 0R51 in series with the L and this should correctly dampen the L .
* Note : the equation shows increasing C2 increases the dampening , so this lowers the value of R needed to dampen the L .
.
Fantastic. Thank you!
Couple of dummy questions: I always though the R was to go in parallell. But in series you say, new to me and so I guess I gotta read on 🙂
Then, given 147.000uF after each L (my case), let me ask it practically: is there any point whatsoever in considering daming that ringing?
Regards,
Andy
Couple of dummy questions: I always though the R was to go in parallell. But in series you say, new to me and so I guess I gotta read on 🙂
Then, given 147.000uF after each L (my case), let me ask it practically: is there any point whatsoever in considering daming that ringing?
Regards,
Andy