TioFrancotirador,
from your experience, best way to supply power Is to use Ultra BIB for analogue (Pre Amp, op amps, analogue section of chips) and Reflector D for digital sections. Is there recomendation on how much uF to use at output of regulators, and local at load. Do you use cs and rs for transformer windings. Is Mundorf M Lytic first choice for cap at regulator Input.
from your experience, best way to supply power Is to use Ultra BIB for analogue (Pre Amp, op amps, analogue section of chips) and Reflector D for digital sections. Is there recomendation on how much uF to use at output of regulators, and local at load. Do you use cs and rs for transformer windings. Is Mundorf M Lytic first choice for cap at regulator Input.
Last edited:
I finally got it working for my pi3!
Salas ultraBiB is used for the first stage regulation: output is 8Vdc.
Second regulation is a LT1963 low noise reg to split into 3 voltages: 5V, 3.3V & 1.8V for pi3.
I replaced M1 with IRF9630 as suggested by Salas; I also replaced the R1 with a 0.15 Ω resistor. Transformer is 50VA, 2x9Vac secondary winding.
Improvement? Definitely yes!
Before, LT3045 reg was used in the first regulation stage and it was already a wonderful improvement over the wall-wart.
With UltraBiB, the dynamic and soundstage and everything gets audibly better. Full of energy for rock songs and powerful enough for big performance. Background is very silent now.
I am very happy with it!
Salas ultraBiB is used for the first stage regulation: output is 8Vdc.
Second regulation is a LT1963 low noise reg to split into 3 voltages: 5V, 3.3V & 1.8V for pi3.
I replaced M1 with IRF9630 as suggested by Salas; I also replaced the R1 with a 0.15 Ω resistor. Transformer is 50VA, 2x9Vac secondary winding.
Improvement? Definitely yes!
Before, LT3045 reg was used in the first regulation stage and it was already a wonderful improvement over the wall-wart.
With UltraBiB, the dynamic and soundstage and everything gets audibly better. Full of energy for rock songs and powerful enough for big performance. Background is very silent now.
I am very happy with it!
Attachments
How was the comparison done ; did you measured the improvement in dynamic brought by the new shunt ? How were you able to do that ?
Did you measure with an oscilloscope the background noise improvement? How much improvement was observed? (mV before; mV after)
Good work! I would be glad to "see" the heard improvements.
I don't have an oscilloscope. Everything is by ears.
I have a short playlist of music for testing all the time;
I switched between 2 regulators and listened to the same music.
I know my ears would adapt after listening to the same thing for one or two days. Instant switching of some component becomes very obvious in terms of SQ.
I remember my instructor said spectrum analyser was to be used to check the pattern of frequencies: capture the pattern if you like the SQ and use that as a reference.
Though I never have such expensive equipment.
I have a short playlist of music for testing all the time;
I switched between 2 regulators and listened to the same music.
I know my ears would adapt after listening to the same thing for one or two days. Instant switching of some component becomes very obvious in terms of SQ.
I remember my instructor said spectrum analyser was to be used to check the pattern of frequencies: capture the pattern if you like the SQ and use that as a reference.
Though I never have such expensive equipment.
Almost 4 Ampere constant current setting! Now that's some real dedication to run the Raspberry Pi 3 computer from a shunt regulator
That 50VA traffo must be having a pretty bad time, though..
Understood ; thank you for the detailed answer. And yes, ear, unfortunately, is directly tied to a complex processor; nothing they render can be trusted in terms of test results or quality.
Instant switching of sources of sound with identical amplitude @1Khz , with less than 10ms silence between source changing and no pop's is the only way we could observe differences.
Is very important in comparison to have the sound sources identical in amplitude otherwise our brain would prefer the music that is a little bit(~2-3db) louder.
Along with a good friend we used this method for comparing sound sources.
Findings are that unless is devised a solid way of testing/compare sound sources, ears are mostly useless and are influenced on how much we spent on equipment (not only money but also time and emotional investment).
I say this, because i proved to my friend (and later him to me) that we liked more the sound of a different device although when using the proper testing method (by not allowing brain to add the emotional color) as described above they sounded identical or even worst.
Instant switching of sources of sound with identical amplitude @1Khz , with less than 10ms silence between source changing and no pop's is the only way we could observe differences.
Is very important in comparison to have the sound sources identical in amplitude otherwise our brain would prefer the music that is a little bit(~2-3db) louder
.Along with a good friend we used this method for comparing sound sources.
Findings are that unless is devised a solid way of testing/compare sound sources, ears are mostly useless and are influenced on how much we spent on equipment (not only money but also time and emotional investment).
I say this, because i proved to my friend (and later him to me) that we liked more the sound of a different device although when using the proper testing method (by not allowing brain to add the emotional color) as described above they sounded identical or even worst.
Hello Salas, I've recently been using one of your shunts to power my AK4495 DAC. However on switch-on I am getting switch on spikes of a few volts higher than the set voltage, before coming down.
I'm running it at 5V, but with the standard 220uF reference cap, it spikes up to 7.9V. Adding an extra 470uf in parallel, it now spikes at 7.0V, which is safe. I changed the 0r33 resistor on the input to 10V, but that didn't help.
The regulator sounds great - better than the Paul Hynes shunt that it replaced - so it's obviously working well. The LEDs are at normal brightness, so reference current looks normal. Any ideas what I should check?
Thanks,
Dan
I'm running it at 5V, but with the standard 220uF reference cap, it spikes up to 7.9V. Adding an extra 470uf in parallel, it now spikes at 7.0V, which is safe. I changed the 0r33 resistor on the input to 10V, but that didn't help.
The regulator sounds great - better than the Paul Hynes shunt that it replaced - so it's obviously working well. The LEDs are at normal brightness, so reference current looks normal. Any ideas what I should check?
Thanks,
Dan
Last edited:
It normally tests softly without an over voltage peak on single event capture mode when the digital scope is set to very slow. So that's an odd occurrence. Does it do it on a resistor dummy load as well? Can you verify? Trying to firstly understand if it's a malfunction or a momentary oscillation with a particular client circuit. So to analyze for helpful ideas.
It behaves the same with or without load.
I did find I had 270R in R9 - just changed it to 680R (the closest higher value I had) and while the spike is still there, the voltage does ramp up more slowly, up to about 3V, then rises sharply to 7V before dropping down to 5V.
Is it worth going to 1K for R9?
I did find I had 270R in R9 - just changed it to 680R (the closest higher value I had) and while the spike is still there, the voltage does ramp up more slowly, up to about 3V, then rises sharply to 7V before dropping down to 5V.
Is it worth going to 1K for R9?
Last edited: