Dear All,
Advice please; what controller do you recommend for 'modern' 009 mechanisms, in particular the Minitrains Feldbahns and Alcos? Is anyone using Gaugemasters? I assume then 'N' Gauge setting on some controllers is the most appropriate?
Kind Regards,
Andrew Emery
Crimea,
Russian Federation
This is not, I stress, a scientific test but I have just tried a Gaugemaster Combi versus a Minitrains controller using a Minitrains Gmeinder for a layout I am building.
In my humble opinion the Minitrains was the winner because it allowed much lower continuous speeds and less risk of stalling. This was a layout using four inch curves and the Gmeinder worked a treat with the Minitrains but was much less stable with the Combi.
That is what I have chosen to use, It is a pulse width modulated controller rather than straight DC and it seemed to me to do much better for the industrial setting I am modelling.
Are there any similar benefits to using a 'pulse' controller for run of the mill OO9 locomotives fitted with GF chassis and newer models like the Skarloey as well as my Minitrains Baldwin?
Doug
Following a query by Harry Mantheakis to test his Canadian-made Varipulse controller (http://www3.sympatico.ca/kstapleton3/851.HTM), I have recently compared a number of controllers, using a variety of chassis.
I made a simple reference controller that produces either a flat output voltage, or a half-wave rectified output with a 50Hz sawtooth on two settings - the kind of thing we did 30 years ago. The diagram is here if you want it: http://www.009dutch.nl/oddphotos/simple_controller.jpg (http://www.009dutch.nl/oddphotos/simple_controller.jpg). Since I don't have an oscilloscope, all I could measure was the effective output voltage reproduced on an electronic multimeter - the mean voltage of all its samples within the sampling period.
Here is a photo of all the controllers tested. (I later added similar tests of the Minitrains PWM controller):
(https://forum.009society.com/proxy.php?request=http%3A%2F%2Fwww.009dutch.nl%2Foddphotos%2FIMG_2257.JPG&hash=fb8d78abac629547b24a1492c3ee0f0b57b49f22)
I spent a few days noting down results which may shed some light on this. The test was discussed on the NGRM forum. The report is here: http://www.009dutch.nl/oddphotos/controller_test1.pdf (http://www.009dutch.nl/oddphotos/controller_test1.pdf).
Here follows a summary of my conclusions:
Grafar 4MT - old style chassis (T&D Hunslet)
The Grafar 4MT old style chassis performs surprisingly well on all controllers but it has a Tri-ang like growl. It has a five pole motor that goes slowly on all controllers, but it needs keeping down as it has an excessive top speed. Note that the best performance coincides with the lowest effective starting voltage.
Minitrix dock tank (pugbash)
This is definitely the worst chassis tested. It needs feedback control to be kept in check during starting and crawling. It is noisy and has a high starting voltage. The only controller it works reasonably well on is the KPC, with full feedback.
Fleischmann 0-6-0 industrial diesel (tender drive)
This is another chassis that needs a good controller, although it is far better than the dock tank. Its main problem is unsteady low speed
Interesting study Ted!
I agree that you would need an oscilloscope to see what really happens - but even with a nice picture of the saw tooth diagram of a pulse-width controller it is more important what the locomotives are doing with it.
I played with pulse-width and pulse-frequency controllers in the late 70th - and for me there was a big difference between a normal DC controller and those pulse-width circuits.
Regards
Gerhard
Sticking my oar in - I use my own home-made controllers - because I like to dabble in the dark arts!
My current preferred controller is a PWM device that runs at around 20kHz pulse repetition frequency and based on a modern version of the 555 chip. The Minitrains controller is also a high frequency PWM which runs around 16-19kHz if I recall.
The received wisdom is that low frequency PWM is not good for "modern" low inductance motors, due to heating. (Low frequency means at and around mains frequency). On the other hand, there's nothing against using high frequency PWM on "old" motors - so that's a win-win (possibly)
There's quite a bit of interest in these cheap PWM motor controllers you can get via eBay, but I have not tried them, so cannot comment one way or the other.
Among my controllers I also have a pure dc controller based on a three-terminal voltage controller; this is also a good runner.
If you are interested in my controllers, there's a link from my web site address (below)
REMEMBER there's no excuse for dirty track and wheels!
Essentially we're repeating a similar discussion on NGRM. I agree that a high frequency PWM controller is kinder on most motors than low frequency pulses.
Hugh - what I would be most interested in is whether your 20KHz controller handles long track lengths such as on my layout. In a given situation there may be as much as 30 metres of track linked to the controller, and all the associated wiring, in which no provision has been made to keep the live and return leads apart. The capacitance of the total may have been a factor in the poor performance of the Minitrains controller on my layout described on NGRM last year.
Therefore I tested it on Nixnie to prevent capacitance from playing a part in my test. But I found no real improvement on the flat DC control performance. It would be interesting to see what happens there with a scope.
In the NGRM thread several interesting controllers came up. One was a design by a now-defunct company named Tractronics, of which a full spec can be found here: http://webuser.hs-furtwangen.de/~hoenig/2012/Karlsruhe2012/CC.pdf (http://webuser.hs-furtwangen.de/~hoenig/2012/Karlsruhe2012/CC.pdf)
This operates on 60Hz (American mains frequency) and produces the following wave output form:
(https://forum.009society.com/proxy.php?request=http%3A%2F%2Fwww.009dutch.nl%2Foddphotos%2Ftractronics_graph.jpg&hash=deb07e9ac2c45648420fb146692a87337cd12ce7).
At low throttle, only the tops of the rectified sine wave appear above the 'horizon' - at 120 per second. And note that at a medium setting, the voltage is reduced (the recified sine wave has a lower amplitude). The article claims that these controllers won't generate heat at long periods of shunting. Not sure what they would do to a modern motor.
This is fascinating. Thank you to everyone who has responded so far, and to anyone who has more to add. My next challenge will be to get a suitable controller sent to me here in the Crimea, bearing in mind that the UK is obeying Obama's instructions to the British Government that they must sanction us. There's always a challenge ahead.
I never guessed you're that far from home! If it's difficult to get a controller out there, maybe it's easier to source the parts and build one yourself. There must be some electronic parts supplier out there. The controller I built is a very simple one. Not ideal, but it works.
(https://forum.009society.com/proxy.php?request=http%3A%2F%2Fwww.009dutch.nl%2Foddphotos%2Fsimple_controller.jpg&hash=62a4c86fc4726be51d628b48430c5c6502dfd5de)
(https://forum.009society.com/proxy.php?request=http%3A%2F%2Fwww.009dutch.nl%2Foddphotos%2FIMG_2251.JPG&hash=5a8a9b5cd530ef6b31b3bf85a7af9f51e12256b5)
(https://forum.009society.com/proxy.php?request=http%3A%2F%2Fwww.009dutch.nl%2Foddphotos%2FIMG_2254.JPG&hash=03a8f8d3cc63a382a237ab85d1a1dc8a9b1e4ba6)
Thanks Ted! Much appreciated.
Andrew
Thanks Andrew. A few additional notes:
Actually, S2 in the diagram, which switches in the full wave rectification (it completes the diode bridge) and the flattening capacitor C2, is a centre-off DPDT switch. In centre-off mode it switches out one diode and the big capacitor. The result is a saw tooth due to the half wave rectification and C1 (47uF).
When pushed through the centre, it will switch in a second capacitor parallel to C1, which has a bigger value (220uF). This slightly flattens the saw tooth produced by the smaller value of 47uF. Some motors like that better.
Note that this is not an advanced controller, but it works moderately well. I ran my layout using two very similar controllers (only using now obsolete transistors) for many years.
Having read back through the thread, and after a question by a forum member, I decided to test a few selected chassis on the four controllers that were the subject of this test, and make a short video of them all so you may get an idea of the performance of each controller. Probably that will give a better idea than that list with technical data.
First, I tested the simple DC controller I built myself, set at full wave rectification and full smoothing - the output is supposed to be almost flat, like a voltage regulator. The Minitrains and Kato-fitted locos perform well considering the type of power they are fed, but note the jumpy behaviour of the final test loco, which has a Fleischmann 7000 chassis:
https://www.youtube.com/watch?v=FZt3QrQn6zY (https://www.youtube.com/watch?v=FZt3QrQn6zY)
Next, the Minitrains controller, which I tested on Nixnie to prevent interference by a great track length and a lot of wiring. The modern locos perform reasonably well, but the Fleischmann 7000 again doesn't like this controller - it picks up suddenly, then flies off:
https://www.youtube.com/watch?v=9IxuPsTCab8 (https://www.youtube.com/watch?v=9IxuPsTCab8)
The next one is Harry's Varipulse controller. The inertia control takes a bit of getting used to, but it can be set to zero if you want to. Note the vastly improved performance at low speed running and crawling.
However, the Minitrains Egger replica 0-4-0 is slightly sticky, the Plymouth suffers from slight cogging and several locos, if closely observed, show slight defects in their transmission. Not exactly binding, but they have slight stiff points. The Fleischmann runs far better than on any of the previous two controllers:
https://www.youtube.com/watch?v=ltnHk3BFba0 (https://www.youtube.com/watch?v=ltnHk3BFba0)
Finally, one of my trusted and slightly modified KPC controllers. I fitted it with a series resistor which somewhat extends the slow shunting range on the control potentiometer. At first sight there is little difference with the Varipulse. However, if observed closely, the feedback loop keeps the speed in check on virtually all the chassis. There is less jerking as the wheels complete one revolution, although the Plymouth suffers from cogging more than on the Minitrains controller:
https://www.youtube.com/watch?v=rnqf_toddcM (https://www.youtube.com/watch?v=rnqf_toddcM)
Here is a final video testing the Varipulse against a KPC switchable feedback controller. This is quite a bit longer than the other four, to the point of being boring. However, it is possible to see the effect of feedback control on a loco as opposed to straight PWM control without feedback.
https://www.youtube.com/watch?v=EGgGeYXdIAs (https://www.youtube.com/watch?v=EGgGeYXdIAs)
This video shows a test series of a fair number of my locomotives, run on the Varipulse PWM controller, then on the KPC switchable feedback controller. Each time the performance on the Varipulse is shown first, then a banner flies by saying 'KPC' which means the second controller is taking over.
Locomotives tested:
00:10 - Backwoods 'Atlantic' - very smooth on both controllers.
01:00 - Fleischmann 7000 chassis used under 'Greta'. Generally smooth and quiet.
02:00 - Arnold 0-6-0 (ex 1970s BR80) used in the Darjeeling B class. Noisy and cogging compared to the other chassis.
03:30 - Fleischmann 7218 0-6-0 diesel used as a tender drive. Also noisy and cogging at a crawl.
05:00 - Liliput D12 jackshaft diesel used under 'Lyn' - very smooth.
06:05 - Minitrix dock tank (pugbash) - note wheelslip due to cogging under both controllers.
07:30 - Kato 11-103 used under 3D printed tram loco.
08:37 - Minitrains Brigadelok.
10:00 - Minitrains 0-4-0 Baldwin saddle tank - note severe binding in chassis, almost gone using the KPC.
11:09 - 1960s Arnold 2-6-4 re-fitted with Mashima 1220. Slight cogging but very smooth on both controllers.
12:35 - 1960s Arnold 2-6-4 with original and very noisy motor, cogging a bit.
13:55 - 1970s/1980s Liliput 0-6-2 chassis, slight binding on Varipulse, but otherwise surprisingly smooth.
15:00 - Fleischmann BR56 tender drive re-fitted with Mashima 1015. Slight cogging, tends to suddenly unstick from a crawl, less so on KPC.
16:40 - Graham Farish 4MT 2-6-4 used under the Tralee & Dingle Hunslet. Slightly binding on the Varipulse, but not bad.
There are several detail differences in performance between the two. The Varipulse has no feedback control, but has an adjustable inertia setting. From close observation it seems most chassis also run a little quieter on the KPC. In addition, locomotives with a chassis defect (binding of the rods or the gearing, internal resistance, of which I have a fair number) run better on the KPC. The feedback loop reacts to the motor slowing at a binding point in the chassis, or if the loco encounters more friction, giving a quick boost to the motor.
Summing it up, the Varipulse is a good second, I would say not a bad buy at all, and it has inertia control if you want it. But I'm still very happy with my KPCs...
I wonder if it would be possible to make a controller where the frequency of the pulses got higher as the speed was increased?
The idea being to give the "kicks" at low speed required to get "difficult" motors to start, but going on to the "better" high frequency pulses when we've got the loco in motion.
I shall have a think about this.
There was such a controller in the past - impulse frequency controller. It produced a rectangular DC signal created by a Schmidt trigger. The lower end could be set to a minimum frequency, then you could change the frequency up into the kHz range. This brought the single rectangular impulses so close to each other that you get more or less a DC signal. Problem was the noise. Every engine worked also as a loudspeaker... This was a circuit I found in a German book - which I lost during the 10 times I moved house. I built it and tested it in the late 70th. I also put the output to my fathers stereo set.... :-)
Will try to find the book later on the net - today everything is available on Google ....
Quote from: hughn on February 19, 2016, 11:42
I wonder if it would be possible to make a controller where the frequency of the pulses got higher as the speed was increased?
The idea being to give the "kicks" at low speed required to get "difficult" motors to start, but going on to the "better" high frequency pulses when we've got the loco in motion.
I shall have a think about this.
Not a bad idea, Hugh. If it can be switched on and off so it won't damage coreless motors you'd have the best of both worlds.
Will you come to Sparsholt? If so, we could try and hook up one of your PWM controllers to Dunalistair.
Gerhard: your father's stereo set ??? ?? What did he say? ;D
Quote from: tedp on February 19, 2016, 14:22
Not a bad idea, Hugh. If it can be switched on and off so it won't damage coreless motors you'd have the best of both worlds.
Will you come to Sparsholt? If so, we could try and hook up one of your PWM controllers to Dunalistair.
I have a diary clash at the moment for 9th April...
Just in case I can get to Sparsholt, what type of connector do you use to connect your controllers to your layout?
DIN 270 degrees, looking into the rear end of plug (where you solder the leads) the AC input is on the left, the centre one is unused, the two on the right are for the DC output.
Found the book - as I said Google finds everything... Modelleisenbahnen elektronisch gesteuert. Band 2: Impulssteuerungen, NF - Zugbeleuchtung und Peripherie - Elektroniken. Author Knobloch, Winfried. ISBN 10: 3790502952
For those who are not so familiar with German: Model railways - electronically controlled. Part 2: Impulse control, LF train light, periphery circuits.
@ Ted: Well my father didn't hear it as he was working... It is fun to play with a frequency generator and create all kind of sounds.