@Adriatic posted:I think the supply style of the new unit isn't compatible with the design.
The constant pure sign of old causes the thermal coil to react better than a pulsed or "on demand" version because it has less real on time per second for power regulation. Pure unfiltered sine is more constant (Pulse is useful in command and some motors). (I'm not 100% sure exactly what's used on the new, but its its more involved than a rheostat/sliding tap, for sure.)
Results are the same whether I use a ZWC or a ZW
Once the loco is railed, more current is used leaving less to throw at the thermal sw. Are you pushing the transformer limits via lights and accessories? Even the Zs have limits.. Look at the roller spring and that assembly. Weak springs/rust can limit current draw ability.
ZW is hooked ONLY to the test track with an old steamer and the subject tender - definitely not running it to the limit
Search connections for unusual heat once it fades. Heat degradation of a contact is possible too. Flush the roller with wd40 or CRC electric spray, etc.
Since the tender is stationary in my test setup, I did jumper the contact arm (not the roller) directly to the center rail - same results.
I think you could fire that via a capacitor discharge circuit for better timing if you had to.
This is for a friend, not sure he would want to invest the extra gold nuggets $$
