Automatic adjustment of resistance

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
Hi

I have a design requirement and are having an issue of designing a method to achieve it
I have resistive elements that change with temperature and generally start with slightly different resistances in any case, what i need to achieve is a way to balance the resistances or a method to maintain the same volt drop/current input to the resistive elements to maintain an even load and output.

The power supply I am using is a constant current supply across the elements however due to the resistance differences they get loaded significantly differently.

has anyone worked on or produced something that can overcome this issue in the past? my initial thinking was to use inductors and dividers so that if one element was drawing more current then it would add inductance to itself relative to the other elements potentially based off volt drop but I'm not 100% on how I would go about this or even if it makes sense to do and there being a better way to achieve the same outcome

Thank you

JD
 

BobTPH

Joined Jun 5, 2013
8,813
You need a separate constant current supply for each element.

But even that will not guarantee constant power, since the resistances change. To achieve constant power you need to monitor both voltage and current and keep their product constant. Not easy to achieve with analog circuits, but a microcontroller could easily do it.

Bob
 

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
You need a separate constant current supply for each element.

But even that will not guarantee constant power, since the resistances change. To achieve constant power you need to monitor both voltage and current and keep their product constant. Not easy to achieve with analog circuits, but a microcontroller could easily do it.

Bob
Thanks for the reply Bob

As you can tell I'm not well versed on this side of things, I would say I'm intermediate at best when it comes to electronics, would you happen to be able to point me in the direction of any reading material/design concepts to study for something like this? I'm assuming you would set up a controller connected to many individual constant current controllers? the current on this device is very high overall which is part of the challenge, power input can be up to 16kW total however there are many individual elements pulling roughly between 3 and 5A each at 24V so you can see there are a lot of these elements to try and control

Thanks again
JD
 

BobTPH

Joined Jun 5, 2013
8,813
Is constant current close enough or do you really need constant and equal power?

16 KW at 24 V is 666A! This is orders of magnitude outside my expertise.

But, that said, if constant current is good enough, you are better off to put some number of elements in series snd us a higher voltage.

Bob
 

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
Is constant current close enough or do you really need constant and equal power?

16 KW at 24 V is 666A! This is orders of magnitude outside my expertise.

But, that said, if constant current is good enough, you are better off to put some number of elements in series snd us a higher voltage.

Bob
Thinking about it, its going to be equal power I need to achieve....
yeah that's part of my issue, I believe I can utilize higher voltages in series however I need to see where the limits are on this as the elements have a rated voltage however I believe this is because of the current limits of the devices as appose to potential internal arcing issues. If I can up the voltage to bring the current right down then that is definitely the way I will approach it but will still need to maintain 3A across the elements. part of the challenge is also controlling runs of elements separately but I can do series and each series with a constant supply I suppose as a method for controlling that

Thank you
JD
 

crutschow

Joined Mar 14, 2008
34,285
Thinking about it, its going to be equal power I need to achieve..
Then you need to monitor both voltage and current across each element.
I can do series and each series with a constant supply
If they are in series than each will get the same current, unless you start with a current above the maximum and bleed excess current around each element as needed by each element's voltage and current.
Far from a trivial task.
 

BobTPH

Joined Jun 5, 2013
8,813
Then you need to monitor both voltage and current across each element.
As I said in post #2. I was trying to simply if constant current was good enough, but apparently it is not.

This is not an easy project.

Sanity Check: You said 16KW and you said each element is 3-5 A at 24V. That means 133 elements if they are all run at 5A. Is this right?

Can I ask what the application is?

Bob
 

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
As I said in post #2. I was trying to simply if constant current was good enough, but apparently it is not.

This is not an easy project.

Sanity Check: You said 16KW and you said each element is 3-5 A at 24V. That means 133 elements if they are all run at 5A. Is this right?

Can I ask what the application is?

Bob
Thank you both

That's correct for the largest device being 133 elements in the current design, this is for a project that i am building that i have a patent going through at the moment for so don't want to delve to much into the project as a whole but i have managed to build a small scale version consisting of 40 elements as a proof of concept and prototype and one of the issue i have ran into is this uneven loading of the elements which obviously worsens as they heat up. the operating temperature wont exceed 100degC but is more that enough especially when they are in series for the resistances to drift and at the moment i only have the design for a single constant current device or devices that are commoned together by modbus and voltage sensing to give the output power required for all elements

i could see this being way out of my abilities to design the component required to maintain the balanced power so thought i would hop on here with the wealth of knowledge and maybe help others if we came up with something and they have a similar issue

at the moment the prototype version is running off of one single constant current supply which has proven one aspect but this unbalanced loading is proving to be an issue

Thanks again
JD
 

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
Ah, so a secret project that uses 16KW. Hope it is targeted for industrial use, and not home.

Bob
That is the plan if the proof of concept and prototype works out haha! But it would be good to have freely available outcomes like this on the journey to see if i wasted a lot of hard earned money on the project, not a total waste then if it doesn't pan out
 

crutschow

Joined Mar 14, 2008
34,285
Would it be possible to monitor temperature near the elements instead of power, and adjust the element current for even temperature?
That would likely be easier.
 

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
Would it be possible to monitor temperature near the elements instead of power, and adjust the element current for even temperature?
That would likely be easier.
I think the main issue from looking at the design so with that method is the number of thermocouples I would need to install and monitor then control to, the other issue is lack of space for these but I'm sure I could design them in from a ground up approach. The cool thing about doing it that way i suppose is i would be able to get a matrix of temperature readings which would help with the R&D process of the machine.
 

LowQCab

Joined Nov 6, 2012
4,029
You've got a Patent Pending,
so You don't want to say what it is supposed to do,
but You don't know how it works so You need help figuring it out ?

In order for a Patent to be enforceable, it must be as "Broad" as possible,
otherwise some sharp guy will pick your Patent to pieces and walk right around it.
.
.
.
 

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
How tightly are you trying to control the power/temperature?
Not incredibly tightly, for example on of the series runs was at 3.5A and another was 5.2A, this was using a constant current device connected to a bus bar that then feeds the elements but I belive the voltage was divided due to the difference in resistance
What I need to really achieve is say 4A plus or minus half an Amp, running at the same voltage e.g 24V, if that makes sense
 

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
You've got a Patent Pending,
so You don't want to say what it is supposed to do,
but You don't know how it works so You need help figuring it out ?

In order for a Patent to be enforceable, it must be as "Broad" as possible,
otherwise some sharp guy will pick your Patent to pieces and walk right around it.
.
.
.
I understand but this part of the project isn't the patent pending part, this is just a method of powering the device I've built correctly for test purposes
 

LowQCab

Joined Nov 6, 2012
4,029
Are You saying that the Device will work properly without all of the
parameter controls that You need for testing or "proof of concept" ?
.
.
.
 

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
Are You saying that the Device will work properly without all of the
parameter controls that You need for testing or "proof of concept" ?
.
.
.
Yeah, so the device works to a degree but currently I am struggling to keep even power due to changes in resistance over the matrix of heating elements I am working with which may be impacting performance and the outcome of the device, if I can keep the distribution even it may improve the results
 

LowQCab

Joined Nov 6, 2012
4,029
This is sounding like the practicality of the device is heavily dependent
upon even Heat, or even Power distribution.
So there's no point in the whole concept until You get that worked-out.

Has it occurred to You that this is the reason that this idea is not in common use ?
.
.
.
 

Thread Starter

JDuffy93

Joined Sep 12, 2021
14
This is sounding like the practicality of the device is heavily dependent
upon even Heat, or even Power distribution.
So there's no point in the whole concept until You get that worked-out.

Has it occurred to You that this is the reason that this idea is not in common use ?
.
.
.
This is something that I am trying to figure out as part of testing, as I said the device works and this is just part of getting further results.
Luckily I don't give up hurdles, got through them so far

Anyway back to the original question of methods for accounting for changes in resistance either through analogue or digitally if that's okay
 
Top