Custom built fan controller for insert

[paulm81]

We have an Osburn 1600 insert that came with the house when we bought it three years ago. The insert works great and really cuts our oil consumption, but I've always been annoyed by the non-thermostat fan. At night, we fill the insert with wood, but used to turn the fan to low speed in order to not have it blowing a lot cold air in the morning. This was a compromise that still had problems, not putting out as much heat as high speed when everything was hot, and still blowing cold air and wasting power by morning. This is my solution:


It's a custom designed and built controller that senses the temperature at the top of the insert and sets the fan speed accordingly. The brains is a PID temperature controller from ebay that came with a sensor. I drilled and tapped a small block of aluminum to thread the sensor into and installed it on top of the insert just in front of the flue exit. It is wedged in place with two small pieces of high temp silicone gasket that seems to be withstanding the heat so far. It was pretty fiddly getting in place in such a narrow space! I put some heatsink grease on the block for better conductivity.
The toggle switch has three positions, center is automatic mode, down is manual low speed, and up is manual high speed. When we start a fire, we start the sequence with the black push button. The fan stays off until about 85*C (trial and error to find useful heat on low speed) and at 125*C (more trial and error), it switches to high speed. As the fire dies down, the fan goes back to low speed at 125*C and then shuts off as temps drop below 85*C. At this point, the power is also removed from the controller, shutting everything off. This was important to me as I didn't want to have to remember to turn the controller off every morning.
So far it has worked very well. My wife thinks I should remove one of the old fireplace grills and mount the box in the opening for aesthetics, but I didn't want to make it a permanent installation. I also had to remove the cord from the fan unit and install a four wire cable to the control box.
I used a program called TINA to design and test the circuit before actually building it. Below is a screenshot of the schematic:
It took a lot of head scratching to figure out a circuit that worked with a minimum of parts with the simplest possible user controls. The T-stat, warm, and hot contacts are all combined in the PID controller. The manual and hi/lo switches are also combined in one toggle switch.

Being able to see an actual stove temp is kind of interesting; the hottest we have seen so far is just over 300*C. Hearing the fan drop down to low speed is also a good reminder to throw some more wood in.

 

[Jon1270]

Coincidentally I was just trying to puzzle out how to do this very sort of thing today. One potential issue I've been wondering about: how do you keep the fan from cycling between on and off, or between hi and low speed, when the temperature is near the set points? I would think the fan turning on or switching to a higher speed could push the temperature down below the set point. Are you seeing any of that behavior, or is the aluminum block maybe massive enough to slow the PID's responses?

 

[paulm81]

The only time I have noticed any cycling is if the temperature drops down to the setpoint and the new piece of wood takes a bit to get going. The top plate of the stove is fairly thick and I think this really dampens the temperature fluctuations. It would be different if the sensor was on the thin flue pipe. There is also a hysteresis setting in the controller, mine is 2*C. Basically, the controller goes from low to high at 125*C but doesn't drop back to low until 123*C. I can't remember if I could actually change this setting or not, maybe not as it's a very cheap controller.

 

[TheBaron]

Wow awesome engineering batman. That's great...

Our unit has a crude snapstat but it doesn't let us go from low to high then back to low as the load burns like your solution does. Sounds like an ideal solution unless we get into variable speed fans, DC etc.

 

[Jags]

Very nice job. I use a stove stat (have for years), but it does not allow for hi/low fan speeds. You get that thing figured out and install it in a nice oak or cherry box and you could very well have a marketable item.

When we start a fire, we start the sequence with the black push button

Would it be possible to automate (or have it auto - reset) this part?

 

[Ashful]

how do you keep the fan from cycling between on and off, or between hi and low speed, when the temperature is near the set points?

Increase the hysteresis setting. I like the way the heavy top plate gives a long time constant, pretty much eliminating this problem, even with the hysteresis set as low as 2C.

Sounds like an ideal solution unless we get into variable speed fans, DC etc.

PID controllers are actually made for variable control, and can give you near-perfect critically-damped response, if you have the know-how to set them up. Standard second order system, Control Theory 101 stuff, but it's always easier to figure the coefficients in a text book case than in real-world applications.

Great job, Paul!

 

[paulm81]

Very nice job. I use a stove stat (have for years), but it does not allow for hi/low fan speeds. You get that thing figured out and install it in a nice oak or cherry box and you could very well have a marketable item.



Would it be possible to automate (or have it auto - reset) this part?

Two problems with marketing such a thing:
1. Obtaining CSA or UL approval since it is an electrical item
2. Replacing the cord on the fan unit with a four-wire cable is not very DIY friendly

The entire process could be easily automated if I left the controller powered up all the time. This would probably be simpler than the way I have it now and since the controller only draws a few watts, the extra power consumption would be minimal. I'm stubborn though and wanted everything off in the morning!! lol

 

[paulm81]

Increase the hysteresis setting. I like the way the heavy top plate gives a long time constant, pretty much eliminating this problem, even with the hysteresis set as low as 2C.

Yup, it gives a slower response, but in this case it's irrelevant as the heat output depends on the temp of the heavy plate.

PID controllers are actually made for variable control, and can give you near-perfect critically-damped response, if you have the know-how to set them up. Standard second order system, Control Theory 101 stuff, but it's always easier to figure the coefficients in a text book case than in real-world applications.

Great job, Paul!

Even the cheap PID controllers advertise an auto-tune function, not sure how well they work. I just installed a three-phase motor and VFD on a metal lathe at work and the VFD's auto-tune feature worked great. In the stove's case I'm not sure it would be much of a benefit, as we are not trying to control the temperature, rather just responding to it. And even then, most of the time the fan would be running full speed anyways.

 

[Jags]

Two problems with marketing such a thing:
1. Obtaining CSA or UL approval since it is an electrical item
2. Replacing the cord on the fan unit with a four-wire cable is not very DIY friendly

The entire process could be easily automated if I left the controller powered up all the time. This would probably be simpler than the way I have it now and since the controller only draws a few watts, the extra power consumption would be minimal. I'm stubborn though and wanted everything off in the morning!! lol

Nothing wrong with doing it "your way". That is why you get to decide how it is built.
Again, very nice.
And welcome to the forum.

 

[Ashful]

Even the cheap PID controllers advertise an auto-tune function, not sure how well they work.

Cool! Back when I was doing PID stuff (late 1990's / early 2000's), the ones I was using typically had to be set up manually. Figuring out the PID coefficients was the toughest part of the job.

 

[TheBaron]

I really like this.

I guess you would want to connect directly to the fan motor and not just use the stoves power cord... (A standard PC cord) user would leave the fan on high on the stove and your device would go inline with power feed and turn down the power to slow the fan?

 

[Ashful]

I guess you would want to connect directly to the fan motor and not just use the stoves power cord... (A standard PC cord) user would leave the fan on high on the stove and your device would go inline with power feed and turn down the power to slow the fan?

Assuming these blowers are run by induction motors, you can't reduce voltage to slow the fan. You need to pulse width modulate, or vary the frequency. Most AC motors have a narrow range of frequency they can run without overheating.

If the blower is running from a universal motor (unlikely), then it can be made variable speed. The OP's fan is likely a dual-winding induction motor. Switch one winding out, or parallel/serial configuration, to get double or half speed. That's why he needs the dual-speed switch, and four wires to the motor.

 

[TheBaron]

Thanks Joful. I figured it wasn't that easy!

I suppose if you didn't need it to switch speeds then a simpler approach would work.

 

[paulm81]

. The OP's fan is likely a dual-winding induction motor. Switch one winding out, or parallel/serial configuration, to get double or half speed. That's why he needs the dual-speed switch, and four wires to the motor.

Exactly. I wouldn't be surprised if it's one winding with a tap part way from one end. In any case, it's important to set up the controller so that there's no chance of both windings being energized at the same time. This is why I used a relay to switch between one or the other, instead of a separate relay for each speed. There is also voltage coming back on the un-energized wire which can backfeed stuff if you're not careful!

 

[Swedishchef]

Super nice fan controller.
Well thought out!
I think you have something going for you...this could become an interesting item.

Where in Ontario do you live?

Welcome to the Forums1

Andrew

 

[kallsop]

The Osburn 2000 has a variable speed fan, set by a knob. I was considering building a blower motor controller including a room thermostat, to adjust the blower to maintain a constant room temperature, also with a stove temperature reading to prevent blowing cold air around. If it were a cat stove with much longer burn times, I'd have given the project higher priority. When you are reloading wood frequently, the controller is not so compelling, because "set it and forget it" isn't possible when you have to manually reload.

 

[paulm81]

Thanks everyone for the positive feedback and the warm welcome! I have no intention of commercializing this product; that takes all the fun out of building it in the first place if you have to consider costs, marketing, liability and all that crap! This product would be most useful if offered direct from manufacturers.

Andrew, I am in southwestern Ontario about 2 hrs west of Toronto, on the so-called "west coast of Ontario" because Lake Huron is only 15 min. away. True western Ontario is about 24 hrs of driving away!

Kallsop, do you know how speed is varied on your current setup? What kind of controller did you have in mind?

 

[Swedishchef]

Hey Paul,

I know all about Ontario. I drove from BC 6 years ago to where I live today. So you're really in southern Ontario! Chatam-Kent region?

 

[paulm81]

A little north of Chatham-Kent, I'm in southern Bruce county.

 

[Swedishchef]

Very nice area.

Welcome to the forums. Everything you will want answered about stoves will be answered by the PROS of this forum. I only wish I had found the forum sooner!!

 

[EvilRoySlade]

Have any of you researched higher efficiency blower units. With these inserts you really need the blower going to get good heat. I have not checked amp draw of my fan so I don't know the watts used but running all day for months can add up. It's in my future plans if my blower ever fails.

 

[Swedishchef]

EvilRoySlade: I don't think a blower fan takes much wattage. To my knowledge most fans are 40-75 watts. So if you want it 24 hours a day at 75 watts, for 120 days it would give you 216 Kw. Not very expensive to run considering it is an insert and not a radian/convection stove.

Andrew

 

[paulm81]

Have any of you researched higher efficiency blower units. With these inserts you really need the blower going to get good heat. I have not checked amp draw of my fan so I don't know the watts used but running all day for months can add up. It's in my future plans if my blower ever fails.

I just measured our fan with a Kill-a-watt meter: On low speed it draws 28 watts, high is 50. The controller adds 2 watts to that. Ours maybe runs an average of 12 hours a day so at 52 watts, that is 0.62 kWhr per day. Our incremental hydro rate here is about 15 or 16 cents per kWhr, so the fan costs us about 10 cents a day to run. Well worth it for the heat it puts out!

 

[EvilRoySlade]

That being said, I am going to modify my previous statement. When my fan dies I will be looking at quieter designs seeing as the energy use is insignificant relative to the rest of my house. I'm sure someone makes a fan that can move the same CFM with much less noise. Unrelated to the controller topic, sorry.

 

[HaTaX]

This is way cool to see, glad this thread got bumped!! I've done some digging around on the internet and have had a hell of a time finding someone that makes a digitally controllable rheostat I could use to get many different speeds and use them with a temp curve. Right now I've got one of those rheostats that the blower plugs into and that controls the speed of the fan quite nicely, but it's manual...

Initial thoughts on doing it on the cheap, a small arduino board to drive a RC servo and monitor a temp probe. The RC servo would hook to an adjustable rheostat and adjust the fan speed, kind of rigged but the cost on all the parts would be cheap. I'd have to program the arduino to do what I wanted, but as a bonus it could hook up to the internet and I could see the status of my stove wherever I was at. (Could even rig an email alert for danger temps or reloads, hmmmm)

Okay, this is getting pretty nerdy for the hearth forums so I'll stop here...