Honeywell S8610M intermittent pilot control valve

_big_

I have a 21 year old Superhot SG-270.

It mostly works but occasionally refuses to fire for a random amount of time. I've observed that during these misfire periods, there is demand for heat, the boiler tries to light the pilot, but fails. Then there is about a 5 minute wait period for any accumulated unburnt gas to dissipate. Then it will try again. I've seen these un-light episodes last for an hour or two.

We've had periodic failures like this for at least a year. One of the strange observations I made was that the boiler is more reliable when I leave the front cover of the boiler off. Interesting.

Was investigating a non-fire episode yesterday, and I noticed that the pilot inspection hole cover plate was in the "inspect position" and not in the "cover the inspection hole position". Even more interesting.

The boiler room is a small closed room with makeup air vents. The makeup air vents are to an inside corner of the house exterior. And would definitely be influenced by wind direction.

At this point, I think the theory should be obvious. When the wind blows in the "right" direction, the pilot is blown out by the wind current. And while you might assume the front cover of the boiler would reduce these wind currents, its not a solid plate and could very easily direct the wind currents into the pilot inspection hole.

I've placed the inspection hole cover to the "run" position. And it ran fine last night. This is not necessarily conclusive, but is certainly evidence supporting the theorized root cause.

I have a couple questions for the HeatingHelp crowd.

Its possible we just didn't notice the problem, but It seemed to run fine for 18+ years. I'm wondering if the intermittent pilot control valve is aging and not throwing ignition lightning bolts like when it was young pup. Anybody know anything about the behavior of aging intermittent pilot control valves?

Is it possible the ignitor/pilot assembly has grown some low grade fault that is bleeding off lightning bolt energy?

The pilot control valve is a Honeywell S8610M. Apparently, these are no longer available. The Resideo VR8345M/Q/K is listed as a replacement. I would kinda hate to step away from the original equipment, but that might be necessary. Any comments about the Resideo?

This boiler is the sole source of heat in the house. Reliability is quite important. Would consider proactive parts replacements.

EBEBRATT-Ed

Resedeo is just made by Honeywell they just renamed it.

I would remove the pilot assembly and clean it. Cobwebs dirt and dust and soot. A little bit of that can screw things up.

Are you sure the pilot is not lighting or is it lighting and not being proven by the control? Makes a big difference in troubleshooting.

Make sure you are getting 24v ac to the gas valve when the pilot tries to light. Make sure the pilot assembly is clean and screwed in tight . Inspect the spark cable and remove both ends and plug them back on a little corrosion can stop the spark.

Karl Reynolds

The Honeywell S8610M on my Allied boiler gave up after twenty-one years. You can replace it with an S8610U3009. See page #2 of the instructions with the S8610U. You may need to set a switch or two. The VR8345 that you refer to is a gas valve. The new S8610U and a good cleaning like Ed recommends and you should be back in business.

_big_

Finally got to the bottom of my boiler problem. I'll record my experiences here as a possible benefit to others in the HeatingHelp community who might pass this way.

First, I'll explain how the S8610M (and similar) works. It is an "intermittent" pilot controller. This means there is no "standing" pilot and thermopile (old school). When there is demand for heat, the S8610M turns on the pilot gas and fires its sparker. Nominally, the pilot will be set alight by the spark.

The S8610M applies a voltage to its electrode (while not sparking) and looks to sense a current. I think the threshold is 2 micro amps. Burning gas is somewhat conductive, unburning gas is not. This process is called "proving" the pilot. Once the pilot is proven, the S8610M will open the main gas valve.

In my case, you could see the pilot light, but then no main gas valve actuation. The S8610M would then enter its 5 minute unburnt gas dissipation period. This might happen as many as 10 or more times in a row, and then it would get lucky, prove the pilot flame and open the main gas valve. This failure to consistently prove the pilot went on for at least a year. We never had a cold house. Occasionally, we would have cool DHW.

At this point there were only two possible culprits. The S8610M, or the pilot assembly itself. The S8610M is a sealed box. Not much you can do with it. I did remove the pilot assembly, cleaned the electrode, inspected etc. No obvious problems.

Concluded it must be the S8610M. Ordered the Resideo from Amazon. Installed it, no joy. Same basic behavior. One of the nice features of the Resideo is it has an LED upon which it will "morse code" various status messages. When the main gas failed to come on, the Resideo would indeed report a pilot proof failure. Confirming the suspicions.

One note about the Resideo… RTFM. There was a minor wiring change from the original Honeywell S8610M.

Must be the pilot assembly. I removed it and got a bit more aggressive with it and discovered some melted insulation. That must be it. Time for a new pilot assembly. Well, the best Amazon could do was 2 days out…

Meanwhile, my more aggressive treatment of the pilot assembly had made the problem worse. Generally speaking, none of the plumbing supply houses around here will even talk to a consumer. So I called a plumber hoping that he could get me a new pilot assembly quickly.

Plumber arrived quickly. He did his own analysis and also concluded I needed a new pilot assembly. He retired to his truck to see if he could get one. But on his way back to his truck he made a comment about how he saw the sparking action.

This comment got me thinking… without removing the pilot assembly, I used a screwdriver to adjust the electrode so that it would be more in the burning pilot flame. Voila. That did it. Its worked perfectly for over a month now.

Plumber couldn't get a new pilot assembly faster than Amazon. Dropped that effort. Amazon has delivered the new pilot assembly, but I haven't installed it. If it ain't broke, don't fix it.

While the plumber was on site, I asked him several other questions about the boiler and its associated systems. There are some things I don't like about how it was originally set up. His answers were quite useful. The best part is, he only charged $30. He didn't do any real work, but really helped me out. I would have gladly paid him $100.

Jamie Hall

At some risk of seeming odd… that misaligned electrode should have been caught and adjusted by whomever is supposed to be servicing the equipment…

EdTheHeaterMan

The "Proving Flame" function of the S8610 is called Flame Rectification. That means the flame is actually a Rectifier in an electronic circuit that is changing AC current to DC current that will operate the flame safety circuit and allow the S8610 to take the next step to open the main gas valve.

For the flame to become a "Rectifier" the AC electric current needs to enter the flame from a small electrode and complete the circuit to the other side of the flame at a larger surface. The larger surface is the flame hood of the pilot burner. The return path of the DC (after rectification)to the control is the chassis ground. If the electrode is dirty or not in the flame correctly, the AC current will not enter the flame. 

If the Pilot burner hood is not properly grounded to the control because the mounting screws are loose or corroded, or the connections between the pilot burner hood thru the boiler hardware, to the mounting screws on the boiler cabinet jacket, to the mounting screws to the S8610 that has a ground connection, or the ground wire spade connector is compromised by corrosion after 20 years of sitting in the boiler room, then you may end up replacing the pilot burner, the S8610, and a couple of other things until you happen upon the loose cabinet screw that makes the ground connections happen.  And that is only by chance so you have no clue what fixed it.  (ask me how I know this).

I have resolved a rusted boiler Ground Connection in the past by attaching a ground wire directly on the screw that holds the pilot burner on the main burner and attaching it directly to the ground terminal of the S8610. This eliminates any bad ground connections in the boiler assembly.  

Here is an excellent video from Gary McCreadie explaining Flame Rectification. https://hvacknowitall.com/blog/that-flame-is-not-being-honest

There is a Goodman Training video attached below Gary's video that explains the importance of a good ground for Flame Rectification to work.

I hope this helps with solving some similar problems with flame sensors.

Steamhead

@EdTheHeaterMan , that's exactly what I would have done!

_big_

Here's some description of physics behind "flame rectification". https://en.wikipedia.org/wiki/Flame_rectification

Gary McCreadie's discussion says you can measure the sense current by inserting a milliammeter. Maybe yes, maybe no. In my case, there is a single electrode for the spark and the current sense. I wouldn't recommend inserting your milliammeter in this wire. I don't know of an easy way to measure the sense current for pilot burners that don't have distinct spark and sense electrodes.

As for adjusting the location of the electrode relative to the pilot flame, its difficult to predict where the flame will emanate just by looking at the burner. Get the pilot burning and then visually check to make sure the electrode is squarely in pilot flame. The Wikipedia article suggests the strongest rectifying action occurs low in the flame.