Texman 182g Mods

Clearly there is a transient from the dsi causing this, what's interesting is that using the relay makes it worse but it still happens without using the relay.  That appears to indicate that the logic circuit is getting whacked 2 different ways. If moving the tstat away from the dsi module eliminates the problem when not using the relay part of the problem may be that a transient voltage is being induced (without any actual connection) in the logic circuit by proximity to the dsi module or to the conductors running to the electrode from the dsi module. You could try wrapping the dsi module in foil and see if that stabilizes things with and without the relay in the circuit. My tstat is right next to my dsi module btw. 

Originally posted by texman that is the exact stat that is setting in my amazon shopping cart that is was thinking of trying to see if it helped.  Seems like a better stat may help that is more isolated.  Not sure about it being any better but its worth a try. since the dsi spark is basically a direct short to ground, using the ground terminal on the heater body is probably the worst place to use.  It should only make this worse. I'm using the ground screw in the little jbox on the water heater also. There is definitely some sort of surface voltage jumping around with the noticed changes from proximity to heater body and module. Its likely not a direct application of voltage. Rapidly changing currents and voltages cause electromagnetic fields which can induce voltage and current in nearby circuit loops. The induced voltage is proportional to the field strength and the area of the circuit loop so it gets larger if you have a bigger area enclosed by a loop.   The back of that module is open and has an unused adhesive strip. Mine too.  i did have to remount the module because the zip ties from factory broke. i used some solid copper insulated 20 ga wire wrapped around the styrofoam layer. so that module is setting next to that.  I am now thinking that could be contributing to the surface voltage spread for sure.  thoughts on that? Seems doubtful but I suppose the wire loops you added could be increasing the induced voltage in the adjacent logic circuit (either in the board itself or its power supply wires).  I don't have wires wrapped around my styrafoam so that is one difference.   Just to set the base parameters in case i am doing/using something that is causing this: I am using 18 awg solid copper wire (thermostat wire) that is separated from the sheathing as conductors.  still has the wire insulation, just not the sheathing cover.  i have heavier stranded if that would help.  I don't think heavier wire would make any difference, but I have my 12V tstat supply wires in a single 2 conductor cable up to where I end the +12V conductor where pick off the +12V, from there my -12V supply wire carries on to the ground point. That could make a difference as the high voltage dsi cable is adjacent and it could be creating a larger induced voltage in your power wires if they form a loop. Because mine are together I don't have a loop. You could try wrapping the two power wires around each other to create a twisted pair. In electronics cables that are used to carry small signals so need to be protected from voltage transients are usually either twisted together or shielded or both. my latest testing for water heater ignition is being done on an empty water tank.  i don't think a few seconds of heating is going to hurt anything but does it change the grounding somehow without water?  don't think so, but wanted to mention. I don't think so either.  plugged into shore power too. shouldn't matter i think. Agreed. i may get another stat to test, i need a spare anyway in case one fails while in the boonies. just not sure if i should get the one you have or something that is encased.  If its encased in plastic it won't change anything, the electromagnetic fields won't know its there, it would need to be in a metal enclosure.  most wire seems to be rated for 600volt max.  the dsi spark is around 20kv right?  how does that contribute to this?  Looks like the dsi cable is a high voltage cable, like you might have for a spark plug. That should keep it from arcing but it won't stop it from creating induced voltages in nearby circuits. That's why you sometimes get static on car radios when you turn the engine on.  I definitely need to find a different ground.  let me know what you fine. and thanks for all of your help. Looks like mine is in the same place as yours is. You could reattach it to any negative 12V wire in the trailer instead. If you have a totally isolated 12V source like another 12V battery you could connect the tstat to that and see whether that got rid of the problem (when the relay isn't connected). If it still didn't then you know that you are seeing induced voltage so twist the supply wired together and move the tstat and wires farther away from the dsi board and cable until it is stable.  PS just thought of this: the temp sensor is in direct contact to the metal tank which is grounded.  Mine is pressed up against the metal tank also, but its possible yours isn't isolated and mine is. Check the resistance between the temp sensor case and the two leads.

Don't try to measure resistance between positive and negative terminals of a power supply like a battery, you can blow the fuse in your meter. If you want to measure resistances between power terminals of a device, you have to disconnect it from all power first. If you say disconnect your water pump from the battery and measure between its positive and negative terminals that will give you the resistance of the internal windings of the pump motor, that's probably what the 22 ohms was. Perhaps interesting but not going to help in this case. Similarly the 42 ohms is probably the internal resistance of the dsi board and/or the gas soleniod on the water heater. 

Great! Glad everything is working well. Obviously there's something different about the stats. The new one you got is the same I have I think and it works fine for me as well.