This "noisy" boiler sounds awfully like my experiences with a steam plant, reported in the Oct 2010 issue of Model Boats. The gas burner had its own characteristic but acceptable "roar" when fired out of the boiler. When placed in the boiler, a horizontal centre flue type, a horrible loud sound was produced, a sort of rasping horn effect and not something to be welcomed at the pond side.
Adjusting the ceramic burners position could reduce the noise a little but removing the brass funnel which was plugged into the flue outlet on top of the boiler eliminated this sound effect totally. This lead me to the conclusion that I was dealing with a Resonance Tube problem. The modest sound made by the burner was setting up a standing wave inside the flue and funnel.
Luckily the nice shinny brass funnel would have affected the stability of my planned model, a slim early TBD, and had to be removed. The replacement funnel section was made from thin aluminium sheet, same diameter but a different length. This was sufficient to stop the resonance and the horrid noise.
Had I still access to suitable measuring equipment then the frequency of the offending noise could have been measured. From the speed of sound, remembering that the resonance is occurring in the hot flue gas not cold air, the wave length could be determined. By changing the funnel length by around 1/4 wave length it should have moved the resonance effect from a maximum to a minimum. But, it was a lot easier with the aluminium funnel, to use the good old "cut and try" method!
Looking up "combustion driven oscillations" on the Internet can give you more information, For those with a musical bent, try "Pyrophone".
Glynn Guest