fafo/jeol-t330a

Fork 0

Add thermal sensor inputs to monitor diffusion pump temperature #5

New issue

Open

opened 2024-09-29 21:43:55 +00:00 by rahix · 4 comments

rahix commented

2024-09-29 21:43:55 +00:00

Owner

From https://git.fa-fo.de/fafo/jeol-t330a/issues/1#issuecomment-164: succbone should get some temperature inputs to continuously monitor certain temperature zones on the diffusion pump.

From my point of view, 3 measurement points make sense:

1 - Heater, located next to the OEM temperature sensor, on the DP sidewall, just above the heater covers.
2 - Chiller, located right in the middle of the cooling coils, where the OEM temperature shutoff sits.
3 - RP Connection, located next to the flange leading to the rouging pump.

For reference:

From <https://git.fa-fo.de/fafo/jeol-t330a/issues/1#issuecomment-164>: succbone should get some temperature inputs to continuously monitor certain temperature zones on the diffusion pump. From my point of view, 3 measurement points make sense: - [ ] **1 - Heater**, located next to the OEM temperature sensor, on the DP sidewall, just above the heater covers. - [ ] **2 - Chiller**, located right in the middle of the cooling coils, where the OEM temperature shutoff sits. - [ ] **3 - RP Connection**, located next to the flange leading to the rouging pump. For reference: <img width="1027" alt="image" src="/attachments/f29361bc-5f75-488b-9778-bb1dbabc01ce">

image.png

572 KiB

rahix added the

succbone

label 2024-09-29 21:43:55 +00:00

q3k commented

2024-09-29 21:53:29 +00:00

Owner

Do you know of any off the shelf thermocouple sensor that we could interface to, eg. via a fieldbus or at some other high level? That might be the easiest and most reliable option.

Alternatively we could hack something together based on thermocouple amplifiers and ADCs, or even better integrated chips. But I'd really prefer skipping the R&D on this one and just getting straight to the point.

Do you know of any off the shelf thermocouple sensor that we could interface to, eg. via a fieldbus or at some other high level? That might be the easiest and most reliable option. Alternatively we could hack something together based on thermocouple amplifiers and ADCs, or even better integrated chips. But I'd really prefer skipping the R&D on this one and just getting straight to the point.

rahix commented

2024-09-30 01:16:23 +00:00

Author

Owner

Do you know of any off the shelf thermocouple sensor that we could interface to, eg. via a fieldbus or at some other high level? That might be the easiest and most reliable option.

There are many options, but they are gonna be expensive. >>300 EUR easily.

I think our best option is to buy some cheap off-the-shelf transceiver modules:

MAX6675-based boards go for around 4 EUR each, but need SPI. So only an option if we have the CS-pins to spare.
MCP9600-based boards are I2C based, but seem to be a bit more costly - ~25 EUR per board.

We can get some of those modules with a thermocouple included, but we need to make sure the wire length is long enough (should be >= 2m I'd say).

> Do you know of any off the shelf thermocouple sensor that we could interface to, eg. via a fieldbus or at some other high level? That might be the easiest and most reliable option. There are many options, but they are gonna be expensive. >>300 EUR easily. I think our best option is to buy some cheap off-the-shelf transceiver modules: - MAX6675-based boards go for around 4 EUR each, but need SPI. So only an option if we have the CS-pins to spare. - MCP9600-based boards are I2C based, but seem to be a bit more costly - ~25 EUR per board. We can get some of those modules with a thermocouple included, but we need to make sure the wire length is long enough (should be >= 2m I'd say).

rahix commented

2024-10-01 04:34:48 +00:00

Author

Owner

One more "middle-ground" option would be a thermocouple transceiver.

Just for comparison: Cheapest ready to use transmitter with...

MODBUS TCP: https://www.aliexpress.com/item/1005005952927177.html (180 EUR)
MODBUS RTU: https://www.aliexpress.com/item/1005004134571951.html (40 EUR)

RTU means we need to add an RS485 interface to succbone. Wouldn't be too difficult, just needs a uart + a direction GPIO. We can then tweak the devicetree to make it work.

Honestly don't see the advantage of RTU over just using the SPI/I2C modules for this current setup, though.

One more "middle-ground" option would be a thermocouple transceiver. Just for comparison: Cheapest ready to use transmitter with... - MODBUS TCP: https://www.aliexpress.com/item/1005005952927177.html (180 EUR) - MODBUS RTU: https://www.aliexpress.com/item/1005004134571951.html (40 EUR) RTU means we need to add an RS485 interface to succbone. Wouldn't be too difficult, just needs a uart + a direction GPIO. We can then tweak the devicetree to make it work. Honestly don't see the advantage of RTU over just using the SPI/I2C modules for this current setup, though.

rahix referenced this issue

2024-10-04 23:24:11 +00:00

Invest into the MODBUS ecosystem #9

rahix changed title from ~~Add thermocouple inputs to monitor diffusion pump temperature~~ to Add thermal sensor inputs to monitor diffusion pump temperature

2024-10-04 23:24:31 +00:00

rahix commented

2024-10-04 23:27:26 +00:00

Author

Owner

We decided to go with a MODBUS-based approach now, as documented in #9. As a part of this, we decided to go with PT100 sensors instead, as hardware is a bit more available for those.

This means, three high-temperature PT100 sensors needed, as detailed above. The selected transmitter hardware can support up to 8 channels so we can think about monitoring more locations as well.

We decided to go with a MODBUS-based approach now, as documented in #9. As a part of this, we decided to go with PT100 sensors instead, as hardware is a bit more available for those. This means, three high-temperature PT100 sensors needed, as detailed above. The selected transmitter hardware can support up to 8 channels so we can think about monitoring more locations as well.