Buy a probe thermometer for your pipe, drill approx a foot above the collar, also use same hole to test draft, you want .05wc" while the stove is running on high setting.
As far as intermittent smoke smell, does the gasket have an equal knife edge mark all the way around? is there any discoloration beyond the knife edge mark, I know my gasket looks pretty dirty on the inside knife mark and its white after the knife mark (3 seasons on it) If you do take the gasket off, also check the glass retention bolts that hold the window, a bolt or two could be loose.
The one thing that needs to be remembered here is the BK base air input (when the t-stat is totally closed) is much smaller then other stoves, so if you hit temp and the t-stat closes then your essentially running on the hole, if your draft is minimum you can experience different issues, the more common would be cat stalling, but if you have a marginal set up like yours you could have smells due to turbulence of smoke (2) 90's fairly close to each other (elbow then crock into the unlined masonry chimney)
The other issues is flue gas condensation, since the King is designed to smolder and burn smoke for heat production on lower settings your flue gasses are much colder then other stoves (why an insulated liner is highly recommended even on interior chimney's), cooler flue gasses are more stagnated in nature, and if you have a leaky chimney (due to gaps at the black pipe connection or a poorly sealed clean out door) you'll have even slower issues. No one asked yet, but what kind of cap do you have?
Excellent explanation of the working system and contributing factors!
One question though, why do you say i have a unlined masonry chimney? There are beautiful clay tiles lining it, and from camera inspection the joints are almost perfect aligned and almost no trace of cement hanging inside.
As far as signs of leaking around the gaskets there are none. I have been going through this stove with a fine tooth come trying to understand the forces involved in pushing gases "through" the door gasket, also the glass is secured properly.
No leaks from the pipe or the connection at the collar or thimble, I have verified this 100% with a simple sniffer tool made with foil and shaped like a long funnel. This allows to pin point fumes from a three foot distance keeping myself out the the heat turbulence from the stove. The joint to the thimble is air tight, caulked and secured with two screws.
About the draft, when the stove is on high the draft measures between .05 and .06wc, when the stove is at 20% it measures between .04 and .05 wc about 4 inches above the collar. I might note that similar values are measured in the hole above the cat.
These measurements tell me the draft is sufficient and makes me suspect something very difficult to grasp is causing localized high pressure around the the door gasket.
These facts lead me to believe this theory:
- gases are exiting around the door through the gasket
- WC >= .04 above collar
- opening the door slightly the fire rages and no more gases exiting (if draft is issue how is this possible?)
One thing you said(King is designed to smolder) sparked a thought, is it possible the gas leak is happening only when the flame is extinguished? So in the slow mode it would be more prevalent, in the mid range intermittent, and in the high range when the flame is pretty much constant, un-detectable. All this is dependent on ambient temperatures however lets assume in the mid operating temperature range the thermostat control is moving from closed to some percentage open to maintain the set temperature, when it is closed the flame is out and the coals smoldering, when open the flame is re-established. This seems to track with what I have experienced over the trouble shooting process. Very interesting!
