Unimig Razor Cut45 - info

Hi All,

I've been looking for a while for info regarding the Unimig Razor Cut45 mainly the volt divider and cnc port pinout and I have not found much anywhere. So I thought I'd drop this here so it might help someone in the future. It seems that there is a lot of different info getting around, but anyway, this is what I have found with my 6 month old Cut45. I have 2 cut charts that both appear to be very outta whack (for my machine) but I'll attach them anyway.

If anyone else has any good info (manuals, parts lists, cut charts etc) please add it so as to create a bit of a library all in one place, as long as the moderators don't have an issue with this thread.

BTW I am a sparky so I took all the necessary precautions when I had the unit open, you SHOULD TOO which may mean using someone else to do the work.

So when I bought the unimig I opened it up and found 2 resistors listed as R87 and R88 which are listed on an attached doc i found somewhere on this forum. R87 is 100k and R88 is 2k ohms which works out to roughly 50:1 (51 actually), so I thought that was the ratio I had. I also proved that the volt divider volts didn't exceed ~6V. When I fired it up with my arduino uno THC attached (with a programmed 50:1 ratio), I found that the torch voltage did not match the THC display.

So I set about proving what the cutter was actually putting out before I went any further. I used 2 multimeters and found that my unit was set up as a 20:1 ratio. I did find on a doc that 20:1 was the default which this test backed up. I did this via the 2 procedures below 'resistance tests' and 'voltage tests'.

Resistance Tests::
I used a multimeter to test the resistance between the following points (please see pic of divider):
- the main return (work clamp +ve) and cnc port pin 6(+ve) -- this was a dead short as expected
- the main return to cnc port pin 5(-ve) -- this showed 2k ohms
- the torch supply to cnc port pin 6(+ve) -- the showed 42.5k ohms (not what I expected)
- the torch supply to cnc port pin 5(-ve) -- this showed 40.5k ohms (not what I expected)

Voltage Tests::
I connected 1 meter on the torch supply and main return and the other on terminals 5 and 6 on the CNC port. I roughly received about 130VDC and 6VDC respectively when activating the torch without cutting. (the voltages fluctuate a bit so that's why the ratio is a little off.) 

The resistance tests above indicated that there is more resistance in parallel with the 100k ohm resistor. One of the docs shows that a 20:1 ratio requires a 68k ohm resistance in parallel with the 100k and if you calculate the resistance of those 2 in parallel have a guess at what you get..... 40.5k ohm. This is the same reading as the last resistance test above.

So I had a better look (a few times in fact) and I finally found another resistor (68k) with no name on it on the opposite side of the PCB, directly behind the R87 resistor. To get an eye on this resistor you have to remove an insulating plastic sheet that is glued and pinned (nylon pin) to the PCB. Once the 2 nylon pins are removed from the PCB the sheet can be peeled back to expose the glue now holding it on. I used a screwdriver to start removing the glue from the sheet, it came away easily but was stuck in a fair few places.  Once I had access I double checked that I had the correct resistor by doing a few more tests back the torch and cnc port pins. Then I cut one leg of the resistor off and did the resistance and voltage tests again and sure enough, I now had a ratio of ~50:1 (51:1 actually). So see below for some pics which hopefully might help someone out.

Hopefully, that wasn't too confusing.

 

How are you measuring torch voltage to want 50:1? This modification is probably not required.

If you are using a THCAD10, 20:1 is preferred. This means your full scale (10 volt) will allow you to sence 200v and the THCAD overvoltage protection extends to 500 volts.  If on piercing  the voltage  exceeds 200 volts briefly, it won't matter as it is not at a time accurate voltages are required for torch height sensing. eg. before the pierce delay expires.

Note: neither of the cut charts seems right to me in my limited testing. So far I found 83V, 1000mm/mm seems to work well with nearly no dross on 6mm mild but I have limited access to scrap at the minute so that's about all I have managed. I tested the plasma without the THC and at 1.5mm cut height I was reading 84-87 volts so I tried a little lower which seems to go well. Sounds low but the results appear good so who am i to argue.

If anyone has any good starting cut charts i'd love to see them.

No thcad RodW. Although, I should have ordered one from mesa when I ordered my 7i96s but..... I already had an arduino version built before I decided to go with the mesa board.

I am using an arduino uno along with some optos, regulated supply etc that I made up from a design by a bloke who calls himself swolebro. The program is a bit different now from his though.

For anyone that interested you can see a simulation of the program here on the wokwi simulator
wokwi.com/projects/339317146336100947
 

I suspect your voltage sensing is not calibrated correctly. I ahve not used a Razr but others on the forum have. I would expcect to see the voltage 1round 110 volts.

Potentially. although I have done some testing using a calibrated fluke multimeter alongside the arduino and they were fairly close. But my plan is to run some more tests as soon as I can find some more scrap.

Also the program does not update the arduino display as regularly as it does for the height control.

The unit says it is a max of 90V 98v which is obviously not correct but looking at the charts I have found so far, some of them mention 150v cutting. When using the max function on my meter I pick up what I believe are spurious spikes up to 300v, but I only get 135V typically at a max with the meter showing live measurements, so not sure how those charts came about. I'd love to have access to a data recorder/scope to get a better pic. A THCAD would have helped a heap, live and learn. I can't warrant spending the same money on postage as the unit.

A lot of the reason I put this out there is to try to get find some more info.

Hi Andy,

I thought he must have been on here at some point, he has a few nice videos on youtube regarding linuxcnc, hal files etc.

I reached out to him recently about using his code in my own THC version and he told me he's actually sold his plasma table which might explain his absence. He managed to make the THC work fairly well from all reports.

I started kicking myself as soon as my mesa card arrived as I really should have ordered a thcad at the same time. The post to Australia effectively doubles the price of the unit, which goes against the original aim of this whole project.

Now, after spending $350 au on the mesa 7i96s, the machine (not including plasma) has cost just a touch over $1000 au. So to add a thcad with postage will now be a 5th of the total value of the machine.

Does anyone stock them in Aust???