CaptureWidget - Page 4 - diyAudio
Go Back   Home > Forums > Design & Build > Equipment & Tools

Equipment & Tools From test equipment to hand tools

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 14th March 2013, 01:41 AM   #31
davada is offline davada  Canada
diyAudio Member
 
Join Date: Jul 2004
Location: Fort St John, BC Canada
Hi Demain,

That's a big system. It's going to take some time to bring this all together.

Do you have measurements of USB power from various laptops?
Is it really clean enough? There must be some switchers and or charge pumps somewhere I would think.

Why don't we make it a variable stepped gain block instead of a fixed gain. It seems redundant to go the expense of multiple ADCs just to drive the noise up 15dB. You've seen what happens when you increase the gain on the EMU from a point of reference. If we extend the measurement range by increasing the DR and SNR then it's unlikely amplification would be necessary. It doesn't really work anyway. It just drives the noise floor up.

If possible I would really like to operate the input op amps with a healthy +/- 15V. They just work better at a higher voltages and it broadens our range of choice. These supply voltages are rather dangerous for an ADC supplied by 3 and 5V. So we must have protection. I really don't like shunt protection directly on the signal path particularly considering our target. I've thought of different form of protection you can give some thought to. If we use a window threshold detector at or near the ADC drivers we can throttle the voltage regulators supplying the op amps to a safe level. We can trigger slightly above the maximum input level. We can keep the rail voltages with in an operational level so long as it's with the maximum rating of the ADC,s. I'm sure we can come up with some non intrusive crowbar at the input to force a fuse to open for those emergency situations.

Cheers,
__________________
David.
  Reply With Quote
Old 14th March 2013, 01:55 AM   #32
diyAudio Member
 
dirkwright's Avatar
 
Join Date: May 2011
Location: Virginia
You can study the circuit diagram for the PCM4222EVM. They used 15V supplies for the OPA1632 and 5 and 3.3V for the digital, all on the same board.
http://www.ti.com/lit/ug/sbau124/sbau124.pdf

I thought you were just designing an ADC and was not going to include a DAC?
  Reply With Quote
Old 14th March 2013, 02:26 AM   #33
davada is offline davada  Canada
diyAudio Member
 
Join Date: Jul 2004
Location: Fort St John, BC Canada
Quote:
Originally Posted by dirkwright View Post
You can study the circuit diagram for the PCM4222EVM. They used 15V supplies for the OPA1632 and 5 and 3.3V for the digital, all on the same board.
http://www.ti.com/lit/ug/sbau124/sbau124.pdf

I thought you were just designing an ADC and was not going to include a DAC?
Hi dirwright,

What Demian has presented has been on the table for quite some time. He's just presenting it. The ADC is a good starting point. It all has to be assembled in peaces and proven out. It all begins with ideas and discussion and we might find that some of our ideas just don't work. We have to test them. Our aim is to exceeded the performance of the PCM4222EVM.

I already have it mind to try the TI stuff like the OPA1632 and such. But I don't have anything on the bench to work with. Still waiting on some boards.
__________________
David.
  Reply With Quote
Old 14th March 2013, 02:57 AM   #34
davada is offline davada  Canada
diyAudio Member
 
Join Date: Jul 2004
Location: Fort St John, BC Canada
An alternative to galvanic isolation of the I2S, clocks and control lines could be to use s/pdif or more like AES. TI has some DIX transceivers at a reasonable price and footprint which exceed our needs. Aside from handling the audio stream they also provide a path for control data as well. This would simplify isolation at the cost of additional hardware. I think the cost would balance out anyway. This would provide us with a balanced transmission line for a board to board link. Since it's just a board hop we shouldn't have problems with the known issues. I think they come with PLL as it is.
__________________
David.
  Reply With Quote
Old 14th March 2013, 05:10 AM   #35
1audio is offline 1audio  United States
diyAudio Member
 
Join Date: Mar 2004
Location: SF Bay Area
Blog Entries: 3
The block diagram is a bit of a wakeup call. I have these DC-DC converters now. I'll check to see how well they work and how noisy they are.

Actually SPDIF would make the USB interface simpler. It could be as simple as an AB1.2 with an spdif receiver. AES in practice is SPDIF with higher voltage, using the same receiver. We would need to set up an SPDIF receiver and have it auto lock to the external device. The Widget code and the driver would need to understand the sample rate coming from the device (its not unusual for pro level stuff), For ADC applications jitter on the line is far less of an issue. However the ADC device still needs power, clocking, sample rate control etc. as well as the analog stuff. The nice part is that it can be switches and the Widget doesn't need code to manage it.

Input overload etc. is another chapter to figure out later.
__________________
Demian Martin
Product Design Services
  Reply With Quote
Old 14th March 2013, 05:32 AM   #36
davada is offline davada  Canada
diyAudio Member
 
Join Date: Jul 2004
Location: Fort St John, BC Canada
Quote:
Originally Posted by 1audio View Post
The block diagram is a bit of a wakeup call. I have these DC-DC converters now. I'll check to see how well they work and how noisy they are.

Actually SPDIF would make the USB interface simpler. It could be as simple as an AB1.2 with an spdif receiver. AES in practice is SPDIF with higher voltage, using the same receiver. We would need to set up an SPDIF receiver and have it auto lock to the external device. The Widget code and the driver would need to understand the sample rate coming from the device (its not unusual for pro level stuff), For ADC applications jitter on the line is far less of an issue. However the ADC device still needs power, clocking, sample rate control etc. as well as the analog stuff. The nice part is that it can be switches and the Widget doesn't need code to manage it.

Input overload etc. is another chapter to figure out later.
I play around with the input overload ideas when I get things set up.

For now just get the widget assembled and working. Digikey as usual beat Canada post even though the widget board went out first.

If the clocks are on the ADC board can the widget get the clock off the SPDIF?
I was thinking the widget would handle the control for the sample rate through an SPDIF foldback. That's why I suggested the DIX over just a DIT.

I thought AES differed by specifying transmission line impedance and that it's balanced over a shielded twisted pair. Doesn't matter the TX does both.
__________________
David.
  Reply With Quote
Old 14th March 2013, 03:53 PM   #37
1audio is offline 1audio  United States
diyAudio Member
 
Join Date: Mar 2004
Location: SF Bay Area
Blog Entries: 3
Actually there are two AES formats, but he common one is 110 Ohms differental at 2V vs 75 Ohms at 1V. Not well known is that the comparitor in the receiver will have more jitter with more overdrive. Actually for the ADC side there is no real downside using Toslink since the jitter in that path really only affect data integrity.

The links are one way so there is no reverse path for control or clock. The Async USB module ensures lowest jitter on the DAC side with local precision clocks. While it almost feels like a cop-out this solution combined with the active balanced output may be a good solution. Its also much quicker to make. The ADC can be based on the reference design, which needs some simplification.
__________________
Demian Martin
Product Design Services
  Reply With Quote
Old 14th March 2013, 06:34 PM   #38
davada is offline davada  Canada
diyAudio Member
 
Join Date: Jul 2004
Location: Fort St John, BC Canada
Quote:
Originally Posted by 1audio View Post
Actually there are two AES formats, but he common one is 110 Ohms differental at 2V vs 75 Ohms at 1V. Not well known is that the comparitor in the receiver will have more jitter with more overdrive. Actually for the ADC side there is no real downside using Toslink since the jitter in that path really only affect data integrity.

The links are one way so there is no reverse path for control or clock. The Async USB module ensures lowest jitter on the DAC side with local precision clocks. While it almost feels like a cop-out this solution combined with the active balanced output may be a good solution. Its also much quicker to make. The ADC can be based on the reference design, which needs some simplification.
Which would use the least amount of power the optical or Xformer coupling?
For the ADC there would have to be two links id the the rates are to be controled for software. One in and one out. We can get this in a single DIX package and use the receiver for control data. TI has demo boards for these. Something to play with. For FFT we don't necessarily have to have a continuous uninterrupted stream and ismall delays don't matter. For a DAC it's a different matter. We can approach each differently if need be.
__________________
David.
  Reply With Quote
Old 14th March 2013, 06:54 PM   #39
1audio is offline 1audio  United States
diyAudio Member
 
Join Date: Mar 2004
Location: SF Bay Area
Blog Entries: 3
To keep life simple and end up with reusable blocks I would use an SPDIF-AES interface on the ADC and not bother with computer control. None of the software I know of can control the ADC automatically (except for the Juli@ in XP on Praxis). The Windows sound engine makes it a mess. Just use manual controls, which is how you would use it for a recording interface. Adding a Toslink output as well is trivial so why not?

However if we can get someone to figure out using a pic/arduino/etc. to control the ADC from a serial port equivalent that has attractions.

In measuring I usually get it all working at 192K and leave it. Digital measurements can benefit from other sample rates but the analog stuff really doesn't.

If the ADC box has a switch with 6 positions for sample rate and an internal-external clock switch I would consider it done for now. A separate control for input sensitivity is important and the issue of calibrated input is something to address. This is where a really stable low distortion oscillator comes into play with a known output level. Although it could even be a square wave counted down from the master clock and still have a known peak value pretty easily (a la Praxis).
__________________
Demian Martin
Product Design Services
  Reply With Quote
Old 14th March 2013, 09:13 PM   #40
borges is offline borges  Norway
diyAudio Member
 
Join Date: Dec 2003
Location: Oslo, Norway
SPDIF TX is already in place. See Q N K T C USB-I2S Module and Analog Board It implements trafo-coupled SPDIF. Toslink or AES/EBU (the physical side at least) should be easy.

Do we need SPDIF RX for music data? (I believe an SPDIF protocol analysis tool is too much to ask.) SPDIF RX is tricky, mainly because the data master is also the clock master. Now, I have experience with the SRC4392 async sample rate converter, but part of the point may be bit-correct capture which it would not give unless used as a clock regenerator. And if the SPDIF RX is the clock generator, that means the USB interface and PC must adapt to that sample rate.

So which side determines the recording clock? I must admit I know very little about the ADC side (device -> host) side of USB audio. May the device change the sample rate or request a sample rate change?

For DACs the host sets the sample rate (at least that's what is done today). I believe it would be inconvenient to have a DAC and an ADC run on different sample rates. I'm not too keen on a DAC-less board even though an ADC is the priority.

Demian, in your sketch you have the clocks on the digital side of the isolator. Is that correct? IMHO the clock is a very analog signal and belongs on the same side of the isolator as the converter chips. The digital interface handles jitter much better than the human ear :-)


BÝrge
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 07:31 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2