Some help with a circuit I'm working on...hopefully

[kevlar129bp]

Hello all,

Long time, no post. Hope everyone is well. I'm sure hoping one of you brilliant minds can assist or critique an analog circuit I'm working on.
This circuit will scale the NEGATIVE dc voltage (ground referenced) of my plasma cutter, down to a usable ADC level. I've been simulating it
with the Falstad circuit simulator and it seems to work, but I would much rather have a second, third, fourth, etc. opinion on it. If you would like,
I can post the Falstad file in a code block as well.

Thank you all for your continued help in my crazy projects,
Chris

[Charlie]

What is the range of voltage you expect from the torch? (-300 V +/-??)

1) What is the relay for? Is there some function you did not mention?
2) The inductors are not necessary.
3) You will need positive and negative supplies for your op-amps in this configuration. You could avoid that by connecting your divider resistors to +12 instead of GND.

[mark_s]

Using an opto-coupler could make things simpler. A series resistor from the torch node to the LED, then to ground clamp node. You did not say what the function of your circuit is, and accuracy required?

[kevlar129bp]

Thanks for the replies guys,

As far as the voltage expected from the torch, it will range from a maximum of 0 VDC to a minimum of -300 VDC... or visa-versa, depending on how you want to look at it. The relay is implemented to delay the activation of the scaler due to crazy stuff happening on ignition of the plasma. I'll also use it in the PIC code as an interlock for the scaler itself. I realized, after I posted this, that my son had a copy of ISIS from school. We ran a simulation on it as well, with the same results... -300 VDC in, 2.048 VDC out. I am no analog guy, can you explain why I need a split supply on the differential amp? Love to be educated on this.
In regards to the accuracy, the ADC will be 10 bit so... 300/1024=0.29296875 (1/4 volt or so).
Thanks a bunch guys,
Chris

[mark_s]

I found this page doing what I suggested. Opto-couplers are safer for high voltage and isolate noise.

In this diagram you would need to change the series resistor to some higher value. Where it say's AVR connect to your PIC adc.
Try simulating it

http://www.ustudy.in/node/7519

[amgen]

and.... lookup linear opto couplers, plus your code can correct for some non-linearity.

[kevlar129bp]

Hello all,
The opto route looks very interesting, I have to say. I'll have to do some further studying on that. As far as my second post/reply, I apologize because I believe I misspoke in calling the op amp a "differential", instead of an "inverting" like it is. The simulation shows the "Clamp" to be an active participant but it is, in fact, ground. As far as the voltages on the "Torch" voltage divider, they will range from -100 to -275 vdc. The sim shows it as AC but, I think I mislead you fine folks. I was simulating it with an AC signal, although I neglected to mention it has a -150 DC offset. Sorry for not being more clear. Now that I hope I have given you guys the info you needed to begin with (my bad)...since the inverting input will never see positive voltage, therefore the op amp is not expected to invert it into a negative voltage output, do the same split supply rules still apply? Just for the heck of it, I removed the DC offset from the AC signal in ISIS (swings above and below ground) and the o-scope showed what I predicted...when the AC swung positive, the DC output was at ground...when the AC swung negative, the DC output followed the sine wave (+). Maybe I'll whip it up on the breadboard, just for the heck of it...
Any more thoughts or input is always appreciated!
Thanks guys,
Chris

[mark_s]

Hi,

If its for controlling a plasma torch z axis. See pages 4, 5 and 6

http://www.eaziform.co.uk/files/THC.pdf

[kevlar129bp]

Hey Mark,
Thanks for the link. Cool info to stash in my file of goodies. The circuit they are referring to in the pages you pointed out, is for an "arc ok" or "ok to move" signal. My Thermal Dynamics machine has a built-in "ok to move" dry contact to signify arc transfer. What my machine does lack, as I'm sure you have gathered is the voltage divider circuit. In my travels, I cam across a cool thread on the Parallax forum...check it out:
http://forums.parallax.com/showthrea...Height-Control
Head to post #17 and take a look at the schematics. They have been my wealthiest resource to date.
Look forward to all of your input,
Chris

[kevlar129bp]

Hey Mark,
After reading the pages that you pointed out more in depth, I digress. They do seem to be using a current transformer to sense voltage...Which is very weird, in and of itself because plasma machines are constant current devices. I would love for one of the circuit gurus on this board educate me further on this scheme. It sure has me scratching my head All of the other THC's out there do it with a scaler (divider)... Oh boy, I'm baffled now!
Thanks Mark,
Chris

[mark_s]

Thanks for the link!

I am working a new cnc machine now. In the future, when time allows I want to build a plasma cutter,
so I will be interested in your progress.

[mark_s]

Hey Mark,
After reading the pages that you pointed out more in depth, I digress. They do seem to be using a current transformer to sense voltage...Which is very weird, in and of itself because plasma machines are constant current devices. I would love for one of the circuit gurus on this board educate me further on this scheme. It sure has me scratching my head All of the other THC's out there do it with a scaler (divider)... Oh boy, I'm baffled now!
Thanks Mark,
Chris

It's beyond my knowledge. You may want to ask over on the cnczone forum. Someone can probably tell you if this method works?

[kevlar129bp]

No problem Mark, and thanks for yours. I actually have had a THC on my radar for about 6 years now, but a friend of mine just got his table set up and asked me if I would go back to the drawing board for him. What a guy won't do for his friends huh! I do have to admit, it is self-serving too. Heck, I'll get a THC out of it too. Here's some progress of the PCB's ready to go to OSHPark. RS232 and 20x4 lcd display through the RJ-45 jacks up top. All menus and settings will be controlled through a 24 detent encoder with a built-in push switch. The scaler and gpio boards are set up as daughter boards to the main board as well. I guess I should breadboard the project, now that I have the PCB's done
Any input on this thread is still more than appreciated.
Thanks guys,
Chris