I replaced Jumo thermostat at cookie factory on baking machine. Sadly no one remembers old temperature setting.
I need to fix old controller just to be able to read its settings. Currently I can read some segments but some are not shown.
To me it looks like that Epson SED1510 chip gets correct power and data signals since I can read error message when probe is disconnected.
But upper row is missing a lot of segments.
What can go broke on these TN displays? Used IPA to clean zebra strips and the panel but no improvement.
When I was young, I remember that one old repair men used multimeter on mega ohm range to test the segments on displays like this.
I tried but I am not getting any to be shown dark, even the ones that I saw working when connected to the controller PCB.
Mechanical contact on zebra strips seems good. There are no signs on mechanical damage on the glass.
Attached two pictures of the display part.
What's on the LCD top right corner? Looks like a little chip came off.
Apparently it's an overheating protection device, not a thermostat. So temperature setting should be a bit higher than max temperature setting on that oven. If you replaced it with the same device, you could try simply swapping LCD controller PCB/LCD (should take only a few minutes on site), or transferring EEPROM IC to the new controller. As DavidAlfa said, it looks like LCD has a chipped corner.
How feasible would it be to find another idendical thermostat with a good display? That's probably going to be the easiest, assuming that is a chip in the glass that it appears to be.
Checked and indeed glass is chipped on the corner. Conductive strip doesn't cover that part and there is no pad under it.
This device is already 20 years old and is replaced with new series with same name but newer hardware.
Main use it is to be safety device to limit overheating. But I think that designer of the machine uses one feature of the controller that is called warning temperature.
One can set warning temperature and the controller will open and close relay contacts with hysteresis around this temperature.
But if temperature goes above set alarm temperature relay goes off and never goes on unless manually reset.
This warning temperature and hysteresis around it is what I am trying to find out.
On the PCB there is connector to connect adaptor to PC and t read settings from the PC.
But since there are two versions of the hardware old plain TTL and newer with Modbus probably there are two versions of PC software.
I have one and can not connect and guessing that is wrong one. I am truing with simple ttl converter and I don't have old Jumo adapter. Lot of variables that lead to failure.
I can find old hardware on ebay but shipping time is month+ so not benefit from that also. One simple part replacement ended in endless story of cookies that are not good enough. But that is the price of 20years old equipment and no-one carrying to read parameters properly in those 20years.
You say different HW versions but aren't LCD the same? Because old and new controllers look basically the same from outside and at least in yours it's just a plain LCD with no electronics.
Also I don't see any mention of Modbus in latest specifications. Datasheets from 2004 (model 70.1140) and 2015 (model 70.1160) look almost the same.
I have looked at ebay and price is not a problem but shipping time is.
We are outside EU and shipping usually takes 2 weeks and one week for custom procedure.
As a last option I will order several controllers to get displays. Think it will be faster to write SW to decode spi stream from display controller.
Now I am looking to use cheap usb-ttl converter and this software from Jumo to get the data.
Protocol and the adapter internals are not available.
I am seeing modbus in logs and error details. It is not RS485 but I thin protocol is based on Modbus.
From tests I can see that software sends some byte to the controller and waits for up to 5 bytes back.
After that I guess it expect to communicate with modbus formatted packets. I tried to send dummy data with nucleo and I am getting errors that modbus crc is not ok. Need to find out what original communication adapter does. Maybe it inverts the data or adds some pull up to the lines.
On the controller side there is 87c51 micro and the lines for communication go directly to the micro(to RX and TX line).
But did you actually take the new controller apart? As I already said, even though HW is different, LCD is probably the same.
Yeah, sniffing the SED1510 traffic would be easy.
You only need to check for address set and and memory write commands, and take in count that every memory write increases the address, then it's only 32x4 bits.
You'll be able to extract the LCD commons and segments from the pcb and the controller pinout, however you don't know how every segment translates into the LCD.
You"ll have to figure out...
One thing I'm still slightly confused about though, wouldn't it be fairly easy to figure out the temperature by making an educated guess and then baking a batch of cookies and then tweaking the temperature a bit if they come out over or under baked? Baking cookies is typically not a precision process.
Seems next cookies are going to be extra-high carbon content Oreos
