PMX 45 (nonxp) voltage divider question

[robertspark]

I know these sorts of questions have been posted before, but I was wondering if I could get a better (clearer) answer than a home brew guess

The PMX 45 (nonxp) has a 50:1 voltage divider output, and from the schematics within the service manual the voltage divider is created (as would be expected) from two resistors

What I would not have expected was that the resistors are listed as 100k and 2k.

While the ratio is 50:1 the voltage divider output is not 50:1, because the total resistance is 102k, so therefore say for simplicity 1000V / 102000 ohm = ~9.8mA
9.8mA X 2,000 = ~ 19.608V

So the actual voltage divider ratio is 1000 : 19.608, or 51:1..... Is it not?

The reason why I ask is because it like to change my PMX 45 from a 50:1 voltage divider to 20:1 voltage divider for a different THC

The question is for a 20:1 divider, do I go with:
100k & 5k (2 X 10k resistors in parallel)?
Which is actually a voltage divider ratio of 21:1 (eg for simplicity 1000v across the voltage divider would give ~ 47.61v across the smaller resistor, so 1000 / 47.61 = 21:1


Or should I be looking for (simplicity) 1000 v across the divider and 50v across the smaller of the two resistors so the smaller resistor (retaining the existing 100k ) should be closer to 5,263 ohms (4700+560 ohm in series)

Yes I know that the voltage is just a number (setpoint) but I'd like to be close to the book numbers for voltage setpoints thanks

[tcaudle]

The correct formula is
Vout = VIn * R2/R1+R2 where Ri is the top resistor connected to the voltage to be divided and R2 is in series and to the voltage ground. The Vout is at the junction of the R1 R2 connection and the same ground
This all works perfectly until you throw in other components or you hand any kind of load across the Vout side
Rather than take a chance of blowing up an expensive control board in your 45 I think I would consider building my own 20:1 divider and use the raw volts (Hypertherm Field service document shows where to hook external voltage divider.) You will also find its not a pure DC signal and it has lots of switching noises that need some filtering to be able to use. Be mindful of the voltages involved (they are lethal ) when the torch is fired. Be careful of power dissipation. E^2/R. Raw volts can reach 175 to 180 volts on open circuit 45 . So the lower the voltage ratio the lower the resistance and the wattage goes up at the square of the numbers. You take the voltage drop across each resistor and divide it by it reistance to get watts of power

[robertspark]

Thanks Tom for taking the time to reply.

Vout = Vin * R2 / (R1 + R2) for Hypertherms Resistances of 100K and 2K still provide a 51:1 voltage divider and not 50:1 as listed in the book.

Am I missing something?

Again, correct me if I'm wrong but a 50:1 voltage divider, I'd put 50V between the input and should expect to get 1V between the output pins.
Likewise, for a 20:1, if you put 20V on the input of the divider, you'd expect to measure 1V between the output of the divider?

http://eagleplasma.com/cutters/manual-pm45.pdf

2018-11-08 15_42_03-Powermax45 Service Manual.png (35.39 KiB) Viewed 1984 times

2018-11-08 15_42_03-Powermax45 Service Manual.png (35.39 KiB) Viewed 1984 times

2018-11-08 15_44_00-Powermax45 Service Manual.png (22.36 KiB) Viewed 1984 times

2018-11-08 15_44_00-Powermax45 Service Manual.png (22.36 KiB) Viewed 1984 times

The open circuit voltage from the manual is 275V which aligns with the varistor on the LHS of the screenshot attached.

The brunt of the voltage drop is taken by the 100K resistor to provide the major voltage drop.

________________________________________________________
current through the 100K resistor at 51:1 (100K+2K) at OCV (275v) is 2.696mA, which gives a power dissipation through the resistor of 0.73W

changing the circuit to ~20:1 (100K+5260 {4700+560}) at OCV puts the current at 2.6125mA. which gives a lower power dissipation through the 100K resistor of 0.68W

As can be seen on the service manual screenshot the 100K resistor is rated at 3W so no issue there.

[tcaudle]

Open circuit voltage is not 275V.
You need to run the numbers for 20:1 if that is you goal. You will raise the voltage drop across either resistor you elect to change so the wattage will go up for that one.

I suspect any schematic is not really the exact values that are in the circuit. They will be 1% values. all you have to do to adjust the ratio is to slightly change the lower R2 value.

[jimcolt]

The idea of the voltage divider is to provide an analog feedback voltage to control torch to work distance, as well as to limit the highest voltage output to protect the THC electronics in the event that open circuit voltage happens. (this only will occur if there is an issue with the consumables in the torch) The difference in actual torch height that would occur if the voltage divider had a ratio of 50:1 as compared to 51.1:1 is inconsequential to torch performance in terms of cut quality or consumable life. I agree with Tom Caudle....if you have the need for exactly a 50:1 divider ratio.....access the raw arc voltage and build a more accurate voltage divider. Keep in mind that consumable wear in the torch as well as air pressure and cut speed variations all affect the voltage / cut height relationship far more than a small error in the voltage divider output ratio does! Jim Colt Hypertherm

[robertspark]

It's just interesting that something quoted as 50:1 may not be, when it connects to another device that requires a quoted input of 50:1 .

Whilst the effect may be realativly minimal, it would be nice for someone to admit that it's actually 51:1, and the manual quoted values are taken as such from the CPC.

Say I intend to cut at 140v, I set the 50:1 THC at 140v, and the output voltage at the CPC will be 2.8v..... but the raw volts will actually be 142.8V..... ok, only 2.8v out but it will mean that for the first cuts the torch will be too high and I'll need to reduce the setpoint to ~ 137v to align with the book values.

I am fine with raw arc as I've been using it for a while, I was just trying to make better use of the onboard voltage divider and was curious about the math because 1:50 seemed more like 51:1 so I was wondering if the same applied to 20:1 and it should be 21:1

It's like when Tom says that ocv is 175 to 180v on the raw arc, and you print an extract of the service manual that says ocv 275v, and CPA 225v..... And still comes back and says the ocv is not 275v

But when you look at the circuit it's clamped with a varistor rated at.... 275v....

I ain't making this up its from the service manual link and extract posted earlier.

Thanks all but I've not gleened or corrected any info that I didn't have when I posted this thread and posed the question.... Is the pmx 45 50:1 or 51:1 with a 100k 3w resistor and a 2k resistor

[Rodw]

If you've got any doubts about the divider network, I recommend you grab a couple of 48 v power supplies in series to get 96 volts (somewhere close to cutting volts). Disconnect the plasma divider board and hook this voltage up to the raw torch voltage input and measure the output. Also make sure you measure the output when its connected to the CNC control box.

It wasn't a Hypertherm, but I found that on another machine that the control panel torch voltage sensor also had some scaling resistors in it and the plasma divider resistance was counted as part of that which distorted the torch voltage reported by a factor of about 1.5.

I personally think we should be sampling torch voltage to set the setpoint rather than keying in a voltage in a cut chart that may or not be perfect on the day you choose to cut on for any number of reasons. I had a perfect 16:1 until I connected to the control box and it changed to 24:1. If you do this, it does not matter if the divider is 50:1 or 51:1, the system will just seek to keep whatever it read at the start of the cut stable.

[islander261]

Either sampling voltage at the start of the cut or using the actual cutting voltage from test cuts.

John

[robertspark]

Thanks Rod, no I've not got doubts over the divider network, I was wondering about changing it from a 50:1 to 20:1 divider {which can be easily done} but then I noticed that with a 100K & a 2K it probably wasn't as per the manufacturers published data a 50:1 but a 51:1 {and questioned my own calcs ( I am an engineer, and started in electronics, moved on to industrial electronics, electrical and now consulting})

I have a THC with sample and hold so the number makes no difference, but I kind of like to start with the book value when I start with cutting something new to avoid waste and will adjust settings accordingly.

I kind of expected someone in the know {Tom / Jim} to say something like "50:1 or 20:1 is the resistor ratio and its sometimes incorrectly referred to as the divider ratio" ..... or "it is 50:1 because although it looks like 51:1 with a 100K and a 2K you need to consider the parallel resistance that your THC imposes on the voltage divider so hypertherm build in this allowance to compensate for the secondary parallel resistance of the THC" ...... or "So when hypertherm do 20:1 we use a ratio like 21:1 to again allow for the THC's additional imposed parallel resistance of the THC which is added to the smaller resistor."

All of which would make perfect sense.
_________________________________________________________________________________

I've got another question regarding ohmic sense.

At the moment I've got an old steel wire armour cable gland earth tab which I've filed out to act as an ohmic tab
https://www.cmp-products.com/wp-content ... TAGS-1.jpg
so it looks something similar to this (but with a soldered lead) {where I got the idea from}
http://www.dynatorch.com/images/ohmic1.jpg

{yes I know there is the official ohmic cap https://images-na.ssl-images-amazon.com ... SX466_.jpg} but I could not be bothered to buy one..... I'll probably upgrade to a duramax torch at some point as I have some pmx65 consumables which will fit the duramax torch}

Whilst perusing the PMX 45 service manual schematic I noticed that the cap appeared to be wired back to the pilot arc module (2x red wires size 18AWG, connecting to J17 on the main board, hence I wondered if I could not just tap into the ohmic sense at this point (J17) and avoid wiring back to the torch and having my ohmic sense interface attached to the table {I am doing a wiring tidy up exercise at the moment given I've built a new cabinet for all my cnc drives, motion controllers and power supplies}

my ohmic sense interface: https://i.ebayimg.com/images/i/32184642 ... -l1000.jpg
my thc interface board: http://neuroncnc.com/assets/components/ ... PCB800.png

I was thinking about putting my thc interface board within the casing of the PMX45 and linking the ohmic sense within the casing (with the interface module too within the casing) for a bit of neatness, and adding a few pins to the CPC for a neat finish.

which negates all the discussion about voltage dividers as I'd just use the raw voltage and the filtering of the THC divider board {which uses 10:1 output and is less noise sensitive on the analog input to the THC}

I've had a bit of skittles luck at present ..... everything I touch turns to xxxx!
https://www.youtube.com/watch?v=C3Zma2_n5CA

Hopefully I'll get a bit of time this weekend and a clear workbench to open up the 45 and look at the space available inside to include the THC and the ohmic sense

[jimcolt]

Ohmic sense operates from the shield, which is isolated by the threads (dupont vespel composite) on the retaining cap. If you connect for ohmic where you point out you will be at power supply positive (in relation to the electrode) which is at the same potential as the work clamp when the torch is idling and in pilot mode, then is floating when the cutting arc transfers from electrode to workpiece / plate. Ohmic sense relies on the floating shield and contact with and a continuity reading to the workpiece to locate the surface of the workpiece. Jim Colt Hypertherm

[Rodw]

I started to make an Ohmic sense on another unshielded torch. Then after some thinking, I became worried that the torch voltage could arc across my proposed isolating relay contacts so gave up on the idea in the absence of the isolated shield Jim mentions. As Jim says, the shield is not shown in the circuit diagram.

[robertspark]

Thanks Jim & Rod.... I'll keep my fly lead then and swa gland earth tag

I use the minithc ohmic sense as I got one when I got a minithc from Denis http://minithc.com


Andrew also does them at neuron
http://neuroncnc.com/products/004

Both will deal with high voltage fine and won't be killed. I like Denis' as it has an adjustment which I don't think the neuron ohmic has

[Rodw]

Hmm, the minicnc one looks the goods for my 24 v hardware. It even has a PNP output which is prefect for my Mesa hardware. Thanks Robert!

[robertspark]

Denis may be on annual leave at minithc, he tends to disappear for a few months, and his listings on eBay disappear. Think he goes south for the weather haha. He is good to chat with and very helpful. Suggest a chat before purchase if you are interested. He has a few clips on YouTube regarding the THC not sure about the ohmic.

I have 24v for most of my boards but I've added 12v for the neuron and some Arduino boards I use for various other things now (fast I/o gates)

[Rodw]

Old thread but: Changing the links on the Hypertherm board so it is a 30:1 is perfect for the THCAD-10 board's 10 volt range (300 volt full scale). It is also perfectly valid to run say a 200 volt full scale as the THCAD has good overprotection. This give you greater resolution in the usable range as anything over say 150 volts is not really of any interest. If you do get unexpected voltage values, try changing your backshell resistor and use the raw arc voltage.

[jimcolt]

This thread was originally about a Powermax45 (non xp)…….which has a fixed value voltage divider set at 50:1. The newer Powermax45xp has an optional voltage divider with 5 different selectable settings (50:1, 40:1, 30:1, 20:1 and 21.1:1) Jim Colt

[jedmundson]

Robert if you are still available, I just joined because I was looking for someone who converted down a 20:1 to 50:1 voltage output. Were you able to do it by stepping down the 20:1 output? Thanks

[robertspark]

Robert if you are still available, I just joined because I was looking for someone who converted down a 20:1 to 50:1 voltage output. Were you able to do it by stepping down the 20:1 output? Thanks

unless you know what the current resistors are you are better just to install a 50:1 voltage divider network off the two raw voltage (torch and clamp) connections.

what machine is it you are working on?