SparkFun Forums 

Where electronics enthusiasts find answers.

For the discussion of Arduino related topics.
By beefy
#168033
Hello Guys,

I'm a bit of a beginner to the Arduino so if I'm talking backwards at any point, feel free to be direct with me.

I've got a plasma cutter for cutting steel plate. To establish the arc, the following happens:

1). A dry relay contact closes and tells the machine to cut.

2). The air solenoid energises (24v DC)

3). The air flow / pressure causes the torch nozzle to move away from the electrode and a pilot arc is struck. This pilot arc "ignites" the main air flow establishing the plasma jet to the steel plate (which has the ground electrode attached).

Regarding 3). I want to data log the arc voltage from the moment the machine gets the signal to start, up to the main plasma jet being established. There's probably a whopping half a second of time for this to happen so my main concern is sampling fast enough to give me a bunch of measurements I can make a graph from.

This arc voltage goes as high as about 230 volts DC so I was thinking of using an isolated supply for the Arduino with a common ground with the arc voltage. As for the sampled arc voltage I would use a resistor network to get 5v at the analogue input pin when the arc voltage was at 230v. I'll probably also throw a zener across this input for extra protection.

Now for the sampling / storing / uploading the data. My sketch would monitor the start signal and when it arrived, sampling would immediately start. I was thinking I could just take the analogue input readings and store them in an array of a given size. The Uno only has 2K of internal memory (SRAM) so I would reduce the resolution of the analogue read to 8 bits instead of the default 10 bits. This way I only need one byte in my array to store one sample, effectively doubling how many samples I can store (8 bit resolution is fine for my needs). Once the array has been filled up, my sketch would send the array values to the PC.

If the sampling rate was too fast and filled up the array before the main plasma arc was established I would simply put a delay between each sample/write to array task.

How does that all sound. I'm still learning the programming so just wanted a general opinion if my methodology is on the right path, before I go and waste my time learning to do something that won't work.

Cheers,

Keith.
By Mee_n_Mac
#168076
How fast a sample rate do think you need ? Without using lower level commands, a 10 bit conversion using the Arduino analogRead() takes about 120 usec. 8 bits will take less time. You can also get memory to add and faster A/Ds but at some point I'd just get (or rent) a DSO.
By beefy
#168144
Thanks Mee_n_Mac,

I was starting to give up on this forum, it took 2 days just for my post to get here. Is that normal.

A sampling rate of at least 1Khz (1000 samples / sec) should be OK. With 2Kbytes of ram (minus of course what might be used in the sketch) that should give me plenty storage space for single byte readings. If the Arduino samples too fast and my array fills too quickly, I'll have to put a delay between each reading to lengthen the total sampling time.

Ha ha, yes I could rent a DSO but then I wouldn't learn how to use my new Arduino a little bit. It seems like it could be a very useful tool in my workshop.

Cheers,

Keith.
By waltr
#168158
New users on this forum are screened before their posts are allowed. This does prevent lots of spam that would other wise show up here. You shouldn't have any issue now.

Nice idea but due to the very high Voltages I don't suggest trying to connect your Adriuno as a first project. One little oversight in the circuit and your Ardiuno will go up in smoke and this will not be an easy circuit to get right even for experienced engineers.
Do however work to the code to do a logger with the Ardiuno and also do much more reading on how to interface low Voltage analog circuits to such a harsh high Voltage. Remember that at a steady state you will have about 230VDC but during the phase when the arc strikes there will probably be Voltage spikes several times that.
By beefy
#168166
Thanks Waltr,

I've got a couple of ISO122P isolation amplifiers, so I might try and use one of those. The "spiky" signal did have me wondering so maybe the isolation amplifier will help with that too.

Keith.
By jremington
#168167
This arc voltage goes as high as about 230 volts DC so I was thinking of using an isolated supply for the Arduino with a common ground with the arc voltage.
Do not use a common ground in this situation. The Arduino MUST be entirely optically isolated from the arc circuitry, with kV of isolation.
By beefy
#168181
jremington wrote:Do not use a common ground in this situation. The Arduino MUST be entirely optically isolated from the arc circuitry, with kV of isolation.
Seems quite a few others have used this technique (common ground). Could you give reasons why you think the Arduino MUST be completely isolated. I will learn nothing if all I am told is don't do that.

I'm often attached to that ground by the way. It's the steel that is being cut with the plasma torch and the torch ground lead is attached to that. As long as the Arduino power supply is isolated then won't the Arduino simply be floating at the same ground potential as the torch ground.

Keith.
By jremington
#168183
There are lots of possible problems. Two that immediately spring to mind (because I've encountered them more than once) are:

1) If the Arduino power supply is on an AC circuit, especially if a different branch circuit than the arc, there will very likely be a ground potential difference between the branches. Ground currents can introduce many problems, including frying the Arduino and its power supply. If connected to a laptop, the laptop goes too.

2) The arc will generate large amounts of radio frequency radiation which will induce RF currents into long leads, INCLUDING the ground lead. Those currents will fry the input pins of the Arduino. It takes only a volt or so at a few mA to melt the protection diodes on the input pins.

If your Arduino is battery powered, you might be able to get away with your approach. However even in that case, I would put spike protection circuitry on all Arduino inputs, as used in automobile electronics. On the other hand, as learning experiences go, Arduinos are inexpensive.
By beefy
#168195
Thanks very much Jremington,

the voltage is DC, but I was completely ignorant about the radio frequency radiation. My ISO122P isolation amplifiers are starting to look like the way to go.

My cnc electronics is directly interfaced to the HV connections with a dedicated PCB. I might remove that and see if I can identify it's components and circuitry and copy it.

Keith.
By waltr
#168215
the voltage is DC
But it can still be connected to the AC Mains common which is the possible issue jremington listed in item 1.
By beefy
#168216
Thanks Waltr,

I emailed the maker of my cnc electronics. They said after lots of experience they decided to have complete isolation from the arc voltage. Their circuit is proprietary so they won't give me any real details.

I'm going to go with an isolation amplifier fed from the middle of a voltage divider network. I'll see what design pointers I can pick up from looking at my cnc high voltage interface card too. There may be a circuit I can trace that's on the HV side and feeds the input to its isolation amp or whatever it uses.

I'm starting to see where the cost of this stuff comes from LOL.

Keith.