There has been a lot of discussion about tweaks and improvements that people have made to my original sound card interface PCB. The thread about the original is here:
http://www.diyaudio.com/forums/tubes-valves/155405-test-measurement-interface-soundcard.html
This got me wondering of it would make sense to do an update/improvement to the original design?
I thought it might be best if I started a new thread for the purpose of gathering suggestions for improvements.
I'm not committing to do this yet, but if there is enough interest, it might make sense.
So please post your suggestions. Especially valuable are those that you have implemented on the original board, and found an improvement in performance.
Thanks!
Pete
http://www.diyaudio.com/forums/tubes-valves/155405-test-measurement-interface-soundcard.html
This got me wondering of it would make sense to do an update/improvement to the original design?
I thought it might be best if I started a new thread for the purpose of gathering suggestions for improvements.
I'm not committing to do this yet, but if there is enough interest, it might make sense.
So please post your suggestions. Especially valuable are those that you have implemented on the original board, and found an improvement in performance.
Thanks!
Pete
My suggestions.
- True balanced input with XLR
- Power supply improvements. I experienced common mode switching noise originating from the Murata DC-DC converter (as mentioned in the original thread).
- Shielding.
Input protection. See the ideas populating the Quantasylum 400 thread.
I second the notion of balanced input. You can configure OPA4140's in parallel for very low noise.
Much of the switching noise can be eliminated with a high quality 3-legged regulator, albeit at the expense of some overhead voltage.
I second the notion of balanced input. You can configure OPA4140's in parallel for very low noise.
Much of the switching noise can be eliminated with a high quality 3-legged regulator, albeit at the expense of some overhead voltage.
Balanced inputs: with a restricted common mode voltage range (like +/-15V) balanced is not difficult. But if something larger (like 200V) is needed, it gets difficult, and expensive. If I recall, Audio Precision actually used transformers and some fancy circuitry to linearize them. What is needed (and what are we willing to pay for it)?
Protection: The existing unit is protected pretty well, at least the input is. I've applied 200V to the 200mV range without any problem. What more is needed?
Pete
Protection: The existing unit is protected pretty well, at least the input is. I've applied 200V to the 200mV range without any problem. What more is needed?
Pete
* Balanced input
* Higher input impedance option
* 20mV range (Lowest on my wish list)
* Flexible reference connection option using other channel input from sound card.
* 1:1 transformer isolation option on sound card IO connections (compact Lundahl audio transformers) to deal with conducted noise from the PC. (No ground connection between PC and interface, interface floats and is grounded to the DUT)
* common and diff mode filtering on input/output sides of DC to DC converter
* Higher input impedance option
* 20mV range (Lowest on my wish list)
* Flexible reference connection option using other channel input from sound card.
* 1:1 transformer isolation option on sound card IO connections (compact Lundahl audio transformers) to deal with conducted noise from the PC. (No ground connection between PC and interface, interface floats and is grounded to the DUT)
* common and diff mode filtering on input/output sides of DC to DC converter
* balanced input
* 20 mV input range
* Sound card output switchable attenuation: useful when measuring for crossover distortion (very low high resolution input signal for amp required)
* USB control to facilitate automated measurements (for example to measure distortion vs power)
For this USB control part, I would be willing to write the software (using http://www.waitingforfriday.com/index.php/Open_Source_Framework_for_USB_Generic_HID_devices_based_on_the_PIC18F_and_Windows)
Maybe USB control can also lower the cost as no front panel with knobs and switches is necessary. PIC controller can be powered by USB (including the small relays for switching), analogue section of course needs a separate power supply to avoid damage to the pc.
* and maybe a complex one: the bob Cordell distortion magnifier (I have a “separate” one in the making)
* 20 mV input range
* Sound card output switchable attenuation: useful when measuring for crossover distortion (very low high resolution input signal for amp required)
* USB control to facilitate automated measurements (for example to measure distortion vs power)
For this USB control part, I would be willing to write the software (using http://www.waitingforfriday.com/index.php/Open_Source_Framework_for_USB_Generic_HID_devices_based_on_the_PIC18F_and_Windows)
Maybe USB control can also lower the cost as no front panel with knobs and switches is necessary. PIC controller can be powered by USB (including the small relays for switching), analogue section of course needs a separate power supply to avoid damage to the pc.
* and maybe a complex one: the bob Cordell distortion magnifier (I have a “separate” one in the making)
This would be very convenient when using it as a general-purpose troubleshooting tool, but I wouldn't want this feature to significantly degrade performance as a precision measuring tool. I'd rather build up two versions, or have some sort of switchable input, than compromise noise or distortion by more than a dB or so.
(And could a 'scope probe even be used with a balanced input, which seems to be the other most popular request.)
The DC-DC converter seems a little expensive, since it is only powering a half dozen IC's. Are there other viable options?
Dale
Ok so im going to try getting all the components and building the old version.
Should i hold on and wait for the improved version?
What would the waiting time be for pcb boards to be ready for shipment?
Will most of the components still be the same so if i buy double now then i could make the old version asap then when new pcb is ready make that one too.
Kind regards
Richard
Should i hold on and wait for the improved version?
What would the waiting time be for pcb boards to be ready for shipment?
Will most of the components still be the same so if i buy double now then i could make the old version asap then when new pcb is ready make that one too.
Kind regards
Richard
No guarantee that I will do this at all, and if I do it will be at least a few months. So I wouldn't wait.
It's not yet clear how similar the BOM would be. My guess is that it would be very different.
Pete
Thanks for your reply pete.
Could you pm me a parts list i will need please
Kind regards
Richard
Could you pm me a parts list i will need please
Kind regards
Richard
Ok mate i will get the parts together then buy your board.
I like that you listed the place to get the parts too 🙂
Kind regards
Richard
I like that you listed the place to get the parts too 🙂
Kind regards
Richard
Wim in principle that is a nice idea, but my fear is that you re-introduce connection to a PC with all the noise and switching junk taht entails.
Jan
Hi all, i just designed in past box for planning interface based on pmillet, but not finished. only mechanical. But i have not time to finish 🙂
Attachments
Last edited:
- Status
- Not open for further replies.