Thank you very much for sharing.
I wonder if I can use the dual NoNoiser for powering a phono preamp +24V/-24V <1.0A
That is the phono section of my preamp that now is powered via split psu with 4X4700uF filter caps followed by LM7824/7924 fixed voltage regulators.
Last edited:
Do you mean I should alter the value of R5 to 175K to get the 24V output?
I am a bit confused, the schematic you posted is different than the one Trileru has uploaded with the files.
Last edited:
I think the schematic is the same.
You need to alter R5 to 175K for 24Vout. The NoNoiser uses a different scheme to set the output voltage.
Also you might have to lower C3. LTSpice shows that for a similar response as a LM317 + denoiser (1.9k with 220uF) you can adjust C3 for 24Vout NoNoiser (R5=175k) to 3.3uF. Maybe stick a 10uF if you don't have a smaller electrolytic.
edit:
you can check Elvee's post for the math for setting Vout with the NoNoiser.
on the dual pcb I forgot to adjust R1's value to 2.7k for NoNoiser, in the project's schematic. if there's something you're not sure about (PCB part values for different configurations) it's best to ask here.
You need to alter R5 to 175K for 24Vout. The NoNoiser uses a different scheme to set the output voltage.
Also you might have to lower C3. LTSpice shows that for a similar response as a LM317 + denoiser (1.9k with 220uF) you can adjust C3 for 24Vout NoNoiser (R5=175k) to 3.3uF. Maybe stick a 10uF if you don't have a smaller electrolytic.
edit:
you can check Elvee's post for the math for setting Vout with the NoNoiser.
on the dual pcb I forgot to adjust R1's value to 2.7k for NoNoiser, in the project's schematic. if there's something you're not sure about (PCB part values for different configurations) it's best to ask here.
Last edited:
You can de-noise a Maida Regulator, but need a high voltage PNP for the CCS -- something like a KSP92.
I want to try something to see if it helps low voltage output, say 5V. The CCS becomes harder to use for lower Vout.
I'm not sure how critical it is for the CCS itself to be powered from Vout. If you have another higher voltage available in the case (say 12V-15V for opamps) then you could use that to power the denoising circuit. The CCS has some (small) PSRR as well.
I wonder if this trick could be used for Maida to get around HV parts or HV waste as heat.
I'm not sure how critical it is for the CCS itself to be powered from Vout. If you have another higher voltage available in the case (say 12V-15V for opamps) then you could use that to power the denoising circuit. The CCS has some (small) PSRR as well.
I wonder if this trick could be used for Maida to get around HV parts or HV waste as heat.
Here's a sim for what I was proposing. I though about powering the CCS from the input of the regulator but there seems to be some strange interaction that I don't understand.
But if I power it from a separate LM317 (stock application without Cadj), there seems to be an improvement on 5Vout vs just resistor. I don't think the CCS can even be used for 5Vout.
But if I power it from a separate LM317 (stock application without Cadj), there seems to be an improvement on 5Vout vs just resistor. I don't think the CCS can even be used for 5Vout.
Attachments
Another thing you should consider is the fact that for different denoiser transistors (Q20/Q24 in the attached photo) there seems to be a different DC offset on the denoiser side of the coupling capacitor (C32/C37).
For lower Vout you might need to rotate that capacitor. And I guess that might eat into the CCS performance as well?
edit: I exagerated the current with 33R.
For lower Vout you might need to rotate that capacitor. And I guess that might eat into the CCS performance as well?
edit: I exagerated the current with 33R.
Attachments
Following the schematic you last uploaded at post #2,340 and adjusting the values of R1 to 2.7K, R5 to 175K and C3 to 3.3uF (as advised in order to get 24Vout) do I have to lower the value of C4 too or keep it at 100uF?
Last edited:
hi Elvee
Thanks for such a great solution!
You mentioned here the most secure way of stabilizing compensation for the 337.
Should I modify the denoiser as follows?
Thanks again
That's my recommendation, based on a (non-exhaustive) number of samples. Your mileage may vary, but the values should normally be safe enough, unless your 317 is really an odd, exotic beast.
Also read Trileru's posts about the subject, they are highly informative
I said to myself that i'll measure original d-noizator but i have been busy for quite some time. Finally i found spare time and tested it with my simple LNA and HP 334A as measuring equipment. Conditions for testing we're a bit unusual.Input voltage around 11V,output voltage 6.3V(maybe a bit too low for d-noisator work properly), current 0.6A. Unusual test conditions are cause i'm gonna use it as for ecc88 heater so i wanted to test it in real working conditions. I know it's overkill but it's cheap,simple and you can get parts everywhere so why not used it and it was perfect time to measure it.
Results we're a bit disappointing. Input ripple voltage was 0,25Vrms and output 7uVrms so after i do math i get around 91dB ripple rejection, which is more or less 10dB better then normal lm317(80db typical ripple reduction). Not sure if result is bad cause of low output voltage or something else. I'll test it under same current consumption but around 15V output voltage and see what result i get.
Results we're a bit disappointing. Input ripple voltage was 0,25Vrms and output 7uVrms so after i do math i get around 91dB ripple rejection, which is more or less 10dB better then normal lm317(80db typical ripple reduction). Not sure if result is bad cause of low output voltage or something else. I'll test it under same current consumption but around 15V output voltage and see what result i get.
The typical 30dB improvement applies to a 12V configuration. At 6V, with the values unchanged, there will be a degradation of a few dB, but not 20dB or even 10dB.
I suspect that some additional ripple leaks through a non-ideal GND configuration, a magnetic coupling of the ripple current to the regulator or the denoiser, or something similar.
To keep the same 30dB performance at 6V, you can scale all the resistors of the denoiser: halve them in your case.
I suspect that some additional ripple leaks through a non-ideal GND configuration, a magnetic coupling of the ripple current to the regulator or the denoiser, or something similar.
To keep the same 30dB performance at 6V, you can scale all the resistors of the denoiser: halve them in your case.
This is PCB, i placed one extra transistor in parallel to LM317 if extra current is needed but in tested configuration i used plain Lm317 dnoizator so R10,11 and T1 are not used. Maybe mistake could be that i placed dnoizator parts in between of power paths and 0,6A is not small current so there could be interference from magnetic coupling.
I tried to keep power gnd away from local lm317 and dnoizator gnd so those two meet only at output capacitor.
Thanks for info and i'll try test this PCB with smaller output current at different output voltages and see what results i get.
I tried to keep power gnd away from local lm317 and dnoizator gnd so those two meet only at output capacitor.
Thanks for info and i'll try test this PCB with smaller output current at different output voltages and see what results i get.
