Speed control

A place for discussing battery-electric locomotives, whether they're diesel, steam or even electric outline
Post Reply
SteveC
Cleaner
Cleaner
Posts: 19
Joined: Thu Aug 07, 2014 7:30 pm
Location: Richmond, Surrey

Speed control

Post by SteveC » Fri Aug 08, 2014 8:04 pm

I hope this might be useful to someone. I may be new to garden railways, but I've been doing electronics for over 50 years. I built myself an HGLW loco using their body & 4WD chassis. Even using their 2 cell suggestion the loco was just too fast! So I thought lets try RC, there just didn't seem likely to be enough room. So lets use the KISS principle and make a simple manual speed control. Most of the ones I've seen stall at low speed so PWM had to be the answer so here's my solution for what it's worth. I use a 4.8V NIMH pack of AAA cells.

http://www.tocsin-bang.info/graphics/dc ... roller.png

jay1976
Fireman
Fireman
Posts: 326
Joined: Mon Apr 29, 2013 4:32 pm

Post by jay1976 » Fri Aug 08, 2014 10:40 pm

I got a hglw loco(I have a few other bits from david as well) you can fit rc to them..but need a micro receivoir and speed controller.. I have a speed sontroller under the bonnet,the rc sits under the chassis by the rear wheels and batterys(4X AA rechargable with charging socket)which sit in the cab with a man sitting on top of the battery pack..and thats the small diesel oco with all axle drive.. I get a days running from this set up as well as slow speed,my son who is 4 can control the loco ok and the loco is pretty strong even down to going flat out into the buffers a few times when he gets distracted and falling from 2 foot high onto the grass(not his fault a twig on the line caused this) so far no problem from mine..and easily pulls 30kg of rolling stock

steampig
Cleaner
Cleaner
Posts: 79
Joined: Sun Jun 01, 2014 2:05 pm
Location: North Gloucestershire

Post by steampig » Tue Nov 11, 2014 4:14 pm

Clearly this works for you.
However, you would do better with a Schottky type diode such as 1N5817/8/9 rather than a junction diode such as the 1N400x series.
The problem is that the diode conducts while the transistor is off (to maintain the current through the motor) and is then subjected to full reverse voltage when the transistor turns on (to 'top up' the motor current). A junction diode treated this way will actually pass a brief but very large and potentially destructive reverse current (due to the lingering presence of 'minority carriers' when it was conducting), whereas the Schottky diode physics is different, there are no such carriers, and it turns off immediately without a reverse current spike.
If you find the diode (and/or transistor) running hot, or even going pop, use a Schottky type diode. They only cost pence, like the junction ones.
And have a lower forward voltage too :)

SteveC
Cleaner
Cleaner
Posts: 19
Joined: Thu Aug 07, 2014 7:30 pm
Location: Richmond, Surrey

Post by SteveC » Tue Nov 11, 2014 10:06 pm

I agree entirely about the use of Schottky diodes. I have used this circuit many times for other motor control purposes (including mechanical, Baird type TV). I've personally never had a problem, but certainly your suggestion is preferable.

steampig
Cleaner
Cleaner
Posts: 79
Joined: Sun Jun 01, 2014 2:05 pm
Location: North Gloucestershire

Post by steampig » Sun Nov 16, 2014 10:58 am

I've finally spun some numbers and withdraw any (implied) criticism. It looks like your ne555 is producing pulses at about 100Hz only, so any 'free running' current in the motor(inductance)-diode loop has probably decayed to zero long before the BD139 switches on again, so there will be no reverse conducting spike in the diode, as it is by then not conducting.

I'm afraid I was thinking of what would happen at the 25-30kHz* I use to control lego and other small toy motors, with a 1000uF cap (low ESR for choice) logically across the battery but as close to the motor loop as possible to stop inductance in the battery leads spoiling things. In these circumstances, almost the full motor current persists through the diode when it is reversed by the transistor switching on; it also happens 250-300 times more often.

I actually use a pic12f509 to drive a bc337/327 H-bridge (with 1n5817 bypass diodes). The pic just has a enough legs left to decode r/c servo pulses (or about a kilobaud of my homebrew digital r/c signal) to decide speed and direction.

I fell out of love with ne555s after trying to run one 5cm from a 433mHz 1-10mW transmitter module. No amount of lead decoupling would stop it from going bananas. Two transistor RC multivibrator - no problem.

*) Because of the unfortunate mechanical/electrical resonance interactions encountered when trying to run at low speeds with what are affectively low-rate torque pulses in the presence of backlash and a rubber-band drive! After that you go supersonic to avoid whistles and wails.

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests