Scalefour Forum

Having consigned very old, rather dodgy, units to recycling facilities during a house move, I now want to renovate my test track and connect it with a new DC controller. As this would be for loco testing only I would prefer pure DC with no feedback or such sophistication - adding a couple of cheap analogue meters would seem sensible - twitching needles can give a lot of information! (All the meters I've seen are now 0 to something, rather than minus something to plus something, so the supply would need to be 0 to +12V and the meters followed by a DPDT switch.)

I've read the "Which Controller?" thread https://www.scalefour.org/forum/viewtopic.php?f=26&t=1328&start=25 but quite a lot has changed in the last few years (Maplins gone as well as Pentrollers etc. and DC chargers are now much more common).

Does anyone have more up-to-date suggestions?

Off the shelf, I'd probably get a Gaugemaster panel unit for what you propose, then add a direction switch down-stream to solve the meter issue. The one I used to really like isn't available anymore, it had two knobs, one normal "speed" knob, and another which altered the overall voltage range.

( I'd skip the Pictroller based on my evaluation of them. There are a number of serious annoyances in the way it works which makes it pretty much useless for a test track setup. ).


DIY, might look at doing something with an Arduino and a motor drive H-bridge. That could be coupled to various input controls, including a speed potentiometer, but also buttons/switches which could alter the nature of the output. Won't be pure DC, the H-bridge will be doing high frequency PWM, though depends on the design chosen as to whether there is any feedback adopted. Power for this from either a DC power brick, or some batteries if portable is required.

Ammeters - choose carefully, or arrange appropriate current shunts around them, so they give as much needle deflection as possible. I find an ammeter to be extremely useful in diagnosing the cause of tight spots in loco running.

I usually test with an old AMR which has had the feedback reduced before purchase. Often though I check with a Gaugemaster handheld, the cream coloured one, not the black. I also use a Modelex, recommended to me by Roy Jackson, and this runs engines very quietly. Not goo much fancy electronics on the last two, I think, so good for testing.

Philip

Dc controllers will often turn up at exhibitions on second hand stands and provided the mains supply is OK - assuming that is supplied, will be a perfectly adequate source of control. You may even find something on the Bring and Buy stand at Scaleforum.

Terry Bendall

I use a smooth DC controller based on a LM317 chip
http://www.ti.com/product/LM317?utm_sou ... Lv-fD_BwE#
circuit from MRJ No. 11. I use this in association with an analogue ammeter. This combination sorts out most problems of stickyness/poor pickups etc.

Gents,

Many thanks for the informative and useful replies. Plan now is:
1. extend the shopping list for Scaleforum - if this fails
2. get detailed info on Gaugemaster and Modelex units - if this approach fails
3. get a copy of the MRJ article and then DIY.

Third possibility seems the most likely at present, but the least favoured as I’d rather spend time on models than on ancillary bits.

I appreciate you want something plain but you mentioned Pentrollers earlier. It's successor, the Pictroller is still available and you can turn off the braking.

See Malcolm's Miniatures.

Tim V wrote:I use a smooth DC controller based on a LM317 chip
http://www.ti.com/product/LM317?utm_sou ... Lv-fD_BwE#
circuit from MRJ No. 11. I use this in association with an analogue ammeter. This combination sorts out most problems of stickyness/poor pickups etc.

This would be a better version to use for amateur application:

https://uk.rs-online.com/web/p/linear-v ... s/1686809/

regards

Alan

Tim V wrote:I use a smooth DC controller based on a LM317 chip
http://www.ti.com/product/LM317?utm_sou ... Lv-fD_BwE#
circuit from MRJ No. 11. I use this in association with an analogue ammeter. This combination sorts out most problems of stickyness/poor pickups etc.

Also note follow-up comments and a letter in MRJ 13 which adds a few details to the article in MRJ 11, and corrects an error in the circuit diagram (orientation of a capacitor).

David B wrote:I appreciate you want something plain but you mentioned Pentrollers earlier. It's successor, the Pictroller is still available and you can turn off the braking.

See Malcolm's Miniatures.

I don't see the Pictroller as a "successor" to the Pentroller. Its only a successor in the sense that it occupies the same physical shape, with the exterior connections in the same place, and the same case shape for handhelds. The Pictroller's control of motors seems to be very different to the original Pentroller design.

The East Anglia group owns several Pentrollers. I arranged a group evaluation of a Pictroller and found that it causes "burps" to a loco, making it move when setting direction, prior to moving the speed control from zero. ( That alone was enough to make the device unacceptable for a layout which requires locos to stop and start in public view. This behaviour is reported by others on various reviews. ). Low speed control wasn't as good as we'd have wanted. It was tried over a range of motors including: Mashima, some modern RTR Hornby/Bachmann, and at least one coreless motor of Swiss origins. The end result was deemed to be unacceptable by the group, and other controllers were investigated.

From that evaluation, I would say only try a Pictroller if there is an option to return for a full refund if it doesn't meet your needs.

I have an old Hammant Morgan controller from my youth which I have been using to run motors in etc. Is this a bad thing? Using the Protocab system I have tended to use this to bed everything in prior to connecting up to the RC setup.

H&M controllers are pretty basic and will not be kind to motors that expect smooth DC. There's a thread over on RMWeb where someone tried the "DC" output of an old H&M with an oscilloscope and it had gone completely mutant, nothing like DC at all. I think there was even one phase where the voltage went negative. Granted that was a duff unit where one side of the rectifier had failed, but how unbroken is your old box? If you don't know and can't test, I'd suggest replacing it.

Guy Rixon wrote:H&M controllers are pretty basic and will not be kind to motors that expect smooth DC. There's a thread over on RMWeb where someone tried the "DC" output of an old H&M with an oscilloscope and it had gone completely mutant, nothing like DC at all. I think there was even one phase where the voltage went negative. Granted that was a duff unit where one side of the rectifier had failed, but how unbroken is your old box? If you don't know and can't test, I'd suggest replacing it.

At the moment my other option is a basic starter set Hornby controller from 2010 ish .... would that be equally bad? Don't really want to spend money unnecessarily as use will be very limited - but equally don't want to trash my nice shiny new motors.

I now want to renovate my test track and connect it with a new DC controller. As this would be for loco testing only I would prefer pure DC with no feedback or such sophistication

Note the original request. Sophisticated devices like Pentrollers/pictrollers don't come anywhere near this requirement, neither do old traditional controllers which provide a rectified sign wave.

You can build a simple controller using an LM317 as mentioned above for little more than a few pence, power it by a 9V battery until you are happy with it then use a plug in 12v regulated PSU.

Le Corbusier wrote:
At the moment my other option is a basic starter set Hornby controller from 2010 ish .... would that be equally bad? Don't really want to spend money unnecessarily as use will be very limited - but equally don't want to trash my nice shiny new motors.

R964 & R965_18652_Qty1_1.jpg

You really can't tell unless testing with an oscilloscope. Fortunately, somebody has already done that for us; see http://www.scottpages.net/ReviewOfControllers.html. The verdict was that the R965 was actually one of the best cheapo controllers - and this despite the output not being smooth DC.

Back in my childhood, I was a member of a school club that operated a layout in the school physics-lab. We were so potless that we couldn't afford controllers, so we used one of the lab power supplies, connected direct to the track. This provided very smooth DC at selectable steps of (IIRC) one volt.

Alan, Keith, Nigel, Tim

Having done a little bit more hunting around, I've decided to go with the d-i-y approach - thanks for the useful advice. One question now is, could someone let me have a copy of the article from MRJ 11 and the follow-up letter in MRJ 13 (when these came out I wasn't into electronics so didn't keep copies.)

Another option for a DIY route is to join https://www.merg.org.uk/index.php
Model Electronic Railway Group
and purchase the Pocket Money Project number 10 for £1.39, which is a speed controller based on the LM310 chip.
Steve

At last it's complete (apart possibly for adding buffers)!
Decided on the MERG kits in the end as I was able to buy these when attending the Missenden weekend.
Adding meters to the set-up was delayed by my ordering one incorrectly and having to wait for a new one: all mounted in a box with the following set up:
- input either 12V DC, or 16V AC and MERG rectifier (DPDT switch)
- voltage regulator (MERG);
- voltmeter (0-15V);
- ammeter (0-500mA) range seem OK at the moment but I could swap it if needed;
- DPDT switch with centre off for direction control;
- DPDT switch to feed either the rolling road rollers or to the tracks and an external output for layout testing;
- DPDT switch to choose between feeding rolling road & tracks from the internal controller or from an external one (eg DCC)
- external output and input for an external controller

The DCC Concepts rollers can, of course, be placed on the tracks as intended and pick up their power that way but having seen a better test system at an exhibition I decided to power each roller separately. I tapped each roller unit 10BA and attached tags and wired them to plugs that fit the back of the test track. Half the units have the wiring on one side, and half on the other, so that it is still possible to place the units as close together as intended by the design. I tapped the units after assembling them, but it would have been better to do them first.
A small piece of 10thou plastic attached to the bottom of the roller carriers with bluetack insulates the rollers from the track and also lifts them clear of the chairs - as has been reported elsewhere, I found the spacer unit is a little deep and hits the chairs of (my) bullhead track. I can now switch individual roller sets on and off which I think will be a great help in identifying any pick-ups that are not functioning well. My initial set up is for 4 axles but there's room in the box, and the basic wiring, to add more (up to a 4-6-0 with 6 wheel tender).

So far I've only tested one P4 loco which I felt ran acceptably. On the rollers two things became apparent: firstly when running forwards one set of rollers rocked back an forth so I guessed at a problem with a wheel being not really concentric, however this did not happen in reverse. (Such behaviour was not visible when the loco runs on track.) Close examination now suggests that the cause is a bit of slop in the horn guides so I've learnt that I'll need to take more care with my next build. The second discovery was that the ammeter needle did flicker, sometimes at random, but also at times regularly in time with the wheel revolution - not the expected indication of something sticking, but poor current collection: the current dropped at times during each revolution. (The loco had not run for 2-3 years and was not cleaned before testing).

One surprise I had was to discover a small N gauge loco drawing very nearly 500mA whereas my P4 one takes around 120mA (at full power on the test unit).

I expect to make much use of this set up, so very many thanks for all the suggestions made.