Iso regen

[scan80269]

As usual, thanks for this John. Just a quick question, I thought the Intona used only linear regulators, not switching? That's what's mentioned on their features page.

 

Not asking to be a smarty pants. Genuine query.

 

I'm only going off their website. Obviously if you have one and are looking with your eyes directly at a switching regulator then you win ! :-)

 

Cheers

Not John, but looking at a pic of the Intona circuit board, I see a total of six 3-terminal regulators in the SOT-223 package. At least several of these are LD1117 LDO regulators, which are linear. John is right in that there is a separate small power transformer used to couple power from the upstream side to the downstream side to feed the chips on that side. A transformer obviously does not work with DC so the circuit utilizes a switching frequency, which can be a source of noise. The AC voltage on the secondary side of the transformer is rectified and capacitor smoothed to DC to feed three LDO regulators on the downstream side of the board.

 

One can, in theory, make the Intona perform better by disconnecting the power transformer and externally feed in clean DC to power the downstream side, but it is far easier to wait for ISO REGEN to become available.

[scan80269]

3 hours ago, R1200CL said:

 

So why not just set LPS-1 to 7 VDC, and insert a resistor that drop the voltage to 5 on the SU-1. You are in any case modifying your SU-1, right ?

(You bypass the original SU-1 PS and use the LPS-1, or have I misunderstood how the SU-1 works)

 

You want the resistance as high as possible. I'm sure John can confirm this and suggest a value.

 

I suppose you already has verified powver consumption? (Thought I'm not sure if the ISO regen numbers has been released.)

 

Using a resistor to drop voltage from 7V to 5V is not recommended.  The voltage will change depending on device current draw (per Ohm's law).  Even using 3 diodes in series to realize a ~2V drop will not be all that good.

 

There should be no need to mess with 7V.  Set the LPS-1 to 5V to power the ISO REGEN, do VBus injection via USB cable into the SU-1 (to power the XMOS receiver module on upstream side of the moat) and finally the SU-1 downstream logic via a modification to bypass the internal linear PSU and run in 5V via a DC jack.  If you want to do this then you need a 3-way DC splitter cable with three 5.5/2.1 DC plugs wired in parallel.  Like John said, this scheme will bypass much of the isolation that ISO REGEN and SU-1 offer with their designs.

 

A more sensible power solution would involve at least two supplies, one for ISO REGEN, another for SU-1 (2 rails).  Perhaps a battery can be used to power the ISO REGEN.  The ultimate configuration would involve three LPS-1 supplies, one for each rail, to take full advantage of the double isolation.

 

[scan80269]

How about a cable like this:

 

https://www.amazon.com/QVS-HDT-0-5F-Shielded-Video-Network/dp/B01BSNT8SC/ref=sr_1_7?ie=UTF8&qid=1491253375&sr=8-7&keywords=HDMI+cable+6"

 

[scan80269]

45 minutes ago, tboooe said:

The plan is to power the SU-1 and supply VBUS with the LPS-1 but I have not received mine yet (I just placed an order for it a few days ago).  For now, I am using a Teknet 5v battery and it works just fine.  I wanted to verify that i could get the SU-1 working in this configuration before buying the LPS-1.  The SU-1 uses very little power by the way.  If the LPS-1 sounds better than my battery or linear power supply I will need to buy another one for the IsoRegen.  Ouch!

I measured a maximum of 111mA of current consumption at +5V for the downstream (I2S, coax, AES, etc.) side of my SU-1.  The current reading was highest when streaming 24/192 PCM out I2S (to my Spring DAC; I don't have any 24/384 content), so perhaps the 24.576MHz oscillator is a tad more power hungry than the 22.5792MHz oscillator??  The two oscillators are power-gated on the SU-1 so only one gets powered at a time depending on the incoming sampling rate.  I tried playing DSD256 (44.1K-based 11.2896MHz) but the current reading became slightly lower vs. 24/192 PCM.  Please note my SU-1 has been modified with a Belleson Superpower regulator so the current reading can well be a bit different from a stock SU-1, even when powered with the same DC +5V power source.

 

Instead of using two LPS-1, perhaps you can keep just one to power the downstream side of your SU-1 exclusively, and use a decent linear power supply or battery to feed the ISO Regen when it arrives.  The USB input (upstream) side of the SU-1 has galvanic isolation (i.e. moat) from the downstream side, so powering the upstream side with another LPS-1 may not yield much of an incremental benefit.  I suppose it can be tempting to try it anyway, though.