Theory meets practise....

Dear ???

The issue is most likely current-draw related.

Circuitron Tortoise motors are designed to draw a max of 4 milliAmps when in-motion,
and 20 milliAmps under stall conditions.

http://www.circuitron.com/index_files/tortoise.htm

Fulgurex motors, as best as I can find, draw around 200milliAmps when in motion.

http://www.clag.org.uk/fulgurex.html 

So yes:
- Both Tortoise and Fulgurex run off 12VDC
- Both use the polarity to determine the direction the motor moves in
- Sure, the Fulgurex limit switches "cut off the motor" at each end of travel,
and thus they do not draw power under "stall" condition at either end of throw

BUT

while in motion they draw significantly more than a Tortoise when stalled...!!!

Now, given that NCE clearly state that the Switch-8 MK2 has a max output per channel of 40 milliAmps

https://ncedcc.zendesk.com/hc/en-us/articles/200526429-Switch-8-Mk2

it becomes clear that a 200milliAmp motor (Fulgurex) is going to exceed what the Switch-8 can supply...

Ergo, Switch-8 is not compatible (as supplied * ) to driving Fulgurex motors.

I hope this helps...

Happy Modelling,
Aim to Improve,
Prof Klyzlr

​* if you were to bolt-on some form of buffer circuit between the Switch8 and the Fulgurex,
like the H-bridge currently being shown on PQE's "Digital Circuitry" thread
http://model-railroad-hobbyist.com/node/28026

they could be made to work,

but then we're waaaaay beyond "plug n play" as NCE intend...

relays or transistors

Thanks Prof Klyzlr, I totally forgot to look at the current specs. This really helps.

I've done more looking around but I can't see any other accessory decoders that clearly indicate they would support these Fulgurex and also have support for rotary toggles; most of them clearly indicate they only accept momentary push buttons (e.g. SMD84 or Switch-Kat); there's the Digitrax DS64 that claims it can work with anything (although it's doesn't clearly indicate its mA output) and in the user manual its button control doesn't clearly state whether it would work with rotary switches.

Bottom line I don't see anything easier than add a "buffer" as you say behind the Switch-8, even if it's extra work.

Just thinking aloud, two obvious way to go around the current limitation would be to use the Switch-8 outputs to drive relays or transistors.

Thanks for linking to PQE's excellent article. I was going to point out 4 transistors would be needed, which is cumbersome, what you and he call the H-Bridge configuration, I didn't know that had a name. And then he points to the L293D. I was not familiar with that device, it seems like it would perfect for that job.

Looking at the TI datasheet for the L293D, it looks like I'd use 2 drivers per output, so four L293D will cover 8 turnouts for one Switch-8.

The TI datasheet for the L293D from  http://www.ti.com/lit/ds/symlink/l293.pdf is a bit confusing since it labels Vcc1 as a "5-V supply for internal logic translation" then on the next page gives it a max rating of 36 V. The ST datasheet for their L293D has a rating of 4.5 - 36 V. Ideally I could use the 12 V DC which already powers the Switch-8s, it's easier. Otherwise I'd need an extra LM7805 -- or better a switch mode equivalent like the R-78C5.0-1.0 or the 7805SRH-C. However a trick I've already done with the NCE AIU cards is that they already have a 7805 on board so I just tap that to get 5 V for infrared sensors. The Switch-8 also has a 7805 on it so I could use its 5 V to power the L293D if needed.

Thanks, that really helped.

Relay interfacing

Dear Ralf,

Unsure where you are located, but if you want to pursue a simple relay interface  solution,
http://www.parts-express.com has a 12V 2A DPDT relay with a 960 Ohm coil.

http://www.parts-express.com/nte-r40-11d2-12-2a-12-vdc-dpdt-security-relay-pcb--075-134

http://www.parts-express.com/pedocs/specs/075-134-nte-r40-11d2-12-specifications-46465.pdf

Given that the Switch-8 supplies 12VDC at max 0.040A

and a 960 Ohm relay coil will only demand approx 0.012A from a 12VDC supply,

this relay should be able to hang off the Switch-8 without stressing the outputs,
and connect a suitably-grunty aux 12V supply to the Fulgurex.

Happy Modelling,
Aim to Improve,
Prof Klyzlr

The L293D...

...has 2 VCC imputs one for the internal logic and one for the item being driven. Pelsea explains this in his posts on simple logic control.

A simple way to do it without electronics would be to use a relay with a diode in series with the coil. The BD8 would power the relay with the power one way but when reversed the relay would drop out. The contacts would be wired as a reversing switch powering the Fulgurex from its own supply. Simple. 

Trying the L293D

Thanks Prof Klyzlr and Graeme. 

I did look at Digikey yesterday, plenty of SPDT or DPDT relays to choose from with 12 V coils that draw about 10-12 mA, they would work well with the Switch-8.

For now I'm going to explore the L293D, mostly because I'm not familiar with these and it's an interesting experience to learn something from. I can make a little PCB with 4 of these and terminal blocks, it should be fairly compact and well adapted to my use case.

NCE Dual Relay

I'll leave these here in case anybody has a similar issue, I found there is an "easier" solution for those who want a "plug'n'play" experience.

NCE sells "Dual Relay" boards which are simply 2 SPDT relays on a small board with terminal blocks and diodes. In this case, one would simply use one such board per Fulgurex to control.

https://ncedcc.zendesk.com/hc/en-us/articles/201576969-Dual-Relay-Board

I extracted this from their PDF that explains clearly how to use it in this case:

https://ncedcc.zendesk.com/hc/en-us/article_attachments/200591915/Dual_relay.pdf

 

The L293D is limited to 7 V on Vi and Ven

Looking more carefully at the L293D datasheet on both TI and ST, I see that Vi and Ven are 7 V max, even though Vss (logic supply) can be up to 36 V. That means I can't just dump the 12 V of the Switch-8 outputs on the inputs of the L293D. 

Oh well, back at using relays then

Relays work

We've been using relays to control Fulgurex machines on Tehachapi for over 20 years.  They have proven highly reliable.

However, the circuitry is different from those suggested in this thread because we chose to run the Fulgurexes at about 1/2 voltage.  They work just fine if a bit slower.   The circuit involves half wave + or - being fed to one side of the Fulgurex motor with a common ground on the other side. 

The relay that does this is SPDT spring loaded.  The spring loaded contact drives the Fulgurex to the normal route, while the powered contact drives the machine to the reverse route.  FWIW, another similar relay circuit is used to select whether the machine is in CTC or local control.

Yeah, admittedly old fashioned relays instead of solid state, but effective and reliable.  Maybe even cheaper than solid state back when the (surplus?) relays were purchased.