Pulsing BBR

Hi all,
My friend and I have recently built a 5" BBR from 2 drums and split fire bricks.
Everything is exactly to spec. The riser is square with the rear chambered to just above the height of the port. From the height of the top of the fire box upwards it is a 5minute riser. No ramp  the same as the 2017 BBR on Peter's site.
The thing heats the house very well, however it is pulsing very badly like others I've seen. A few flames per second pulsing out of the primary air.
I read somewhere that Peter said these are usually caused by a restriction down the line. The flue is also 5" so it should be fine.
The possibility that comes first to mind is perhaps the transition area fell barrel to flue needs to be larger than 5" diameter?
Opening the door a little stops the pulsing, but that is no doubt due to poorer efficiency of the burn and therefore less air requirement and poorer draft so it no longer suffers from the restriction causing the issue.
There is definitely an efficiency problem because the bricks never clean like my ones at home do shortly after lighting it. They stay black.

Am I on the right track with the transition area? Or is something else worth checking first?

https://photos.app.goo.gl/mBLiN2BHGrBVzEQu9

A video with my friends son explaining to me what is happening haha.

https://photos.app.goo.gl/ubZmrYvEDixn2VF76
https://photos.app.goo.gl/z7sVDbi4GzpQKcZ7A
https://photos.app.goo.gl/KKhu4oKZgSVgjaC99

Also, the riser is very off center the way he has placed it in the drum.

An inefficient burn and off-center riser could have created a constriction in airflow between the top of the riser stack and the top of the barrel.

I have seen fly ash build up very quickly in this area (sometimes within just a few weeks) and your 3" gap is now down to 1.5", or less.

Yes, there should be a restriction down the line. What springs to mind first: this 5-minutes riser's diameter is 5" I assume. So the square of 5" is quite a bit larger than the circle of 5", the transition forming a restriction. Riser should be uniform in shape and csa overall, there should be no steps.

Sometimes people apply the same rules for J-tubes to batch boxes and creating a top gap that's way too small for such a device. This top gap should be at least equal to the riser's diameter, preferably twice that.

The exhaust opening close to the bottom of the barrel might be not adequate especially when it's situated too low. In most instances the solution is mounting a reducer, giving the gases more room to come around the 90º corner. This reducer could be 8" to 6" and then to 5", the size of the flue.

And then there's the issue of the main inlet. It's a combined one in this case cause there's a floor channel. In the video this inlet looks quite cramped, it should be at least 30 cm² for this size of batchrocket.

Hope this makes sense.

Thank you, Peter.
The air intake is correct I size. There is a 50x50mm opening above the floor channel plus the area for the floor channel. Calculated using the specs from your site.
I suggested to him yesterday that he should try a reducer at the bottom of the barrel. The first thing I noticed when insane the heatet is that there is only a 5cm gap between the bottom of the flue and the bottom of the barrel.
Also there is at least 30cm gap between the top of the riser and the top of the barrel.
If I understand correctly, you're saying the 5 minute riser could be too small due to the square portion at the bottom? The square brick portion is 125x125mm and them we chamfered the rear corners.

The csa of a 125 mm dia circle is smaller than the csa of a 125 mm square, that's all. It's still smaller even if you chamfered the corners, sorry to say. Your best bet would be to install a reducer at the exhaust side is my quess. And/or enlarge the inlet opening slightly so it's running smoothly all the time. In case all that doesn't help, the riser is the remaining culprit.

You know, sometimes a small deviation on its own won't do any harm, but the accumulated effect of several deviations could trigger bad behaviour.

I understand.
Thanks again for the responses.
He's taken the heater outside and it ran beautifully without the bell over the top.
A reducer will go in first as the most likely solution.
Have a great day.

First of all, read this. May be it can shed some light.   http://donkey32.proboards.com/thread/1406/calculating-ring-circumference-projection-gap

I have very little to add to what Peter said.

Remember, that my stoves haven't ever been testo tested.  

But i like to run the primary air a bit bigger than Peter's recommended values.  

I think I understand.

So the flue exit is at the bottom of the barrel. The nearest object to it is the barrel floor at roughly 2". This is the "gap"? Since there is a 12" space between top of the riser and the bell ceiling. Also at least 7" between the riser exterior and the bell wall.

The system is 5" with a CSA of 19.6.

Circumference of the 5" exit flue is 15.7. Multiplied by the gap is 31.4 square inches. Which is just larger than 1.5x the CSA.

An 8" exit will be 50 square inches or 2.5x the system CSA.

Benen, the gap is the distance between heat riser and the flue's edge.

So if i get it correctly, 7 inches.

Tho, you have a restriction, in the  2 inches at the bottom. That means that only the two sides and the top of the flue's edge have plenty of space. Nothing much to worry about mind you. In fluid mechanics  it is often said that the boundary layer, between stalling against the pipe's wall, and full speed is about 3cm. So gases in your 5 incher have only a tube at the middle, of 56mm, to go at fulle speed. Well, this is all verry theoretical. Since hot gases behave differently depending on their temperature.

I would try running it with  two or three times the recommended air intake size, at first. And if it still pulses (oh yeah! :D )  i would enlarge that flue entrance.