A wiring puzzle

[Izzy]

Recently I have been making a new small circular layout in 2mm/2FS. This uses 6v hacked servos for points and signal along with a 12v Mitsumi motor to power a working level crossing. The power source for all this are 4x AA rechargeables which give between 5.2v-4.8v. Having just finished wiring it all up I have encountered a very odd situation which I can’t fathom out. I just wondered if someone else might be able to spot what’s wrong. Maybe a real simple mistake that I just can’t ‘see’.

 

Here is the wiring dia. A bit crude but should hopefully be clear enough.

 

 

The level crossing was made based upon that featured in the second Finescale Model Railway review and built by John Watson. Micro-switches are used to limit movement at the ends of rotation, with diodes to allow reversed current to change movement direction. This is done using a DPDT. It all works well with a 50ohm variable resistor tapped at 38ohms to get the slowest speed the motor will reliably work at.

 

The hacked servos used for the points are controlled through pairs of SPDT’s working in unison to both allow change of direction and limit movement. Again it all works as it should.

 

The hacked servo used for the signal is controlled by a biased centre-off DPDT. This is where strange happenings have been encountered which I am at a loss to explain. It’s power is taken off the same resistor reduced feed for the crossing with another 25ohm added to get the speed down to the lowest possible, so 75ohm total.

 

To prevent being able to pull the semaphore signal – which protects the level crossing – ‘off’ when the gates are open to road traffic, and to also prevent the gates being closed to trains when the signal is ‘off’, micro-switches have been fitted into the feeds to both, and controlled by the other. In other words when the signal is pulled off the power is cut to the crossing (MS3), and vice-versa when the crossing is open to road traffic power is cut to the signal (MS4).

 

Now when the signal is pulled off the speed is normal, but….. if the feed to the crossing is not cut using the micro-switch (MS3), then in reverse it can hardly move. And it must be in the line to the diodes. Put the switch in that to the motor and the problem still exists. This is how I first encountered it. Only by moving it to the other line did it not happen.

 

Okay, with the MS3 where it is the issue doesn’t happen. But I am concerned to know why this should be needed, and whether any other problems exist with it I haven’t yet spotted/encountered. Is it a short circuit caused by the two sets of DPDT’s when set in a particular order. Or an errant wiring circuit enabled by same. Are the batteries being drained/damaged? They don’t seem to be, their charge seems stable and long lasting, but…..

 

Any thoughts gratefully received.

 

Thanks

 

 

[Nigelcliffe]

Are you sure the diagram represents what exists in the actual wiring ?   There's at least one discrepancy;  discussion of taking a 38ohm tap from a variable 50-ohm, and elsewhere adding a 50 to 25 to create 75 ohm.  Either the resistances reported are wrong, or the diagram is wrong.  

 

I think what is being seen is the combined resistance through the two circuits changes when the switches are operated.  (Resistors plus motors treated as a resistor).    

 

If you put a DC volt meter over the motor terminals, then operate the switches you should see the voltages change.   

 

Dropping voltage via resistors is likely to lead to this sort of behaviour as the load changes depending which motors are in circuit.  The fix is to bring both motors back to the 5v, with their own independent resistors to drop the voltages (or, use a different way to drop voltages!).  

 

 

- Nigel

[AndyID]

43 minutes ago, Nigelcliffe said:

Are you sure the diagram represents what exists in the actual wiring ?   There's at least one discrepancy;  discussion of taking a 38ohm tap from a variable 50-ohm, and elsewhere adding a 50 to 25 to create 75 ohm.  Either the resistances reported are wrong, or the diagram is wrong.  

 

I think what is being seen is the combined resistance through the two circuits changes when the switches are operated.  (Resistors plus motors treated as a resistor).    

 

If you put a DC volt meter over the motor terminals, then operate the switches you should see the voltages change.   

 

Dropping voltage via resistors is likely to lead to this sort of behaviour as the load changes depending which motors are in circuit.  The fix is to bring both motors back to the 5v, with their own independent resistors to drop the voltages (or, use a different way to drop voltages!).  

 

 

- Nigel

 

Yes, the voltage drop across R1 varies with the the current passing through it (Ohms Law). If both devices are drawing current the results will be unpredictable 

[Izzy]

 

Thanks for the resitor thoughts Nigel & Andy. Sorry, I didn't draw the R1 resistor qute right, to show the tapping of the 38ohms off it. Ammended one below.

 

I didn't think the resitors an issue because the two motors are not powered at the same time. The signal being through a centre off is of course also electically dead most of the time. I need the crossing to have a plain DPDT though as it takes 20 seconds to complete a movement, to open or close the gates, and I don't want to have to hold a biased switch all that time, too boring!

 

It's also the fact the power/voltage drop to the signal servo only happens in one direction that has foxed me.  Since the MS3 switch has solved the issue I'll leave what exists alone. But I will remove the crossing circuit from the resistor temporarily, and feed it the full 5v, and see what effect that has.

 

Thanks again

 

 

 

 

[Izzy]

I have now tried to dispense with the resistors since taking them out of circuit certainly proved they were the issue when the circuit cut-out switch wasn't enabled, indeed further testing proved the power provided did tend to vary quite a bit each and every movement, not being constant, and drive the motors concerned on reduced voltage instead. This was done by taking feeds off the 4xAA pack, so 2xAA for the Mitsumi crossing motor and 1xAA for the signal servo one. Sadly the voltages weren't quite enough in either case, 2.5v & 1.25v but using 3xAA & 2xAA respectivly, 3.75v & 2.5v, meant they moved too fast. So for the moment, as it now works, the resistors have been kept in conjunction with the switch, as it will remain anyway.

[AndyID]

34 minutes ago, Izzy said:

I have now tried to dispense with the resistors since taking them out of circuit certainly proved they were the issue when the circuit cut-out switch wasn't enabled, indeed further testing proved the power provided did tend to vary quite a bit each and every movement, not being constant, and drive the motors concerned on reduced voltage instead. This was done by taking feeds off the 4xAA pack, so 2xAA for the Mitsumi crossing motor and 1xAA for the signal servo one. Sadly the voltages weren't quite enough in either case, 2.5v & 1.25v but using 3xAA & 2xAA respectivly, 3.75v & 2.5v, meant they moved too fast. So for the moment, as it now works, the resistors have been kept in conjunction with the switch, as it will remain anyway.

 

You could use diodes to reduce the voltage. They produce a reasonably constant voltage drop although it does vary a bit with current. The drop across each diode will be somewhere between 0.7 and 1.0 volts depending on the current. Your average 1 amp silicon diode will work - something like a 1N4001 or similar.

[Izzy]

Thanks Andy, I did try that approach. But after three pairs in series didn’t get there I sort of gave up with that.

[AndyID]

2 hours ago, Izzy said:

Thanks Andy, I did try that approach. But after three pairs in series didn’t get there I sort of gave up with that.

 

That's a pity. I'm wondering if the batteries are holding up under load? They should have quite a low internal resistance but those servo motors have very little resistance and draw a lot of current.

[Izzy]

I did wonder that.  But this particular hacked servo is a JP7.5gm with a small coreless motor and 620-1 reduction and can operate on far lower voltage with less current draw than the others, which are TP MG90’s with cored motor + 220-1 and work just fine. They draw much more under load, twice as much actually, at 5v 970ma versus 560ma of the JP. With no load it’s 120ma v 80ma.

[DCB]

On 31/05/2021 at 00:11, AndyID said:

 

That's a pity. I'm wondering if the batteries are holding up under load? They should have quite a low internal resistance but those servo motors have very little resistance and draw a lot of current.

There is a massive difference between decent high quality AA rechargable batteries as supplied as original equipment by Nikon etc for their cameras and the dross you can buy elsewhere,  DIY stores etc for between 50p and £5 a pop.  My cameras drain a DIY store AA in about 6 shots while an OE Lasts all afternoon, 200?   If you want 5 volts there are numerous USB chargers on the market.   Rechargables aren't great for models as they discharge so rapidly if shorted,  They need a fuse in circuit, polyfuse maybe.  Diodes are good for controlling voltage but the steps are too big at 0.7 volts.  Optimum is to use a variable voltage power supply like a Morley which delivers a set voltage over a wide range of amps. 0.01 to 1 or more, but that costs serious dosh.    I sympathise, my semaphore sgnal driver took ages to set up to get the right resistors to make it move realistically then every time I used it it needed tweaking again. I gave up and let it work too fast but work every time.

[kevinlms]

2 hours ago, DavidCBroad said:

I sympathise, my semaphore sgnal driver took ages to set up to get the right resistors to make it move realistically then every time I used it it needed tweaking again. I gave up and let it work too fast but work every time.

The solution is not that difficult, you need a REGULATED power supply, that will properly supply the right voltage. Juggling with resistors never works.

[AndyID]

4 hours ago, kevinlms said:

The solution is not that difficult, you need a REGULATED power supply, that will properly supply the right voltage. Juggling with resistors never works.

 

Yes, better to use a LM317 or similar. Cheap as chips these days too 

[Izzy]

Thanks for all the thoughts regarding re-chargeable batteries and resistors. Having used the former for over a couple of decades in various digital cameras I would agree that early NiMH's were very problematic in that they just couldn't hold on to their charge unless in-circuit in a camera, and then not for long either. But today I find that almost all of them have no issues in this respect. Enloops were I seem to recall the first to crack this but the other battery makers soon caught up with their various brand makes. I currently have several Maplin branded sets which are good for around 12months stored.

 

I wanted to used batteries so I could keep mains power requirements simple and separate from the DCC system I use which I just plug into whichever layout I am using at the time. I have one layout where it is also used to power the accessories, via a VR which becomes a pain when I want to test something and need to plug it in and power it up to do so when working on the layout.

 

This is now the second layout to use batteries, and resistors where needed for voltage reduction, and apart from this wierd situation with the one servo motor, which I wanted to understand why, I have had no issues whatever. Sadly I seem no nearer to solving it apart from the fact it's connected to using the resistors paired to drive two motors. I also feel it must be connected to the level crossing circuit involving the diodes somehow.

 

 

 

 

 

 

[Izzy]

Just to update I have now found no choice but to remove the resistors and do some reworking to use back-to-back diode pairs as voltage droppers as an alternative.

 

As had been stated fixed resistance works okay with newly charged batteries, but thanks to the diode control circuit for the level crossing as soon as the voltage started dropping from the initial 5.2v to 4.8v with use, and because there wasn't enough 'overhead' the motor would no longer start, the resistance needing to drop in line with the voltage reduction which of course it doesn't do.

 

So 4001 diodes in pairs have been used instead. Two sets each at 3AA's for the crossing and 2AA's for the signal. Two taps off the 4AA pack used for the hacked point servos. This seems to be working okay by giving some leeway for the voltage variation.