[johnjen]

Nice build on your LPS, with plenty of extra space for 'upgrades' like adding additional bypass caps for 'faster' PSU response.
So when the itch to sling hot solder around takes hold… hahahahahaha

I too have made several LPS's for my AOIP setup and have replaced 5 SMPS's.
And along the way I measured the digital hash that SMPS's create which was substantial.
And after I made the LPS's I measured their noise as well, which is substantially less.

Then the kicker was, I measured the noise at the LPS's when it was running an ethernet to optical converter and found that digital devices contribute a substantial amount of noise that LPS's can't fully suppress.
IOW just running any digital device generates a fair amount of noise and the SMPS adds even more.

So the take away is to use as few digital devices as possible and avoid SMPS's where ever possible.
And the noise that SMPS's (and all digital circuits) create can 'sneak' back into the ac distribution system especially the ground circuits that directly feed into all of our audio gear, which then modulates the entire circuit for all such operating devices.
This is why I shut down the modem, router, screen etc. when I want 'optimal' SQ.

And granted the SQ improvement is not huge but is easily discernible in my system, especially late at night when the entire house is quiet and still.

And if you understand how SMPS's operate (they generate noise right in the middle of the audio band) which is filtered out (meaning dumped into ground), as well as they inject this switching noise back into the hot and neutral ac power lines.

And while some claim that that SMPS's can be designed that don't impact the SQ of the device they power, my take is if I don't create this noise in the 1st place (or create substantially less) then there is less that can 'contaminate' the ac mains supplying power that then feeds the rest of the audio gear.

And that's my story and I'm sticking to it… hahahahahahahah

JJ

 

[gefski]

Nice build on your LPS, with plenty of extra space for 'upgrades' like adding additional bypass caps for 'faster' PSU response.
So when the itch to sling hot solder around takes hold… hahahahahaha

I too have made several LPS's for my AOIP setup and have replaced 5 SMPS's.
And along the way I measured the digital hash that SMPS's create which was substantial.
And after I made the LPS's I measured their noise as well, which is substantially less.

Then the kicker was, I measured the noise at the LPS's when it was running an ethernet to optical converter and found that digital devices contribute a substantial amount of noise that LPS's can't fully suppress.
IOW just running any digital device generates a fair amount of noise and the SMPS adds even more.

So the take away is to use as few digital devices as possible and avoid SMPS's where ever possible.
And the noise that SMPS's (and all digital circuits) create can 'sneak' back into the ac distribution system especially the ground circuits that directly feed into all of our audio gear, which then modulates the entire circuit for all such operating devices.
This is why I shut down the modem, router, screen etc. when I want 'optimal' SQ.

And granted the SQ improvement is not huge but is easily discernible in my system, especially late at night when the entire house is quiet and still.

And if you understand how SMPS's operate (they generate noise right in the middle of the audio band) which is filtered out (meaning dumped into ground), as well as they inject this switching noise back into the hot and neutral ac power lines.

And while some claim that that SMPS's can be designed that don't impact the SQ of the device they power, my take is if I don't create this noise in the 1st place (or create substantially less) then there is less that can 'contaminate' the ac mains supplying power that then feeds the rest of the audio gear.

And that's my story and I'm sticking to it… hahahahahahahah

JJ

Some of us are “on restriction” as to the number of general purpose devices in the home that we are allowed to shut down.

 

[johnjen]

Yeah every once in a while I get caught and have to 're-enable' certain 'services' after a frantic call goes out "Why is the internet not working????"
hahahahahahahahahahahaha…

JJ

 

[Tjj226 Angel]

Changed a SMPS for a linear PSU on a mac mini

Verrrrrrry interesting.

Im really curious how this made a difference. Any computer power supply needs to provide exceptionally clean voltage to things like the CPU, RAM, and other assorted components. Ideally the voltage coming out of the SMPS that apple gives you should be pretty darn clean and then it should go through a series of on board voltage regulators before it hits different components on your PC.

If the mac mini was perfectly engineered, I would say that the linear PSU should make no difference what so ever. There should be so much on board filtering, regulating, and general safe guards that you could plug the dirtiest power in the world into it and it shouldn't make a bit of difference.

The fact that you did hear a difference makes me think the mac mini wasn't perfectly engineered (which doesn't surprise me). The questions then become, where are the deficiencies in the mac mini, and how exactly the the linear PSU fix it.

If we could figure out what the linear PSU fixed, then there might be a way to circumvent the problem in the mac mini all together and improve things even further.

 

[gefski]

Verrrrrrry interesting.

Im really curious how this made a difference. Any computer power supply needs to provide exceptionally clean voltage to things like the CPU, RAM, and other assorted components. Ideally the voltage coming out of the SMPS that apple gives you should be pretty darn clean and then it should go through a series of on board voltage regulators before it hits different components on your PC.

If the mac mini was perfectly engineered, I would say that the linear PSU should make no difference what so ever. There should be so much on board filtering, regulating, and general safe guards that you could plug the dirtiest power in the world into it and it shouldn't make a bit of difference.

The fact that you did hear a difference makes me think the mac mini wasn't perfectly engineered (which doesn't surprise me). The questions then become, where are the deficiencies in the mac mini, and how exactly the the linear PSU fix it.

If we could figure out what the linear PSU fixed, then there might be a way to circumvent the problem in the mac mini all together and improve things even further.

Here is Uptone’s sales pitch on it. Note that part of their claimed improvement is changing the fan control from PWM to linear. For my system, the improvement is real (believe me I’ve purchased many ”side grades” or worse through the years). Whether the improvement is because of the linear PS, the fan control, just getting the PS out of the computer, or what % combination of these or other factors I have no idea. I agree with you that it seems there are always ways to improve things even further.
https://uptoneaudio.com/products/mac-mini-dc-conversion-linear-fan-controller-kit-mmk

 

[johnjen]

One thing I learned a ways back is that the ground is the ONE common connection to EVERYTHING, as in all voltage regulators, IC chips, analog ground, PSU ground etc. etc, it's ALL tied into ground.
And since it is used as part of the filter to deal with the noise generated any/everywhere, then some of that noise is going to propagate to everything else via this one common connection to EVERYTHING.

And as I mentioned previously, its FAR easier to deal with a 'lack' of noise than it is with additional noise generators, all with their own noise 'signatures'.
And granted this generated noise does diminish as it propagates out from the noise source(s), even so these various sources of noise do interact (harmonically mix) which in turn adds more noise on the entire systems noise floor and in 'unpredictable' ways.

And don't get me started on what we think ground is vs. what it really is, 'cause that's a whole nuther topic, one that most are blissfully unaware of.

JJ

 

[Tjj226 Angel]

Here is Uptone’s sales pitch on it. Note that part of their claimed improvement is changing the fan control from PWM to linear. For my system, the improvement is real (believe me I’ve purchased many ”side grades” or worse through the years). Whether the improvement is because of the linear PS, the fan control, just getting the PS out of the computer, or what % combination of these or other factors I have no idea. I agree with you that it seems there are always ways to improve things even further.
https://uptoneaudio.com/products/mac-mini-dc-conversion-linear-fan-controller-kit-mmk

If you say it's a real upgrade, then it's a real upgrade. Im not gonna deny anyone's experience without ACTUAL SCIENTIFIC EVIDENCE (looking at you sound science).

The flip side, is that I don't think that the fans have anything to do with this. People report the same kind of improvements that you talk about when using a linear power supply on a raspberry pi which has no fans or attached devices that would radiate high frequency hash.

I would be MUCH more curious to look at the power going into your storage drive and ram.

 

[johnjen]

So the latest mods I've made to the Purp-Amp are replacement 71A cathode bias resistors and new bypass caps.
These are accumulating hours as I type this.

The previous resistors were running WAY to hot (250ºF) and the caps were 'suspect' and were a bit too large in value.
So the new Caddock 25watt TO-220 resistors are running at ≈ 109ºF at the top plate and I've added 0.22µf bypass caps to the new ASC 60µf (instead of 100µf) bypass caps.

It will take a few more days before they settle in enough to figure out what the changes/SQ benefits are to the amp, but there are certainly no downsides, at least thus far that I can determine.

JJ

 

[johnjen]

So tonight S/S (Spooky/Scary) has come to the fore in the Purp-Amp.
It is about on par with the other S/S examples I've heard before, with perhaps a few new subtle nuances here and there.
But then it's still rather early in the blossoming process.

But that this degree of S/S is in evidence so soon, again, is a very good sign.

And plans are afoot to install the newest version of these Rod Coleman DHT filament regulator boards.

JJ

 

[johnjen]

So a statusupdatereport… thing.

I've left the Purp-Amp alone and just listened for over a week now, aside from adjusting the bias into narrower margins of voltage difference, and it has stabilized and blossomed and set new levels of just about every descriptor I use to describe those sonic traits I deem noteworthy and highly valued.

And I did this to see just where on the scale of tweakdom the amp is at the present time, but also because there is a meet this weekend and I wanted the amp to be a wowser, as much as it could be, anyway.

And I may have stumbled upon the 'proper' bias settings for the amp, sorta, finally.
But right now the amp does seem to deliver a satisfying experience, especially when cranked up into the low-mid 90's and higher.
That is when the bass has kick, punch and smack, all rolled into one.

And I can now, barely, hear a SQ shift in the mids when in the 'correct' phase.
It used to be that only the bass gave me much of an audible clue when I was in or out of phase.
But now there is a slight difference in the other percussive 'voices' such as piano, cymbals, drum smacks etc.
It sounds like the difference between blowing a positive pressure wave, vs. sucking a negative relative vacuum, even in horns and even acoustic guitars etc.

This is related to the notion of dual counter rotating identical geometric patterns finally coming back into perfect alignment and thus creating less and less of a Moiré pattern effect.
Think less fuzzy, a more sharply focused and accurately repeated, original pattern (ie. less distortion).

And now to tweak the bias settings, yet again.

JJ

 

[bcowen]

This is related to the notion of dual counter rotating identical geometric patterns finally coming back into perfect alignment and thus creating less and less of a Moiré pattern effect.

Does Google have an English translation for this?

 

[Tjj226 Angel]

Does Google have an English translation for this?

Even I have no idea what he is talking about and it's my design : P

 

[johnjen]

I wondered what sort of reaction I would get with my dual counter rotating morié pattern generator analogy.
hahahahahahahahahahahahaaha

I'll write up a more illustrative explanation in a bit.

JJ

 

[bcowen]

I wondered what sort of reaction I would get with my dual counter rotating morié pattern generator analogy.
hahahahahahahahahahahahaaha

I'll write up a more illustrative explanation in a bit.

JJ

I like pictures. Words are hard.

 

[johnjen]

I'll meet you half way, with pictures from within your mind, as in a thought experiment where you imagine and choose your own pictures…
hahahahahahahahahahahahaha

JJ