I did better than that, Dave. I was just hoping someone had a quick answer.
I first tried the 10A input on my DVM but must have blown its internal fuse at some point in the past - I need to check that now too. I then put an amp clamp on it so as not to blow up my DVM. It checked out around 200 mA. This was pushing the low end of the amp clamp in its lowest (40A) range. But, as it was "safe" at less than my DVM's V,Ω,mA input limit of 320mA max, I was able measure the current directly.
The stack I checked had a steady state current draw of about 135 mA at 18 VDC with an initial reading around 180 mA. The stack started to smoke around 10 - 12 VDC, with decent smoke at about 16 VDC and excellent smoke at 18 VDC. Current draw was about 100 mA at 12 VDC. The resistance must increase quickly as the stack heats up. It measures around 33Ω when read directly with the DVM. A LGB 5006/5007 combo was the power source for the testing.
For those who may be curious, I'm planning to use an ESU 54610 LokPilot V4.0 Multiprotocol decoder and 54670 "Power Pack". The specs are 1.1A continuous for the motor, 250 mA max per function output with a total function output of 280 mA. (FL, RL, F1 and F2). I would have used a LokPilot XL but wasn't sure it would fit. It was a challenge soldering the three leads for the 54670 to the decoder (fat fingers, unsteady hands, poor eyesight, and tiny wires and connection pads) but I think it's OK after checking with a 10x loupe. I don't believe I fried anything by applying too much heat but the proof will be in the testing!
Phil S.