DIY Temperature Controller for Heating Blankets

[Robbie Collins]

I'm thinking about building a temperature controller using a PID unit off of eBay. Does anyone have experience doing this?

[Ned Knepp]

Steampunk, Cyberpunk. One pumps electrons and the other pumps water vapor but it's all gona end badly in the end. ( Or, at least my mom used to say.)

Is it possible that you need to remove the binding first then attack the fingerboard?  ( can't say since I've never really had to worry about wrecking someone else's high dollar guitar... and that why I'm a hobbist! ) 

[Robbie Collins]

You want steampunk? Check out these guitars:

http://www.tonycochranguitars.com/index.html

[Ned Knepp]

I do like steam punk  but I've never tried to make anything. Wonder if I can work up an acoustic that's steam punk. Maybe a resonator?

[J R Emmett]

Excellent observation Ned.  SSR's are great from a duty cycle perspective and provide excellent temperature control but they do fail closed. 

I've designed hundreds of heater controller circuits for industrial use.  Experience (and a few corporate safety audits) have driven the requirement to provide redundant over temperature control IF the application has the potential to be run unattended or can start a fire. 

 

A mechanical relay based solution is best  as those systems have an open circuit failure mode.  The inexpensive solution is to use a mechanical limit thermostat.  A device like this attached to the heater or heated surface  would be wired in series with the hot wire supplying voltage to the heater:

 

http://www.newark.com/jsp/search/productdetail.jsp?SKU=16M4626&...|

 

It provides a redundant temperature limit that disables the heater if it exceeds the prescribed temperature.  The bad part is that it will reset itself so a fault condition can be repeated so this is not perfect.

 

There are adjustable bulb and thermocouple type devices available as well.  The company where I used to work required a manual reset, relay based controller that would shut down the power supply at the mains like this:

 

http://www.watlow.com/products/controllers/series-lv-limit-controll...

 

They aren't cheap.  If you are absolutely NEVER going to operate the heater without continuous monitoring and have a plan to deal with a fire should it happen - you probably don't need this level of redundant safety.  If you are going to let it run continuously, you may want to consider adding a safety control.

[Ned Knepp]

J.R., 

 I've seen these thermostats in old Laser printers and copiers in the heater circuits. Never knew what they were. I may actually have a couple of them laying around somewhere if I didn't haul them away with the hazards waste I tend to collect.  

  I break down my old computer system to recyclable and non-recyclable parts which is usually hazardous waste so I collect that and take to our local HW station. Several of my friends know that I do this so sometimes I end up with several system to do at once. I'm pretty good at it now.

[Robbie Collins]

I finally got all the parts in from China:

Yesterday I cut the holes in the Enclosure with a Dremel and a Gyros ST carbide wheel. If you haven't tried them they are a couple of inches in diameter and very robust. Lasts much longer than a little Dremel wheel. Today I wired it up using wire I stripped from a IEC AC cord and some connectors from Harbor Freight.  I found an old aluminum Pentium heat sink, cut it to size with the bandsaw, and installed it on the solid state relay base with heat sink compound.  I then attached it to the bottom of the enclosure using the screws that held the fan. I drilled some holes under the heat sink for ventilation.

Here's it plugged in. You may notice that there are two fuses. One that's part of the AC input/switch/fuse combo and a standalone fuse.  The input fuse is 250v 5A and basically protects against surge and excessive draw from the controller or relay.  The other fuse is wired between the relay and the heating blanket and protects the relay from excessive draw from the blanket. I have a 120v 6A fuse which by Ohm's Law W=V*A = 720w which is about 13% more than the largest side bending blanket from LMI draws.

My blankets still haven't arrived. When they do I'll start posting on the Trini Lopez Fingerboard thread I started.

I couldn't find a panel mount K-type female jack for the thermocouple so I used a banana plug unit. I'll cap the thermocouple ends with some male banana plugs. Next I'll use my Brother labeling unit and print out some legends. Also need to figure out how to light up the blanket switch on the front. It has three posts...two silver for the switch and a third brass one. Would that go to neutral?

[J R Emmett]

Awesome - very nice construction of the control box and a professional wiring job!  One source (although a bit expensive) for all things temperature control that I've used for years is Omega Engineering.  Great source for thermocouples, plugs, sockets, controllers and wire. 

 

http://www.omega.com/

 

Just one note on thermocouples - for absolute accuracy, all connections from the controller to the thermocouple should be the proper type of wire and sockets.  You may see a degree or two of temperature offset in your controller, which should not matter for this application.  Also, a general rule of thumb for temperature control accuracy is that a +/- 2C error in any system like this is typical (and acceptable).

 

When you get your heaters, you will need to tune your controller.  The Omega site has pretty good technical information about PID controllers and setting parameters.  Most modern controllers have an "auto tune" (no vocal talent jokes, please :-) ) feature that will get you very close.  If you run into issues (slow to reach temperature, unstable temperature, etc) send me a message and I'll do my best to help.

[Robbie Collins]

J.R. - how does the Autotune feature work? I assume this an auto-calibration feature? I set Autotune=1 in the setup menu. It seems like this would be a one-time deal.

What's confusing is that auto-calibration would need feedback from a second sensor or for the operator to input the actual blanket temperature read with, say, an infrared thermometer.

[J R Emmett]

Auto tune is not about calibration or accuracy of temperature.  If you want to check accuracy - place the thermocouple probe in a bucket of ice water (should read near zero degrees C) and then in a pot of boiling water (should read near 100 degrees C).

 

Autotune sets up the parameters for the PID controller to provide precise temperature control - particularly in the transient (warm up) state.  If the PID parameters are incorrectly set, you will see gross setpoint overshoot  (a few degrees is normal) during warm up and you may also see an oscillating temperature that never stabilizes around the setpoint.  In the "old" days, we used to enter the parameters manually and iterate them until we had a stable temperature with low overshoot.  Autotune was a godsend as most controllers will develop better parameters than I ever could.  Check your manual - it should have instructions about how to use this feature.  Typically, you set the parameter to 1 and then turn on the heater and let autotune do it's thing as the system warms up.  If you power cycle the temperature controller, I suspect you will find the autotune parameter set to 0. 

[Robbie Collins]

I finally received the heating blanket from LMII and was able to test my controller.  I set it to 121C (250F) and it took the blanket about 3 minutes to get to that temp. Unfortunately, I can't lay my hands on my infrared thermometer to check it.

The unit is has two readouts for PV and SV.  PV means Processed Value and is the thermocouple reading.  SV is the Set Value or the temperature you want. When the unit boots it gives the general parameter settings on the PV and SV LED readouts. Mine said it was configured for a K thermocouple which is what I'm using.

Setting the temperature requires pressing the Set button which highlights the right hand side digit. Pressing it again gets the 2nd digit, etc.  I set it to 121C then turned off the unit. The unit rebooted with SV=121.

Wiring the DC output to the SS Relay, I couldn't figure which screw (4 or 5) was positive. The SSR is marked + and - so I assume it cares.  I checked the output with my Fluke and got 13.68v but swapping test leads got the same value. I thought it would read +13.68 or -13.68 depending on whether the positive lead was on the positive output, etc.

If SV > PV then the controller should kick on the relay, but the red relay LED wasn't on.  So, I swapped the control wires on the SSR, played with the internal menus, power cycled the unit and the SSR came on. So, I don't know whether it was wired backwards or not.

I turned Autotune on, but truthfully I don't know what it does. Auto-calibration?

So far, so good! Tomorrow I will attempt to remove a fingerboard and will report back. Any ideas on the initial temp for removing a bound fingerboard?

The mess around the rear connections is hot glue. My dremeling wasn't very exact.

[J R Emmett]

I like the temperature conversion label  :-).  Unfortunately, the only metric units that I think in are Temperature and Mass (grams).  Your controller may have a setting to display temperatures in either unit. 

 

I'll offer a link to this page as it contains converters for nearly everything.

 

http://onlineconversion.com/

 

Great job on the project Robbie.  Look forward to seeing the test reports.

[Nate Scott]

Nice controller box build, Robbie. I'm collecting parts to make one for myself. What are the dimensions of your enclosure? Source?