Controlled combustion � a beekeepers smoker.

[petertr]

While experimenting with using an electric hob in a microwave to generate smoke for cold smoking (the smoke to be piped into a whisky barrel) because the deadline was drawing near (the ham and bacon to be smoked is now finishing curing) it seemed a good idea to have a backup plan.

In jest, I said to my 12yr old, that the backup plan was for him to sit puffing our bee smoker into the barrel for 12 hours a day.

Sometime later it occurred to me that a design like an automatic beekeepers smoker might be worth trying.

Our bee smoker is the most widely used traditional design � a combustion chamber (round tin body), tapering at the top to a nozzle and with hand operated bellows to blow air into the chamber and produce a puff of smoke. Common fuels are Hessian sacking, cardboard, wood shavings or sawdust.
The hardest part is getting the smoker going �initially you must build up enough hot embers to maintain combustion. This can be done by fast burning an initial small amount the fuel (puffing with the lid open) then adding the rest and closing the lid.
Air convection gives a trickle of smoke but if not �puffed� often enough, at least 2 or three times every 5 minutes, the smoker goes out. It is a controlled combustion � it is the amount of air entering the combustion chamber that sets the burn speed. My small smoker smoulders 25 to 30 minutes from full. With unlimited air the fuel would go up in a quick fire.
Clockwork beekeepers smokers are available but they�re only designed to run for an hour or less.

How to make an automatic large scale beekeepers type smoker � cheaply? �.
Possibly - use a PC (or two) and a few tin cans and copper pipe and a few spare plumbing fittings.

Our loft has a few old obsolete PC bits and pieces; I want two small electric fans � processor coolers and a 12V supply to run them � a PC case with PSU.

The combustion chamber is two large tin cans joined together end to end, each is 4� diameter by 8� long. The can ends are all removed except the bottom end this has two 2.5mm holes drilled for air to enter.
This stands on the lid of another slightly larger tin can (a cleaned out 1 litre paint tin). This has four 5mm holes near the centre (below the combustion chamber), and has one of the fans mounted on its side to blow in air. This needed the can side pushing in to flatten it a little, a 28mm diameter hole cut with tin snips and 4 small self tapping screws to fix the fan on (these came with the fans).
I wanted to be able to control the fan speed. Some PC cooling fans have a speed control, this one didn�t so I wired a 200 ohm wire-wound potentiometer in series to do the job.

I needed a detachable lid for the top of the combustion chamber, I couldn�t find a close fitting tin (that would fit over) so had to fashion a lid from scrap metal.
I cut a 15mm diameter hole in this which would take the end of a 10mm copper pipe compression fitting I had spare. The 10mm pipe (internal diameter around 6mm) is to take the smoke away, but for initial testing I left it off.

To make the setup more stable I put over an outer tube � an old section of 6 inch clay field drain. An alternative might have been the PC case � when smouldering there is not much heat. The other reason for this arrangement is that the incoming air is warmed as is drawn down between the outer tube and the combustion chamber. This (secondary) combustion method means that a slower fuel burning rate can maintain the combustion. For a small barrel smoker there�s more than enough smoke at the slowest burn rate possible.

Lighting:
With the lid off and the fan at maximum I dropped in a few pieces of burning newspaper, each 2-3 square inches, on which I had earlier poured a few drops of vegetable oil. Then rapidly threw in some wood shavings, a few a first then more as the fire builds. If too much is put in before hot embers have built up then it smothers the fire and goes out. Once a 2-3 inch layer was burning the fan was turned down (low flames instead of a roaring furnace) and another 2-3 inches of fuel, a mix of shavings and small (half to one inch) pieces of wood was added. To control any over rapid build up of flames the lid was put on for a few seconds to starve the fire of air.

Finally the lid was left on and the fan setting adjusted until a steady stream of smoke was produced. Too slow a speed and the smoke gradually tails off as the heat dies, when this happened I turned the fan up and removed the lid just until flames started appearing again. Too fast then the smoker won�t run as long and flames might flare up - the smoke coming out then being much less dense.
Once the right speed is set more fuel was added, this time using 1 inch slices off a 2inch branch. It�s best not to put in too much fuel until the speed is right, just in case it goes out and needs relighting.

I half filled the combustion chamber, it�s now been smoking for over 3 hours and so far only about half of what I put in seems burnt. With any luck one new load per day will suffice.

As it stands the smoke is too dense for smoking (want it thin and wispy).
Still to do:
Try putting 10mm smoke pipe into large hole in bottom of barrel, see if heat in smoke will cause enough convection drawing in fresh air at bottom.
If not them use second fad to blow in fresh air � mix smoke into air feed.

[saucisson]

Another interesting idea Peter, any chance of some pictures? I concluded that if I could keep something self combusting it produced too much smoke for my use, but if this device can fine tune the air flow or excess smoke is less of an issue for you then you could have a winner.

Dave

[petertr]

I�ve finally managed to put together some pictures of my Cold Smoker build, it�s been running for the last two days. I tweaked the design a little and have added details of these.


Traditional cold smoker designs allow fresh air to be freely drawn over the open fire and into the smoker such that the smoke emerging is thin and wispy - not dense. If insufficient fresh air then the smoke is too hot and, even with dry wood, too humid, 80 % relative humidity is ideal. This is because the process of burning gives off carbon dioxide and water vapour. A lower smoke temperature gives better smoke diffusion but needs longer smoking times. it is a slow process and hams can be smoked from 2 to even 6 weeks.
I was aiming for a medium scale DIY cold smoker that (as I want to do hams) would run as long as possible without attention and was cheap to produce � ideally using mostly materials to hand (junk). Having been given a large cask all I�ve had to buy is a 100 mm Flanged Spigot at �2.40.

My smoker consists of a combustion unit that generates dense smoke and a mixing unit that combines the smoke with a large amount of fresh air. The smoking chamber, a Whisky cask (Hogshead size) stands on three concrete blocks that make room for mounting the mixing unit on the underside of its base.
A 4 foot length of 10mm copper pipe carries the dense smoke from the combustion unit to the mixer. Due to the slow airflow and rate of burn the water vapour from the burning process tends to condense in the copper pipe. The pipe run was shaped (it bends easily by hand) with a definite bend at the lowest point. A condensate drain hole (3.2mm) was drilled in the bottom of the pipe at this low point.


A burner design where airflow relies on convection is sensitive to wind, local turbulence and air pressure variations, e.g. chimney downdraught. The smaller the fire the worse the problem, it tends to either go out or flare up and get too hot. We often get strong winds here so I�ve used fans to give the controlled airflow needed. Both are PC processor cooling fans, their small size being an ideal fit with the other smoker parts. I had a scrapped 486 PC and another scrap motherboard so that gave me two small 12 Volt fans and a supply of 12 Volts.
I�ve put the PC case in an outbuilding and ran the 12 Volt supply to the smoker using bell wire.

The combustion unit is basically a vertical tube through which air is blown, it is made in two parts- the blower and the combustion chamber. This makes it easier to remove the combustion chamber to empty it.

The combustion chamber is two large tin cans joined together end to end, each is 4� diameter by 8� long . The can ends are all removed except the bottom end this has three 3.2mm holes drilled for air to enter. Note that the size and number of these holes, together with the blower fan speed, controls the burn air flow (I started with two smaller holes for the test run and drilled extra as required).
To join the cans I wrapped a 2 inch strip of scrap lead flashing around the join and twisted some galvanised wire tightly round while tapping the lead to form it round the can rims.

The combustion chamber stands on the blower unit which I made from another slightly larger tin can (a cleaned out 1 litre paint tin). This has four 5mm holes near the centre (below the combustion chamber), and has a fan mounted on its side to blow in air. The can side was pushed in to flatten it a little and a 56mm diameter hole cut with tin snips. Four small self tapping screws, that came with the fan, fix the fan on.
I wanted to be able to control the fan speed. Some PC cooling fans have a speed control, this one didn�t so I wired a 200 ohm wire-wound potentiometer in series to do the job.



I needed a detachable lid for the top of the combustion chamber, I couldn�t find a close fitting tin (that would fit over) so had to fashion a lid from scrap metal. I used a 6inch circle cut from scrap lead flashing, this was tapped into a lid shape using the can base as a former.
I cut a 15mm diameter hole in the lead lid, and in one of the removed tin lids, to take the end of a 10mm copper pipe compression fitting to act as a pipe flange. The fitting sandwiches the two lids together, the tin lid going under the lead to act as a heat barrier and stiffener. The 10mm smoke pipe is just left loose in this fitting so it can be lifted on and off.

I used a length of 6 inch diameter clay field drain as an outer sleeve around the combustion unit (including the fan). This makes it more stable, also the incoming air is warmed as is drawn down between the outer sleeve and the combustion chamber. This (secondary) combustion method means that a slower fuel burning rate can maintain the combustion. For a small barrel smoker there�s more than enough smoke at the slowest burn rate possible.

SMOKE CHAMBER
This was an enjoyable task - in a small workshop the whisky fumes were quite overcoming.
The cask lid was first removed by cutting around it. Sixteen � inch holes were drilled in the lid and covered with perforated zinc. Across this are three batens at right angles that hold the lid planks together, and create an air gap; over these is an oversized rain cover. I used galvanised steel sheet for the cover, with the edges bent down to shed water, and screwed a wooden handle, cut from a short length of (2� x 1�), to the centre batten.








A 100mm diameter hole was cut in the cask base and over this stood a baffle on � inch spacers. The baffle spreads the incoming jet of smoke-air.




I�ve arranged supports that allow for hanging meat and using a grid shelf.
Two 10mm copper coated steel rods (an electrical earth rod cut in half) go through holes drilled near the top of the cask. The rods are parallel and across them rest notched wood battens, for hanging meat, and/or an old BBQ grid shelf. Meat is tied to a batten then lifted in or slid along the rods together.









The mixing unit adds a relatively large amount of air to the dense smoke. Air from the fan is formed into a jet by the nozzle and smoke from the combustion unit is added at this point. Where air speeds up to go through a constriction the air pressure drops (the Venturi effect) this helps draw smoke from the combustion unit � rather than try to blow it back.

I�ve fixed a 100 mm Flanged Duct Spigot to the bottom of the barrel where I�d cut a 100mm diameter hole. After cutting its top rim off, the tin can I used fitted exactly over the spigot. I�d bought this spigot for my mk1 smoker design (electric hob & gutted microwave), otherwise I�d probably have fitted the can direct cutting 3 inch holes in its base and in the barrel.

The fan is fixed, with self tapping screws, over a 45mm diameter hole in the bottom of the can. A piece of fine mesh to stop insects is sandwiched between the can and the fan.

I used part of a plastic juice bottle for the nozzle, one that was a reasonably tight fit over the fan. I drilled a 10mm hole in its neck for the smoke pipe and another level with this in the can wall. Where the pipe enters I used a 10mm copper pipe compression fitting (drilled out to let the pipe go right through) but just a tight hole in the can would have done.
MIXER AND BLOWER



LOOKING DOWN THROUGH THE CASK BASE HOLE



Burning
Test run
Initial tests of the combustion unit alone used Pine shavings and chunks up to 2 cubic inch. Little combustion airflow was needed, otherwise there was a tendency for it to flare up if the lid was removed. A surprising amount of liquid tar ran down from under the lid.



With the mixing unit and smoke chamber connected tests using Plum wood chips and chunks revealed that there was insufficient air flow to maintain combustion. The number and size of the holes in the combustion chamber base was increased (to 3 x 3.2mm).
The smoke from the combustion unit is about 10� C when it reaches the far end of the copper pipe.
On the test day the (damp) air temperature was 8� C and in the barrel was 6� C. This must have been due to the cooling effect of evaporation from the whisky soaked wood. An inspection light was left to warm the barrel overnight with the fans running to dry it out.







The first smoking day it went out after around 2 hours, the fuel hadn�t settled so emptied and relit. Eventually the fuel ran out after between 8 and 9 hours, fans switched off at 9hr point. Second day, still burning at 9 hours when fans switched off.
I would expect that a full load of shavings only would last less than an hour, and a chippings only maybe 2 � 3 hours. I�m experimenting with the best wood size mix to use, currently using Plum wood:
To light � a handful of shavings, burning scrap of paper then two handfuls of shavings added a little at a time.
Getting going � a few handfuls of chippings (from cutting with chainsaw)
Fill to top � mix of wood chunks and more chippings, starting with small chunks and building up to about 1 cubic inch. Then put on lid and attach the smoke pipe.
I�ve run the burner fan at full speed for the first 2 hours or so, to get the combustion chamber hot, then reduced it slightly to give less smoke and a longer burn time. On the test day there was cold damp drizzle all the time, the fan had to be kept at maximum.

A full load will burn for over 8 hours but needs an occasional poke with a stick to settle the fuel, every hour or two or when the smoke is becoming thin, otherwise it can go out. As it burns a cavity must form under the fuel, if the burning fuel falls to the bottom then the heat wouldn�t reach the remaining fuel. The ribs in the can walls or the seams at the join may be the problem.

Smoking
We were using fresh back and streaky bacon, from our own pigs, that had been dry cured.
Day 1. Air temperature was 4.8� C and in the barrel was 3.9� C Not sure if this drop is due to further whisky evaporation or the back & streaky bacon drying off. Not ideal as relative humidity will be higher than outside. The bacon colour does not seemed to have changed yet, probably due to the low smoking temperature.
Day 2. Continued smoking bacon. To increase the temperature I�ve added a 21Watt car bulb, temporarily loose just under the baffle. Its fed off the PC 12V via another length of bell-wire. Now air temperature 7.5� C and in the barrel was 9.3� C
A test batch of Cheddar was also done � 1/5th not smoked, 1/5th smoked 1hour, 1/5th smoked 2 hours, 1/5th smoked 3 hours and 1/5th smoked 4 hours. Wrapped and will taste after 5 days.
The bacon has been smoked for around 17 hours. It smells very smoky but didn�t appear obviously darker, however comparing it against unsmoked bacon showed a light golden brown tint. We properly tested some streaky bacon at lunchtime. The colour and smell ran right through the slices and the result got 10/10 all round.

Modifications - perhaps
If the heating lamp continues to be needed (whisky finished evaporating ?) it will need to be mounted somehow. To get a bit nearer 12�C maybe I�ll need either increased bulb wattage or reduced fresh air flow (a second fan speed control).
The tin cans used for the combustion unit would be better replaced by a stainless steel tube, smooth so fuel might settle better and will last longer. I�ll be on the look out for off cuts of chimney section.
Having removed the cask lid it, and before the wood shrinks as it dries, I should strengthen the cask by screwing the metal bands onto the oak planks. Stainless steel screws are best, others will stain the oak.



PARTS LIST
Smoke chamber. . . . . . . . . . . . . . . . . . . . . whisky cask
Insect mesh. . . . . . . . . . . . . . . . . . . . . . . . perforated zinc, fine plastic mesh or microwave door mesh
Rain cover. . . . . . . . . . . . . . . . . . . . . . . . metal or plastic sheet or thin ply
Two support rods. . . . . . . . . . . . . . . . . . . . . copper coated steel earth rod1.2m x 10mm
Two small 12V fans. . . . . . . . . . . . . . . . . . . . . PC scrap
12 Volt power supply. . . . . . . . . . . . . . . . . . . . . PC scrap or 12V dc supply mains block.
DC fan speed control. . . . . . . . . . . . . . . . . . . . . 200 ohm wire-wound potentiometer (volume control)
Tins for blower unit and mixing unit.. . . . . . . . . . . . . . . 1litre paint tins
Tube for combustion unit, with base and removable lid. . . . . . . . . Two 8inch by 4inch diameter tins
10mm copper pipe, 4 foot length. . . . . . . . . . . . . . . Central heating pipe
Flange for 10mm pipe. . . . . . . . . . . . . . . . . . . . . 10mm compression fitting
Air jet nozzle. . . . . . . . . . . . . . . . . . . . . . . . Plastic juice bottle
Baffle plate 6inch diameter. . . . . . . . . . . . . . . . . . Copper sheet or thin plywood
100 mm Flanged Duct Spigot (optional)

[petertr]

One image link wasn't right.

The initial small quantity of shavings and chipping flares up, as more fuel is added this turns to dense smoke.

[Patricia Thornton]

I know little about smoking food but I found your post extremely interesting and quite amazing. Congratulations.

[jenny_haddow]

Very impressive!

[saucisson]

A superb piece of work peter, I'm very impressed. It's funny we are both using plum tree. Can you estimate how much smoke goes into the barrel as compared to out of the top of the combustion chamber?

Dave

[petertr]

Can you estimate how much smoke goes into the barrel as compared to out of the top of the combustion chamber?

I�d say at the moment around 10% of the smoke is lost from around the combustion chamber lid. It was much less until I dropped the lid! - being soft lead it bent a little.
Here�s two more test photos with just a pipe and with the test mixer running.




A little smoke can also escape from the condensate drain hole, but a film / meniscus of liquid seals this hole after a while.

I felt that, at minimum fan speed to maintain combustion, I had more smoke than was needed, for the size of enclosure, and was concerned that the smoke produced was too dense � hence mixing in air.
This was guesswork � I couldn�t find any figures for recommended air/smoke changes per hour. Traditional cold smoker designs allow fresh air to be freely drawn over the open fire and into the smoker. Once a wood burning fire gets properly going there�s not much smoke in the rising heated air (unless using very damp wood), I was aiming for a mix like that coming from the chimney of our lounge wood-burning stove.
If anyone has figures I�d be very interested.

Are you at all worried about the lead causing any contamination?

The lead doesn�t come into contact with the smoke that goes to the cask, only what escapes. The lead lid, which has a rim, also has a flat tin lid sandwiched under it. Also the lid never gets very hot, just warm to the touch, as the burning stays in the bottom quarter of the combustion chamber. There�s a possibility of some condensate on the lead running back inside but as there�s no sign of the lead corroding I guess it should be OK.

After some use the lead lid may fracture at the fold and would need replacing.
Really the tin cans and lead are a lash up to test the thing, now I know it works I�ll try to find a better long-term solution.
Unless I see something better I�ll use a steel or stainless steel tube for the combustion chamber, weld on a base, and use another steel plate for the lid with a rim of the next tube size up welded on. The rim would make sure that the lid was put on properly, the lid weight alone would be enough to make the seal needed. At most this would cost �20 for materials.

If anyone has other ideas for improvements here or elsewhere, again, I�d be very interested.

[adwindrum]

Fantastic set up there! I just use an old woodburner and a cupboard. I dont burn continuously as it goes out after 3 hours and I dont fancy night shifts, but it doesnt seem to effect the end product. A borrowed gas blow torch gets it going. Being minimal effort allows me to be happy smoking for a couple of days or more if needed. (I did overdo my first attempt at salmon the other day though (much more sensitive than pork)...going to have to use for cooking...doh!
I use green oak sawdust from a sawmill that lasts and lasts.
Just one note about yours...I was wanred off using chainsaw sawdust as it contains a lot of oil. Reading your bumph you seem to only use it for getting the thing going so no problem if thats the case.

[petertr]

Living half way up a hill in Banffshire, and wanting to have it outside, I needed something that would keep going if (when) we get a gale, and heavy enough to stay put.

I was wanred off using chainsaw sawdust as it contains a lot of oil.

I wouldn't have used chippings from my old chainsaw as this gets through a lot of chain oil but the new Husqvarna uses much less. In the 1 - 2 seconds taken to cut each slice only a tiny amount of oil can be used. The chippings are sort of narrow slivers, a chainsaw that produces sawdust is badly in need of sharpening.
I'd say that most of any oil left is probably smeared on the edges of the cut slices.

I found that whole slices were too large to burn, they each had to be split into around 1inch squares. Next time I might try splitting the branches lengthways into pieces small enough either to feed into our garden shredder or to cut on the saw bench - both being oil free.

[opus]

Wow!! Wanna come here and build one for me?