Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Pid

Status
Not open for further replies.

xanadunow

New Member
You are all invited :)

After having captured you minute attention with this statement; what IC’s would you recommend to capture the AC current sensing.. and the temperature sensing (thermistor).. ?

I do have some samples coming from TI (I had already received some INA128P samples an the INA1101HP have just been posted). I Have also secured a couple of OP177FPZ – would you have any better suggestions?

The AD620 series had also crossed my mind..

I am also being gratuitously rewarded with some A/C current sensor samples from the Coil people.

The AC current sensor will only be used to “back-off” the integral part of the PID controller, as I do want the “I” not to “kick-in” until the desired temperature is almost reached to avoid it’s saturation. The “D” part of the circuit will be present but not employed, as in this application I am dealing with a considerable “time” delay within the process (heaters). The “D” would be maintained only for the purpose of the versatility of the controller.

The temperature - is of the paramount importance.

A 0.1 deg C would be appreciated the most. I have seen the Digikey having on offer an excellent temp sensor, but at the price outperforming the total cost of hardware intended for this project. I do believe, a thermistor is accurate enough to maintain even better sensing if “tuned” to the temp range required.

The employed sensors may sound like an “overkill” for a system with a considerable time delay, but I do not feel it as being hurtful to provide “the best” from sensors maintaining the versatility of the PID at the same time.

I do look forward to your input..

Regards,
xanadunow

PS. I do keep the use of SMTs for the simplicity of testing - not.
 
Last edited:
One more thing. A generic thermistor is just not very linear. You would have to compensate for these nonlinearities with software, or some other means. An off the shelf self linearized device is the easiest way to go. The cost per device may offset the time required to linearize a thermistor.
 
Thanks Mike

The resolution does sound just right. I was not aware of these chips before.

Should you want to endavour on some clicking , the limor does provides so much of the resource information - that I felt being "in love" the moment I saw it.

I would like to explore this tmp.. thank you!

Regards,
xanadunow
 
Last edited:
No software..

No software Mike, I am looking for all analog solution - the tmp chip is not a challenge here.. I just need to get my hands on it first..

Regards,
xanaduow
 
I can arguably, linearise the thermistor to "match" the foreseen range of the temperature(s) that I am after.

Thermocouples would be my choice if the temp was in the high hundreds of degs C or so, this really - leaves me with thermistors.. As it is, I am just less or more, on the boiling point of the water.

Regards,
xanadunow
 
A simple diode gives a -2mV/deg IIRC..:)

Enjoy your New Year.!


EDIT: added image
 

Attachments

  • diode_t1.pdf
    143.1 KB · Views: 188
Last edited:
Do Enjoy, only 40 minutes left - here..

It is not my New Year to enjoy, it is yours too Eric. It is ours.

Let Us all enjoy it in the good spirit.

2mV/deg is comparable with other sensors.. but - the diode is a nonlinear device (to start with..).

xanadunow
 
It is not my New Year to enjoy, it is yours too Eric. It is ours.

Let Us all enjoy it in the good spirit.

2mV/deg is comparable with other sensors.. but - the diode is a nonlinear device (to start with..).

xanadunow

hi,
Look at the image in my edit... :)
Its far better than a thermistor for linearity.
 
can I..

trust you with this picture? I do suppose, next you would be telling me is that you are going to win Ashes again, as if - once is not good enough? .. just joking.. we have been just trashed by RSA..

xanadunow (with just 24 minutes left)
 
Last edited:
there is no comparition seen by me to acknowledge this, and - I know to make the thermistor linear exactly wher I need.. it..

xanadunow
(with just 3 minutes in this year)
 
A simple diode gives a -2mV/deg IIRC..:)

Enjoy your New Year.!


EDIT: added image

I cannot see the PDF, I dunno why, but adobes crashes when I open it.


Well, this was very common in the 80's hobbyst projects, to measure the temp with leakage current or forward voltage of a diode, or even the EB junction of a small signal transistor.

I think that's a good idea do buffer the signal with some FET OPAMPs.
 
I cannot see the PDF, I dunno why, but adobes crashes when I open it.


Well, this was very common in the 80's hobbyst projects, to measure the temp with leakage current or forward voltage of a diode, or even the EB junction of a small signal transistor.

I think that's a good idea do buffer the signal with some FET OPAMPs.

hi,
Just downloaded the pdf as a test, all OK.
As you say, a good OPA is required.
 
That Acrobat file crashes on me, too. Can anybody else open it?
 
I did get this file to open finally Eric.. the swing of the voltage does make me think I would need few diodes in series to sense. I have created an image (in my mind only) of a a better resolution - say 10mV per 1 deg centigrade.. this is a value I would fill much more comfortably to work with.. I am sure I have seen these already, but I was probably not happy with the accuracy or the repetivness of the result.. I may need to make the "one only" dedicated approach if I have to, but I would like not to - if I can..

A number of diodes or thermistors in series would give me the resolution per degree that I want but I need to check up some more on the linearity and combined error if a serial connection is used (i.e. more than 1 sensor of the type in series).. let me think about it please. I can also scale-up the output but it would not change the resolution.. (I did mention the .1 deg C). After all.. two sensors connected in series will double the effective error/tolerance.. 10mV/deg C is achievable but what is the best if we want to maintain the .1 deg C accuracy..

xanadunow
 
Using diodes in series would not work. Each diode are different, so you'd have a crazy temperature sensor.

Eric's suggestion is good, use it with a FET OPAMP, it should work very well.

Or use a LM35 instead.
Analog, linear and gives 10 mV/°C, and it's cheap. NATIONAL SEMICONDUCTOR|LM35DZ|PRECISION TEMP SENSOR, TO-92-3 | Farnell United Kingdom
actually diodes are very reapeatable temperature sensors, at 2mV/C as I recall. The only problem is you need to keep the current constant since the voltage changes with current as well.

Dan
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top