PLC Voltage Regulation or Control?

Thread Starter

HighVoltage!

Joined Apr 28, 2014
181
What hardware do I need?
Im trying to compile a list of parts needed for this project...I think Im going to avoid using a PLC for now even though I am highly tempted to! When using thermocouples as your input and current in mVDC (converted AC current to DC using a current transducer)...unlike PLCs which have special thermocouple input cards, does ARDUINO or AVR need TC voltages programmed "by hand " or are they already programmed in them? Curious question since not familiar with programming that much.
 

strantor

Joined Oct 3, 2010
6,782
Why not use the right tool for the job?
PID temperature controllers are in no short supply; You can find some on Amazon for under $20!
(however investing in a Eurotherm/Barber Coleman, Omega, or Wattlow wouldn't be a bad idea)
They (most) already have a thermocouple input, so no need to convert to mV or design a circuit or buy an expensive PLC analog card.
They have the PID math & programming already (which is not so easy to program in arduino).
They typically have a PID autotune feature which really relieves a huge headache.

You could try using a dual output PID temp controller, with the heating output controlling as the "UP" button, and the cooling output controlling as the "DOWN" button.
However, that is a departure from the norm as far as temp control applications which I have experience with. I cannot guarantee that will give a good result, as in the typical application, the power is switched by the controller (via a relay/SSR/contactor), and when the controller stops "pressing the gas pedal," the heating power stops flowing. Your application would demand the controller to be in more of a "cruise control" mode, where it just makes small adjustments. There is probably a controller out there that would work but I cannot say which one.

If it were me, I would:
get rid of the variac
get rid of the 120V control motor
get rid of the thermocouple>mVDC apparatus
Install a SSR in place of the variac
install a cheap single output PID temp controller
Hit the auto-tune button
DONE!
 

Thread Starter

HighVoltage!

Joined Apr 28, 2014
181
Why not use the right tool for the job?
PID temperature controllers are in no short supply; You can find some on Amazon for under $20!
(however investing in a Eurotherm/Barber Coleman, Omega, or Wattlow wouldn't be a bad idea)
They (most) already have a thermocouple input, so no need to convert to mV or design a circuit or buy an expensive PLC analog card.
They have the PID math & programming already (which is not so easy to program in arduino).
They typically have a PID autotune feature which really relieves a huge headache.

You could try using a dual output PID temp controller, with the heating output controlling as the "UP" button, and the cooling output controlling as the "DOWN" button.
However, that is a departure from the norm as far as temp control applications which I have experience with. I cannot guarantee that will give a good result, as in the typical application, the power is switched by the controller (via a relay/SSR/contactor), and when the controller stops "pressing the gas pedal," the heating power stops flowing. Your application would demand the controller to be in more of a "cruise control" mode, where it just makes small adjustments. There is probably a controller out there that would work but I cannot say which one.

If it were me, I would:
get rid of the variac
get rid of the 120V control motor
get rid of the thermocouple>mVDC apparatus
Install a SSR in place of the variac
install a cheap single output PID temp controller
Hit the auto-tune button
DONE!


I have already thought about this. I have experience using PIDs especially Watlow Brand (All the same). However, since the metal I am heating by current induction, once a PID controller turns-off power source (once it gets to temperature), the temperature would plummet pretty fast. Therefore, there will be a constant chasing of temperature and triggering of ON/OFF. If you read my earlier posts, a temperature must be set/controlled in order to keep temperature stable. This is good for smaller projects which Ive done that arent as involved.
 

MaxHeadRoom

Joined Jul 18, 2013
28,619
Couldn't you do this empirically by taking tests and use an initial high current to raise to temp, and then decrease to a fixed level.
Typically induction heating is not an overly long process, the necessary power is usually estimated or observed and a fixed time span of power on is achieved.
Very long induction heating times usually indicate insufficient power.
Max.
 

strantor

Joined Oct 3, 2010
6,782
I have already thought about this. I have experience using PIDs especially Watlow Brand (All the same). However, since the metal I am heating by current induction, once a PID controller turns-off power source (once it gets to temperature), the temperature would plummet pretty fast. Therefore, there will be a constant chasing of temperature and triggering of ON/OFF. If you read my earlier posts, a temperature must be set/controlled in order to keep temperature stable. This is good for smaller projects which Ive done that arent as involved.
This thread has me guessing. What are you really dealing with? Is it really a legit high frequency induction heater or is it some sort of 60hz homebrew resistive heater? I've been assuming the latter all along because I've never heard of such a backward sounding setup on a true induction heater. Induction heaters are usually ran by a complex digital inverter at several khz, with no silly stepper motorized variacs or stepdown transformers. How many khz does your induction heater run at, and why can't you just make the changes you need in the inverter software?

If this is a 60hz homebrew resistive heater as i suspect, but has a very low thermal response time, you can still use a PID controller; you just need one with an analog output, and you need an SCRpack/SSR with a phase angle firing board on it.
 

Thread Starter

HighVoltage!

Joined Apr 28, 2014
181
Couldn't you do this empirically by taking tests and use an initial high current to raise to temp, and then decrease to a fixed level.
Typically induction heating is not an overly long process, the necessary power is usually estimated or observed and a fixed time span of power on is achieved.
Very long induction heating times usually indicate insufficient power.
Max.
You nailed it Max. Thats what I do...Start with a high current then lower it to a fixed level until a satisfied stable point is reached. Therefore, if a progeam can be written to control the whole set up would be great since different alloys are used for this.
 

Thread Starter

HighVoltage!

Joined Apr 28, 2014
181
If you are more familiar with PLC, then I encourage you to use what you are familiar with.

For the thermocouples my team used these boards from Adafruit: https://www.adafruit.com/products/269
They also provide tutorial and software example, so you can just read through and copy the software.

Interesting. I read its compatible with any microcontroller...what MC do you recommend? What cable do I use to sync to PC?
 

Thread Starter

HighVoltage!

Joined Apr 28, 2014
181
This thread has me guessing. What are you really dealing with? Is it really a legit high frequency induction heater or is it some sort of 60hz homebrew resistive heater? I've been assuming the latter all along because I've never heard of such a backward sounding setup on a true induction heater. Induction heaters are usually ran by a complex digital inverter at several khz, with no silly stepper motorized variacs or stepdown transformers. How many khz does your induction heater run at, and why can't you just make the changes you need in the inverter software?

If this is a 60hz homebrew resistive heater as i suspect, but has a very low thermal response time, you can still use a PID controller; you just need one with an analog output, and you need an SCRpack/SSR with a phase angle firing board on it.
Strictly current induction...secondary of transformer
 

MaxHeadRoom

Joined Jul 18, 2013
28,619
Is this induction heater operating at mains frequency? If so this is usually very large mass ferrous items, and I would think it very hard to control heat by sensing due to the large hysteresis in heating/cooling.
What is the final desired effect? A end result temperature of the total mass or molten stage?
Is it possible to photo the item undergoing heating?
Max.
 

strantor

Joined Oct 3, 2010
6,782
Is this induction heater operating at mains frequency?.
He deftly evaded the question when I asked about frequency, let's see what kind of result you get. If this is going to be one of those cloak and dagger top secret type threads where the OP won't answer direct questions and only provides enough detail to generate more direct questions to be evaded, count me out.
 

Thread Starter

HighVoltage!

Joined Apr 28, 2014
181
He deftly evaded the question when I asked about frequency, let's see what kind of result you get. If this is going to be one of those cloak and dagger top secret type threads where the OP won't answer direct questions and only provides enough detail to generate more direct questions to be evaded, count me out.
I already explained if you read my first post. The operating frequency is at 60Hz.

Variable Transformer > Transformer (Step Down) > Secondary of Step down XFMR will be loop of material (Temperature will need to be stable for at least 4 hours)
 

strantor

Joined Oct 3, 2010
6,782
The operating frequency is at 60Hz.
Secondary of Step down XFMR will be loop of material
Then there is no reason (that I can see, given what I know) that you can't eliminate the variac and use a PID controller with a phase angle fired SCR straight into the primary.

Here's a sleek solution that could talk to a PLC

Here's a cheaper solution that would be used in conjunction with an inexpensive analog PID controller like this.

If you want to maintain your manual functions via the variac, you can still do that with a switch to select between variac control or PID control, but you will need to get a more advanced PID controller that has an Integral Hold input to avoid integral windup and over/undershoot when returning to PID control.
 

Thread Starter

HighVoltage!

Joined Apr 28, 2014
181
Then there is no reason (that I can see, given what I know) that you can't eliminate the variac and use a PID controller with a phase angle fired SCR straight into the primary.

Here's a sleek solution that could talk to a PLC

Here's a cheaper solution that would be used in conjunction with an inexpensive analog PID controller like this.

If you want to maintain your manual functions via the variac, you can still do that with a switch to select between variac control or PID control, but you will need to get a more advanced PID controller that has an Integral Hold input to avoid integral windup and over/undershoot when returning to PID control.

The step down pulls roughly 100A...can it handle that? I know PIDs usually turn ON/OFF things (pulse) to maintain a temperature. Can the pid be used to regulate the voltage going into stepdown?
 

strantor

Joined Oct 3, 2010
6,782
The step down pulls roughly 100A...can it handle that? I know PIDs usually turn ON/OFF things (pulse) to maintain a temperature. Can the pid be used to regulate the voltage going into stepdown?
sure. If you get a PID controller with an analog output (i.e. 4-20mA), and a phase angle fired SCR with an analog input, it will control the voltage to the transformer by phase angle control. It works like this:

Instead of pulsing on/off every few seconds like the PID controllers you're probably used to, it controls the ON time during every single half-cycle of the AC waveform, thereby delivering consistent RMS voltage (although non-sinousidal) to the transformer without any interruption.
 

Thread Starter

HighVoltage!

Joined Apr 28, 2014
181
sure. If you get a PID controller with an analog output (i.e. 4-20mA), and a phase angle fired SCR with an analog input, it will control the voltage to the transformer by phase angle control. It works like this:

Instead of pulsing on/off every few seconds like the PID controllers you're probably used to, it controls the ON time during every single half-cycle of the AC waveform, thereby delivering consistent RMS voltage (although non-sinousidal) to the transformer without any interruption.

So the Phase angle fired SCR will be wired directly to line voltage replacing variac (control), the PID receives the temperature signal, lets say once PID decides a raise in temperature is needed, a signal is put out (4-20mA output signal) indicating ON in which I assume will be the ANALOG INPUT of the PAF SCR. How is this controlling (Raising/lowering) the voltage when its just putting out an ON/OFF signal? Meaning you somehow must pre-set your desired voltage = your desired output current? Can it handle up to 100A? Also, for initial start-up, the temperature is going to be close to ambient meaning the PAF SCR will start high at max voltage to try and get temperature quicker onto "load" (can limits be set lets say for "Max Voltage" = safety?...If not, I believe its unnecessary to stress equipment/wiring). Does the PAF SCR learn like the PID's autotune mode?
 
Last edited:
Top