A nice tool that I carry at work is an infrared thermometer. I have not calibrated my feigner tips yet. I checked the temperatures under the hood.
The measured temperatures follow:
1 Weller solder pencil 462 F
2 Tabletop 72 F
3 U1 86F
4 U2 77F
5 U3 104F
6 U4 102F
7 U5 82F
8 U6 82F
9 U7 95F
DT
Just For Fun!
The measured temperatures follow:
1 Weller solder pencil 462 F
2 Tabletop 72 F
3 U1 86F
4 U2 77F
5 U3 104F
6 U4 102F
7 U5 82F
8 U6 82F
9 U7 95F
DT
Just For Fun!
Ordered all of the parts this morning for the interface.. Inevitably some parts were out of stock. The converter was out of stock at Mouser, but in stock at Digikey..
I'm going to use OPA2604 on the input because that is what I have on hand for the time being. I may have an OPA2134AP lurking around somewhere and may give those a shot regardless. Input noise is a couple of rtHz lower in the 2134 than the 2604 - in practice I suspect the input protection network resistors would be a significant noise source in most cases. (I haven't calculated their noise contribution, but I expect their noise contribution to make using a significantly quieter op-amp a non starter which is actually good since I have not found any except for bipolars like 5532/LM4562 which have significant input currents.) The tradeoff I think is worthwhile to be able to withstand a couple of hundred volts rms accidentally applied to the input. (My old design would not survive such a major overload.)
Pete, I have a couple of thoughts/questions:
I am building mine without the meter option as I don't really need it. (I may add it later) I have determined with my current choices of op-amp and INA/DRV amps that the total quiescent current is about 25mA - right at the minimum acceptable value for the switcher, so I am thinking of adding a couple of resistors to load the supply and bring the current up to 30mA or so..
I was wondering whether the dc offset null is actually needed if the meter is not installed? I assume the 10 ohm resistance in the offset null circuit significantly reduces cmrr in the INA134. I have ordered the parts for the offset null in case I do need it.
I will match the input stage resistors to maximize cmrr.
Now all I need is a monitor..
I'm going to use OPA2604 on the input because that is what I have on hand for the time being. I may have an OPA2134AP lurking around somewhere and may give those a shot regardless. Input noise is a couple of rtHz lower in the 2134 than the 2604 - in practice I suspect the input protection network resistors would be a significant noise source in most cases. (I haven't calculated their noise contribution, but I expect their noise contribution to make using a significantly quieter op-amp a non starter which is actually good since I have not found any except for bipolars like 5532/LM4562 which have significant input currents.) The tradeoff I think is worthwhile to be able to withstand a couple of hundred volts rms accidentally applied to the input. (My old design would not survive such a major overload.)
Pete, I have a couple of thoughts/questions:
I am building mine without the meter option as I don't really need it. (I may add it later) I have determined with my current choices of op-amp and INA/DRV amps that the total quiescent current is about 25mA - right at the minimum acceptable value for the switcher, so I am thinking of adding a couple of resistors to load the supply and bring the current up to 30mA or so..
I was wondering whether the dc offset null is actually needed if the meter is not installed? I assume the 10 ohm resistance in the offset null circuit significantly reduces cmrr in the INA134. I have ordered the parts for the offset null in case I do need it.
I will match the input stage resistors to maximize cmrr.
Now all I need is a monitor..
Last edited:
Almost done
I have mine almost completed.... just waiting on the 0.1uF caps. The original p/n was out of stock and I forgot to look up a substitute, doh! They're scheduled to arrive on Monday and should be completed shortly after they arrive.
I'm planning to use mine with a Digigram VX Pocket card and balanced cables to/from the interface. if all works well, I'll probably build a second for stereo use. Overall, very well done project, I'll post some feedback once I start get some use time in.
I did the full build, buying all of the parts. Between Pete, Mouser and Digikey the costs were around $270 USD.
Regards, KM
I have mine almost completed.... just waiting on the 0.1uF caps. The original p/n was out of stock and I forgot to look up a substitute, doh! They're scheduled to arrive on Monday and should be completed shortly after they arrive.
I'm planning to use mine with a Digigram VX Pocket card and balanced cables to/from the interface. if all works well, I'll probably build a second for stereo use. Overall, very well done project, I'll post some feedback once I start get some use time in.
I did the full build, buying all of the parts. Between Pete, Mouser and Digikey the costs were around $270 USD.
Regards, KM
I managed to chop about $130 off of that by not implementing the meter, and through careful shopping/raiding my parts bins.
Haven't got off so lucky elsewhere though as I had to build/repair a PC, purchase a keyboard/mouse (directron.com) and a nice used Dell 15" monitor on eBay and an M-Audio Audiophile 24192 which was fortunately on sale and heavily discounted at Amazon last week, Still a good deal. I will probably configure the other available channel for TS measurements a little further down the road.
I bet the DC/DC converter will be fine... the voltage might creep up a little but I suspect not enough to worry about.
Definitely don't need the offset null without a meter. Actually you'd probably be just fine without it even with the meter...
Pete
Lars -
December 9th or 10th, via first class mail (not priority mail). They say that can take three weeks... maybe more with holidays.
Pete
The power house fire in Kirkwood, CA sent me home early, and I have now had time to check my build.
I just completed the measurement you described above: unit set to 200mV range, input measures 200.7mV on my Fluke, U7 pin 6 measures 501.x on another, the unit's display shows 1 for over.
Now set to 2V range with the imput signal still at 200.7mV, U7 pin 6 goes to 50.3mV and the display shows 0.401V.
In all ranges, for all combinations of switch settings and for all input voltages that I have tested the display always shows exactly twice the "actual" voltage.
The output signal to the soundcard has minimal noise and distortion levels in line with expectations.
OK, things are good there.
Next thing to check is that pin 6 of U5 should read 50mV DC under those conditions.
If so, verify that R14, R15, and the trimpot R9 are the right values. If they are, there should be 20mV DC on the input to the DMM.
Pete
OK Peter
Fingers crossed. Have never lost anything coming from the States before so I am not that worried
No worries, if it doesn't show up I'll send another. Or maybe deliver it in person - I may be in Scandanavia end of February...
Pete
The mail system was exceptionally slow this December. I send my boards out using Priority Mail which is the fastest. Even this was erratic. I mailed 3 boards out the same day in mid December. One went to California, and another went to Singapore. Singapore took 8 days, but California took 17!
I got a small LCD monitor from an Ebay seller in Washington State. It was sent first class mail, and it took 23 days. Good thing that I wasn't in a hurry.
There is 48.9mV DC on pin 6 of the AD536A. R14 is 10K, R15 is 34.8K and the trimpot R9 is 5K ... but Input Hi on the MDMV0000 actually sees 39.8mV DC.
I'll gladly send a picture and/or poke around some more. At this point I'm baffled.
I've stared at the circuit diagram too long.
My math for the voltage divider between the AD536A RMS-TO-DC converter and the MDMV voltmeter comes out real close to a factor of 1/1.25, depending on the trimpot setting.
That corresponds perfectly with the voltages I am actually seeing, but I assume that the divider should result in a factor of 1/2.5. Am I seeing things that aren't there?
Looking at the pictures of the board in your article it seems that you used R14 - 27K and R15 - 15K [instead of R14 - 10K and R15 - 34.8K in the schematic].
I am now using R14 - 24.9K and R15 - 12K, because I had those handy, and I'm getting correct readings on the MDMV voltmeter within the limits of my ability to calibrate mind you.
I am now using R14 - 24.9K and R15 - 12K, because I had those handy, and I'm getting correct readings on the MDMV voltmeter within the limits of my ability to calibrate mind you.
DOH!
I apologize profusely to all. These values were left in the schematic from a time when I was running the RMS converter at a lower level. I decided later to up the gain in the first stage to get more accuracy in the RMS converter. The problem is, I forgot to go back and update those two resistors that scale the output for the DPM in the schematic and BOM!
R14 should be 15k, and R15 should be 27.4k (or any ratio that gets you close, the pot can adjust it some).
You could also leave R14 at 10k and change R15 to something between 4-5k (4.7k would work). The control will be a little more touchy but you should be able to calibrate it OK.
I have updated the files on the web site and it's uploading now. (It might take a while because there are some PDF books transferring too.)
Again, very sorry to all for the confusion on my part!
Pete