A novel way to massively improve the SQ of computer audio streaming

[[email protected]]

On 10/8/2019 at 5:14 AM, Superdad said:

The technique is hiding in plain site in the data sheet.   John has wired LT3045s both ways and measured about half the output impedance when it is done right.  For some reason Alexy at ldovr does not believe it. Oh well...

possible reasons

it is almost useless
it is harmful

i belive that you can get lower impedance with ILIM circuit, but i prefer have lower noise 

[[email protected]]

On 10/15/2019 at 7:12 AM, sandyk said:

 

At those kinds of already very low noise levels , a LOW and FLAT Output Impedance over a very wide bandwidth is more desirable with PC Audio than the ultimate in noise level. This applies just as much to Digital areas as it does to Analogue Audio areas.

( My apologies for the Off Topic response.)

Alex

Again, In my humble opinion, ILIM drop compensation is USELESS... I have only following statements from @Superdad
1. Output impedance for LPS 1.2 is a "FEW" mOhms
2. ILIM Circuit is a “easy way to cut the output impedance nearly in half”

So – what is the “few mOhms” in figures??? Should I consider it as 10mOhm with ILIM? And <20mOhm without ILIM? You need to add also 2 connectors and feeder cable resistance and easily can have additional 50-100mOhms impedance at the point of load - difference will be 50 vs 60mOhms or even 100 vs 110mOhms between ILIM/No ILIM options.  

[[email protected]]

1 hour ago, sandyk said:

Alexy

I suggest that you discuss this with Alex Crespi.

I am only reporting my own observations

Regards

Alex K.

i will be happy to discussi it with @Superdad but very often he is ingoring my questions
i really want to know more details about measurements they did, see diagramms available if any, equipment used for the measurements etc

 

[[email protected]]

On 10/24/2019 at 9:03 PM, mourip said:

Consider starting another thread for technical arguments or else contacting folks by PM?

well, i decide to start new thread

 

[[email protected]]

4 hours ago, seeteeyou said:

For owners of Chord DACs who don't have HMS yet, it doesn't hurt to give HQPlayer 4 a try since this HMS owner reported similar performance between the hardware and software solutions recently

https://www.head-fi.org/threads/hugo-m-scaler-by-chord-electronics-the-official-thread.885042/page-650#post-15365237

 

https://www.head-fi.org/threads/hugo-m-scaler-by-chord-electronics-the-official-thread.885042/page-652#post-15369740

 

https://www.head-fi.org/threads/hugo-m-scaler-by-chord-electronics-the-official-thread.885042/page-653#post-15370054

 

https://www.head-fi.org/threads/hugo-m-scaler-by-chord-electronics-the-official-thread.885042/page-653#post-15370114

 

HQP settings were already linked above, though @ray-dude and others might also share their own experiences at some point?

 

It's getting more affordable than ever, one NUC for running Roon Core, another NUC (i3/i5/i7 ones are working just fine according to @dmance) for running HQP, and then add RPi 4 (or even Allo USBridge Sig) as NAA. Just get rid of the battery inside Chord Mojo and use it as desktop DAC.

Hi

I have limited abilities to comment here due to my internet access limitation. 

Please note that above calculations valid only for the "measurements were taken in still air on a 4 layer FR-4 board with 1oz solid internal planes and 2oz top/bottom planes with a total board thickness of 1.6mm. The four layers were electrically isolated with no thermal vias present." but not valid for dxp

Very efficient heatsink enclosure (7.32C/W), Copper Heatbridge with thermal trasnfer tapes 10x35x5mm (<2C/W), Unique Thermal Vias Pattern and PCB Layout, all this allow you to have upto 4W dissipated power  for single stage and up to 6W for dual stage PSU (with equal heat dissipation for each board).

By other word it is safe to feed 5VDC out single stage even with 9VDC in if output current do not exceed 1A
 

 

[[email protected]]

18 hours ago, kennyb123 said:

It’s coming up on three weeks since I placed my order for a pair of those boxes and yet my order status is still processing.  Website indicates a 1-2 week lead time so holiday travels would certainly explain the delay.  Sure would be nice to get a status update from @[email protected].  

 

i am apologies for the delays, but i can start shipments by the last week of January only , i can not get enclosures faster, Chrismass and New Year holidays affecting deliveries

[[email protected]]

5 hours ago, RickyV said:

 

Has anyone already done this? It is use as a second regulator for the ER. All through holes connected. Especially the SET through hole.

 

Looks nice

Yes,  someone already tried.... 

[[email protected]]

3 hours ago, austinpop said:

the DXP-1A5DSC units can and do improve regulation, so can improve the SQ of an existing PSU. But the driving PSU's quality still matters. So, using a DXP-1A5DSC with an el cheapo Chinese LPS will not be the equal of the DXP-1A5DSC with a higher grade PSU. Sorry - I don't have direct experience comparing LPS-1 and LPS-1.2, both with 

Hi Rajiv

 

You are right - driving/feeder is important for any PSU, even Uptone LPS-1.2 is sensitive for the feeder type. Several users confirm that its works better with a Linear PSU as a feeder vs stock Meanwell SMPS (in therory it should not make a difference because the dual stage linear regulator of lps1.2 is powered by isolated supercap bank/and not by the drive/feeder PSU). It might be interesting if someone can make a direct comparison of cheap Linear PSU (like STUDER900 15W Regulated PSU) followed by DXP-1A5DSC vs Uptone LPS-1.2 with stock SMPS

 

3 hours ago, austinpop said:

The DXP-1A5DSC cannot help you with current capacity, and the innate ability of a PSU to deliver current. So with LPS-1 or 1.2, you're still limited by the 1.1 amp capacity. I would go further and say I prefer LPS-1.2 only when they are run below capacity, at < 1a.

Again - it is true, tripple LT3045 can provide 1.5A max (and dual LT3045, like LPS1.2 - 1A only and not 1.1A, @Superdad confirms that LPS1.2 can provide 1.1A, but  accordind to the regulators specs TPS7A4700 and dual LT3045 used in LPS1.2 is rated for 1A only and higher rating is purely marketing, internal LT3045 current limiter configured for 0.7A, so dual lt can provide 1.4A and triple LT3045 upto 2.1A, but can not provide listed specs for the noise and PSRR

 

4 hours ago, austinpop said:

The DXP-1A5DSC will limit you to 1.5A, irrespective of the current capacity of the PSU. You also need to be careful about the voltage drop from input to output - Alexey's chassis is rated to dissipate 5W max. I believe. Check with him before you buy.

it is abit complex, all internal parts of DXP are rated for 100C+ operation (125C for LT3045, X7R MLCC and Tantalum caps are rated for 125C, Polymer Caps for 105C), 3W per stage is a safe margin, but with 6W of dissipated power the chassis temperature can reach 70C+ and 6W per stage can burn the internal parts and might require expensive repairing (like 1$ parts replacing)

 

4 hours ago, austinpop said:

you cannot run LPS-1 (or 1.2) in parallel. @Superdad has posted about this quite often.

you can not parallel LPS1/1.2, but can parallel 2xLT3045A  LDOVR boards and feed them with 2 LPS PSU independently

[[email protected]]

6 minutes ago, Superdad said:

"Purely marketing" ?!  Are you accusing us of lying?  Would you like me to post photos of an UltraCap supply happily cranking out 1.15A for hours on end?

Alex, as per LT3045 specs it is 1A MAX, probably you can send an email to Analog devices and ask to revise the specs as per your measurements

 

 

 

and about the current limit, can you confirm how many ampers can be provided with 215Ohm ILIM resistor (and it is not 0.55A per IC)

 

 

 

[[email protected]]

11 hours ago, Superdad said:

Really Alexey? 

 

Yes

I am able to get 0.7A from a single LT3045 (internal current limiter configured for 695mA) but is it the right way to use it? My 5seater car can easily carry 7PAX, but i dont think that i can sell it as a 7seater van.

Even your beautiful  LPS-1 (Supercap followed by single TPS7A4700) rated for 1A only, but LPS1.2 (supercap followed by the same TPS7A4700 and dual LT3045) rated for 1.1A. I can not find any logical explanation why the old LPS-1 is rated for 1A only...

 

[[email protected]]

8 hours ago, Superdad said:

I was not kidding when I said you could easily need several a 5~7 amp supplies. I'd be kidding if I suggested you attempt it. 

5-7A!!!  Are you serios?? 

 

i think problem not with the current limit but with the way of external regulator connection, if onbord regulators not disconnected from the load (reverse current protection can cause the problem). 

 

@RickyV How did you connect Lt3045 regulators to the BS??

[[email protected]]

22 minutes ago, Superdad said:

 

Well the near 20-year old 16-port with integrated Gigabit copper PHYs Broadcom BCM53343 is not a low-power design. It needs multiple supply rails: 3.3V, 2.5V, 1.8V, and if a lot of its ports are used it will add up a LOT of current. We see switches with 5-10 amp DC-DC buck regulators (fronting several sub-regulators all the time.

For the Melco switch, they blocked off (turned off) 8 of the 16 ports, and it is near certain that part of the reason they set the registers for 4 of the 8 ports to be 100Mbps-only was to save power.

So even if I slightly exaggerated, there is slim chance that a single LT3045--or even a paralleled pair--is going to be enough for some of the rails of that switch. 

Yes

 

as per switch specs max power consumption about 11W for all rails

And probably 2A is not enought for 1V rail but it is impossible to have 5+A @3.3V

 

 

[[email protected]]

28 minutes ago, Cooler said:

Thats actually a half solution, as I remember there is no direct connection between eR case and chip, which actually generates so much heat.

Exactly, all this heatsinks absolutely useless, melting point of aluminium about 660C

[[email protected]]

2 hours ago, barrows said:

Of course the heat sink on the case is helping, if it was not drawing away heat, then the sink would remain at room temperature!  I realize that this approach is not the "best" way to cool the internal chips, as these chips are not in direct contact with the case.  But, one has to consider that the air inside the case is transferring heat from the chips, to the case, and hence to the heat sink on top of the case.

Well, I told it is useless because the way used not very efficient, heatsink attached can reduce the case (and internal temperature) for a few degrees, but the ic still can be very hot because it is not attached to the chassis and the heat dissipated not transfered to the chassis/heatsink, below you can see an example of proper design, but I don't think it can be implemented with the chassis used 

 

As you probably know - air is a good thermal insulator (with very low thermal conductivity) and even 0.5W of dissipated power can cause local overheat for the IC (not a case temp, which basically depends on total dissipated power inside). And if you drill some ventilation holes and allow air circulation, you can reduce internal and case temp much more efficiently