I had a HSV slowly deteriorate on me till eventually it was fully evident. I thought my tune was chaniging over time, and had to start dumping more fuel as a/f's were leaning out. Put a new HSV in, and I was pig rich. It took a couple months to go from normal to erratic to total failure

Just another option in addition to what Richard has listed. Doesn't explain why it occurs in only cold weather, but mine was a bit erratic to begin with before total failure. If you're on straight meth with the age of the kit, it's something to consider.

 

This is another possibility, but not strictly temperature related. We can also add a temperature compensated circuitry to add more duty cycle to the HSV.

Do you have the failed valve? Perhaps ypu can send it to me for close examination.

 

Richard,

I did send it back to you, it was about this time last year. I never heard what you learned from that examination??? It came from Houston, Texas.

 

I will look back.

 

1- can be tested, lets work on a way to try this. I'm willing to buy some hardware if needed.
2- somewhat unlikely due to being in the engine bay, pretty warm in there...could test this somehow I bet, let me think about that one more
3- even with increased viscosity, would the flow be lowered? Not necessarily, if the pump is still up to the task. Pressure would likely rise though, which would go back towards something with #1
4 - Interesting! I am intrigued by this one...I have the valve in the cabin though, so it would have to be really really cold, the reduced flow does not get better if the car is driven on a long drive and really warmed up (I will be mindful of this and verify)
5 -not the case as I've seen the AFR lean out on my wideband, about half of a point.

 

Due to the relatively large coefficient of expansion of plastic, could the tubing be shrinking smaller in one of the fittings between the tank and pump allowing a partial loss of prime of the pump? The symptoms sound like other I've seen that have a priming issue, but yours is related to the cold temps. Maybe the vacuum of the pump is drawing air in somewhere, reducing flow? You check for small bubbles in the line before the pump.

 

I have a HFS-1, so this can't be the problem, can it, unless it is simply not opening far enough.