Analog Reading from a voltage divider thats is part of a -350V supply

Thread Starter

Turdnugget

Joined Mar 3, 2012
23
Attached is a schematic of the the basic configuration for a hypertherm powermax45 plasma cutter Output and its analog voltage signal provided to the interface port for CNC operation.
In the service manual schematic there is not a indication that the 36khz power transformer secondary is going to ground. The positive voltage of the rectifier is referenced to ground via work clamp to work piece which grounds through the table to earth.

Im lost on how i should build this part of the circuit with regards to the micro controller ground ref point for the analog reading ... As well as the best approach to filtering it

CPC-5(-) and CPC-6(+) are the CPC ports pins for the CNC interface.. So these are the signals which i must filter and apply to a micro controller as a analog reading.

10.JPG
 

Thread Starter

Turdnugget

Joined Mar 3, 2012
23
I drew this to help simplify the basic idea of what i need help with.. This is from the service manual for a Hypertherm Powermax 45 it is the circuit inside the machine.. If there is a whole community working on plasma CNC electronics please tell me where.. I have seen other open source designs but i would like to make one as well that is a little different.
 

Thread Starter

Turdnugget

Joined Mar 3, 2012
23
The way I see it is the micro controller needs to have its own isolated supply and the CPC5(-) connection which is the center node
of the voltage divider needs to be tied to the micro controllers isolated ground plane. Then CPC-6(+) connection would go to a filter and then a analog pin to be read

11.JPG
 

MikeML

Joined Oct 2, 2009
5,444
The way I see it is the micro controller needs to have its own isolated supply and the CPC5(-) connection which is the center node
of the voltage divider needs to be tied to the micro controllers isolated ground plane. Then CPC-6(+) connection would go to a filter and then a analog pin to be read
That will be just peachy until the first time you forget to unplug the programming cable. What are the "outputs" of the microcontroller tied to?

You need galvanic isolation between the plasma cutter and the micro...
 

Thread Starter

Turdnugget

Joined Mar 3, 2012
23
So then something like,
ACPL-790B-000E
Precision Isolation Amplifier
Applications :

Current/Voltage Sensing in AC and Servo Motor Drives
Solar Inverters, Wind Turbine Inverters
Industrial Process Control
Data Acquisition Systems
Switching Power Supply Signal Isolation
General Purpose Analog Signal Isolation
Traditional Current Transducer Replacements
 

AnalogKid

Joined Aug 1, 2013
8,299
What are the voltage range, min and max frequencies, and wave shape of the signal across the 2K resistor? Analog Devices and others make galvanically isolated signal amplifiers and buffers for this type of application.

ak
 

PeterCoxSmith

Joined Feb 23, 2015
148
I've been working on an isolation amplifier that uses the ACPL-C79B to galvanically disconnect the source from the controller. That circuit will be available as a DIN rail module in a couple of months. If you want to design your own circuit the ACPL chip is a good place to start as it has a bandwidth of 200kHz.

As an alternative you can design a non-isolated circuit based on a differential amplifier like AD8278ARZ. That is just an example, you need to select a component that has the bandwidth you require and can accept the negative common mode voltage.

You also might like to read this...
http://forum.allaboutcircuits.com/blog/cursns4-current-shunts-need-an-amplifier-difference-or-isolation.693/
 

Thread Starter

Turdnugget

Joined Mar 3, 2012
23
Thanks MIKEML for pointing that out. That is a good point!

AnalogKid,
I hooked the o-scope up to the plasma today and read the output while cutting with it.as well as open circuit.Attached are some of the scope shots showing exactly what I'm dealing with as a signal... The O-scope is a tektronix that has 4 channels with isolted grounds.. and when i was using it it was running on a battery.. So the There is no conflict of scopes ground with the plasma cutters.

PeterCoxSmith,
I believe that will be the chip i use!
When i was doing my scope readings, 1 of the tests I set the scope to AC mode for the channel. Then i turned it to 500mV/DIV. There is about a 18Mhz Ring with a delta 2.5v happening in 6.8uS periods (+- 148khz)

So This ACPL0C790B chip will actually help act as the filter to the 18mhz ring since it has 200khz bandwidth. If anything the signal should just need to have a RC lowpass filter on the output side of the isolation amplifier.

The ACPL-C790B
21.JPG

I drew up the circuit real quick that seems right to me. 23.JPG

This scope shot shows the scope Channel in AC mode @500mV /DIV with 100ns/DIV. t shows the Frequency of the ringing from the power supply. The ringing is close to 18MHz. with a deltaV of 2.5v.

24.jpg

This Next Scope shot shows the Channel in DC mode.. So the initial higher analog voltage state is the time that the plasma is establishing a conductive path to the work piece being cut.. The ringing is the 18MHz. Take note that the Time Domain is 50mS in this shot
17.jpg

This Final Scope shot Shows the frequency of the reoccurring supply ringing which is around 150khz
26.jpg




19.jpg 18.jpg I20.jpg

Any thoughts or opinions which may help to improve the performance/ stability/ accuracy are welcome :)
 

PeterCoxSmith

Joined Feb 23, 2015
148
Some thoughts:
The ACPL-C79B has a max input of 200mV so your 7V needs to be divided down some more.
And yes I've used a filter to limit my bandwidth to 80kHz.
Peter.
 

DNA Robotics

Joined Jun 13, 2014
527
I am also interested. I don't think my Chinese Cut 50 plasma torch has the voltage divider signal. It looks easy enough to add and even trim it to less than 5 volts.

You might check http://www.cnczone.com for the whole community of people doing this type of thing that GopherT referred to.

If you start a topic there, please keep us posted or leave a link to it here.
 
Top