Thanks! I will try and let you know.
BTW, in the simulations for the original schematic: I changed C103 to 47uF and then added a 1mF between C103 and 0V(GND). This way the ripple decreased from 70-50mVp to 5-2mVp. Is it safe to try? I use load of 20mA, not more.
BTW, in the simulations for the original schematic: I changed C103 to 47uF and then added a 1mF between C103 and 0V(GND). This way the ripple decreased from 70-50mVp to 5-2mVp. Is it safe to try? I use load of 20mA, not more.
This add-on replaces only the C103? You write B_Q101: this means connecting the Add-on Dnoizer to the three nodes, but leave in place the R102. Right?
Do I let in place also the rest, like R104-R105-P106? I assume yes, else how do I adjust the voltage?
If I keep the R104-R105-P106 but no C103, then R102 can go. Right?
I will start simulations.
I have made progress with the simulations using Dnoizer in A77. I really like the results!
How should be implemented the Kelvin connections in this case? Sorry to ask this kind of questions, my head is very unfocused these days. Thanks for help!
I observed in simulations, the hum is correctly displayed as 100Hz.
But, the original problem was 50 and 150Hz (plus higher harmonics but less pregnant).
Hence, I think the original problem has not much to do with the stock regulator (as 'ideal device') but instead it picks up 50Hz from environment (220V lines, 120V, lower V lines, Transformer, motor(s), capstan regulator, relais and... potential Ground loops)
Can I use a 24VAC derivation and make a separate 21V additional regulator (Dnoiser or NoNoiser type) for the Repro boards, just near them? Idea is to avoid the long distribution DC wiring.
Or, this derivation will still remain contaminated by noise&hum in the first regulator plus all pickup from environment?
How should be implemented the Kelvin connections in this case? Sorry to ask this kind of questions, my head is very unfocused these days. Thanks for help!
I observed in simulations, the hum is correctly displayed as 100Hz.
But, the original problem was 50 and 150Hz (plus higher harmonics but less pregnant).
Hence, I think the original problem has not much to do with the stock regulator (as 'ideal device') but instead it picks up 50Hz from environment (220V lines, 120V, lower V lines, Transformer, motor(s), capstan regulator, relais and... potential Ground loops)
Can I use a 24VAC derivation and make a separate 21V additional regulator (Dnoiser or NoNoiser type) for the Repro boards, just near them? Idea is to avoid the long distribution DC wiring.
Or, this derivation will still remain contaminated by noise&hum in the first regulator plus all pickup from environment?
from simulations apparently won't work at all. Seems to me it creates a loop via ground which upsets badly the non-Dnoizer regulator. probably the idea to use one winding for two parallel rectifiers+regulators was completely silly.
There is a chance to use one of the unused 32V speaker amp trafo windings. Or both windings. Then I will have a super clean +/- 30V rails for a new tape preamp on steroids 🙂
I don't know... it seems too wild.
Don't worry about Kelvin connections, just connect the add-on across the output posts.
If 50 and 150Hz is present, it is most probably caused by magnetic leaks, and the denoiser won't be very helpful.
You cannot connect the AC and DC sides of two supplies using the same winding, as you discovered
Thanks!
BTW speaking of different voltages for Dnoizer, earlier this thread I think it was promised a formula. Can anyone point for me the Post number with it?
BTW speaking of different voltages for Dnoizer, earlier this thread I think it was promised a formula. Can anyone point for me the Post number with it?
They are just the regular design equation for the 317 regulator. They can be found in any datasheet. The denoiser itself remains unchanged whatever the voltage
Okay, I understood.
But I meant to ask this: for any output voltage between 4.25V and 40V do we use the Dnoiser with unchanged values and still obtain the 'optimum' in all cases?
If not, is there an equation to tell us the optimum of Dnoiser settings (values) vs. the output voltage?
You answered with yes, it does not depend on output voltage. It will work very good. Clear.
My question was if the 'optimum/best performance' needs a specific design target for component values.
But I meant to ask this: for any output voltage between 4.25V and 40V do we use the Dnoiser with unchanged values and still obtain the 'optimum' in all cases?
If not, is there an equation to tell us the optimum of Dnoiser settings (values) vs. the output voltage?
You answered with yes, it does not depend on output voltage. It will work very good. Clear.
My question was if the 'optimum/best performance' needs a specific design target for component values.
Last edited:
off topic, today I started to redesign the routing of power and signals cables in A77.
The Dnoizer deserves its best chance (my best efforts with cabling) to shine also here, especially since I can use it also with the non-317 regulator.
Simulations of line and load regulations for Dnoizer are blowing good.
(I still keep in mind the wild solution, to use the 24VAC for 21V on Speed board, and the 32VAC windings one for Repro boards and one for Oscillator+Rec boards. That's why I am asking for optimum, if any)
The Dnoizer deserves its best chance (my best efforts with cabling) to shine also here, especially since I can use it also with the non-317 regulator.
Simulations of line and load regulations for Dnoizer are blowing good.
(I still keep in mind the wild solution, to use the 24VAC for 21V on Speed board, and the 32VAC windings one for Repro boards and one for Oscillator+Rec boards. That's why I am asking for optimum, if any)
The ideal values are slightly (ever so slightly) dependent on the supply voltage, and the standard circuit is designed for 12V, but unless you go to extreme values, the change will be minimal -1~2dB at most.
I think the subject is covered somewhere in the thread, but I have no idea where.
Try first to have a denoiser actually working in your context, and then you can make tweaks to improve it further and scrape (maybe) 1dB.
As a rule, the higher the voltage, the higher resistances are, but it not proportional, more like square root.
For a 4V output, 1K2 collector resistor and 120K or 100K base resistor should be near optimum, but don't expect any audible improvement, just nicer measurements
I think the subject is covered somewhere in the thread, but I have no idea where.
Try first to have a denoiser actually working in your context, and then you can make tweaks to improve it further and scrape (maybe) 1dB.
As a rule, the higher the voltage, the higher resistances are, but it not proportional, more like square root.
For a 4V output, 1K2 collector resistor and 120K or 100K base resistor should be near optimum, but don't expect any audible improvement, just nicer measurements
At this point i'm ready to give up. Built tombos version and the voltage won't stop oscillating. The positive rail is rock solid but the negative one is driving me crazy. With any load the voltage oscillates between 8.71 and 12.97. I tried increasing the 22nf capacitor to 44,100,200nf and in last ditch effort to 470nf and 2.2uF. The 2.2uF film capacitor reduced it to 13.03 to 11.98 but not gone. Tried inreasing/decreasing the capacitor resistor but nothing. Tried both ONSemi LM337 i had with no luck. If i disconnect the denoiser then the voltage becomes rock solid. Output capacitor is Nichicon 120uF 194 mohm without any series resistance.
The LM337 is notoriously temperamental, but it can be made stable. I tested the negative denoiser with an ONsemi, and it made no problem.
In the VRDN thread, some members have had stability problems with variants of the 337 (some were impossible to stabilize), you can have a look to collect some hints.
Tombo advised a low esr output cap:
https://www.diyaudio.com/community/...grade-any-317-based-v-reg.331491/post-6359232
Can you post the exact schematic + a picture of your build? There could be some clues
In the VRDN thread, some members have had stability problems with variants of the 337 (some were impossible to stabilize), you can have a look to collect some hints.
Tombo advised a low esr output cap:
https://www.diyaudio.com/community/...grade-any-317-based-v-reg.331491/post-6359232
Can you post the exact schematic + a picture of your build? There could be some clues
Ah, vices of my ‘youth’ are hunting me.
I’ve built and thoroughly tested positive regulator only. Later, in this thread, problems were determined with negative regulator by one member and recommendation was made to avoid dienoiser version of negative regulator.
After that experience, I’ve never published anything that was not first build and more than thoroughly tested.
Sorry, but negative dienoiser version can’t be made stable without changing design. Positive regulator is unconditionally stable and two positive regulators can be used for a bipolar supply.
I didn't create gerbers; @Trileru did.
I hope you are talking about the regular denoiser, because the dienoiser cannot be stabilized, as Tombo said. For the regular denoiser, you can find useful tips about the suitable types of 337 in the VRDN thread.
You'll have to browse through a large number of posts, but it is worth the effort
Well, wish i knew it when i ordered my boards years ago. I put the positive and negative onto one board so its smaller. Guess ill just convert it to a normal denoiser. Removing the 390ohm resistor and second bjt. Its better than nothing.
This is a good issue to discuss!
What are the disadvantages of using two positive regulators in series to obtain symmetrical voltages?
Can you explain?
No real disadvantages, as is with connecting two separate batteries to get symmetrical supply.
It would be impossible to arrange bipolar supply on one board sharing common ground with two same regulators, but here we have two entirely separate boards and we can connect their outputs as we need. There is actually advantage that both supply rails will have exactly the same performance and transient behavior, so true symmetrical supply (if that’s important for any reason).