I love to tinker with components. Now I'm going to tinker with an ancient 3021 diesel with an experiment. My "null hypothesis" is: "Converting the LFCM motor with a 60944 5 pole kit yields no more smoothly running effects than retaining the 3 pole armature and adding a permanent magnet, when using the same decoder." I'll post results. But the reason I'm posting in the first place is to ask y'all: How would such an experiment be set up according to research methodology? Kind insane to ask, I know, because you'd have to convert a huge sample number. I guess I'll "know it when I see it" in terms jerky running at slow speeds.

Thanks for listening,
Tom

Thanks,Dale. It was exactly these kind of parameters I was looking for. I guess I'll have to list them all, e.g;

-Brand new 60944 kit with 2 motor covers and a 5 pole armature
-Brand new permanent magnet from ESU, retaining existing 3 pole armataure.
-128 speed steps.

Background reasons for this experiment: I always wanted to take math, statistics, chemistry,engineering, and physics in high school, college, and grad school. They're so interesting and the only way we have to "prove" anything. But I dare not take them, cause I'd fail! I just struggled so hard. My mind tended instead towards writing, literature, and English. But NOW I can experiment without ruining my grade point average, 40 years later. (I just wish my twins had somehow inherited math aptitude instead of my liberal arts aptitude.)

well, I can tell you not many people even know what a Null hypothesis is.... refreshing to see mention of one! So don't sell yourself short! Good engineering also includes the ability to document and describe the results... so your writing is also critical....


Does the 60944 decoder use back EMF ? (I always installed ESU decoders as they were always ahead of mother M)

I suspect the firmware of the decoder is what is most important.

Aha! Thanks, John, for bringing up ANOTHER consideration. And how to "control" for this one, I'm not even sure. To your knowledge is the new v 5.0 ESU decoder "different" from the v 4.0 ESU, in how it handles the slow speed syndrome? Also, I thought of another pesky element that could confound the results. If I install a permanent magnet in the existing LFSCM universal motor, and keep the 3 pole armature, I haven't changed anything except the armature. On the other hand, if I install the components from a Märklin 60944 conversion kit, I install a new motor cover as well! That's a "confounding" variable. Rats.

I'm beginning to think that I need to tweak my null hypothesis to the following: "In a digitally controlled scenario, converting a Märklin LFCM motor with a 60944 5 pole kit does not absolutely guarantee smoother running effects than would retaining the 3 pole armature and adding a permanent magnet. Given enough time and patience by the operator to choose a satisfactory digital decoder for the 3 pole motor and enough time and patience to adjust CV's for optimum slow running, the supposed advantage in the 5 pole solution for slow running will not be observable or measurable."

WHEW! That took a while, but it's better. One thing that worries me is the tension on the motor covers. I've read that too much tension can affect smooth operation. I'll have to be careful to adjust both covers to optimum tension.

Thanks for listening,
Tom

Tommy, I'm not an expert in electronics and every time we have an update of a decoder more gadgets or programming options you have. I was happy with the V 4.0 decoder and to be honest I don't know the benefits a new V 5.0 brings. to me its all more confusing but ESU reckons you can tinker with motor control CV's even more, I haven't been able to proof or disproof any improvements.
your decoder will decide which motor will run better at low speed and unless you're fortunate to have a lokprogrammer changing CV's without a Lok programmer, I'm hopeless.
I've tried to run an ESU perm magnet + 3 pole armature + V 4.0 decoder and had no luck in getting the loco run smooth at low speed.

John

First results of this experiment are in and they are surprising; and I haven't even gotten to the 5 pole yet. A Loksound 5 decoder, ESU #58429, produced extremely poor results with the converted Märklin motor. Autotuning made no difference. There was extreme cogging/jerkiness. It was so bad I was suspicious. So I removed the 58429 decoder and installed a brand new 58513 decoder, the
Loksound 5 XL for large scale. The result was excellent running at step 1. I didn't even need autotuning. It was really surprising and it rendered my original experiment somewhat moot. I saw that a 3 pole rotor, in and of itself, is no impediment to smooth digital operation. Period. But why did the XL perform perfectly as soon as I applied power, but the 58429 did not? I had assumed that vs 5 firmware is the exact same across all decoder sizes, from micro to "regular" to L to XL.

Strange. Any comments?

Tom