Soekris Dam Dac

[rickmcinnis]

I continue to look at the AD evaluation board - $70.00 with freight.

Then there is this kits thing where I see jkeny had asked the fellow a question MANY years ago:
https://www.circuitsathome.com/products ... interfaces

I think I like EVAL board better, though.

I like the ready to go aspect of the INTONA I wonder how much money I could spend getting the AD finished.

I like the idea that it is easy to use outboard power with the AD.

I was worried that the BRAINBOX might not work with the isolator in the chain.

[rickmcinnis]

Nigel,
Take a look at this when you have some time.

http://www.analog.com/media/en/technica ... UG-043.pdf

They mention a common mode choke to use that is no longer available.

Do you think there is a need for such a thing? Could you recommend a replacement value? Do you think this is critical (other than power)?

I would power each end separately which I would think would negate the need.

Any thoughts?

[jkeny]

Rick,
The problem with all these USB isolator devices is that they don't support high speed USB 2

[nige2000]

Put the dams isolator back on with cut i2s legs and bridge cross
All seems a little ott though

[rickmcinnis]

Jkeny,

Are you saying the AD device would not work with the BRAINBOX because of the less than USB 2 abilities? I would not think the umanager needs much in the way of speed but I have wondered if the BRAINBOX requires USB2 I would have a problem.

Nigel,

I did not remove the port isolator. Could not remember yesterday.

Are you saying you would prefer the isolated port AS IS compared to doing something to isolate before the "regular" port?

Of course, this is a case of obsession over a detail.

Have you listened to your DAC with umanager running?

I was getting some strange behaviour last night and found there are parameters one can set for the BRAINBOX within DEVICE MAMAGER. I set the CPU USAGE to minimum and locked down all of the serial port values and it seemed to solve the problem.

One can adjust latency, too, with the BRAINBOX. I figured low latency is meaningless here. I figure setting CPU USAGE low is the equivalent of highest latency. I guess that means it is sending big blocks of data instead of smaller chunks but since I am only changing polarity when needed, it seems low CPU usage is the way to go. Am I missing something?

I would think your speakers would reveal polarity differences relentlessly. Have you tried this yet? At this level the difference it makes is at least as large as our modifications.

[gstew]

Rick,
The problem with all these USB isolator devices is that they don't support high speed USB 2

JKeny,

You know this world better than I... would just an RS-232 connection ever use USB 2 speeds? I suspect not, so I suspect these isolators are quite adequate for this purpose.

And in this case, two sets of isolators (at both the USB and at the DAM) might be best.

Rick,

***EDIT - I just had the thought that with the isolators not oriented the same way and the lines not going to the nearest isolator input, maybe what Soren did was use three of the same isolator chip (2 inputs on one side of the chip going to 2 outputs on the other side), but flipped the orientation of one of them and brought both output lines to that chip? Nige, you traced the flow of the lines, is this right?***


Some thoughts on putting the isolators back for the isolated RS-232...

1. With four inputs and two outputs spread across three isolator chips, they are almost certainly not the same chip as each line is a one-way connection (input->output). Looking at the best picture I can find (HiFiDuino's), I cannot confirm the chip designations. I CAN do this tonight if you need me to do so.

2. Also looking at that picture I do see they are not oriented the same way.

3. Also can see in the picture that the lines may not go into the isolators in the same order as they are positioned. So you may need to put back two of the chips instead of just one. For example, for the isolator chip just across from the Isolated RS-232 ISO RXD IN and ISO TXD OUT locations, one of the inputs looks to be the ISO RXD IN and the other appears to be the I2S BCLK IN. I may be wrong here, have to ring out the connections to be sure.

4. You might also consider using the equivalent ILxxx isolator(s) instead of the SIxxx ones. The SIxxx ones use micro radio transmitters/receivers and people like John Swenson suspect they may spray some interference where the ILxxx ones use a more benign and likely quieter process. Once you find the equivalent, functionally they should be the same.

Now that you have confirmed the INVERT does flip signal phase, I will look at DimDim's Arduino setup to make this easier to do. I REALLY wish Soren could bring that function out to some existing pin instead of needing an RS-232 connection, but when I last looked, he has no spare pins.

Greg in Mississippi

P.S. I used this function a lot on my cMP/cPlay setup... and setup my tonearm outputs and phono stage inputs with plugs that are symmetrical so I can flip phase quickly and easily. I find this a MUST-HAVE for best sound.

[rickmcinnis]

Greg,

I did not remove the isolator for the isolated port. I bet I thought about it, though!

It does go in the "other"direction than the IIS ones and now I see why. I wondered why this was so.

So I have no need to replace anything.

In this application there would not be the problem with cascaded isolators that there would be with the music signal but my speculation is: do we want this connection so close to the music signal?

The fact is I do not hear any artifacts with the regular series connection. Nigel is correct, my audio obsession has resulted in getting shot as soon as my head rose above the edge of the trench but that happened years ago.

I figure with the AD device I can quell the voices that would constantly be asking me if I am hearing "something" from the umanager connection. I have lots of stuff I need to utilize somewhere, so they will power the AD board. Will power each "end" separately - should be pretty damn isolated!

One wonders if the BRAINBOX adapter adds some isolation due to the conversion?

EDIT: I cannot imagine a way for SOEKRIS to switch polarity other than the umanager. Now that I am used to umanager I like it just fine. Simple as can be other than having a computer in the way. I am hoping that the isolater will allow a longer USB cable so I can move it to the side. Might need to lengthen both cables for that matter.

[jkeny]

Yes, apologies to all for the misdirection - I see now what you are trying to do - controlling the Soekris thru it's RS232 port but using a USB to RS232 converter (Brainbox?).

Yes, I doubt the Brainbox uses USB 2.0 for this, never mind high-speed mode - so any isolator should work. But isn't this RS232 connection only needed for uploading firmware? Maybe I'm wrong but couldn't you just disconnect it & test if the sound changed?

The Intona is way overkill for this

[gstew]

Greg,

I did not remove the isolator for the isolated port. I bet I thought about it, though!

It does go in the "other"direction than the IIS ones and now I see why. I wondered why this was so.

So I have no need to replace anything.

Rick, you might need to add the one next to it back in... Looking at the pictures, the ISO BCLK IN appears to go to the isolator you left on the board. OF COURSE, I may be wrong on this, I only have access to pictures right now and cannot confirm by tracing the connections with my meter.

In this application there would not be the problem with cascaded isolators that there would be with the music signal but my speculation is: do we want this connection so close to the music signal?

The fact is I do not hear any artifacts with the regular series connection. Nigel is correct, my audio obsession has resulted in getting shot as soon as my head rose above the edge of the trench but that happened years ago.

I figure with the AD device I can quell the voices that would constantly be asking me if I am hearing "something" from the umanager connection. I have lots of stuff I need to utilize somewhere, so they will power the AD board. Will power each "end" separately - should be pretty damn isolated!

One wonders if the BRAINBOX adapter adds some isolation due to the conversion?

I think two levels of isolation would be good here. And I doubt the BRAINBOX adapter does isolation.

EDIT: I cannot imagine a way for SOEKRIS to switch polarity other than the umanager. Now that I am used to umanager I like it just fine. Simple as can be other than having a computer in the way. I am hoping that the isolater will allow a longer USB cable so I can move it to the side. Might need to lengthen both cables for that matter.

The RS-232 connection and the umanager are just a way to enable commands and data to be sent to the FPGA. He already has command interfaces that bypass the RS-232 and umanager with the input select function and the volume control. INVERT is just sending a yes/no binary type of communication to the FPGA similar to the input select, it could go that way just as well. BUT he has no free input pins to use at this time.

Greg in Mississippi

[gstew]

<SNIP>
<SNIP>

Yup, looking at Nige's modified DAM, it looks like the middle isolator chip handles the output streams (ISO TXD OUT & I2S MCLK OUT) and the upper and lower the input streams.

And since Nige did not jumper one of the input-output sets on the upper isolator chip pads, I bet that is the ISO RXD IN line.

So to use the isolated RS-232, I bet you'll need both the middle and upper isolators in place. You can still bypass the I2S line from input to output. But it won't be quite as nice as not having that device on the board.

An alternative solution might be better? Possibly an add-on board or just air-wiring a 2-way isolator in to handle the two directions of the RS-232 with connections from-to where you need them so that you don't have to add that isolator back in?

Greg in Mississippi