Pellet delivery and control system

Posted: Sun Oct 03, 2010 12:03 am

Here are photos of the pellet drive assembly that I had been planning as an alternate heat source for my cooker. I finally got around to getting all the parts I needed for an enjoyable day of R&D...

I used a 4"x4" stainless tube that I found at the local surplus store for the main body. On the very bottom is a cheap electric charcoal igniter. Inside the main tube, I dropped a hole for the burn cup, and inside the burn cup, I made notches for the igniter and some holes for the auger feed pipe and for some forced air. The auger is about 20" long and is made from the curly part of several cheap dog-tie augers. Auger motor is a surplus item as well.

Here is a shot of the burn cup showing where I placed the various holes and notches. The burn cup is a piece of 4" flue pipe from a old hot water heater. The arrows mark the air holes, there are (4) of them and they are 1/4" in diameter. I placed the holes this way in an attempt to float and circulate the ash up out of the burn cup without blowing hot pellets all over the place.

Here is the auger tube being filled with pellets...

All assembled and sorting out power cord. I use the extension cord for the igniter for about 5 minutes and after the pellets started smoking, I plugged in the fan.

We have ignition !!

I let it burn down a bit before manually turning on the auger motor for more pellets

fire with a nice cup of pellets

getting a nice swirly forge-like flame using a small muffin fan

And here is the entire setup after getting it going

There are still some minor mechanical details to workout (auger to motor setscrew and tweak the angle a bit) but I was able to make a nice pellet fire without ever touching a lighter or hot coals.

Need to work on the control loops for the fan and motor to make it cook. I have a general purpose PID controller that can modulate the combustion fan for temperature control. In the GF design, we have a stack of unburned fuel, and modulate the stoker fan as needed to get more heat.

However, when using pellets, we also need to control fuel delivery at a frequency that matches the demand for heat. There are at least five different fuel requirements that I could think of...
1) for startup,
2) variable meat loads,
3) maintaining temp
4) recovery after a door opening
5) drop to a holding temperature when its done

All things considered, I will probably try to use some kind of average pellet delivery speed (on/off timer) and fine tune the temp by modulating the fan.

Need to make sure the auger is not feeding too much so that the cup does not run over, and make sure it is not giving too little so that it takes too long to get up to temp.

I think a few afternoons of tinkering may give me some idea for the on/off auger timing needed to fill the cup and let it burn down again using full fan speed without blowing out the last burning pellets.

There are simple timers used in pellet stoves , but I may try to use a 555 timer and a couple of adjustable resistors because I already have those parts laying around.

My unit has a 3.75" dia and 3.75" tall burn pot
Using a 3-1/2" muffin fan for combustion blower... unknown CFM.
Auger motor is 25RPM (so on-time will have to be shorter)

Have not measured heat output yet or pellet consumption over time...

Ideas, suggestions specs on your system that works are welcome.

Posted: Tue Oct 05, 2010 10:12 pm

1) for startup,
Startup is a breeze , On my unit I just hit the firepot with a hand torch with the muffin fan running and away it goes .

2) variable meat loads,
I never find that a problem , Maybe a tweak or so with the on / off timer for the auger .
3) maintaining temp
Again , not a problem once you get on to the on off auger timings .
4) recovery after a door opening
5) drop to a holding temperature when its done
My holding is done in a Colemen Extreme cooler with the meat double wrapped . Holds for hours if needed . Trying to get the on / off to work with a low setpoint like 150 or so is a real pain .

I also tried tuning my PID controller .. That was another pain .. Its not like working with gas / steam / electricity etc . Much easier to get on to on and off auger times . ]

As for the muffin fan .. I tweak it with a house hold fan controller and once I hit a nice burn it dsnt change . These pellets need steady combustion air to burn properly . Trying to control cooker temps with the combustion air is a no / no ..

Just my 2 cents .. Looking forward to your progress .. Im always game to new ideas ..
Cool

Posted: Thu Oct 07, 2010 12:34 am

DarrelB,

I read the other day that my charcoal lighter should be removed from the heat or it may explode! Probably a scare tactic, but I have seen 2000F easy in a decent sized pellet coal bed, so I like your thinking on startup.

I may have to use one of those weed burner torches to start my cooker since I have a fully enclosed heat transfer chamber on the bottom of my cooker where the firepot will be. I can get to the firepot using the 2" observation port. I plan to use that port to vacuum out the ash after a cook if needed.

Ok, so I have had a few days to think about this and reflect on how the PID controller works on most stoker assisted GF units, so here is my line of thought on controlling a pellet system using a PID controller.

Auger motor:
For now, I have a fast auger motor (from what I can tell, most pellet stoves are 1 or 2 RPM, and I see a 5 RPM unit ever once in a while)
however, the one I have is a very torquey 25 RPM. That rules out using an off the shelf controller for a pellet grill unless I want to have an exact match for all of the other components (whats the fun in that?)

I have another motor to try from an ice maker out of a home refrigerator. I can check that one and see what the RPM is for reference. Regardless of the speed that I have from the auger motor, I am pretty sure that I will need to slow it down to keep from filling up the cup. I can do this by using a relay to switch the motor on for a few seconds every 2-5 minutes using a pulse from a 555 timer. I plan to power the 555 timer circuit using the 12v SSR output from the PID controller that traditionally drives the fan like most stoker units.

Plan A:
When the cooker needs heat, the PID controller turns the SSR output on and that voltage powers the combustion fan and auger timer circuit.
After a couple of minutes, the auger timer circuit will send a pulse to a relay that turns on the auger motor for a couple of seconds or so - long enough to drive one turn of the auger and deliver some pellets.
The combustion fan will be running and the auger timer will be feeding a turn of pellets every couple of minutes as long at the PID controller is holding the SSR output on and calling for heat (PT<ST).

When the unit gets to temperature, (PT>=ST) the PID controller starts pulsing the SSR output, leaving it off a little more each time. When the SSR output is on for less than 2 minutes, there is no auger feed pulse because of thereduced need for fuel. In theory, this would be the smoking cycle, because the pellets would start generating some with less combustion air...

Eventually, the process temperature will go below the set temp again (PT<ST) and there will be another call for heat by increasing the on time of the SSR output pulses. When the SSR output is held on for more than 2 minutes, there is a corresponding pulse from the auger timer circuit to feed more fuel and this process is repeated for 12-18 hrs while I do something else.

Plan B:
Keep the combustion fan powered on all the time, and have the auger timer circuit powered by the SSR out of the PID controller to keep from overfilling the pot.

Balancing act:
The trick here seems to be matching the feed rate to the burn rate for the pellets. There needs to be enough coals in the pot to continue to burn long enough so that when the PID calls for heat and holds the SSR output on for more than two minutes that the fan will not blow the fire out.

Along the same line, there needs to be enough combustion air to prevent any bad smoke since the pellets seem to smoke a lot more than smoldering lump charcoal. (although there is a lot less pellets in the pot than lump on the grate when compared to the classic GF design)

Other ideas:
I have seen pellet stove schematics using low temp and overtemp circuits for additional security that the fire does not get too hot or go out - are you aware of these circuits on pellet grills or the FEC series of cookers ?

I am considering alternatives to the AC auger motor, but have not seen any DC gear reduction motors, with the exception of automotive windshield wiper motors or a power window motor. I would like to use one flavor of voltage for all of the control circuits, and 12v DC is perfect because I want mobility.

More data needed:
I plan to do some measurements for the following items:
how many seconds to fill the burn pot with auger drive at 100%
how much does a pot of pellets weigh
how long does t take to burn 1/2 a pot of pellets

With this data, I should be able to tell how many pounds of fuel per hour I can burn using my particular fan speed and burn pot diameter, to see how close I am to the FEC specs (or yours) ... I have a few fans to try aas well - AC and DC ones. Once I have a handle on the burn rate and pellet delivery rate, I can build the auger timer circuit.

Thanks for the help

Posted: Thu Oct 07, 2010 10:28 am

What I found when trying to tune the PID controller to do the feeding is big time over shooting . If you dont start pulsing way before the setpoint is reached you will over feed the pot . That gets the inside of the cooker so warm that by the time it cools off below setpoint and starts calling for pellets the fire would be almost or at times out .
I tried tuning a few times and then moved away from that to strickly time based feed . It would be different if you were able to get the same amount of pellets each time the auger ran , but thats not possible .

Here's my setup - PID controller is stricky used as a relay output to a PLC . When the temp drops below setpoint the relay turns on a bit in the PLC which starts the timing logic ( all in the PLC ) These times are easy to set / tune once you get burning .

I do have a safety timer in the logic . If there is no call for pellets for a longer period due to temp being above setpoint , It will give a small shot of pellets , just enough to keep a fire going .

Posted: Sat Oct 09, 2010 9:27 pm

I agree that time based delivery would give the best chances for success, and the controller has to drive a short output pulse and then wait a few minutes to see if another one is needed. That almost demands that the fan and the auger work independently.

Also, the system as a whole will have limitations or difficulties in ramping or holding temp if the system components are not matched. For instance, my 25 RPM auger motor may be OK for a gassification furnace that burns 10 lb/hr, but way too fast for smoking at 225F. This speed probably exceeds the control parameters of my controller (meaning "it'll never work")

Reading the Traeger docs, I found they use a 15 second "on" time for their auger and vary the off time. This was seen in the directions for fine tuning the "smoke" setting on their grills. That tells me they are probably using a 4 RPM auger motor, and the 15 seconds of "on" time would be enough to get one rotation of the auger so there are some pellets dropping. They vary the off time from 55 seconds to like 80 seconds or so, and that varies the smoke temp from 160-180F. Lower than 160 and the fire may go out and the re-light process is a drag.

I would like to know more about the FEC-100 and the speeds and feeds on that unit. They seem have the nicest features on their IQ-4 controller.

Meanwhile, I am off my business trip and able to make some measurements with what I have today, but I have already started shopping for a more robust PID controller. I have also found some 2RPM and 4RPM gear reduction DC motors. If I shop wisely, I may be able to have my unit totally DC powered for portability.

Will look into a PLC... what is that ? a time delay relay or something ? And you run your combustion fan continuously as well ?

Thanks for sharing what you know. Ever smoke a Turkey ?

Posted: Sat Oct 09, 2010 10:41 pm

Quote:

Ever smoke a Turkey ?

Just did a 16# bird today on the FEC100. Cooked at around 260-275 for about 4-4.5hrs. I go more by temp than time. They turn out great and all I do is spray with butter flavored Pam at the beginning and again about 3 hrs in.

Posted: Sun Oct 10, 2010 10:43 am

The fan runs continuously I think you'll see that with all pellet munchers .

The PLC is a software programmable device . Mine is a small unit that is just discrete ( on / off ) etc Some other units can handle analog devices also . That would eliminate the need of the PID controller .

Regardless , its not really needed if you have a couple timers your good to go .

Posted: Sun Oct 10, 2010 7:32 pm

Jim,
I have cooked some good turkeys in my water tank gasser, and use the 275-300 range with good results too. I use cajun seasoning and about a dozen chunks of mesquite. Really hard to beat a nice smoked turkey.

Do you use flavored pellets, like apple or peach ?

Darrel,
I was looking at the intermittent wiper timers in my vehicles today, and they can give power every 15-30S (depends on the manufacturer) and it seems a windshield wiper motor would be strong enough, but still at too high RPM.

I have reached my theoretical saturation point because I discovered the term "Ardunio" and found that whole plug and play development environment with electronic building block modules and source code that they use to make robots. I think I have read a few posts about using the Ardunio parts to make a controller and data logger. (similar to the stoker controller) Seems like something I would like to explore eventually.

Thanks again for the tips, I will do some more hands on testing this week with the parts I have to see if I can get a closed loop sustained burn using a drum as a test vessel.

Pellet fired UDS pics will be up soon.

Posted: Sun Oct 10, 2010 7:37 pm

Quote:

Jim,
I have cooked some good turkeys in my water tank gasser, and use the 275-300 range with good results too. I use cajun seasoning and about a dozen chunks of mesquite. Really hard to beat a nice smoked turkey.

Do you use flavored pellets, like apple or peach ?

I use mainly oak pellets, but this time I used pecan. I don't mix like some people do. I use up a whole bag and then just add another of whatever I want.

Posted: Sun Oct 10, 2010 8:03 pm

Just found this write up on a programmable stoker project for a green egg. Would work for any GF design, but no provision for adding pellets.
DIY pellet systems are a road less traveled for sure...
http://www.atmel.com/dyn/resources/prod_documents/controller_3_04.pdf

Posted: Wed Oct 13, 2010 1:17 pm

If you can do as suggested by DarrelB and just use a couple of timers that would be the easiest. Second easiest is to buy the temperature controller and build it yourself it can control an SSR relay wich could turn on the auger when heat was needed, but your going to need a timer to shut the auger off, or it will just sit there and run till it reaches the set temp. Google Auber instruments. The third more difficult method would be to actually program the PID controller, as I believe Darrell did. It takes some ladder logic programming as I recall. Good luck. Also, if you search this site, somewhere I build the Auber to be a down and dirty stoker, just controlling a fan, but with no connectivity or reporting.

Posted: Wed Oct 13, 2010 10:51 pm

Agreed ! The best way to git -er done is to go as far as I can with the PID controller and other parts that I have for now.

I plan to do manual control of the auger motor for a while so I can get a handle on the timing needed to drive to a temperature and determine if the combustion fan I am using is sufficient.

I am happy with the way the auger assembly is turning out. I got rid of the 2" Tee fitting and swapped out the 25RPM auger motor for an easier to obtain ice maker motor. This one is 18 RPM and designed to operate intermittently - a few seconds at a time is all I need to get one turn of the auger for pellets.

I am using a drum as a test chamber because it is easy to see what is going on with this setup. Once I get a working combination, I will put it on the Fridge cabinet and build a hopper, etc..

I am planning to build a 555 timer circuit that will start when my existing PID controller calls for heat. The timer circuit will be adjustable to give a few seconds of auger drive time and then wait for a minute or two (or three??) for the pellets to burn into charcoal before repeating another few seconds of auger.

This timer cycle will repeat until the PID controller that is monitoring the temperature decides to cut power to the timer circuit. The combustion fan will be running continuously, so the temp will eventually drop to the point where the PID controller will call for more heat and re-energize the timer circuit to feed more pellets.

If this works as expected, it will give me a relatively simple closed loop control for temperature using a PID to drive the home made auger timer instead of modulating the fan as in a stoker setup.

Posted: Mon Nov 01, 2010 11:30 am

OK, well a couple of weeks ago I did something that caused my pellet auger drive to clog up and it actually sheared the shaft of the ice maker motor I was using. I had cross-drilled a hole through the shaft to keep the auger retention nut from backing off the shaft since I had to reverse the rotation of that motor.

I found that mounting the motor to the auger shaft was problematic for the two traditional auger motors that I had tried up to this point. Another problem that I was facing was with the speed of these two motors. They were 25 RPM and 18 RPM, requiring very short "on cycles" and presenting a small window of error for process control.

I did some shopping around and found this 2 RPM DC motor that has plenty of torque. The motor is small and circular with an offset output shaft that has a nice flat spot on it. Here are some photos of how the new auger motor is coming together.

I cut the dog-tie auger material on a chop saw, and got a 1-1/2 inch washer and welded it on as centered as possible. Then on the back of the washer, I welded a small piece of 1/2" angle that I notched and drilled so that the coupler would connect to it. The a slot in the coupler has some "wiggle room" to compensate in case anything is off center. It works kind of like a universal joint.

I had to make a coupler out of some stock I found at the local surplus store... I drilled it to fit the shaft, then cut threads for a set screw. Next, I rounded the nose a bit and drilled and tapped a hole for a bolt that will hold the auger captive to the shaft.

The next shot shows how I put the motor on the new 1-1/2" feeder pipe. ( I was using a 2" pipe mainly because I have a 2-3/8" hole saw, but the 1-1/2" pipe is the correct size for the auger so there will be no chance the auger could ride up over the pellets and cause problems.

I cut a couple of inches of the feeder tube away so the auger can pull in the fuel, and added a couple of sloped bottom pieces for the pellet bin.

To mount the motor, I just welded a length of 1" angle to the bottom of the feeder pipe so the motor lays in the cradle and can be be held in place with a couple of hose clamps. I used a large washer to close off the auger pipe end, but the washer on the end of the auger does a pretty good job keeping the pellets away form the motor. The extra washer is just extra.

Looks like the bin will be something like 9"x12" or so. Not sure how tall I will make it yet, and I may make it wider to match the width of my exhaust cover that I have on my Fridge cabinet.

Posted: Fri Nov 12, 2010 9:31 pm

I spent the past couple of weeks looking for a way to regulate the speed of my auger motor so I can have closed loop PID temperature control that would also drive the pellet auger.

The basic problem is that unless you have a REALLY slow auger motor, the PID controller would drive the auger too much and cause the pellets to overfill the burn cup. There may be a way to program the PID parameters or auto-tune it, but my research shows that could take a really long time, and there may be limitations with the parameters that my particular PID controller offers... so I wanted something easier to understand and adjust.

I think I have found a great solution. I located a repeat cycle timer (turns a device on for a short period of time and turns it back off for a longer period of time, then repeats the cycle automatically). The Futurekit FK432 hat two knobs to adjust "on" and "off" times, here is a link to their site:

http://www.futurekit.com/2009/index.php?option=com_content&task=view&id=226&Itemid=174

My system runs on 12v DC power provided by a "wall wart" adapter for now. The combustion fan is on a manual switch so I can turn it on or off for lighting or making smoke. The red wires carry 12v DC.
The combustion fan uses yellow wire to indicate power is switched.

I took my JLD7100 PID temperature controller and set it for RELAY output. (no need for a solid state relay SSR).

I use the RELAY output of the PID to control the 12v power to the FK432 cycle timer circuit using yellow wire,

The output of the FK432 timer provides 12v power to drive the 2RPM auger motor using orange wire.

The timer is set for 20 seconds of "on time" and that delivers 1 revolution of the auger. I have the off time set for 2 minutes just to show the on/off times are independent.

In the center of the photo is a quarter for size comparison. The components are not really that big. Starting above George Washington's head, we have the FK432 timer, and going clockwise, a 2RPM 12v DC auger motor, a 12v DC fan, the JLD7100 PID temperature controller, a K type thermocouple and the 12v DC power adapter.

System operation:
Here's how the system works when it is first powered on.
The PID measures the temp, compares it to the set value of 225 and calls for heat. RELAY out from the PID energizes the timer circuit, which picks another relay to turn on the auger motor for 3 seconds and the PID abruptly removes power as it evaluates the change in process temperature.

Since there is little (or no effect) the PID calls for heat again, driving the output longer. This causes the cycle timer circuit to remain powered on as long as the PID is calling for heat, so the cycle time starts counting the seconds again to see if it can reach 20 seconds and turn off the auger motor.

If the PID cycles are less than the 20 seconds that the timer is set for, the motor will run unitl the PID removes power. After about 6-8 PID cycles (about a minute or so) the PID figures out that there is a long way to go with the temperature, and the short "on" times are not getting there, so it starts driving the output for a long time. The cycle timer circuit remains powered on, and counts up to 20 seconds then shuts the auger motor off. It waits for 2 minutes and then picks the relay to turn the auger motor on again for 20 seconds.... and this cycle repeats as long as there is power to the circuit.

This is as far as I have tested, but it works as expected : )

When the PID senses that the temperature is near the set value, it will drop the output and the cycle timer circuit will turn off.

The pellets in the pot would be burning down now, and the PID is monitoring and waiting for the temp to drop enough so it can call for heat again....

And this cycle of cycles repeats for 12-16 hrs - until we have good BBQ.

So, now I have some more fab work to do on my pellet hopper, and soon I will be firing up the pellet powered burn barrel again with closed loop controls this time !

Edit:
Added another picture showing the system powered on. I was measuring the "off" time for the motor and snapped the photo just as the cycle timer circuit kicked on to drive the motor again:

Posted: Fri Nov 12, 2010 9:57 pm

Great work ,, Looking forward to seeing the first fire .. Cool