A flame that shrinks to a yellow sputter partway through a boil is usually not a clogged jet. It is a seal that has stopped sealing: most often the O-ring at the canister or pump connection, then the threads it seats against, then the valve behind it. The jet sits at the far end of that chain, which is why it is the last part to fail and the first one most of us take apart.
I learned that kneeling on a slab of granite with a pot already on the burner. The stove had lit with its usual hard roar and then dropped into a hiss with a wobble in it, the flame lifting and settling, lifting and settling, like something was breathing on it. My first move was the one everybody makes. I shut it down, pulled the jet, ran the wire tool through it, looked at the sky through the tiny hole, put it back together, and lit it again. Same wobble. Same half-hearted ring of blue.
So I did the other thing, the one I should have done first. I cracked the valve for two seconds, shut it, and ran the back of my finger around the joint where the canister threads into the stove. There was a spot down there that felt cold in a way that had nothing to do with the morning. Gas was escaping before it ever reached the burner.
That is the whole reframe, and it took me a few seasons to trust it: a weak flame is a flow problem, and flow is decided upstream of the jet far more often than at it.
The connection fails in order, and the jet is last
Walk the path the fuel takes. Canister or bottle, then the O-ring that seals the joint, then the threads that hold the two halves together, then the valve that meters the gas, and only then the jet and the burner head. Everything downstream of a leak gets starved. Everything downstream of a clog gets starved too, which is exactly why the two feel identical at the pot.
What separates them is where the loss starts and what it does to the flame. A clogged jet gives you a flame that is small but steady and usually lean — clean blue, just weak, sometimes with a pop when you shut down. A leaking seal gives you a flame that wanders, that goes yellow at the tip or lifts off the burner, and that changes when the canister gets cold or when you nudge the stove on the ground. A seal leak also puts fuel smell where fuel smell should not be after shutdown. If I catch a whiff with the valve fully closed, I stop and I go looking at the connection. That smell is the cheapest diagnostic on the stove.
What a hardened O-ring actually feels like
The first time someone told me to check the O-ring, I looked at it, decided it looked fine, and went back to the jet. Looking is useless. You have to feel it.
Roll it between your thumb and forefinger. A healthy ring is slightly tacky and springs back when you pinch it. An old one has gone waxy or chalky, and when you press a thumbnail into it the dent stays. Sometimes there is a flat facet worn into one side where it has been compressed against the same seat for years. Sometimes there is a hairline crack you can only see if you stretch the ring gently and hold it against the light. Any of those and it is done, whether or not it leaks today.
Material matters less than you would think, and size matters more. Silicone and nitrile hold up to fuel and cold better than natural rubber, which is why the good rings are made from them. But a ring that is a fraction too thick will not seat, and one that is a fraction too thin will not compress, and either mistake produces a leak that looks exactly like a worn-out ring. This is where I stop improvising. The manufacturer's specified ring, in the specified size, is the only version of this part that is known to work in that seat. A hardware-store ring of matching diameter and unknown rubber is not a repair; it is a variable.
A light film of silicone grease keeps a ring supple and helps it seat without twisting. I keep a tiny tube in the repair kit. What I do not do — anymore — is reach for whatever grease is closest. Petroleum-based lubricants will swell and soften some elastomers, and a swollen ring will not sit down in its groove. If the tube does not say silicone, it does not go on the seal.
Threads are a seal surface too
A surprising number of weak flames are dirt. Grit in the threads holds the connection a fraction of a turn short of seated, and a fraction of a turn is all a seal needs to leak. I wipe the threads on both halves with a rag before every trip, and I look at them in good light, because a cross-thread started once tends to get started again. Overtightening is the other version of the same problem: it compresses the ring past the point where it can recover, and it chews the threads that are supposed to hold the alignment.
I do not chase a leaking thread with tape or sealant. Those threads are engineered to seal with the ring that belongs there, and wrapping them changes the depth the fitting seats at.
The valve side
There is a second, smaller seal inside the valve spindle, and it fails quietly. The tell is a stove that holds pressure fine but leaks slowly when the valve is closed, or one that will not shut off completely — you close it and the flame shrinks but does not die, then dies a few seconds later. That is not dirt in the jet. That is gas getting past the valve seal. On stoves with a pump, there is a third seal at the pump cup and a fourth at the fuel cap, and the same rules apply to all of them.
The bubble test tells you the truth without a flame
Here is the procedure I use now, and it takes about a minute. A cup of water with a good squeeze of dish soap. A drop or two smeared around the joint — around the base of the canister where it meets the stove, or around the pump fitting. Valve cracked just enough to pressurize. Then watch. Do not light it.
A good seal gives you nothing. A bad one gives you a ring of tiny bubbles that appear fast and grow, and occasionally one that swells and pops. If it bubbles, it leaks. There is no interpretation required, which is why I trust it more than my own ears. I have talked myself out of a hiss before. I have never talked myself out of a growing bubble.
Do this test cold, and then do it again after the stove has been running a while, because temperature changes the answer. Elastomers stiffen in the cold and seals shrink with them, so a ring that is marginal at room temperature can leak outright on a frosty morning — precisely when you need the stove most. That is also why a weak flame that improves as the canister warms up is a clue pointing at the seal, not the jet.
And to be fair to the jet: sometimes it really is clogged. Dirty fuel, a stove stored with fuel sitting in the line, a trip in dusty country. If the bubble test is clean at every joint and the flame is still lean and small, then yes, go back into the jet. I am not telling you the jet never fails. I am telling you it is fourth in line, not first.
Storage is part of the seal
The most common way these rings die is not use. It is sitting in a hot garage for eight months with a canister threaded on and the valve shut, the ring compressed flat against its seat the entire time. I store mine with the canister off and the valve open, so the seals relax instead of taking a permanent set. It costs nothing and it is the reason I get years out of a ring instead of one season.
I also stopped storing the stove with fuel in the line. It is not a dramatic difference, but the seals that stay wet with fuel and then dry out tend to be the ones that go chalky first.
Now, when a flame goes weak, I do not start with the tool. I start with my thumb on the ring, a cup of soapy water, and a minute of patience. Nine times out of ten, the jet was never the problem.
