Best thermocouple location for PID controlled steam boiler

[mwills]

i'm prepping to install a PID to my Wega mininova. the motivation is primarily temperature control but also looking to address preventative maintenance (old pressure-stat), noise (old pressure-stat), and temperature stability. my main concern is were to insert the thermocouple. here is a diagram of the boiler:

here are the part descriptions:

as you can see, there are no un-used ports on the boiler. however, if i were to drill and tap a new port (inserting this TC: http://www.espressoparts.com/product/MZ_L530) - any suggestions on where to place the TC? where would be the most 'volatile' region in the boiler? thx.

[AndyS]

as you can see, there are no un-used ports on the boiler. however, if i were to drill and tap a new port (inserting this TC: http://www.espressoparts.com/product/MZ_L530) - any suggestions on where to place the TC?

If you're willing to spend $89 on that thermocouple, then you should consider doing it right and running your PID off a pressure sensor instead. Advantages include no holes drilled in the boiler and faster response. Disadvantage is slightly higher cost ($125 for sensor, $15 for power supply).

Thermocouple-based PIDs are so 2001. Why not do something more up-to-date?

[jesawdy]

A couple ideas...

Just surface mount the TC to the boiler.
Insert the TC into the 'T' where the pstat was located (item 17, no?)
With the use of appropriate tees and elbows, create a port for the TC where the vaccum breaker or pressure relief valve is.

[jesawdy]

If you're willing to spend $89 on that thermocouple, then you should consider doing it right and running your PID off a pressure sensor instead. Advantages include no holes drilled in the boiler and faster response. Disadvantage is slightly higher cost ($125 for sensor, $15 for power supply).

Andy, have you used one of these? That seems relatively cheap.... stated operating range is only up to 185F (or at least for the one you linked to).

[mwills]

andy -

thx for the suggestion. i wasn't aware you could pick up pressure sensors for that price. i'll follow-up with the company on the price and operating temps.

jesawdy -

i thought about surface mounting but was worried about the extra time necessary for temp stability when considering the thermal gradients through the water+boiler walls. ie. how long would it take the boiler walls to respond to a 1deg change in desired water temp.

i am also considering your other suggestion - i think adding a T to the pressure relief would be the most economical and eliminate the need for the TC listed above. such as a sheathed omega TC?

[AndyS]

i wasn't aware you could pick up pressure sensors for that price. i'll follow-up with the company on the price and operating temps.

Please share the results of your follow-up, thanks.

I didn't mention that the PID has to be able to accept an analog input, such as 4-20ma. But you probably knew that already....

[cafeIKE]

Some more suggestions:

- Use a T type TC. It's more suited to the temperature range.
- If you insulate the boiler, the time difference between a TC inside and a TC outside the boiler will be minimal. There are eddy currents in the boiler that circulate cooler and hotter water which may be averaged by the tank wall. Judicious tuning of the PID parameters should be more than capable of maintaining the set temperature.

A TC and an OmegaLok should cost less than half the $89. If you use an external TC, no OmegaLok required.
Check out http://www.omega.com and for PID tuning see http://www.omega.com/temperature/Z/pdf/z115-117.pdf

[mwills]

thanks for the input all - i'll post my results when i'm finished. i'll start with the less invasive surface-mount method method.

[SPG]

When I put a PID on the steam tank of my LM, I was told by Bill C. to put it near the top so that the flash boiling of the water was less likely to hit it. As for response, it is very fast, Just cracking the steam valve drops the steam temp 2-3 degrees. I don't think the external TC mounting would be nearly as quick. I have narrowed the overall range and set the Proportional range very narrowly so that after dropping 2-3 degrees it will be 100% on, with a feathering as it approaches 260 deg. F (my 1.5 bar setpoint). It has a bout .5 deg overshoot when being controlled so aggressively, but since it is not my brew tank, I really don't mind, it is still not enough to notice on the pressure gauge. The pressure gauge just seems locked on 1.5 bar until you crack the steam valve, once you do the temp drops and the element is solidly on just like the pressurestat used to do.

The advantage as I see it: It's quiet, it's more stable and equally as fast, it's solid state, and it has less overshoot than the P-stat.

The disadvantage is cost and installation complexity, neither of which are excessive.

[augustusflavius]

If you're willing to spend $89 on that thermocouple, then you should consider doing it right and running your PID off a pressure sensor instead. Advantages include no holes drilled in the boiler and faster response. Disadvantage is slightly higher cost ($125 for sensor, $15 for power supply).

Thermocouple-based PIDs are so 2001. Why not do something more up-to-date?

AndyS-
Can you give some more input on the use of the AST4100. I have a Fuji PXR4 and want to understand how best to use it with a AST4100 since the setting value would not be temperature related. Have you used a pressure sensor to PID an espresso machine?

[Matthew Brinski]

AndyS-
Can you give some more input on the use of the AST4100. I have a Fuji PXR4 and want to understand how best to use it with a AST4100 since the setting value would not be temperature related. Have you used a pressure sensor to PID an espresso machine?

Click here and you'll find some fantastic info for reference ... http://brewtus.googlegroups.com/w...s%20Compendium.pdf

[AndyS]

AndyS-
Can you give some more input on the use of the AST4100. I have a Fuji PXR4 and want to understand how best to use it with a AST4100 since the setting value would not be temperature related. Have you used a pressure sensor to PID an espresso machine?

You would need a PXR4 that accepts a 4-20ma input, NOT a thermocouple input. If you already own the PXR4, then it's probably too late.

You'd need a 24 vdc power supply in series with the sensor and the PID's input stage.

You'd need to provide a little passive temperature protection for the sensor, probably by simply piping it through a small coil of soft copper tubing.

You would program the PXR4 to read bar rather than degrees, although you could really do it either way.

The main advantage of all this would be a little bit faster reaction time. When you opened the steam valve, your control would instantly react to the lowered pressure. Another advantage is that it would just be more INTERESTING than another ho hum PID installation.