X motor high current

[JWRStation]

I have an interesting fault that is present on a couple of my older loco. For info I test my locos with Nicad DC batteries set in 3/4/5 etc cells in series so as to give a fixed voltage whilst testing. I then monitor speed / current in order to determine the condition of the magnets i.e. high running voltage and/or current for a set speed then magnetism is low. However some of them when running, especially at higher voltages, have very erratic current readings i.e. it will be running at say 250 mA but then suddenly jumps to 2.5 Amps before returning back to 250mA followed by erratically jumping at random levels then settling at 250mA again.

Now my understanding of theory is that the stall current i.e. without any back emf will give the highest possible current as this is then determined by rotor resistance and applied voltage. Most of the older motor have a stall current of 650 to 750 mA with about 12 volts DC therefore 2.5 Amps is theoretically not possible.

Now I know its motor current and not a short between the tracks as I had fitted a decoder to one of my older trains which had previously given good readings only to have it blow due to excessive load (decoder rated at 1 amp continuous). It was only when I retested it rewired back for DC that it displayed the erratic current.

I did do a quick test and when it indicated 2.5 Amp I grabbed it and stalled it on the track and the current immediately dropped down to 650mA i.e. stall current.

My only thought is a short in the rotor windings but the above test showed the windings must have been reading normal when stalled. Possibly when running due to centrifugal force but if so it would be a reduction of rotor resistance from nominally 22 Ohm to 6 Ohm to get the high current and I would perhaps expect the loco speed to jump about. However, speed is constant without any change of running noise..!!!

One final thing, the magnet was not too good and when I replaced it with a Neodymium magnet then the erratic readings near enough stopped though an occasion 1 volt was seen. So rotor winding short it is, or not!!!!!

[Rana Temporia]

If the magnets are weak the motors will draw excessive current and get very hot. I have learned this through experience. The other thing to check is that the bearings are lubricated, just a small bit of oil or silicon grease makes a huge difference to old motors where the original may have dried up.

[Topcat1707821625]

@JWRStation

Have you actually tried measuring the armature resistance? The armature is connected in a delta configuration, i.e. two armature winding ends are connected to each commutator segment /_ . Take out the brushes so that they do not short out the gap between one armature winding and the next, then measure resistance between adjacent commutator segments. They should should all measure the same value, which is the resistance of one armature winding in parallel with two armature windings in series. If any reading is much higher then you have an open-circuit winding and if it is lower you have a short-circuit winding.

Another potential problem is carbon build-up between the commutator segments. The commutator should be cleaned with very fine emery paper (greater than 600 grit) and the insulation between segments should be undercut a fraction of a millimetre below the copper to get rid of any carbon.

[The son of Triangman]

Firstly neo magnets aren't good for such motors, they drag the motor shaft out of alignment causing extra bearing wear. As Topcat points out a resistance test on the motor coils using the commutator segments when the brushes are out will soon tell a story or either open circuit, short circuit or ok.

[JWRStation]

Thanks Gents....

Forgot to mention that I was (retired now) a Marine Engineer for 45 years with many early years spent on DC ships so I know DC motors both practical and theory plus spent many 'Happy' hours in hot parts of an engine room stripping and repairing DC motors.

The following therefore involves a lot of 'humble pie'. The final tests I did on the motor where with it running unloaded. The result was the faster the motor went the higher the current taken up to 20 Amps at full load. This was repeatable, a continuous value and obviously a lot of codswollap as the motor should have burnt out and it wasn't even warm!!!!!

Then, like the proverbial 'ton of bricks' my ageing mind figured it out. DC motors because they have fixed brushes make and break the armature circuit in order to rotate in one direction. Thus for a split second you have a collapsing current through a set of windings in a magnet field; just like the coil on an old points fitted car. This generates a high voltage spike whose value depends upon the speed of current collapse i.e. speed of rotation; reason for fitting a capacitor to prevent this spike interfering with TV's etc. Through a lot of testing, this spike must have broke down the 'smoothing' circuit in the multimeter thus it was reading the voltage spikes as a current and thus as the motor speed changed it read increasing current...!!!

I checked the meter on a loco I had tested a couple of months ago which at the time had stable readings, sure enough the ammeter readings were also all over the place. Double checked with my expensive industrial specification Fluke 77 meter and that was straight as a die.

So the old adage of results are only as good as the equipment certainly held true and I should have clocked this earlier but the Decoder failure did confuse the issue.

On that subject maybe it failed for the same reason with high level voltage spikes because the recommend is that capacitors are not needed when decoders are fitted!!! Perhaps not so, especially on older motors running at relatively high current. I will definitely be refitting capacitors back on my loco's.....

You never stop learning, or perhaps more specifically, remembering what you once knew ?????????

Joe

[96RAF]

See Brian Lamberts web pages for how to carry out a stall current check of your motors, with the ammeter in series with the motor, making sure your multimeter can read up to 2 amps DC current. Many meters top out at 200mA and trying to measure stall current pops their internal fuse. Other meters have a 10A DC current range through an internal shunt.

[Fishmanoz]

Not sure your theory on needing a capacitor is sound as decoders include a suppression function as part of their circuitry so the external capacitor is not needed. If the decoder can handle the motor current, it should be able to suppress the spikes as well without overload.

[Topcat1707821625]

@Fishmanoz

I'm not sure I understand your comment as I don't see what a decoder being able to handle motor current has to do with it suppressing voltage spikes as the two are not connected in any way.

I'm no expert on DCC decoders but the output stage that drives the motor will probably be a MOSFET full bridge (two N-channel MOSFETs and two P-channel MOSFETs).

The motor will be will be connected between the commoned drains of of S1/S2 (D1/D2) and those of S3/S4 (D3/D4).

When S1 and S3 are turned on the motor will turn in one direction and when S2 and S4 are turned the direction of the current fed through the motor is reversed and it will turn in the other direction. Motor speed is controlled by altering the duty-cycle of the pulsed drive signal. Each MOSFET has a diode connected between source and drain to clamp any reverse voltage that might damage the MOSFET, hence no need for a suppressor capacitor, except maybe a very small one to suppress radio frequency interference.

Another reason for not using a suppressor capacitor is that some DCC decoders switch at ultrasonic frequencies. A capacitor has an impedance that decreases with frequency, so the higher the switching frequency of the decoder the more current will flow through the MOSFETS just to charge and discharge the capacitor, causing more power to be dissipated, and thus wasted, in the MOSFETs.

Current handling capability of the decoder has absolutely nothing to do with suppression of voltage spikes from the motor. It depends on the spec of the MOSFETs, their thermal resistance, are they mounted on a heatsink?

[Fishmanoz]

Hi Topcat, thanks for confirming that RF suppression is built into decoders and therefore no parallel capacitor (or series inductor) is needed to achieve it. The point about current draw at decoder PWM frequency is well made too (ultrasonic?).

It’s now some 55 years since I sat in class learning that lot. An interesting time to learn electronic engineering with valves still common along with discrete transistors and very early ICs. Then, having been called up during Vietnam days, seeing that full range of technologies used in one item of military kit (counter-mortar radar) with the level of sophistication/integration running to just one flip-flop on its own circuit card.

And yes, I realise there is there is no direct relationship between output power and ability to suppress RF without its destroying the decoder.

PS. For the mods - just noticed something about TC’s presumable accidental repeat of his post via the blue button - the forum software no longer places the extra carriage returns between the paras when you do this.

[Topcat1707821625]

@Fishmanoz

Yes I boobed there when posting my reply. That blue button with the arrow is so tempting and exactly the same symbol as used for replying in email programmes.

Coincidentally 55 years ago was my first year at university. I think the only solid-state devices we learnt about were diodes and bipolar transistors.

[96RAF]

@TC

An H-Bridge is the standard decoder motor control method.

@Fishy

Your formatting observation reported to Hornby. Maybe if we charged per click folk would stop using the troublesome feature.

[Brew Man]

Maybe if the troublesome feature was removed folk would have no option but to not use it. It does seem to take forever for Hornby to take notice of mod requests.

[Topcat1707821625]

@96RAF

I agree entirely with Brew Man. You scroll down to the bottom of the page and see the blue arrow, which is the symbol for "reply" in 99.99999999% of apps and below that the page number (which is what you expect to see at the bottom of a page), and/or Previous Page - Next Page. Half the time you may not even see the "Post Reply" window.