Need circuit help, Trying to convert a 5v DC standard to 4v Pk to Pk AC?
rwgast_logicdesign
Posts: 1,464
Ok I originally asked this question on the EEVBlog forums thinking it would be more appropriate there, but I haven't gotten any definitive answer, About the best info I got was to look in to a wein oscillator circuit, which may be what I need in the end but im still not sure.
So heres the issue, I have a DMMCheck from voltagestandard.com. I also have a 40,000 count dmm with .05% accuracy. The DMMCheck has a 1ma reference along with a +5v DC reference and some 10ppm .01% resistors. He also builds the DMMCheck+ and PentaRef, which include a 100Hz square wave bipolar 5v pk to pk AC reference. Since this is a commercial business he obviously isnt giving away his secret sauce, He even scratches chip labels
. So I am not sure how he is doing a precision transfer from DC to AC.
My idea is to use a high ohm precision divider to shave the one volt off, then buffer that with an LM385, next I would run that in to a TLE2426 (A TI precision virtual ground based rail split-er) this would give me +2/-2 volts. Next I was thinking something as simple as a 555 could pulse the bipolar voltages. I basically just need a 2 volt or less signal to calibrate my meter, im using the prop for some impeadence testing so Im trying to get as accurate as I can with the AC readings. Im mostly afraid of the opamp or one of these circuits dropping voltage/loading the standard/or creating noise.
Unfortunately I dont have a 6 or 8 digit meter with NIST calibration to be able to verify and trust that my circuit isnt screwing up the original 5.0000v source. So im looking to build something that is ultra low noise and has an ultra High Impeadence input. A 555 is ok because im really just trying to measure a correct uni-polar voltage frequency isn't super critical as long as its good enough to keep an AC voltage stable.
I know most of you guys are coders but I also know there are some really good analog hardware guys out there, @jmg??
P.S. I know an inverter is basically an toroid doughnut with a 1:1 wrap and a 4 way mechanical or transistor switch, but in this case I dont think that it can pull off an accurate voltage transfer
So heres the issue, I have a DMMCheck from voltagestandard.com. I also have a 40,000 count dmm with .05% accuracy. The DMMCheck has a 1ma reference along with a +5v DC reference and some 10ppm .01% resistors. He also builds the DMMCheck+ and PentaRef, which include a 100Hz square wave bipolar 5v pk to pk AC reference. Since this is a commercial business he obviously isnt giving away his secret sauce, He even scratches chip labels
. So I am not sure how he is doing a precision transfer from DC to AC.My idea is to use a high ohm precision divider to shave the one volt off, then buffer that with an LM385, next I would run that in to a TLE2426 (A TI precision virtual ground based rail split-er) this would give me +2/-2 volts. Next I was thinking something as simple as a 555 could pulse the bipolar voltages. I basically just need a 2 volt or less signal to calibrate my meter, im using the prop for some impeadence testing so Im trying to get as accurate as I can with the AC readings. Im mostly afraid of the opamp or one of these circuits dropping voltage/loading the standard/or creating noise.
Unfortunately I dont have a 6 or 8 digit meter with NIST calibration to be able to verify and trust that my circuit isnt screwing up the original 5.0000v source. So im looking to build something that is ultra low noise and has an ultra High Impeadence input. A 555 is ok because im really just trying to measure a correct uni-polar voltage frequency isn't super critical as long as its good enough to keep an AC voltage stable.
I know most of you guys are coders but I also know there are some really good analog hardware guys out there, @jmg??
P.S. I know an inverter is basically an toroid doughnut with a 1:1 wrap and a 4 way mechanical or transistor switch, but in this case I dont think that it can pull off an accurate voltage transfer
Comments
-Phil
If you are running from batteries then the AC is simply the differential outputs rather than trying to make a virtual ground or use split supplies. The way to do this is simply use a dual opamp such as a LM358 as inverted amplifiers with a negative or unity gain as you can voltage divide anyway. If you are controlling this from the Prop then the idea is to drive one PHSx output high while floating the other. This will make the output of the opamp low (or negative) on that side while the other one simply outputs the reference voltage. In this case I have an RC input which can be driven PWM fashion to set the output voltage or you could even generate a sine wave if desired. Once the half-cycle has ended you flip the PHSx pins so now the other one floats and the other one drives the line high.
If you change the gain of the opamps and run them from a higher supply you can even use this to generate up to a 30V peak (60V P-P) AC sine wave using a 32V supply.
Of course a 50cent micro could generate the precise signals too.
@Jackie so are you telling be I could change the amp gain up to 60v Pk to Pk and I wouldn't lose any precision from the DC reference I would have and exact +30.0000 -30.0000 signal transferred from a +5v.0000 signal?! Being able to scale the voltage like that may make using a micro worth it especially if I used hearty op-amps and could get in to the 100v range!