THCAD and Arc Voltages

Well, I'll attempt to keep this short and sweet. I did a lot of cutting recently and I keep having tapered cuts. It occured to me that I never double checked my THCAD with a known voltage to ensure that what is being displayed on PlasmaC represents reality. So, I did some testing and I got interesting results.

I should note that I am running a THCAD-300 with a Hypertherm PMX-65, and the voltage divider is currently set to 50:1 (default).

I guess before I get to the data, I should say that I had a THCAD-300 because I had an everlast previously and I measured raw arc voltage. I am not sure the error with such a wide range of voltage reading ability. I realize that most cutting voltage for the PMX is at most 150v, meaning at 50:1, the THCAD is reading up to 3 volts. It seems a very small percentage of the scale it is capable of, and therefore maybe is some of my problem. The other problem I have is that even at idle, my THCAD seems to fluctuate ~5 volts (on the high side).

At any rate, did some experimenting today. I left the frequency at 1/32. I hooked up a known voltage from a PC power supply, and monitored it with a Fluke multi meter. The entire time it read 5.249v steadily. I'm not going to try to type out the data in a table here, so I embedded a screen shot from Excel below:



As you can see, as the divider amount went up, so did the difference between the high fluctuations and the low fluctuations.

I guess I have a few options, I can either get a THCAD-5, and run 50:1, or I can grab the raw arc voltage from the machine and use the THCAD-300. Hypertherm does have a service bulletin for this.

At any rate, I wonder how to quiet down the THCAD as well. Why doesnt my voltage reading show a steady 0 when the torch is off, and why am I having a hard time getting it to show a steady 5.249v?

The last thing I noticed is that the plasmac.arc-voltage-out in the HAL Watch does not ever match what is being shown in Plasmac, when Plasmac was showing 6v, the pin was showing 4.98v, etc. Am I watching the wrong pin? Or is there something else going on here?

Sorry for being long winded. Please let me know if you need more data, or I am being a knucklehead :laugh:

If you run a 300 V full scale THCAD from a 5V source you are throwing away
almost all the resolution. The THCAD has about 10 bits of full scale resolution
at a 1 ms sample rate (about 1/3 of a volt for 300V full scale) and this can be
increased by low pass filtering but if you only use 1/60 of the full scale range
you only have about 4 bits of basic resolution at 1 KHz, that is only 16 discrete
voltage readings...

This is a bit like trying to read 2.5V on the 1000V range of your voltmeter.
Note that you can modify the THCAD-300 for 5V full scale by replacing all 10
through hole resistors with 4.99K 1%

PCW wrote: If you run a 300 V full scale THCAD from a 5V source you are throwing away
almost all the resolution. The THCAD has about 10 bits of full scale resolution
at a 1 ms sample rate (about 1/3 of a volt for 300V full scale) and this can be
increased by low pass filtering but if you only use 1/60 of the full scale range
you only have about 4 bits of basic resolution at 1 KHz, that is only 16 discrete
voltage readings...

This is a bit like trying to read 2.5V on the 1000V range of your voltmeter.
Note that you can modify the THCAD-300 for 5V full scale by replacing all 10
through hole resistors with 4.99K 1%

I sort of figured that was the case.

So what is better...
1. Read the raw arc voltage with the THCAD-300?
2. Read the 50:1 Arc voltage with a THCAD-5?
3. Read the 20:1 Arc voltage with a THCAD-10?

Also, is it even possible to make a THCAD-300 a THCAD-10?

Edit: Thank you for the explanation!

Yes you can make any input range you like by changing the through hole resistors.

All 300K = 300V full scale
all 4.99K = 5V full scale
all 10K = 10V full scale

These should be 1% or better resistors

PCW wrote: Yes you can make any input range you like by changing the through hole resistors.

All 300K = 300V full scale
all 4.99K = 5V full scale
all 10K = 10V full scale

These should be 1% or better resistors

Thank you sir!

Any recommendations on which scaling and thcad pairing makes the most sense? Is 1:1 always the best option? Or does it not really matter?

I would probably arrange things so you get about 200V full scale
So your 20-1 divider and a THCAD10 make sense
or a 50-1 divider and a THCAD5

PCW wrote: I would probably arrange things so you get about 200V full scale
So your 20-1 divider and a THCAD10 make sense
or a 50-1 divider and a THCAD5

Last question (maybe)... can you explain how the frequency changes things? Should I be shooting for 1/1? Or is 1/32 most appropriate?

It does not make much difference, though the /32 mode can have better noise immunity from external interference in the FOUT signals

A lot of us are using a 30:1 divider (some Hypertherms can be changed to this) with a THCAD-10 and no scaling resistor. That gives you 300v full scale.

Using a 20:1 scale will give you greater resolution in the cutting range (100-150 volts) but I don't think it makes much difference in the real world.

My Thermal Dynamics brings raw arc voltage out onto the CPC port (with 200k resistance so its not lethal) I have run it from that raw arc voltage with scaling resistors per the manual but deducting that 200k resistance from the calculated value and it was still happy.

If you can't change your divider, the THCAD-5 with no scaling resistor would be a good option.

PCW wrote: Yes you can make any input range you like by changing the through hole resistors.

All 300K = 300V full scale
all 4.99K = 5V full scale
all 10K = 10V full scale

These should be 1% or better resistors

Any reason why we couldn't put 10K resistors in parallel with the existing 10K resistors essentially making it 5K THCAD-5?

I want to use it for ohmic sensing.