UpTone LPS-1 Linear Power Supply Review and Measurements

[Blumlein 88]

And then after that, sell your decent DAC for an active speaker with digital inputs and stop worrying about this nonsense entirely.

I think if people understood just how transparent even inexpensive DACs can be, you would see more of this. Digital connection would simply do what analog connection once did with some considerable advantages attached. It is a bit bothersome that I never did find a consumer oriented digital preamp (as in a preamp that does digital input and volume controlled digital output). I am pretty sure if you dropped all analog input, the price of pretty good analog sections is more than the price of good digital handling of the signal. You really should only go analog at the power amp input or in the power amp itself.

 

[watchnerd]

I think if people understood just how transparent even inexpensive DACs can be, you would see more of this. Digital connection would simply do what analog connection once did with some considerable advantages attached. It is a bit bothersome that I never did find a consumer oriented digital preamp (as in a preamp that does digital input and volume controlled digital output). I am pretty sure if you dropped all analog input, the price of pretty good analog sections is more than the price of good digital handling of the signal. You really should only go analog at the power amp input or in the power amp itself.

Doesn't the NAD M12 do digital volume control?

Also, I'm pretty sure all the modern AV/AV pre/pro do digital volume control on their side of the house.

But, yes, this whole luxury DAC + tweaks is just nonsense, hopefully due to being a 'transition' technology and thus will go away, but, sadly, I fear it will stick around.

 

[Cosmik]

There is certainly value in the dual banked floating PSU see DIYAudio: http://www.diyaudio.com/forums/power-supplies/302881-dual-bank-floating-supercap-supply.html

The existence of something doesn't mean it has value.

This whole thing has the feel of a particular measurement being fetishised, and an overly literal 'solution' concocted. It reminds me of when I was a student and didn't want to pay a TV licence fee (that we have in the UK). There was an exemption for truly portable televisions that ran entirely on self-contained batteries. My literal solution (that I never built) was a TV with extended sheet aluminium rear shell containing two banks of batteries in a relay-operated alternating charge-discharge cycle, where I could claim that it literally was running only on its internal batteries. Of course, quite rightly, no one would have been fooled, and no one would have congratulated me on my cleverness. It wouldn't even have improved the picture or sound.

 

[jabbr]

The existence of something doesn't mean it has value.

This whole thing has the feel of a particular measurement being fetishised, and an overly literal 'solution' concocted. It reminds me of when I was a student and didn't want to pay a TV licence fee (that we have in the UK). There was an exemption for truly portable televisions that ran entirely on self-contained batteries. My literal solution (that I never built) was a TV with extended sheet aluminium rear shell containing two banks of batteries in a relay-operated alternating charge-discharge cycle, where I could claim that it literally was running only on its internal batteries. Of course, quite rightly, no one would have been fooled, and no one would have congratulated me on my cleverness. It wouldn't even have improved the picture or sound.

Your goal above wasn't to improve picture or sound rather to avoid a fee and floating supplies don't imply portability and so you made a decision. You contemplated the wrong solution for your problem. If you do have a ground loop or desire to reduce leakage currents, the a floating supply can be very helpful. Batteries and supercaps are neither exotic nor expensive.

 

[amirm]

My welcome to the forum too Jabbr.

On the notion of bank switching, the devil is in the details (of implementation) as we have seen here. Having an active bank being charged next to the inactive one needs to be considered carefully to disallow leakage.

Ultimately though, it may be better to simply use a quiet power supply inside the audio equipment and be done with it.

 

[Blumlein 88]

Doesn't the NAD M12 do digital volume control?

Also, I'm pretty sure all the modern AV/AV pre/pro do digital volume control on their side of the house.

But, yes, this whole luxury DAC + tweaks is just nonsense, hopefully due to being a 'transition' technology and thus will go away, but, sadly, I fear it will stick around.

Yes the NAD and most AV pre's do digital volume control. But then output the signal in analog. The M2 is a bit unclear. It has digital output, but according to the manual it can do sample rate conversion, but does not appear to put out a variable volume digital signal. It appears to be a fixed level signal. The manual could be more informative on the matter however.

 

[jabbr]

My welcome to the forum too Jabbr.

Thanks.

On the notion of bank switching, the devil is in the details (of implementation) as we have seen here. Having an active bank being charged next to the inactive one needs to be considered carefully to disallow leakage.

Looking at the details in the posted link, the PSMN075-100MSE MOSFET switch, claims a LDSS of ~10nA which roughly corresponds to a CDS of ~20pF. If the parasitic capacitances can be kept in this range, the leakage currents should also be kept in this range. With batteries, mechanical relays can be used which can provide higher impedances (than 130 Mohm) if this is necessary.

So anyways with some care, a dual bank supply can be considered truly floating for all intents and purposes.

Ultimately though, it may be better to simply use a quiet power supply inside the audio equipment and be done with it.

There are different types of power supplies that have different properties i.e. "quiet" is defined based on the application, so this belongs in the bag of tricks. Particularly when floating supplies are required for specific purposes.

 

[amirm]

Looking at the details in the posted link, the PSMN075-100MSE MOSFET switch, claims a LDSS of ~10nA which roughly corresponds to a CDS of ~20pF. If the parasitic capacitances can be kept in this range, the leakage currents should also be kept in this range. With batteries, mechanical relays can be used which can provide higher impedances (than 130 Mohm) if this is necessary.

Unless you make the capacitor bank huge, I don't think the relay is a feasible solution. It will constantly click every second or two. With folks leaving their audio system on all the time, that is liable to kill that relay over time and of course the clicking can be annoying if it is audible. You also need to snub the relay contacts.

On the specs for the mosfet, there will be tons of parasitic capacitance everywhere that must be managed. I did a quick and dirty measurements of the LPS-1 with my LCR meter and from what I recall, I measured nearly 300 pf between its two grounds. This is quite a bit more than the disconnect mechanism itself.

 

[jabbr]

Unless you make the capacitor bank huge, I don't think the relay is a feasible solution. It will constantly click every second or two. With folks leaving their audio system on all the time, that is liable to kill that relay over time and of course the clicking can be annoying if it is audible. You also need to snub the relay contacts.

relays would be useful for batteries and yes snub

On the specs for the mosfet, there will be tons of parasitic capacitance everywhere that must be managed. I did a quick and dirty measurements of the LPS-1 with my LCR meter and from what I recall, I measured nearly 300 pf between its two grounds. This is quite a bit more than the disconnect mechanism itself.

The critical spec is capacitance between the drain and source of the switch MOSFET. There will indeed be lots of capacitance everywhere but the critical topology, traces etc are between the drain and source Cds (Coss-Crss) specifically when the drain is "off". I don't know any details of the LPS-1 to know what is predicted but assume that John Swenson had access to the relevant software in his former job.

 

[jabbr]

Real world capacitance value is different to what you see in specs:
https://www.gamry.com/assets/Uploads/Super-capacitors-part-1-rev-2.pdf

That's ok. The supercap can leak without promoting a ground leakage current loop. What I am concerned with is minimizing current from AC mains through the discharging bank ... when the gate is "on" the charging current will freely flow into the charging bank (either battery or supercap). At the same time, when the gate is "off" there will be a 10nA leakage current (significantly) due to capacitance between the drain (connected to the AC main) and the source (connected to the electronics being supplied).

To explain a bit: there will be 4 switching MOSFETS for each bank which is itself stacked in series to get the needed voltage. One up and one down for each +/-. The up/down MOSFETs are in series so this actually minimizes leakage current but there will be extra parasitic capacitance so that mitigates the benefit.

alfe, I'm sure you have that really good SIV software that means your designs work the first time

Always need to measure though, but hopefully the design will predict the measured result

 

[RB2013]

My next audio iteration;

Ethernet based DAC

Balanced Power

Been there - done that - AES67 Dante - Focusrite Rednet 3/Mutec MC-3+ USB AES reclocker/Antelope OCX

Bottom line Hybrid USB/Ethernet GB LAN whooed it.
https://www.head-fi.org/threads/usb...ernet-chain-beats-all-at-least-for-me.829639/

 

[RB2013]

Just found the same guy has LT3045 based reg boards now. Lowest noise 500mA reg on the planet. PSRR is phenomenal as well. Run a cheap SMPS through this board and it would blow away the LPS-1 for only $22.

http://m.ebay.com/itm/LT3045-S-Ultralow-noise-0-8-Vrms-LDO-linear-regulator-0V-15V-500mA-Fixed-out-/252951737282?_trkparms=aid%3D222007%26algo%3DSIM.MBE%26ao%3D1%26asc%3D20150519202348%26meid%3D926fed1b86a0442785e7f06c458883d6%26pid%3D100408%26rk%3D1%26rkt%3D25%26sd%3D252996183940&_trksid=p2056116.c100408.m2460

For drop in replacement upgrade of TO-220 style reg's he made these:

http://m.ebay.com/itm/LT3045-78xx-Ultralow-noise-0-8-Vrms-linear-regulator-0V-15V-0-5A-78xx-Layout-/252948166542?_trkparms=aid%3D222007%26algo%3DSIM.MBE%26ao%3D1%26asc%3D20150519202348%26meid%3Ded26133e3a034327945a85ab33551148%26pid%3D100408%26rk%3D1%26rkt%3D25%26sd%3D252951737282&_trksid=p2056116.c100408.m2460

Amir should order some and directly compare to the LPS-1 connected to ISO-regen/Schitt DAC. Very simple to hook up. Just snip the DC cable of the Meanwell supply and solder it inline.

You mean Alexey from LDOVR? Good friend been building and selling these LT3045 boxes for months...posting about them since Jan...
Late to the game again..DB

Try beating 0.8uv noise and this PSRR chart:

 

[RB2013]

I've been stating on my thread - why do you need a $300 or the VR for much more when yo have a $25 LT3045 (with NO transistor boosters) and a cheap R-Core TeraDak DC-30W LPS (mine modded with ultra low impedance Nichicon HW caps - and 10,000hrs @105 life!) for $150.

Currently running 5 LT3045 boxes with 'Y' cables and custom made 6in quad star silver teflon tinned copper shielded DC cables to the various parts of my USB/Ethernet chain - with the best SQ I've heard from a digital source...or my old $30k analog rig.

All from a $500 Acer iCore 7/WIN10 PC running Foobar. Redbook WAV files.
No need for that $7,440 Mivera contraption - what a joke! Talk about a snake oil salesman...

 

[RB2013]

This is a fun video to watch - don't need a $20K Analyser to see/hear the massive AC leakage of that SMPS energizer

And the AC leakage of your 'Lab' quality LPS is likely not great either - or that very expensive Sbooster you recommend:
https://www.computeraudiophile.com/...-meanwell-teddy-pardo-uptone-js-2-and-others/

For the uninitiated 1mv=1000uv do the math...

Best to go with a cheap R-Core TeraDak DC-30W and a few $25 LT3045's.

 

[Mivera]

I've been stating on my thread - why do you need a $300 or the VR for much more when yo have a $25 LT3045 (with NO transistor boosters) and a cheap R-Core TeraDak DC-30W LPS (mine modded with ultra low impedance Nichicon HW caps - and 10,000hrs @105 life!) for $150.

Currently running 5 LT3045 boxes with 'Y' cables and custom made 6in quad star silver teflon tinned copper shielded DC cables to the various parts of my USB/Ethernet chain - with the best SQ I've heard from a digital source...or my old $30k analog rig.

All from a $500 Acer iCore 7/WIN10 PC running Foobar. Redbook WAV files.
No need for that $7,440 Mivera contraption - what a joke! Talk about a snake oil salesman...

Aren't you a vinyl fan? If so that proves your ears certainly shouldn't be used to determine how something should sound. Unless of course you keep it to yourself

 

[RB2013]

Aren't you a vinyl fan? If so that proves your ears certainly shouldn't be used to determine how something should sound. Unless of course you keep it to yourself

Oh you mean this:

Like 10yrs ago - when you were still wetting diapers. The digital I have today absolutely blows that away for a a fraction of the cost of a stylus replacement on the Dynavector XV1S. But what would you know about great analog sound...great anything SQ.

I see the orgasmic splunge of products on your website - all in the $3000-$7000 range. Funny how Amir never 'measures' them? I guess no one buys them to lend to him?

Anyway you have one thing right - what's so hard about measuring the DC noise on the LPS-1 that Uptone refuses to publish. I mean every DC LDO spec sheet has them and a PSRR graph- if a $20K AP analyser can't measure that what good is it? Might as well be a boat anchor.

 

[Mivera]

Oh you mean this:
View attachment 8328

Like 10yrs ago - when you were still wetting diapers. The digital I have today absolutely blows that away for a a fraction of the cost of a stylus replacement on the Dynavector XV1S. But what would you know about great analog sound...great anything SQ.

I see the orgasmic splunge of products on your website - all in the $3000-$7000 range. Funny how Amir never 'measures' them? I guess no one buys them to lend to him?

Anyway you have one thing right - what's so hard about measuring the DC noise on the LPS-1 that Uptone refuses to publish. I mean every DC LDO spec sheet has them and a PSRR graph- if a $20K AP analyser can't measure that what good is it? Might as well be a boat anchor.

Quit pretending to know what big numbers mean and stick to your vinyl.

 

[RB2013]

Quit pretending to know what big numbers mean and stick to your vinyl.

Blizard-Wizard (or what ever you call your self these days) you had me ROLTF LMAO with your LT3045 - "Hey look at what I found folks" post. Right after posting about the old school TI TPS7A700! Nice Ebay search - is that how you research your $7000 snake oil? You are such a loser - do people really buy that WAY overpriced crap you make. Such a joke.

 

[Mivera]

Blizard-Wizard (or what ever you call your self these days) you had me ROLTF LMAO with your LT3045 - "Hey look at what I found folks" post. Right after posting about the old school TI TPS7A700! Nice Ebay search - is that how you research your $7000 snake oil? You are such a loser - do people really buy that WAY overpriced crap you make. Such a joke.

Yes I was showing folks what they can buy for cheap to clean up ripple noise.

Way overpriced? Well according to one of the most respected guys on Headfi, it's $87000 cheaper than the next best thing!


https://www.head-fi.org/threads/pos...ne-rig-focal-utopia-mivera-superstack.838570/

There's a man who has a real set of ears. No vinyl at his place!

 

[RB2013]

https://www.head-fi.org/threads/xmos-xu208-usb-bridges-the-latest-gen-has-arrived.803111/page-241

Still peddling that "Look at the 12v Daitron. Only 1640uV ripple P-P at full load!. RMS is half. And only 40uA leakage current. There's no budget way to get this level of performance. However only $200 is a bargain compared to audiophile supplies with who knows what actual specs."

Wow only 1640uv of ripple switcher...