Sciencemadness Discussion Board - Cheap PID temperature control: REX-C100

Cheap PID temperature control: REX-C100

Organikum - 12-1-2013 at 06:21

The REX-C100 PID temperature controller is available for about 10.- € inclusive shipping at Ebay. Thats a bargain I thought and got me one.

The version sold so cheap has a relais output rated at 240V and 10A, whats sufficient for most of my purposes as I use it as preprogrammed in PID mode and it works just perfect controlling hotplates. It can be also used on hotplate/stirrer combos like from IKA which follow some standard what is as it seems that power is turned on when a 12V line is switched through. Connection by 5 pole diode connector, soldering the wires to this was the my main problem for me not being a friend of the soldering iron and more important the soldering iron is not befriending me either

Nevertheless you get a PID control or your hotplate for a fraction of the usual price and one with much more options too. (I compare to a IKA DTS-E5).
Not to forget the display of the REX is actually readable from a distance of more then 50cm - something IKA does not bother with at all.....

For a better temperature control by a tighter timeframe of switching a SSD can easily replace the relay or be just added, the relays driving voltage fits to drive the SSD directly and without problems (thats whats told from different sources).

Great.

What I miss and want to have is a connection to a computer/main controller. And there IS a version of this device outfitted with a RS485 connection. Just not mine and I could not find a source selling it (and if, I am afraid it will be multiple the price and uninteresting).
Looks like there is no technical problem, just the RS485 or simpler RS232 transceiver missing, a MAX232 or similar should fix this for cheap. I am entertaining the hope that the manufacturer is not using different software for different versions, but is in the worst case setting different parameters and so switching the Modbus off. But with a bit of luck there are no changes at all, why switch something off whats not connected anyways?
And reprogramming the PIC would be another option, I have seen a flash-programming program working over the serial input is available and the programable code including Modbus will fall from the heavens as usual

Thats what I am missing for a long time and I have not found an affordable solution for. Data aquisition and device control by basically independent controllers (for security and being usable as standalone). Is it possible?

Now if somebody who has a better grip on the electronics and software issues then me as barely literate in these things could have a look? I think this is interesting and useful for a lot of us here, I also recon that the controller can do more then just temperature without excessive modifications. Would be great of course if others would join this project (if it is feasible at all), affordable, modular, integrated data handling and device control was probably never so much in economical reach before.

The chip used is a Megawin MPC89E52AF a 8051 derivate, all necessary tools for programming it and even a micro-RTS plus many drivers for data input/output storage and things I dont understand.

Personally I would prefer Arduinos doing this job, but this will not be possible, not for this price, as if this works I could finally measure temperature at all the different places of a reaction setup I would like to and alone for this five or seven of the thingies are necessary. Yes I am aware of multiplexing etc. but I dont want to go so deep into this and I dont want to use specialiced hardware - the C100 can always also be used as standalone, it can be replaced by a spare anytime (and keeping spares being no issue at the low costs) and the concept of "one simple interchangeable device for one task" is a convincing one to me having learned that always this gives way whats hard to replace directly related to how near the deadline is.
And incorporating Arduinos for other more demanding tasks (motor control for example) should be simple.

Now please lend me a helping hand!

Attached is the manual of the controller, the datasheet of the PIC (Megawin MPC89E52AF a 8051 derivate) and photographs of the dissected controller in special the PCB of with the digital parts.
If more detailed information is needed, please just tell and I will supply it so any possible.

regards
/ORG

PS: Google showed me nothing what indicates that somebody has already tackled this problem, albeit many are tinkering with the device. Or I have not seen or recognized it, whats of course possible too.

Attachment: megawin_mpc89e52af.pdf (526kB)
This file has been downloaded 1447 times

Attachment: cin-REX-C100 en.pdf (800kB)
This file has been downloaded 1323 times

watson.fawkes - 12-1-2013 at 07:46

Quote: Originally posted by Organikum

The chip used is a Megawin MPC89E52AF a 8051 derivate, all necessary tools for programming it and even a micro-RTS plus many drivers for data input/output storage and things I dont understand.

The chip does have a facility for in-circuit programming, but that doesn't mean that you'll be able to use it. I doubt there's a bootstrap loader already on the chip (there would be no point in it) and without one, there's no way to get software onto the chip. There's no JTAG pins on the device, either. So the simple ways seem to be right out.

On the other hand, there are hardware tricks that might be played. There are connector that can latch onto the quad flat-pack packaging on the chip, and in many circuits such things can be used to drive a chip externally. But none of that specialist gear is particularly cheap. Nor is it generalizable, which means that it's unlikely a community would form around hacking this device.

The data sheet also mentions "code-protecting security", so that's another barrier to surmount, if they've turned it on.

All told, it seems like this device is a dead end as a good target of mods.

The only thing that's specifically designed to be hacked on (that I know of) is www.ospid.com.

Organikum - 12-1-2013 at 12:14

The option to reprogram the chip is secondary to my wish to connect a MAX232 or the 485 equivalent and to find out if communication by modbus is possible. I think there are good chances that it is possible and then I got almost everything I want already.

Am I correct that this means to connect the MAX (an additional capacitor is needed here IIRC, what should be no problem) to the correspondending pins of the chip and then just have a look if there is a reaction when the chip gets addressed by modbus protocols?

When you look at the photo I provided you see on the upper right side of the PCB a line of 8 holes, 5 of them are framed and named R-T-I-G-A (top down) the whole is designated "CN3" and that says connector 3, what else. R and T are for sure parts of the serial in/out, but I cannot identify and name I-G-A now, I for sure will find out, but somebody in the know just telling me would be preferred. The frame of "G" is emphasized btw.
Would you agree that this is the connector to a small PCB containing silicon and connectors resembling the missing RS485 Modbus connection? Now whats needed in parts and how to connect this and I can go on for a quick and dirty test.

I never intended to start a community dedicated to this device, alone for me not able and not willing to invest to much time into the workings of it, but if enabling communication is possible by simple means this empowers just by affordability anybody interested to implement some data handling and device control. Maybe in a crude and not very refined way, but it would be a start. The controllers are of marginal interest for me as long I can hook them to bus and this to a computer which stores the data and makes nice diagrams out of it. Setting parameters on the controllers from the computer should be possible, understood. Free and even open source programs for this - serial data processing, visualisation. storage etc. exist no problem on this side. Modbus is an open protocol whats for sure implemented as virtually all other serial protocols.

Maxim and other companies are providing free samples. A chance to get Modbus enabled PID controllers for less the 10.-€ a piece? I will give it a shot for sure as I woould not forgive myself not to, there is nothing to lose.

Anybody going to spare me the dull hours of reading about things I am not really interested in and disclose me R-T-I-G-A and how to go from there to RS232 or 485?

regards
/ORG

PS: "G" is ground. Me stupid.

[Edited on 12-1-2013 by Organikum]

watson.fawkes - 12-1-2013 at 16:08

Quote: Originally posted by Organikum

The option to reprogram the chip is secondary to my wish to connect a MAX232 or the 485 equivalent and to find out if communication by modbus is possible.

You'll be unlikely to drive another chip if you can't write firmware to control it.

Just because there seems to be unsoldered connectors on the board does not mean that there's firmware on the CPU to drive it. Since PCB manufacturing is definitely in the economy-of-scale department, it's rather common to have multiple product lines use the same PCB, but populated differently. This includes the CPU, which is often downgraded for low-end products.

Quote: Originally posted by Organikum

I never intended to start a community dedicated to this device, alone for me not able and not willing to invest to much time into the workings of it, but if enabling communication is possible by simple means this empowers just by affordability anybody interested to implement some data handling and device control.

Either you do the work yourself, or you convince some other single person to do it, typically by paying them, or you boot up a collaboration, a community. Of these, you've said the first won't happen, I presume the second is ludicrous, and so the third seems the most likely.

If you think there's an active-but-unused communication port on the device, put an oscilloscope on it and see.