Ds1820 -> pid -> pwm

Hi Everyone;

I need to do a university project, for "Computer Control. The project consists of performing a PID controller based on a Microcontroller.
The PID wiil work on an Aquarius Water heating process, it will acquire the temperature from a DS1820, compare the temperature with the SetPoint them do the PID calculations, and them output the value to PWM that is connected to a resistive heater. The Man/Machino interface will be performed by a LCD2X20 and 3 Buttons. I need a menu to change the SetPoint, and other to Change the value of the Kp, Ki, and KD variables to see the PID response for different values of these variables!

I'm looking at this threads;
http://www.picbasic.co.uk/forum/show...ghlight=pwm%2A
http://www.picbasic.co.uk/forum/showthread.php?t=12712

It is possible to do this? I think so, however i would like your opinion.
My questions, for now, do you think that this is possible to do with a PIC16F877A@20Mhz??

Thanks everyone!

Quote:

Originally Posted by gadelhas

My questions, for now, do you think that this is possible to do with a PIC16F877A@20Mhz??

Thanks everyone!

Well originally we had 4 PID loops running on a 16F877A with a 20Mhz Xtal so I guess the answer to your question is yes !

As most of your project has already been provided in the sample code in one of your links, there is little work required on your behalf... however I would suggest you didn't rip off the code from that thread and claim it as your own when you submit your project for assestment. This wold be plagerisum (?)

Quote:

Originally Posted by malc-c

Well originally we had 4 PID loops running on a 16F877A with a 20Mhz Xtal so I guess the answer to your question is yes !

As most of your project has already been provided in the sample code in one of your links, there is little work required on your behalf... however I would suggest you didn't rip off the code from that thread and claim it as your own when you submit your project for assestment. This wold be plagerisum (?)

Hi;

Thanks for the answer;

I don't intend to suggest to my teacher that the code is mine. I need to do a report, and i have to mention where i get the source code and all other things, like images, texts..... My teacher is also a user of this fórum.
The meaning of the project is to unsderstand the funcionality of a PID controller, and see the response of the PID by changing the Kp, Ki, Kd variables.

If i complete the project, it's my intention to post here as open source!!!

Hope i can count with your help, i will make the schematic and them post here!!

Should i use a solid state relay to control de Heater Resistor or a Triac??

Thanks once again!

Quote:

....The PID wiil work on an Aquarius Water heating process....

Quote:

...The meaning of the project is to unsderstand the funcionality of a PID controller, and see the response of the PID by changing the Kp, Ki, Kd variables.

I strongly suggest you to remove the fish before experimenting with Kp, Ki, Kd variables.

Cheers

Al.

Quote:

Originally Posted by aratti

I strongly suggest you to remove the fish before experimenting with Kp, Ki, Kd variables.

Cheers

Al.

I think it's better too....

Thanks

Good advice Al :)

@gadelhas
Most of the code in the PID thermostat thread was written by Darrel and Henrik, and contains a means of tuning the values you mention. Later this evening I'll go back through the early versions of the code I have for a single PID channel that can be configured via a PC running Hyperterminal and communicating via the serial port. I'll zip it up and attach that to the post. You can then experiment with the values and compose your report

Hi,
What you want to do is very much doable with pretty much any PIC, a 16F877 will be plenty. Make sure you read thru the complete PID filter thread (if you haven't already) as it covers several things - and provides an updated version of the PID filter in post #57.

The easiest way is to use a solid state relay as it isolates the load from the PIC and is easy to drive (it's basically just a LED). But to give a more detailed answer we'd need to know if the heater is AC and you by PWM really mean phase angle control. Or if it's DC and you intend to use the CCP-module in the PIC to produce a PWM signal - or (if it's either AC or DC) you intend to create a slow speed PWM with a period of several seconds or longer (I beleive this is what Malcolm did) where the fact that a triac or solid state relay for AC doesn't switch off until next zero crossing doesn't matter.

/Henrik.

Quote:

Originally Posted by HenrikOlsson

The easiest way is to use a solid state relay as it isolates the load from the PIC and is easy to drive (it's basically just a LED). But to give a more detailed answer we'd need to know if the heater is AC and you by PWM really mean phase angle control. Or if it's DC and you intend to use the CCP-module in the PIC to produce a PWM signal - or (if it's either AC or DC) you intend to create a slow speed PWM with a period of several seconds or longer (I beleive this is what Malcolm did) where the fact that a triac or solid state relay for AC doesn't switch off until next zero crossing doesn't matter.

/Henrik.

I agree, an SSR is the simplest way to interface the output from a PIC to mains device. In the development of the thermostat project I'm now using zero-crossin opto-triac isolators to drive 16A triacs, but the original prototype used 4A SSR.

As Henrik stated, my thermostat project uses low speed pwm to apply power to the heaters, the result is that the heaters remain at a nice stable temperature rather than the normal method of heating which results in large temperature swings

Here's an example of the frequency

http://www.youtube.com/watch?v=ZAhU8Z7-vI4

Hi Again, and thanks for all of your answers!!

Here is my 1º schematic of the PID Controller. As i say in my revious post, the Buttons, is to set the SetPoint, and change the Ki, Kp, Kd, variables.

http://img802.imageshack.us/img802/8798/pid.th.jpg

Quote:

malc-c
Good advice Al

@gadelhas
Most of the code in the PID thermostat thread was written by Darrel and Henrik, and contains a means of tuning the values you mention. Later this evening I'll go back through the early versions of the code I have for a single PID channel that can be configured via a PC running Hyperterminal and communicating via the serial port. I'll zip it up and attach that to the post. You can then experiment with the values and compose your report

@ malc-c , if possible put the code here as you had mension. I would appreciate.

@ HenrikOlsson, I'm reading and trying to understand your PID routine, i will post some questions soon about it, i have some questions!!
The Heater resistor that i have is 220V AC, i think it's better to do, like malc-c did, with a Solid State Relay, and low speed PWM.

@ malc-c Again. I see your video, and i think the result its really great. If you change the Kp, Ki, Kd, variables, do you see any change to the process? Like, more slow/ fast to achive the SetPoint, for instance?

Thanks once again for everything.
Hope you can understand my english, sorry for all the errors!

Quote:

Originally Posted by gadelhas

@ malc-c , if possible put the code here as you had mension. I would appreciate.

To be honest you could use the code attached to my first post in the thermostat thread http://www.picbasic.co.uk/forum/showthread.php?t=12712
You would need to build / breadboard your hardware to match that used in the included hardware file.

Quote:

Originally Posted by gadelhas

@ HenrikOlsson, I'm reading and trying to understand your PID routine, i will post some questions soon about it, i have some questions!!
The Heater resistor that i have is 220V AC, i think it's better to do, like malc-c did, with a Solid State Relay, and low speed PWM.

I've not tried it, but I would of thought that it would be kinder on the resistive element of the aquarirm heater, especially as with it immersed in water the heater has less chance to cool as a similar element would in air. A SSR is the way to go - bit costly, but a lot safer

Quote:

Originally Posted by gadelhas

@ malc-c Again. I see your video, and i think the result its really great. If you change the Kp, Ki, Kd, variables, do you see any change to the process? Like, more slow/ fast to achive the SetPoint, for instance?

Thanks once again for everything.
Hope you can understand my english, sorry for all the errors!

From memory, changing the values altered things like the amount of over-shoot of the set point, the range at which the output started pulsing, and the length of pulse. The advantage with the code that Darrel and Henrick had put together was that using a serial cable and Hyperterm I was able to change the values via the PC and see what effect it had on my test rig which comprised of a heater in a wooden box with the sensor 6" below a ceramic heater. Over several weeks the values were adjusted and tuned until the temperature set point either maintained or was within half a degree either side (as you can see by the video - in the real environment it's almost spot on)

Working with ac and resistive load, phase angle control is the choice (dimmer principle), PWM works very well only with dc.

Since you will have 50 Hz main, every half cycle will last 10 millisecs, so once you have found the zero crossing moment, you will fire your triac with a delay from 0 millisecs to max 10 millisecs.

where:

0 millisecs delay = 100% duty
.
.
5 millisecs delay = 50% duty
.
.
10 millisecs delay = 0 % duty

Remember that your triac will turn off at every zero crossing, which make phase angle control a rather efficient and easy way to control resistive loads with ac.

Cheers

Al.

Hi Al,

I agree, phase angle control is nice but IMO overkill for a system like this.

The heater itself and the system as whole is most likely quite "slow". Creating a low frequency PWM with a period of 1 or even 10 seconds will likely work just fine.

That way the slight error introduced by the solid state relay (or triac) not shutting off until next zero-crossing can be disregarded. (Max 1% with 1s period or 0.1% with a 10 second PWM period.) Depending on if the SSR is "random firing" or not it'll be a little bit different but still good enough IMHO.

Of course it depends on the power of the heater, the amount of water in the tank and the amount of "temperature ripple" you can tolerate but I don't think the tank will "respond" fast enough that a 1 or even 10 second PWM period won't work.

/Henrik.

Quote:

The heater itself and the system as whole is most likely quite "slow". Creating a low frequency PWM with a period of 1 or even 10 seconds will likely work just fine.

Hi Henrik, Let me say that I could accept your statement only if we talk about an amatour project. But since the OP opened the thread with the following statement:

Quote:

I need to do a university project, for "Computer Control. The project consists of performing a PID controller based on a Microcontroller.

In this case I will expect from him to challenge the highest performace he could think of.
He will be a tomorrow electronic designer and he will have to face an extremely aggressive global market, so I do expect from him something more than "something that will likely work just fine"

Herik, I Hope you will agree with me!

Cheers

Al.

Quote:

Originally Posted by gadelhas

I don't intend to suggest to my teacher that the code is mine. I need to do a report, and i have to mention where i get the source code and all other things, like images, texts..... My teacher is also a user of this fórum.

Quote:

Originally Posted by aratti

In this case I will expect from him to challenge the highest performace he could think of.
He will be a tomorrow electronic designer and he will have to face an extremely aggressive global market, so I do expect from him something more than "something that will likely work just fine"

Al.

Do we have teacher and pupil here ???

Hi Al,
I believe that creating a product that serves its purpose without making it more complicated than it needs to be is key for an electronic designer, today and in the future.

Going by the project description in the first post (ie temperature control (with PID) of a fish tank) I do beleive that low frequency PWM will work just as fine as a phase angle control but will be easier to implement, produce less electrical interference, require less parts and be easier to build - and it'll work on 50 and 60Hz mains without paying special attention. That's good product design in my book but I don't work as an electronics designer, I don't have a degree and I never went to university so what the heck do I know.

If the students gets assessed by what they CAN (possibly) create then by all means, throw everything you got at it. Phase angle control, datalogging, GLCD with trend charts, TCP/IP interface, you name it. But I hope they gets assessed by their abillity to indentify the real needs to solve the problem at hand and then design a product that works - properly of course.

I DO agree that the correct method should be used. If phase angle control in this application offers benefits that justifies the extra complexity then it IS the correct choice.

Well, that's my view on it. 10-4

Sincerely,
/Henrik.

Hi Again Everyone, and once again thanks for all of your answers to the thread!

I can only come to here at late night because i work during the day, and study at night, so sorry for only now answer to the thread!

@malc-c

Quote:

To be honest you could use the code attached to my first post in the thermostat thread http://www.picbasic.co.uk/forum/showthread.php?t=12712
You would need to build / breadboard your hardware to match that used in the included hardware file.

I Allready have that code, however this code don't permite changing the variables with hiperterminal like you said, can you give me the one that does?

Quote:

Do we have teacher and pupil here ???

Yes, my teacher usualy comes to here and MiKroelektronika forum, however he is more a Assembler fan!!!

Quote:

From memory, changing the values altered things like the amount of over-shoot of the set point, the range at which the output started pulsing, and the length of pulse. The advantage with the code that Darrel and Henrick had put together was that using a serial cable and Hyperterm I was able to change the values via the PC and see what effect it had on my test rig which comprised of a heater in a wooden box with the sensor 6" below a ceramic heater. Over several weeks the values were adjusted and tuned until the temperature set point either maintained or was within half a degree either side (as you can see by the video - in the real environment it's almost spot on)

That's exactly what i wont, to see the response of the system, by changing the variables, that is the real porpose of the work!!!

@aratti
I know that the phase angle control it's much better to control AC charges, however, since it's slow speed PWM, because the system is slow, i think, that i have other advantages, like Henrik said! I have PID controllers on the University that i can use, however my teacher made me this challenge, and i accepted!
Please don't judge me wrong, if i have the time, believe me that i would do the phase angle control, and I will put a GLCD to show the curves of the system response, but in the meantime i need to understand the PID, put it to work, and them improve the whole system, if i've got the time! This is not an excuse, but i work, i study and in 3 months my baby will be born, however like i said this is not a excuse. I really appreciate your opinion!

@HenrikOlsson
Why to you have the "pid_I_Clamp" and "pid_Out_Clamp"i'm not understanding the meaning of this variables?

Thanks all for all your support!

Hi Gadelhas,
I've explained the purpose and use of both of those variables in the original thread but I'll try again.

pid_I_Clamp limits the amount of "power" that the integral term of the PID-filter is allowed to contribute with to the total output. When doing your "PID homework" you'll come across the term windup, the pid_I_Clamp provides a way to prevent that.

pid_Out_Clamp can be used to limit the total output of the filter. For example, if you your PWM generator has a resolution of 8 bits there's no need (and can possibly even be bad) for the output value to go above 255. The pid_Out_Clamp allows you to specify the maximum "output swing" of the regulator. Note that if you set it to 255 it can swing +/-255, if your system doesn't accept negative drive you need to take care of that "outside" of the filter.

I hope that answers your question.

/Henrik.

Quote:

Originally Posted by gadelhas

@malc-c

I Allready have that code, however this code don't permite changing the variables with hiperterminal like you said, can you give me the one that does?

I'll configure the EasyPIC5 board for use with a 16F877A and 20Mhz crystal and try the code, probably have to be tomorrow afternoon now as I won't be home from work until late. But I'm sure it does allow you to change and set the values of PID etc as it contains the code to do so (snippet attached)

Code:

;----[handle incoming serial Terminal data]--------------------------------- Term_RX: HSERIN [Char] SELECT CASE Char CASE "R","r" ' Refresh GOSUB TerminalMenu CASE "S","s" ' Enter Setpoint X=1 : Y=22 : GOSUB MoveCursor HSEROUT ["Enter SetPoint ... (340=34.0",$F8,"): "] HSERIN [DEC SetPoints(EditChannel)] GOSUB TerminalMenu LOOKUP2 EditChannel,[EE_SetPoint1,EE_SetPoint2, _ EE_SetPoint3,EE_SetPoint4],Bvar WRITE Bvar, WORD SetPoints(EditChannel) CASE "P","p" X=1 : Y=22 : GOSUB MoveCursor HSEROUT ["Enter P term as HEX4 ... : "] HSERIN [HEX4 pid_Kp(EditChannel)] GOSUB TerminalMenu LOOKUP2 EditChannel,[EE_pid_Kp1,EE_pid_Kp2, _ EE_pid_Kp3,EE_pid_Kp4],Bvar WRITE Bvar, WORD pid_Kp(EditChannel) CASE "I","i" X=1 : Y=22 : GOSUB MoveCursor HSEROUT ["Enter I term as HEX4 ... : "] HSERIN [HEX4 pid_Ki(EditChannel)] GOSUB TerminalMenu LOOKUP2 EditChannel,[EE_pid_Ki1,EE_pid_Ki2, _ EE_pid_Ki3,EE_pid_Ki4],Bvar WRITE Bvar, WORD pid_Ki(EditChannel) CASE "D","d" X=1 : Y=22 : GOSUB MoveCursor HSEROUT ["Enter D term as HEX4 ... : "] HSERIN [HEX4 pid_Kd(EditChannel)] GOSUB TerminalMenu LOOKUP2 EditChannel,[EE_pid_Kd1,EE_pid_Kd2, _ EE_pid_Kd3,EE_pid_Kd4],Bvar WRITE Bvar, WORD pid_Kd(EditChannel) CASE "T","t" X=1 : Y=22 : GOSUB MoveCursor HSEROUT ["Enter Ti term as DEC ... : "] HSERIN [DEC pid_Ti(EditChannel)] GOSUB TerminalMenu LOOKUP2 EditChannel,[EE_pid_Ti1,EE_pid_Ti2, _ EE_pid_Ti3,EE_pid_Ti4],Bvar WRITE Bvar, pid_Ti(EditChannel) CASE "C","c" X=1 : Y=22 : GOSUB MoveCursor HSEROUT ["Enter I term Clamp as DEC ... : "] HSERIN [DEC pid_I_Clamp(EditChannel)] GOSUB TerminalMenu LOOKUP2 EditChannel,[EE_pid_I_Clamp1,EE_pid_I_Clamp2, _ EE_pid_I_Clamp3,EE_pid_I_Clamp4],Bvar WRITE Bvar, WORD pid_I_Clamp(EditChannel) CASE "O","o" X=1 : Y=22 : GOSUB MoveCursor HSEROUT ["Enter Output Clamp as DEC ... : "] HSERIN [DEC pid_Out_Clamp(EditChannel)] GOSUB TerminalMenu LOOKUP2 EditChannel,[EE_pid_Out_Clamp1,EE_pid_Out_Clamp2, _ EE_pid_Out_Clamp3,EE_pid_Out_Clamp4],Bvar WRITE Bvar, WORD pid_Out_Clamp(EditChannel) CASE "M","m" X=1 : Y=22 : GOSUB MoveCursor HSEROUT ["Manual Mode not available. "] CASE "1","2","3","4" ; select channel IF EditChannel != Char -49 THEN GOSUB ClearBox EditChannel = Char -49 GOSUB DrawBox ENDIF CASE ELSE HSEROUT [7] ; bad key, send Bell END SELECT RETURN

It might not work with the version in MCS, but works fine with the version that came with windows XP - maybe google for a similar application

You should get something like

http://micro-heli.co.uk/PPstat1.jpg

Hi;
@Henrik;

Thanks for the answer, i understand now what it means! I will have some more questions, so please don't be mad with me!! :(

@malc-c
Thanks for the code, but i'm not understanding how this piece of code works with the other. Where is the "terminalMenu", "clearBox", "Movecursor" routines?
When i post the schematic of the hardware, was according to my EasyPic6 Board. I will build the circuit and try the codes in the next Saturday. I think EasyPic5 has the same conections that EasyPic6.
Please if you try the code posted, tell me the results, if possible, of course. Thanks again!

Quote:

I will have some more questions, so please don't be mad with me

Don't worry, I'll try to answer/explain what I can.

Quote:

Originally Posted by gadelhas

@malc-c
Thanks for the code, but i'm not understanding how this piece of code works with the other. Where is the "terminalMenu", "clearBox", "Movecursor" routines?
When i post the schematic of the hardware, was according to my EasyPic6 Board. I will build the circuit and try the codes in the next Saturday. I think EasyPic5 has the same conections that EasyPic6.
Please if you try the code posted, tell me the results, if possible, of course. Thanks again!

The code I pasted above was a snippet - It was taken from the zip file attached to the 1st post in my thermostat post you linked to above.

1 Attachment(s)

OK, I've just confirmed that the files in the first post of my thermostat thread compile and display the output in hyperterminal as per the image shown in a post above. This version is nothing fancy, it will display the temperature of 4 x DS18B20's on a 4 line 20 chr LCD and drive 4 outputs which can be connected to SSR's to switch main voltage devices, It also allows the values of the PID to be tuned by entering values via hyperterminal on a PC connected via a serial cable.

Ive re-attached all the files here, inc the compiled HEX - please see the hardware inc file which defines the connections to the LCD etc. You are welcome to use this as the basis for your learning....

Pid

We have a complex temperature control on a 16F777 with 20 temperature sensors (5 DS18S20s, the rest analog LM50 and a couple of MCP9701, plus a Sensirion for ambient T/H.) The poor PIC PWMs 3 sets of fans, reads a flowmeter, drives a heater, talks to another brain and handles two sets of serial out. Just as well, driving temperature, you're not in a hurry.

Temperature with a large thermal inertia has such a long time domain that it's not a great place to demonstrate PID. A few years back I did a motor PID with a PIC that worked great - just one axis and it was slow but moving a 500lb weight. Motor tuning is the quickest way to illustrate the effect of each PID component.

Back to temperature: PID is a wonderful thing but in the real world may need a lot of tweaks. First off we abandoned the Derivative as all-but-meaningless on these timescales and now use a setpoint and hysteresis band system with PI(D) only above the upper threshhold. The temperature moves in a slow triangular waveform inside the band (like any thermostat with hysteresis) and only invokes PI when things get warm. Also we have an emergency threshhold up there that drives everything to safety levels. (We are not allowed to power down.)

Good luck with your project.

Hi Everyone;

Sorry, but only now had time to get to the fórum!

[quote]
Don't worry, I'll try to answer/explain what I can. [/quote}
@HenrikOlsson
Once again thank you for your support!

@malc-c
Like i said before, this weekend i'll try the code posted, and then i put here the result. Thank you for the code!

@kenif
Thanks for your opinion and support, I will take it into consideration!!