I cannot wait to get my hands on this thing.

I made a post about the SMOBOT Kickstarter campaign when it started back at the beginning March. I’ve been obsessively tracking its progress. There are now five days to go and it’s hard to say whether or not they’re going to get funded or not. They need about 50 more people to commit to buying a unit. I have no illusions that a lot of people read this blog, but I’m going to write this in the hope that it might help someone who’s on the fence make a decision. I like the guys behind this. I’ve been using  a pre-production Beta unit for over a year now and I love it.

If you’re not sure what a SMOBOT is, go read the last blog post and watch the video. The tl;dr version is it’s a robotic damper control unit that attaches to the top of a Kamado-style smoker like the Big Green Egg, Kamado Joe, or Primo. A thermocouple tracks the temp inside the cooker and the damper opens and closes to maintain the temperature at the set point. The damper looks like this when it’s operating.

This is a time-lapse of the damper working. In real-time you rarely see the damper visibly move. The individual adjustments are very small. That’s why the whole unit can be powered by batteries.

There are also two food probes that allow you to track the temperature of whatever you’re cooking. And since we live in the future, all the data is available on your local Wi-Fi network and even through a free cloud service. When you’re a nerd like me, that’s pretty damned cool.

What the SMOBOT doesn’t have is a fan or blower. There are other temperature control units out there, but they require you attach a blower unit to the lower air inlet of the cooker. That’s where the airflow control comes from. Clearly the things work because lots of people swear by them, but they never appealed to me. The whole idea of something like a Kamado is that it’s built to draw in air. The things work essentially like a convection oven. Getting air into the things is not the issue.  The issue is how much and when. That’s what the SMOBOT takes care of. It constantly watches the temperature and fiddles with the damper when it needs to, and leaves it alone otherwise.

The data from my last cook. Winds were 18mph gusting to 30mph. They were blowing directly into the damper-opening side of my Big Green Egg. It was like sticking a hair dryer in the lower damper and randomly turning it off and on. To say these were challenging conditions is an understatement. I never touched a thing from the time I put on the food to the time I took it off. I was actually cutting the Kickstarter video the whole time.

I have no idea how much charcoal Eric Reinhart and Curtis Pope have burned developing the algorithm that runs the thing. There’s probably an sixteenth of a degree of global warming with their name on it. Small coal plants in China look at them and say “See?  Why are we getting all the blame? Hey, is that honey-glazed pork?” Whatever rainforests have been depleted, the algorithm is solid. The published control range is +/- 5°F, but I’m convinced the thing takes a 2°F error kind of personally. You can see from that graph up there that the unit never stops looking for that equilibrium point. It’s sort of Terminator-like that way.  Except for the whole traveling-back-to-the-past-to-kill-your-mother thing. I can saw with absolute assurance mine has never done that. Mark that worry off your list.

I really like the design of the production units. I’ve used my trusty Beta unit for a long time, but when I get my new one it’s going to be displayed on a shelf. From that picture I can’t be certain, but I think these units are quite a bit smaller than the Betas were. For one thing they aren’t trying to jam two 9v batteries in there. It’s powered via a USB port now. I’m assuming that’s a micro-USB cable plugged into the side in that picture, which means that whole thing isn’t very big at all. They’ve decided to ship a 4000mAh external battery pack with the Kickstarter package and the estimate is that it should last 48 hours in continuous operation. Worried about weather?  You should be able to toss the controller and battery in a zip-top bag. It doesn’t matter if the damper gets wet. Mine sits out in the weather continuously and laughs at rain. At least that’s what it tells me it’s laughing at. I think it might actually be me it’s laughing at. I’m very insecure that way.

Something that people are really paranoid about is the damper getting gunked up with grease and grit. It can happen, but it doesn’t really cause a problem. If the damper gets bound up with dried grease, you can easily free up the mechanism by lifting up on the arm that moves the damper just a little. It’s not like it’s welded or anything. The bond is pretty weak.  I light my Egg with a torch and I just swipe the flame over the metal part a couple of times.  That melts the grease and then it’s lubrication, not a binder.

They’ve added an indicator light to confirm the unit is actually attached to your Wi-Fi network, which is nice. Once you get it set up on Wi-Fi the first time there’s not much reason to mess with it, but it’s nice to have the visual feedback without having to reach for a phone to check each time. I’ve used the Beta unit at our local Eggfest using a LTE hotspot and it works great.  Cooking at one of those things can be hectic (and loads of fun), and it’s nice when you can resist the urge to fiddle with the dampers, especially since the lid gets opened and closed so much more often than with a normal cook.

The best new feature after the overall size and USB power is that “auto” button.  You’ll usually want to turn on the unit when you’re first lighting the fire. The lid’s up and at that point, though, and you really don’t need the damper working so it powers up in “manual” mode. On the Beta units you had to scroll through a menu to kick it into “auto” mode so the controller would take over.  It was remarkably easy to forget to do.  Or so I hear. Just like I heard a rumor it’s easy to forget to attach the thermocouple to the grate. Don’t think there’s anything they’ll be able to add to fix that.

The coolest thing that’s happened during the Kickstarter is that they’ve been able to add the Char-Griller Akorn Kamado grills to the lineup of cookers the SMOBOT can work with. For whatever reason, most of the ceramic Kamados have top vents that are similar enough that mounting the SMOBOT is just a matter of using one screw. The Akorns — which are made of steel and are typically much less expensive than the ceramic types — use a different type of top vent mount. Apparently a guy with both an Akorn and a 3D printer really wanted a SMOBOT.  After a little back and forth with Eric and Curtis, he came up with a prototype mount that works, so there’s an Akorn version of SMOBOT now. I’m pretty sure the same thing is going to happen with other models of cookers, but probably not in the next five days.

I know I know a lot of people who own some kind of Kamado cooker. Let this be my gentle reminder to you that you need one of these things. It works. It’s easy to use. It lets you sleep when you when you want to cook over night. It’s also just a cool gadget.

[I managed to auto post an incomplete version earlier.  If the version you read sort of ended weirdly, that’s why.]

I’m calling the birthday brisket success. Overall I’d give it a B or B+. There’s room for improvement, but lots of things went really well. Given all the things that I was doing for the very first time, I’m very happy with how it went.

For those of you not scoring at home, the brisket in question was a whole 13-lb packer that I bought at Bill Finke & Sons in Ft. Wright, KY. I’ve always liked going there. They’ve been at it for a long time and they know what they’re doing. It’s a place I can score tri-tip, for example. That’s not easy on this side of the river. They’re old-school butchers in the best sense of the term. I also made the happy discovery that they carry Humphrey Lump Charcoal.  I’ve wanted to try it for a long time. Curt McAdams recommended it to me a while back and I trust his judgement on stuff like this. He’s a pro.  Anyway, the closest to me I knew I could get it was Dayton, OH and it never worked out to pick any up when I was up there for work. Finke’s rearranged the store a little since the last time I was in and I came across it as I was checking things out. Now I have even more of a reason to love the place. Anyway, here’s what the brisket looked like after coming out of the cryovac.

The flat starts on the left on top and runs to the big chunk of fat on the right. The point is the underside below that layer of fat and is what makes up the whole right side you can see.

I didn’t measure the length of the thing, but I believe that cutting board is 19-inches long and the grid on my Big Green Egg is 18.25-inches in diameter. I explained in the preview post about how this is actually two muscles connected by a big strip of fat. I separated the two muscles (typically called the “flat” and the “point”) and came up with two 6-ish lb. pieces of brisket. I didn’t weigh them because their final dressed weight really doesn’t matter that much once things get underway. The important thing for this story is understanding that the very, very, very rough rule of thumb for a brisket at 220F is an hour-and-a-half per pound.  There are lots of variables, but that’s a good starting point. Just don’t plan a bank heist with a brisket being finished at a certain time as a key part of the plan. So instead of estimating a 19-20 hour cook, dividing the flat and point reduced it to a 9 hour estimate. I knew I had no intention of running at 220F, I’d already decided I was going to go 190F through the night and kicking it up in the morning sometime, so I was figuring on a 14-hour or so cook. Here’s what the separated flat and point looked like.

Flat on the left. Point on the right. If you flipped the flat onto the point like you were closing a book, you’d get what see in the first picture.  The thin end of the flat lines up with the thick end of the point. A good big of fat has been trimmed out.

I made up a rub comprised of 2 parts salt to 1 part each of paprika, cumin, chili powder, onion powder and a half-part tumeric and some black pepper . No recipe.  I literally just grabbed what looked good out of the spice cabinet. The salt is the main thing. Lots of very good briskets have been made using nothing but salt and pepper as a rub. I put the well-rubbed chunks in a foil pan, covered it in more foil and parked it in the fridge overnight and through most of the next day.

Here’s the load of charcoal I used.

The wood chunks are pecan, my preferred wood for smoking brisket. Oak is also very, very good but I grew up in a house with a pecan tree in the backyard. It’s what I like.  There’s one little stick you can see in the center sticking up. That’s a 4-inch long “fatwood” fire starter stick shoved through the charcoal pile.  I hit that area with a propane torch for about 45 seconds and there’s a fire. The Humphrey was amazing. I honestly think I could have started the whole fire without the resinous wood starter.

Now we go to data. I won’t deny that Carla and I were having some beer as I was getting this all ready to go. Was that why I forgot to attach the SmoBoT temperature probe to the grid? I don’t know. Causality is such a difficult concept. Mistakes were made. Isn’t it better that we all look forward rather than backwards? We don’t want Al-Qaeda and ISIS to win, do we? Anyway, after a while it really started bugging me that smoke was pouring out the top of the smoker but the grid temperature was stuck at 60F. That’s pretty much what the outside temperature was. That’s weird. It’s almost as if … Oops. When I put on the probe it was quite happy to inform me the Egg was at 320F.  Just slightly higher than the 190F I’d planned on.

Long story short, that graph you see is the cooling rate of the ceramic on a large Big Green Egg.  The two dips were me opening the lid to dump heat. You can also see it was completely ineffective because the heat just came right back up and continued to dissipate at its own rate. It’s as if the thing is working exactly as designed! Sometime between the second lid opening and the time the temp got down to 190F the fire was snuffed.  I put the meat on (that’s why the food probes jiggle) but you can see the temp just kept falling. I thought it would be harder to relight the fire in an Egg than it was. Let’s hear it for propane torches. I relit about 9:47 PM and by just after 10 I was rocking and rolling. And I went to bed. Here’s the graph of the rest of the cook. The flat is Food1 Temp. The point is Food2 Temp.

You can see I was well into the stall by 3AM. I was asleep. Around 4:30AM the fire got wonky. I think it was the fact I really didn’t mound the lump and it took a while for more to collapse back in. SmoBoT handled it while I was asleep. Check out the food temps right around the time all that’s going on. The temperature on the flat is actually falling. I was asleep  Have I mentioned in the last 5 minutes how much I love my SmoBoT?

When I got up I decided to just leave it at 190F. No big strategy. I was just enjoying this view and wanted it to last as long as possible.

If I wanted to be super-critical of myself I probably should have kicked the temp up to 220F around 9 or 10AM. Then I could have gotten the bark going earlier and I would have broken the stall a little more gracefully. Not that I mind the stall. I’ve embraced it. I’m convinced it’s where the magic happens. But it went on a little too long for the flat. It didn’t dry out. But it was starting.

But there are no regrets. I was wanting a day to unwind and just stare into space and I got that. The point was at least an A-  The flat worked well on the awesome Kimmelweck rolls I baked on Monday. So I’m good with a B on the flat. Still better than a lot of brisket I get around here.

When I pulled the flat at 187-ish I wrapped it in foil and stuck it in the only appropriate foam cooler there is.

Before the wrap.

H-E-B is a Texas grocery store.

Resting comfortably.

You can see on the graph I did kick the temperature up for the last couple hours. The fact that the flat rose so much faster than the point should have told me the moisture content of the flat had dropped and I ought to have gotten it off an hour or so before I did. Live and learn.

I’m feeling like this is more about tweaks than anything else.  This was a success by any measure. And a good birthday.

I’ve been planning to cook this brisket for at least a week or so. I’ve thought about it a lot. I prepped the meat last night. I even wrote a post about it. I’ve got this down, right? The plan was to light the fire at 7pm and let it stabilize and put the meat on at 8 or so. A friend I trust had been recommending Humphrey’s Lump Charcoal but I’d never been able to find it around here. Turns out the place I bought the brisket carries it. Proof positive it’s a good place.

So 7pm arrives and I load up the Egg for an overnight cook. I light the fuel using my usual propane torch method and the charcoal lights right up. I mean amazingly so. When you have good lump there’s a sound it makes when it first lights that sounds like glass cracking. I got that sound almost immediately. I was so blown away that I completely forgot to attach the SmoBoT temperature probe to the grate. In a mere 20 minutes I was able to get the pit temperature to 325F. Unfortunately my target temperature for overnight was 190F. I kept looking at the readout and wondering why the it kept saying the pit temperature was 60F when there was smoke pouring out the top. Well, idiot, that’s because you’re measuring the outside air temperature, not the grid temp.

It’s now nearly 9pm and I’m still 44 degrees over my target, but it’s falling. I should have the meat on by 10pm.

There’s a joke that there’s no such thing as a fool-proof plan because fools are so ingenious. Look that up and you’ll find my picture right along with it.

I couldn’t wait. The weather was gorgeous today and I was just itching to light the Egg back up again.  Spring is upon us and I’m really looking forward to doing some long cooks. The trouble with testing stuff like this is you wind up cooking a lot of barbecue. That’s not a bad thing in theory, but it can get expensive. I’m also trying (at least a bit) to watch what I’m eating. Ribs or whatever for a second day just wasn’t a good idea. I adapted an idea from Dr. Greg Blonder, the guy who wrote about the stall that I obsess about so often. In his experiments he used sponges soaked in water and tied with twine to simulate meat cooking.  I rolled up a couple of towels, tied them tightly, then soaked them through. You can see in the picture below how the temperature probes are arranged.  The thermistor-type food probes are in either end of the “meat” while you can see the thermocouple grill temperature probe clipped to the grate at about 12 o’clock.

Click to enlarge.

It looks like my hypothesis about the effect of the open charcoal bowl grate was right. Apparently I was letting in more air than SmoBot could account for. To recap, yesterday I had a bad problem with overshooting the target temperature. Once the temperature was exceeded, it was really difficult to bring the temperature down.  Difficult?  Impossible most of the time. The only way I could get the temperature back down was by shutting the lower damper. As I said yesterday, the country ribs I was cooking weren’t hurt a bit by the temperature variations, but this sure didn’t inspire confidence for letting a fire go all night.

This is the graph of the cook yesterday using the open grate. Click to enlarge.

Below is the graph of my experiment today.  I took out the open grate from the bottom of the charcoal bowl and put the stock grate back in. I didn’t add any new charcoal because I knew I wasn’t going to be leaving the fire going that long. This is what a cook graph is supposed to look like.

This is more like it!  Click to enlarge.

You can see where in the initial heating there came a point where SmoBot was able to gradually bring the temperature right to the target. I made it a point to leave the lid open for several minutes, long enough to let the grill temperature drop below 200. The overshoot that followed was just about the same amount as yesterday, but it was brought under control in just a few minutes. Then it stayed planted at the set temperature of 250. I opened the lid again to simulate … well … opening the lid because that’s what I did yesterday to sauce the ribs. I got some overshoot that was followed by a little undershoot. I haven’t looked at how much fuel remains in the bowl, but I suspect it’s getting down a bit. I started it with only about a quarter of what I’d normally use to cook on. The thing to remember is that there are going to be variations in temperature all the time. Fuel burns out and other lights off. This isn’t an electric element in your oven. The fire is going to change. As long as SmoBot can control the airflow completely, the algorithm is solid.

This graph is also a very good illustration of the stall. Probe 1 was inserted in the towels a little closer to the outside than Probe 2. If you enlarge the image you can see that the temperatures diverge from around 2:15 PM until just before 3 PM. Probe 1 is heating faster than Probe 2 until that point. Then Probe 1’s reading start flattening out because the effects of evaporative cooling have started affecting the probe closer to the surface. A few minutes later Probe 2 starts experiencing the effect and the readings converge again. Both temperatures continue to rise, but slowly. They only would have begun rising again when most of the water had evaporated out. You can see the steam coming off the towel right along the top edge in this picture I took as I was shutting things down. That’s not smoke.  Nothing was burning, though the string was starting to get brown. 

One last note on the open grate. It let a lot of crap fall down into the bottom. Lots of stuff that wasn’t completely burned. I’d cleaned the ash out prior to cooking yesterday, but when I cleaned it out today you’d never have known it. Usually I pull out only fine-to-medium grain ash.  This was great big chunks of stuff that didn’t burn.

I think the only time I’ll use the open grate is when I really want to get the Egg hot, like for pizzas or steak. Otherwise I’m sticking to the stock grate.

Live and learn.

I’m wondering if I’ve out-thought myself.  It wouldn’t be the first time. I’ve certainly done it before. You might even call it my specialty.

You know I’ve been beta testing the SmoBot robotic damper controller. I’ve written more about it here and here. For all of you TL;DR folks, it attaches to the air outlet of most popular kamado-style smokers (like the Big Green Egg). A temperature probe monitors the internal temperature of the cooker and opens and closes the top damper to maintain a consistent internal temperature. I’ve used a couple of different iterations of the hardware and software since November and I can say with absolute confidence that the thing works.

The SmoBot top damper in place.

One key difference between this controller and others is that there’s no blower or anything else that forces airflow. It’s a strictly passive system. The basic idea is that you leave the bottom damper in a set position and then don’t move it. The top damper does all the work. That leads immediately to a question:  how much should the bottom be opened? It’s not a trivial question. I’d argue that the fundamental skill you have to be able to master on an Egg (or any other good-quality cooker) is airflow control.  A self-sustaining fire requires air, fuel and heat. More air supports more fuel combustion, less air supports less. Combustion is where the heat comes from Basic stuff. Go on any of the Egg online forums and you’ll see thread after thread discussing the finer points of damper control. The fact people have such trouble with it is a reason a device like the SmoBoT is such a great idea.

When Eric, the guy who developed the device, first demoed an early version he ran his bottom damper wide open. As the device went through several iterations in hardware and software he changed his practice to running the bottom damper half-open. The idea was that there was still plenty of air available when needed, but wind would be much less likely to affect airflow. That’s what I did when I first got a unit and it worked fine Then I made a change to the Egg. It’s been a bit of a struggle since.

In a kamado-style smoker the lump charcoal is placed in a bowl that sits at the bottom of the unit. The bottom of the bowl isn’t closed. In the case of the Big Green Egg, there’s a grate that looks like this:

You can see the holes aren’t very big. The disc is 9-inches across  and the holes are maybe 3/4 of an inch across each. When you let a fire go for a long time and ash drops to the bottom of the bowl, those holes can get plugged up. Not good for airflow. I replaced my stock grate with this:

No way that’s ever going to get plugged up. But I think it may be too efficient. Here’s a graph of tonight’s cook:

I’m cooking country ribs here.  They don’t take long and they’re incredibly forgiving. Carla had a late night at work and I wanted to have them waiting for her when she got home. I’d normally do these at 225F, but I got a little bit of a late start so I went ahead and set my temp to 250F. That’s where the forgiving part comes in. Anyway, I let the Egg come up to temp to make sure the smoke was clean. I had the bottom damper open about 1/2, a little under if anything. I get 5F of overshoot on the initial climb which is within spec. Anything +/-5F is assumed to be on target. We’re talking about charcoal fires here. 5F is close enough. If you squint you can see that just before I open the lid to put on the meat (the first dip) the temperature had actually settled back to 250 on the nose. Then it got crazy.

When I opened the lid to put on the meat, the cooking chamber got flooded with air and more fuel lit off. That’s to be expected. What should have happened, though, is that once the lid was closed SmoBoT could do its thing and reduce the airflow enough to starve the fire back down. But that didn’t happen. The temperature overshot and stayed overshot regardless of what SmoBoT did with the top damper. It ran a good 10F hotter than it should and showed no trend downward.

I tried reducing the set point to 240F and opened the lid briefly to drop the temperature. I also reduced the bottom damper to about 1/8 open, though the top damper was still partly open. That brought the temperature down.  I  sauced the ribs (3rd dip) and felt guilty about shutting the bottom damper so much.  I returned it to it’s original position. And, as you can see, the temperature spiked like crazy. I quit feeling guilty and shut the lower damper down to this:

That’s the same 1/8 it was open before, but this time the top damper was completely closed when I shut it down. It undershot, but was starting to settle in when it was time to take off the meat so we could, you know, eat it. (Delicious, by the way). No country ribs were harmed in the making of this blog post.

My hypothesis is that the new grate is a mistake for low and slow. Yes it can get plugged with ash, but it takes a long time to happen. I’ ve never managed to do it yet. All that’s needed to clear it is a wire coat hanger. Since Joan Crawford isn’t using my Egg I have no problems doing that. People criticize the stock grate, but I’m beginning to think more thought was put into the design than I initially thought. Something like SmoBoT can only work if closing the top vent will kill the fire. With enough repetitions I’m sure I could come up with a sweet spot for the lower damper with the new grate (and I might have already). But why? The stock grate acts as kind of a restrictor plate. Open the damper full, open it it half.  It’s all the same. The bottom plate is only going to let so much air through. (It occurs to me that the width I had the bottom damper open at the end isn’t that much wider than a single hole in the stock plate. I wonder if there’s some kind of inverse relationship  going on here? )

I’m just going to have to smoke something again this week, but this time with the stock plate in the charcoal bowl and the bottom damper half-open. Oh darn. But it’s for science!