It's done... but....
I found the resistor problem also... raided the junk box and used a 20K and a 12.1K which made the adjustment doable. In any case, I have the unit completed and calibrated. The interface does have a fair amount of self noise, and hum due to a lack of shielding. Using balanced mode is worse.
I've been playing with a shareware copy of Audiotester... I'm not impressed. The UI is horrible (actually, it sux) and it's pretty buggy stuff... crashes frequently. I'm using a Digigram VX Pocket V2 card... which I know is solid as I've been using it with Wavelab for 10 years and have never had any issues with it. Audiotester has issues with all driver modes against the VX Pocket with some that won't work at all. Anybody else have a recommendation for good test code?
Regards, KM
I found the resistor problem also... raided the junk box and used a 20K and a 12.1K which made the adjustment doable. In any case, I have the unit completed and calibrated. The interface does have a fair amount of self noise, and hum due to a lack of shielding. Using balanced mode is worse.
I've been playing with a shareware copy of Audiotester... I'm not impressed. The UI is horrible (actually, it sux) and it's pretty buggy stuff... crashes frequently. I'm using a Digigram VX Pocket V2 card... which I know is solid as I've been using it with Wavelab for 10 years and have never had any issues with it. Audiotester has issues with all driver modes against the VX Pocket with some that won't work at all. Anybody else have a recommendation for good test code?
Regards, KM
You might want to try installing some aluminum foil inside the case to create a faraday cage around the board (ground to one point on the pcb) - this will prevent all electro-static coupling into high impedance circuit nodes on the interface.
Most of the self noise in the front end is probably generated by the 49.9K resistors protecting the front end. You could try reducing their values - you can also look for a quieter op-amp as well, ultimately the ThatCorp INA is a few dB quieter than the BB, but this should be the last thing you change. To do self noise measurements you need to make sure that the inputs are shorted and you are not near any external noise sources. The computer and the monitor (yes even LCDs) spew broadband noise everywhere and if the interface is in proximity that might cause a problem. In my old system I shielded all of the cables to the sound card and found this was good for really reducing spurious pick up from nearby noise sources. Balanced mode operation should take care of a lot of the LF garbage, but CMRR degrades very quickly above a few kHz in most designs unless great care is taken. (I designed this sort of hardware for a living -ATE until the recession took my job. I consult now.)
I've had a lot of problems with Audiotester over the years, version 2.2D does seem better and more stable. I purchased a license long ago. Better software IME costs 5X or more than audiotester. Arta is very good, however it costs several hundred dollars and the shareware version will not allow you to save or print graphs - but works quite well otherwise.
I have noticed with audiotester that it is a good idea when changing sample rates or drivers to close and restart the program before doing anything else - that avoids a lot of issues. I have not had much success with Asio support with this program lately. (I use it for all other audio applications so this is disappointing)
I am not familiar with your sound card, but it is old enough that I am wondering what bus interface it uses. (I see it is pc card) Audiotester runs very poorly on older hardware and is very resource intensive. I get adequate performance with an Athlon XP 64 3000, but large numbers of BINS (>8192) seem to really slow it down. The combination of many bins and averaging further slow it down. This is the fastest machine I own (and it is quite fast) and I would rate audio tester performance as good not great. It runs really poorly on my 2GHz Pentium 4M Compaq EVO..
Last edited:
I need to dig out the Karosene Heater tonight and warm up the garage (currently 48F, with outside temps around 22F). Then I'll try changing the two resistors and see if it corrects my problem I reported in post 135. I suspect this will correct my problem as it seems to be about the same variance from expected value.
One other thing I have seen that causes a lot of RFI/EMI is florescent lights. I run two 48" bulbs over my work bench, and they really create a lot of noise that gets picked up by any antenna. I have to shut them off to take performance measurements.
We used to use a product called "Electoless Nickel Spray" on the insides of equiptment plastic enclosures for shielding. I can't find any information on it any more. It was a spray paint that had a high conductivity. It was mottled gray in appearance once it dried.
I did find this that looks like it might be similar. A tad bit pricy unless you are doing several enclosures. Masking is an issue with it as well.
http://www.action-electronics.com/msds/mg-841-340gconductivepaint.pdf
One other thing I have seen that causes a lot of RFI/EMI is florescent lights. I run two 48" bulbs over my work bench, and they really create a lot of noise that gets picked up by any antenna. I have to shut them off to take performance measurements.
We used to use a product called "Electoless Nickel Spray" on the insides of equiptment plastic enclosures for shielding. I can't find any information on it any more. It was a spray paint that had a high conductivity. It was mottled gray in appearance once it dried.
I did find this that looks like it might be similar. A tad bit pricy unless you are doing several enclosures. Masking is an issue with it as well.
http://www.action-electronics.com/msds/mg-841-340gconductivepaint.pdf
Hello
This bench testing stuff is new to me. I am looking for tools and methods to improve the amplifiers and speakers that I build.
I have used an application a time or two called RMAA by Rightmark RightMark Audio Analyzer. Products. Audio Rightmark
It looks like the tests are automated with fewer user tweeks allowed. I guess not for the Master Bench Tester.
Does anyone have any insights? This application could be an introductory foot in the door to use with the Millett Interface. What do you think?
DT
This bench testing stuff is new to me. I am looking for tools and methods to improve the amplifiers and speakers that I build.
I have used an application a time or two called RMAA by Rightmark RightMark Audio Analyzer. Products. Audio Rightmark
It looks like the tests are automated with fewer user tweeks allowed. I guess not for the Master Bench Tester.
Does anyone have any insights? This application could be an introductory foot in the door to use with the Millett Interface. What do you think?
DT
It looks like it might work, however you might want to download the demo version of audioTester V2.2 which seems to be very popular around here.
audioTester
Kerosene heater won't light. Fired up the propane heater and will work on the kerosene one later after the garage is warm enough to work for more than 15 minutes.
My problem was also the 27.4K/15K problem. Unit is now cal'd and looks good.
audioTester
Kerosene heater won't light. Fired up the propane heater and will work on the kerosene one later after the garage is warm enough to work for more than 15 minutes.
My problem was also the 27.4K/15K problem. Unit is now cal'd and looks good.
Last edited:
Hi Kevin,
All good points... as usual. I have some mu metal sheets from a kit I bought last year... I can try some of that as well. My initial tests were done without any computers nearby... just older analogue test gear, HP and Tek... very quiet stuff. In general, I only use external converters and use digital cards to interface them, but the VX Pocket has been a nice portable setup for mobile operation.
For this little project, my older T43 ThinkPad seems a decent choice. It's the older ThinkPad line with much better shielding than much of the current stuff (no changes when I dragged it out and hooked it up). The VX Pocket has been an excellent performer over the years... very solid and actually sounds very good. It also has SPDIF I/O which adds more flexibility, but you're right... older PCMCIA interface. It does support ASIO which I do like... but like you said, not very useful with Audiotester.
I also have a Digidesign MBOX and MBOX2 which I can also try.... these are USB attached... and additional boxes to drag about. I'm mainly looking for a reasonable test environment that is more portable and simple. The interface Pete did is slick, I like the concept. I just need to dig a bit deeper for software that I like and that works well with my HW.... would prefer to use my MacBook Pro, but oddly very little audio SW is Mac based... who knew.
I'll work on this a bit more once I get back home... out visiting the folks for a bit... and cutting up some Baltic Birch for speaker projects for the next few days... but I have the T43/VX Pocket/Interface with me as well. I think I'll look at the Arta code... thanks for that.
Regards, KM
No, the 1512 isn't really a driver, but the THAT 1646P08 should work fine... in stock at Mouser.
The DRV134PA appears to be out of stock everywhere in the US at the moment, but is available in EU and Asia. the SOIC version is available and could be used with a little adaptor PCB, but that's a bit of a pain.
It's getting VERY hard to design with all through-hole parts...
Pete
Thanks Jack & Pete.
I'm not sure where I came up with the 1512...not what I wanted to copy & paste. I guess I was'nt seeing straight last night.
Newark had the DRV134PA in stock when I ordered online but when it came time to ship they disapeared and the ETA would be May 5th... and that 7 day lead time on the Red Lion panel meter; ordered on December 15th still waiting...frustrating
Thanks again
I'm not sure where I came up with the 1512...not what I wanted to copy & paste. I guess I was'nt seeing straight last night.
Newark had the DRV134PA in stock when I ordered online but when it came time to ship they disapeared and the ETA would be May 5th... and that 7 day lead time on the Red Lion panel meter; ordered on December 15th still waiting...frustrating
Thanks again
Last edited:
Mouser has (had) the ThatCorp equivalent in stock last week when I ordered my parts. The ThatCorp 1646P08 has significantly better performance than the currently unobtainable BB DRV134PA.
I would have gotten the ThatCorp INA in lieu of the BB INA as well, but that part wasn't in stock at Mouser or Digikey.
I'm still waiting for all of my parts, keyboard, monitor and other odds and ends to arrive.. No progress to report.. I haven't even figured out yet how to make it all fit on my bench, something's gonna have to go..
Is it okay to sub a slightly higher or lower value or RUSBF110? Perhaps RUSBF090 or RUSBF135?
Also, what is the largest dimensions that can fit for C2, C3? I haven't received my boards yet
No worries on subbing a different PTC fuse. You could even leave it out (replace it with a jumper), basically it is there as a safety mechanism if you plug in the wrong voltage or polarity of DC adapter.
Same is true for the TVS. Any 1.5KE9.1xxxxx should be fine.
C2 and C3 are 22.5mm lead spacing, and the footprint is 26.8mm x 11.3mm.
Pete
I can tell you what happened, as somebody who's day job is in the semiconductor business... when the economy turned south everybody cut fab and assembly production way back. It turned out the the downturn really was not as big as prople predicted (especially in Asia). So suddenly standard lead times are out to 26 weeks.
Now we routinely get new bodily orfices created by angry customers.
Pete