The only way to measure the track voltage on a digital system is with an oscilloscope. The wave shape and the gaps in the waveform that the mfx system uses for the reply messages will foul up any DVM meter reading. Using a true RMS meter will also not give a true reading for the same reason.

Sorry if I sound a bit too basic and simplistic, but the first thing to check is how much smoke fluid you inject in the generator...

2 ml max at a time. No more.

I measured MS1 output using oscilloscope. It is 19.6 volt peak (39.2 Vpp).
While the output of MS3, which I believe should be designed the same as CS3, is 18 volt peak (36 Vpp)
MS1 produces higher voltage than MS3, although the difference is just a mere 1.6 V
As per my experience; 16-22V Seuthe smoke generators (no.11, 12, 23, 24) are fine with MS1, but some may not perform well with MS3. The 10-16V Seuthe (no.10, 20, 21, 22) perform very well with MS3.
Do note; in digital operation, the smoke generator is not directly wired to the rail, but getting the voltage from decoder, therefore the actual voltage (and current) going into the smoke generator will depend on the decoder as well.

(not producing much or no smoke)

Little or no smoke have nothing to do with track voltage.

Check CV settings or filling procedure.

By the way - CS3 60VDC has 18,4 V on track and 18,3 V when one loc is running.

Torstein

It depends on the track voltage and on the shape of the track voltage.
Make sure red and brown wires are not swapped - and red wire goes to B.
MS2 produces a lower track voltage than the CS2/CS3 because the MS2 has a bridge rectifier on the input side, the CS2/CS3 does not (the bridge rectifier in the CS2 is bypassed when using DC power). And power supplies are 18 V for MS2 and 19 V for CS2/CS3. The difference of 2+ V track voltage can be relevant.

For a two-rail user with an MS2 and a Trix loco the issue was resolved by adding a dummy MM loco to the loco list. With the dummy MM loco the track signal became asymmetric and the smoke generator received more power, but the absolute track voltage did not change.

Depends on the smoke generator.
For H0 gauge Seuthe generators #10, #11, #20, #24 the limit is 0.2 ml, not 2 ml.

Power is one thing, voltage another. IMHO an external booster should not improve the situation as long as the internal booster of the controller is far from the limit and fed with a switching-mode power supply.

I gave up smoking years ago lol

I have seen Seuthe smoke generator works better with ac power on the track.
With digital decoder it translate to analog power to the smoke generator.

Sure you can measure track voltage with a digital multi-meter - if the multi-meter has True RMS with a sufficient frequency range. Multi-meters for ten bananas do not have this feature and their AC range only works for sine-wave AC.

SOLVED – A big thanks to HO (from NRW wherever that is) for his input on this issue. Considering how complex the problem could have become, it was a matter of experimentation with the track wiring. Yes we had reversed the red and red/brown wires and by changing these to the correct format of RED to B on the underside of the track, like magic our smokers sprung to life and choked the room with smoke. Yaaaaaaaaah what a relief. But as suggested, the old analogue smokers are the ones that are working the best. The #11 still does not work, in the Big Boy, so it now has 2 x #10 and wow can it smoke.

We then set too and reversed the polarity on all our 61 turnout decoders. The m83 under table units were easy, but we had to pull up 18 under track turnouts to swap these also as these are the old model and polarity sensitive. This is probably were we went wrong in the first place as we have had to do this 3 times now as we progressed from MS1 to CS1 and CS2 and now onto CS3. Wecwere lazy and probably swopped the wires instead or rewiring all the decoders again. Each time we have had issues with this polarity problem!

But now I ask another tricky question. I thought I was still driving Marklin AC trains and therefore AC driven decoders, which to my level of knowledge should not have any need for polarity sensitivity, as AC has no polarity!! If they are actually AC how are they polarity sensitive? Or has Marklin now actually gone DC with all its polarity problems that AC avoided in the old analogue days?

Anyway, this is the first thing to address with any smoker problem. We now spend more time filling smokers than blowing whistles and its great. At last we have a use for the 1 litre Seuthe fluid we bought last year.

Thanks again for everyones comments, but especially HO.

My simplistic understanding (someone correct me if I'm wrong) is that digital is DC but with some imbedded pulses (i.e.: not pure DC) to send information to the decoders. When converting Maerklin AC locomotives to digital, you take out the old field coil and replace it with a permanent magnet which actually makes it a DC motor now. With all the semiconductors in the digital decoders etc., polarity is now important. Using the term AC to describe Maerklin (or other manufacturers') components currently seems to indicate that you are actually dealing with a three rail system, that is to say the locomotives or rolling stock requiring power have a slider/pick-up shoe. Whereas DC means two rail. The big advantage of three rail, in my point of view, is the ease of creating reverse loops (but two rail has modules that deal with that too).

Cheers,

Harold.

We are going a bit off topic but one advantage with three rail is the extra rail for the return which helps with interior lighting although we rely on the pick up for the other power from the track and this also needs to be cleaned. No doubt reverse loops is a plus without the need for extra wring etc.

I use only the Seuthe #10 or #20 for all my locos using my MS2. The digital versions just don't give enough smoke for me. I use the smoke function continually and also have some older locos where the smoke unit is wired directly to the track. They're never off, yet I've never had a smoke unit burn out on me. Maybe I've just been lucky.

Chuck