Phisolator

Ha ! what would this be eh ?

The USB Isolator (PCB) I have been implying in between quite some lines.
But didn't I say that it did not want to work ?
yes I did. Well, sort of.

Because of unreliable results, I had to take out the prototype I had been using for two months or so, out of my chain. I mean, I still think it is without sense to have me listening to some better SQ for one or the other reason, with that making all SQ observations on behalf of you all, moot.
It didn't want to work - bad luck.

Sadly (for me), I missed the sound from it too much, so things were itching. D*mn, I *had* to get the thing to work !
So I can see by means of some invoice that from of Sept 1 we have been working on it again, quite continuously. Try this, try that. Shortcut this, let explode that. And yes, you may have seen me writing about blowing up something in my NOS1a only a small week ago. That was my USB Interface.
But exactly that thaught me how to go about with it all, and now it works.
AND HOW !

I have it now built in my NOS1a (per yesterday morning in definitive version);
The sound is super special and in general I'd like to call it "sparkling". But I think that it is better to express it as : all is 10 times more emphasized. Could it get more crazy again ?
So Yes.
The sound is so super extraordinary clear that it is jaw dropping. I'll also insert this text which I yesterday (when I wrote the main part of this post) scratched because I thought it was overdone. But yesterday evening I ran into "the thing" again, so I re-inserted that text :

This I had running yesterday and I was blown out of my socks. So ?

Think a moderate louder level (so really not crazy loud). Then there's this hit on the snare throughout the 8 minute or so track. Say once per 1,5 - 2 seconds this hit.
Side note : yes, this undoubtedely is working in combination with the B'ASS and with the notice that the last reincarnation of that was developed with this same Isoator in my chain, with same effect but 10 times less emphasized (different configuration back then). So :
At the first hit on that snare I was blown out of my socks. Not because so crazy loud or dynamic, but because it was so realistic for loudness in the middle of the other instruments. So yes, it *was* dynamic and yes it *was* crazy loud. But now this :
At the second hit (just 1,5 second later) I had landed back into my socks, but was blown out of them again.
And again. And again. And 450 seconds and 300 hits later, the same was still happening. Minutes earlier I already started to explicitly wonder how it could happen that I kept on being surprised. I mean, after 10 hits you know how they go, right ?
No, somehow not. I don't know what kind of dimension I have entered, but it must be about some vision on the artists, playing their instruments, and although I don't want to "see" at all, it just happens and now this new dimension makes it super much more interesting, if you want it or not.
Didn't I tell the other day about seeing a drummer jump in the air to let land the sticks on to the drum with his full body weight behind it ? that. But that was a few days ago and now this is worse (better). Yesterday I saw that all the time throughout, never mind what was playing. You now see *how* it is played.

Thus, I initially scratched this text, but during last night's listening session it kept on occurring to me as *the* feature of the Phisolator; The text deserves to be there ...

Another very clear virtue is the "shortness" of all. Super super dry. Just what a snare drum likes to have. But a good vibrating bass no different.
So Yes, try to envision how sparkle emerges from super dry. This is just the twinkles etc. being sufficiently isolated; no smear happens now.

Like with the versatility of the B'ASS and offering people the means of going back to the original situation, here too I would like to offer the facility to compare old with new. With an external box (small case) that facility would be available naturally, but this is of course not so when it is built in the NOS1a's case. Besides that, USB is so fragile that it better not be mangled with by means of rambling in the case. It should go by means of official connectors, you just swapping cables.

Setup

At the place where now 2 USB inputs are visible in the NOS1(a)'s case, one USB terminal will be added and the one for the not-operative display will now also be utilized for the Phisolator setup.
Mind you, the configuration is such that you can keep on using the normal connection, electrically the same as it always was. Why is this ?

1. So you can go back to that situation, might you want that or migt another USB Isolator come available;
2. So you can compare the two different means;
3. So you can compare outside of the B'ASS upgrade.

Ad 3.
Yes, of course we are going to combine this upgrade with the B'ASS upgrade ! But either changes the sound so much that I think you really want to know which causes what (hence #2).

With the notice that nobody should be worried by additional connectors or pieces of USB wire/cable (like you also are not worried about the additional connections the Intona implies), this is how the connections go :

Envision 3 USB terminals, A, B and C.
C = Direct Input (coming from PC).
B = Output from internal Isolator.
A = Input to internal Isolator (coming from PC).

We have a direct Connection from PC to the regular USB Interface, as how it always has been (and include an Intona in the path if you like, only to indicate what "direct" means in this case). So the PC connects to terminal C in this situation. Thus :
PC Cable -> C Input -> USB Interface.

We now take the provided small USB Cable (explicitly not a Clairixa) and loop through from connector B to connector C. Meanwhile we put the USB Cable from the PC to connector A. Thus :
PC Cable -> A Isolator -> B Output -> Very short USB Cable -> C Input -> USB Interface.

The lot is set up such that no user mistakes can be made with the connections, with the knowledge that the only mistake possible is that you put a USB Cable from A to B which would connect the output of the Isolator to its Input. The means to protect from this mistake will be the length of the provided USB Cable (think 10cm/4") and the positioning of the A, B, C terminals. Thus, if A and B are outside with C in the middle, then the very short cable will be as short as the distance from B to C requires, that being too short to go from A to B. A to C is not possible anyway because no connectors will fit to both (it would require an USB-B to USB-B cable and normally they don't exist).

Technically

What the Phisolator does, is first of all isolating the signal galvanically. Nothing different here from e.g. the Intona. What the Phisolator does NOT, is "processing"; all it does is passing on the USB data (back and forth) and take care of the (handshaking) protocol.
What it does foremost is recondition the downstream signal (from Isolator to DAC). Still it does not really re-generate the signal; it operates quite similar as the NOS1a's internal isolation but now not in the i2s path but in the USB path (might people recall, there is no reclocking in order in the "a" version as such). The "a" means also implies that the power regulation is completely separated on both sides of the isolation barrier (the output/clean side does by not any means depend on the input/dirty side, for voltage as well as for ground). A difference with the "a"'s internal isolation is that this regulation is again lower noise and the lowest possible these days (2016).

Because of the in-DAC application, the grounding could be made as special as I could think of, and then referring to the now three different electrical circuits (each galvanically isolated from the other) - PC to USB Input, USB Input to i2s, i2s to D/A + Analogue out.
What I regard crucial is that where the signal is electrically regenerated for sure, the means doing this now only needs to drive the total of 20cm or so of USB wire with no further downsides of radiation of any kind (I check that these days by means of meters for it). This (no radiation) is to be compared with isolation transformers which for DC Voltages are true AC switchers (at a rate of say 80KHz to maybe 500KHz, depending on the type). There is NO isolation transformer in there.

The future

Believe it or not, but the Phisolator configuration is future proof. How ? well, it can be updated and it can be updated by yourself. Not that I know of anything at this moment, but when time progresses we will see that most proably I will be able to think of new means within the same setup. Thus, we'll have a new PCB produced, we'll apply some soldering to it and for maybe 150-200 euros you will have the upgrade with low shipping costs because so small. You just replace the Phisolator with the new one, at the time.
So only the first time is more expensive because of the initial assembly work. Any next time is 10 seconds of work (no soldering) plus a few minutes for taking off and back on the lid.

I am confident that the future of improving SQ is in the USB Interface.

Price

Let me first notice that there's a crazy expensive chip on there - 80 euros. So there it starts.
PCB + further parts maybe costs 40 euros (including the additional USB terminal for in the chassis) so let's say that 120 euros is the cost price of the Isolator parts. Soldering ? maybe 30 minutes per board.

The soldering / assembly in there ... I say 60 minutes for everything and this includes the changes (USB wiring) to the existing USB Interface and changing the output voltage of the PSU in the left leg.

We must make that very short USB cable of which I say it takes (apart from a 1$ USB cable) 30 minutes, including testing it (en masse).

Testing (analyser) will be 30 minutes,

Because things can fail, I count 30 euros for that (now one out of 10 can completely fail (USB is really fragile stuff) and it is covered for).

Counting the numbers of the blue texts above, I count 150 minutes and at a moderate rate of 60 the hour this comes to a total of 300 euros including the cost of the parts and coverage for failure.
Because I like a bit of profit for all the trouble of designing and trialling and what not what actually can't be changed into $ when keeping it civilised, I add 50 euros for profit. That's 350 euros in total (ex VAT for those liable to VAT).

Availability

There's a small catch here;
In the text above I suggested that we "of course" combine this with the B'ASS upgrade. The PCB's are available already, but the chips have a lead time of 8-10 weeks, except 31 of them which we will have very soon (can be this week). This is only a few more than the current B'ASS orders, also knowing that the PCB's for the B'ASS are in the DHL air plane right at this moment (finally). This means that on a first come first serve base, only a few more orders for the B'ASS/Phisolator combination can be served from that first batch of chips. After that it will be waiting for a next batch of 25 until the end of January 2017.

I am assuming that everyone who is on the list for the B'ASS also wants to be on the list of the Phisolator. Still I obviously like to know it explicitly. So please apply in this topic for the Phisolator, which is also nice for those who may or may not be served by that first batch of 31 (start counting).
For the sequence of upgrading we will use the original B'ASS list, unless you are not on that list and don't want to be either. This also implies that when you are not on the B'ASS list but like to have the Phisolator (and now B'ASS as well, or not, which does not matter), you really need to be the fastest or otherwise someone else beats you to it and you must wait until the end of January at least.

Questions ? don't hesitate.
But man, what am I glad that this worked out !

Peter

	Author	Topic: Phisolator (Read 51337 times)
0 Members and 2 Guests are viewing this topic.