BEMF, experienced in practice
BEMF, experienced in practice

BEMF, experienced in practice

Disclaimer: This article does not aim to explain how Back Electro-Motive Force (Back EMF or BEMF) works. There are numerous resources on the web about it, some examples here or here, which can help understand the matter in a model railway context.

The new locos we added to our fleet were fairly affordable, but they also needed to be digitalised. This wasn’t a major issue with either of them, they were both equipped with an 8-pin plug. For the X200, I bought a decoder specifically tailored for the small space available in the loco’s chassis. For the BR 80, I intended to use one of my spares.

I also have my second track tester sitting around, a GE44 which runs – let’s say not overly well. Research indicated that this may well be due to the factory-supplied decoder, which really only offers basic features. So I also planned to replace the decoder in this one.

You may recall from the first Track Tester Tale that I once bought an NCE 8-pin decoder, which subsequently gave me a lot of problems. At the time I thought the decoder was a write-off, however following some friendly advice on a model railway forum, I kept it around, waiting for a day that I might have the necessary equipment to test and, if necessary, revitalise it (or encounter someone who did). I also bought a cheap decoder of the LaisDCC brand to put into the first track tester, but never followed through because that loco, in the end, was a dud.

In anticipation of further digitalisation work on my fleet, I acquired a decoder tester, which has already proven its worth. The NCE decoder was, in fact, still fully functional. So the math became quite simple. The NCE was to go into the BR 80 until such time a full sound setup was purchased for it. The LaisDCC decoder is more suitable for the GE44 due to its size, the NCE wouldn’t readily fit into that chassis.

The BR 80 is surprisingly easy to open and has ample space inside the chassis. Adding the decoder was a simple job. The NCE still suffered from fragile wires, so I simply replaced its entire wiring harness with that of another, dead, decoder. While the result didn’t look overly pretty, it did the job. In it went, and the loco went on the track.

Where it sat, almost completely ignorant of any power applied to it via the throttle – until I cranked it almost half way, at which point the loco began to roll. Steady and reasonably smooth, and I was een able to dial the power back a bit – but somewhere below 20 speed steps it would invariably come to a standstill.

The motor would buzz when power was applied, but it sounded like it was overcoming some sort of resistance. I opened everything up, applied a modicum of oil and grease in the requisite spots, and checked the motor brushes. That brought a slight improvement, but it was still far from optimal. A call for help on a German model railway forum netted me a tip about BEMF.

Modern DCC decoders use BEMF as a means to provide motor control in slow speeds. Essentially, the decoder dynamically regulates the voltage applied to the motor in order to keep it running at a constant speed. Lo and behold, the NCE decoder does not speak BEMF. NCE instead have added a feature called torque compensation, which is supposed to work similarly. I had played with the settings of that feature, but to no avail.

Fearing that I might need to shell out for another fairly pricey decoder, I checked the LaisDCC’s feature set – and BEMF is included. So I quickly swapped the decoders, and with BEMF enabled the BR 80 starts properly at and runs smoothly at low speeds.

Why BEMF helped so much with this particular loco, I am not sure. It might be due to the type of motor but at the end of the day the fact that I got this loco to run satisfyingly is enough for me. Plus, I learned a few more things about DCC, decoders and what to look out for in the future. Wins all around.

The BR 80 is locked in for a sound upgrade, which will entail a new decoder at any rate. But in the meantime, I want it to run, and it will probably be busier than the GE44. So for now, the LaisDCC decoder will remain in that one, while the NCE goes into the bin. The GE44 will have to wait and sit on its siding until money for another decoder is available. Such is life.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.