next topic »

[Reply #15 on:]

I've modelled this successfully in ltspice with a SINGLE supply, but you've confirmed that in the real world I need a split supply - thankyou.

I still don't think this is necessary -- see my previous post. Especially if you keep the non-inverting input at slightly above 0V to get away from the 0V power rail. I bet you this will work:


http://www.ruggedcircuits.com

[Reply #16 on:]

RuggedCircuits, firstly my apologies. With the myriad replies I somehow overlooked your first post. Hmmmm - this is certainly what ltspice implied also but I wanted to check. About 50% of people I've asked say I'll need a split supply and 50% say I'll only need a single supply. So.... I'm just going to build the darn thing with a single supply and see. Thanks for the circuit - I'll build mine and yours just for fun.

[Reply #17 on:]

Another possibility is a current mirror with pnp transistors with the common at +5V. Use the 0..-7 V signal to draw current from the mirror and load an appropriate resistor from the mirror to arduino ground and measure the voltage across it.
The drawbacks are that the transistors should be matched, the resistors should be of good precision and most importantly a significant part of the range is lost, but it is a simple circuit and perhaps worth a consideration.

[Reply #18 on:]

@RuggedCircuits.

Hi - I've just received a little more info from the cutter manufacturers. The 0 to -350V is scaled down to 0 to -7V with a potential divider of 100K, 2K. *My* spice model is just a *classic* 2 resistor inverting opamp with 220K input and 160K feedback. Could you explain your design and why it's a tad more complex than mine - i.e. I'm interested in what your potential divider into the non-inverting input brings to the party if that's ok.

Cheers

Also - again - thanks to all for your varied suggestions but I'm trying to nail/understand this simple inverting opamp method at the moment.

[Reply #19 on:]

*My* spice model is just a *classic* 2 resistor inverting opamp with 220K input and 160K feedback. Could you explain your design and why it's a tad more complex than mine - i.e. I'm interested in what your potential divider into the non-inverting input brings to the party if that's ok.

The non-inverting input is being held just a little bit above 0V because many inexpensive rail-to-rail op-amps don't really like being right up against the rail on the input side. The more high-quality "beyond-the-rail operation" op-amps are OK with that though. You can try it without the voltage divider and see what happens, but my guess is you'll get better results with the slight offset if using an inexpensive RRIO op-amp.

--
Check out our new shield: http://www.ruggedcircuits.com/html/gadget_shield.html

[Reply #20 on:]

I'm sorry this post doesn't help you in the least, but I'm always up for learning something new...

Is it normal for plasma cutters to output a negative voltage for control/monitoring (or whatever)?

Talk about a pain in the butt!  What does this voltage control or do?

[Reply #21 on:]

RuggedCircuits. Thanks. I'll experiment and see where I get. Much appreciate your help.

ru. Actually most plasma cutters don't have any interface brought out to the panel. It's up to you to open the box and splice in your own, which of course is fraught from both a warranty and health and safety point of view (more so here in the UK - less so it would seem in the US as this is very common practice there). Manufacturers are slowly catching on that it's a good thing to provide this as more and more use is being made of small system automated or cnc plasma cutting.

There are several control signals of interest to the automation engineer. One is the ability to remotely turn the torch on/off, another is to detect when you have a good "arc" but the one of interest here is the torch electrode voltage (which by the physics of plasma is negative with respect to the metal you're cutting). By monitoring this changing (or at least a scaled down safer version of it) during a cut, you can arrange for automated (motor driven) Torch Height Control (THC).

Up until relatively recently THC's tended to be analog devices based around comparators so the -ve voltage probably wasn't so much of an issue. Now of course (and not before time) digital electronics is taking over and so hopefully manufacturers will catch up and give us a nice +ve 0 to 5V signal.

Up until now I've used the analog THCs but now I've designed my own for my cnc plasma cutters based around (of course) the arduino so now I just need to interface it to the cutter.

Hope that helps

[Reply #22 on:]

Thanks for explaining whistler, that makes more sense now - I thought it was just a perverse control method!    8-)

[Reply #23 on:]

You can get small (about the same footprint as an ATMega,  but an inch tall)  DC-DC converters that will give you +/-12V from 5V.

Then you can use an simple inverting op-amp to get a positive voltage,  and a voltage divider to scale it down to 0-5V.

From your description,  it doesn't sound like that output is very precise or clean.  I would add a little noise filtering,  and some protection to make sure that the op-amp output doesn't swing above V_cc.

[Reply #24 on:]

You can get small (about the same footprint as an ATMega,  but an inch tall)  DC-DC converters that will give you +/-12V from 5V.

I take it you mean this to be a split supply for the opamp? - If so does that mean you're in the single supply opamp won't work camp?

Then you can use an simple inverting op-amp to get a positive voltage,  and a voltage divider to scale it down to 0-5V.

I did think about a voltage divider on the output of my inverting opamp for getting the 0 to +ve7 down to 0 to +ve5, but figured a gain of 0.7 would do the same thing. Your thoughts welcome - pros/cons?

From your description,  it doesn't sound like that output is very precise or clean.  I would add a little noise filtering,  and some protection to make sure that the op-amp output doesn't swing above Vcc.

Absolutely right. The output from the plasma unit is notoriously noisy - but other creators of digital THCs claim to be able to resolve to around 0.25V which depending on setup can translate to a fraction of a mm - we'll see what happens. I'd be grateful for your suggestions on filtering. I've had quite a bit of success in my experimentation by filtering with software though - but again your thoughts are welcome.

[Reply #25 on:]

I don't know how to make a single-supply op-amp do the the job,  but I'm not an EE:  I have a modest knowledge of electronics from having been a ham,  and from many years of working with the hardware guys on embedded systems.  But there's a lot that I don't know,  and there may be some trick that a real EE learned and I've never heard of.

Doing your own voltage divider on the op-amp input is tricky,  because your signal is already the result of a resistive divider.  Your voltage divider would wind up being a resistance in parallel with the lower leg of that divider,  and would alter the signal.  For a general-purpose solution,  that would work regardless of the possibly-different internals of different torches,  it's better to buffer first,  then divide or use a second op-amp with a gain of < 1.

[Reply #26 on:]

You said:

Doing your own voltage divider on the op-amp input is tricky,  because your signal is already the result of a resistive divider.  Your voltage divider would wind up being a resistance in parallel with the lower leg of that divider,  and would alter the signal.

???

and then you said:

For a general-purpose solution,  that would work regardless of the possibly-different internals of different torches,  it's better to buffer first,  then divide or use a second op-amp with a gain of < 1.

???

but I said:

I did think about a voltage divider on the output of my inverting opamp for getting the 0 to +ve7 down to 0 to +ve5, but figured a gain of 0.7 would do the same thing.