Multiple decoders feeding a traction 'bus'

[Pugsley]

I'm building a 7mm Class 37, with 6 individual, axle hung, traction motors, and I'm now trying to finalise how exactly I'm going to DCC it. The stall current for the 6 motors is 2.4 amps. A bit more detail about the design can be found here, if anyone is interested.

 

I want to power it using one of the Zimo sound decoders as discussed on the Digitrains thread here, but the current consumption is too high. The larger scale Zimo chips don't have as much memory, so that's them ruled out. If all else fails, I'll just have to go with the Loksound XL, but from the videos I've seen of the Zimo one it looks much more responsive.

 

Is it possible to have two chips, with identical speed curves, feeding into one traction bus, to which I wire all 6 motors in series (they're 3 volt motors, and I don't really want to have large resistances inside the loco)? Or is the alternative to have one chip per bogie, again with identical speed curves.

 

Any ideas or pointers would be much appreciated, I realise it's all a bit left-field

[Edwin_m]

I suspect multiple decoders connected (I presume) in parallel would not work reliably. Rather than generating a true variable voltage, a decoder switches the power between full and off very rapidly with the ratio of on to off time determining the motor speed. To control a high-current load the two decoders would have to be not only driving at the same speed but the pulses themselves would have to be synchronised to an tolerance of a very few microseconds. If they are not then for part of the time only one decoder will be supplying the load, which will either increase the motor speed unexpectedly or overload the decoder or both. Also the back EMF is unlikely to work as this relies on the decoder interrupting its output for a short time to measure the voltage coming back from the motor, and this won't work if the other decoder is still driving.

[Nigelcliffe]

Two chips to one power bus strikes me as a way to destroy decoders. ( Someone might show me to be wrong!).

 

However, two chips, one to each bogie, would be fine. Its the same as double-heading (consisting), and setting the speed curves and other motor parameters to the same values is the starting point. You might also want to reduce the BEMF influence (a motor setting) to avoid the two decoders having a fight if one bogie or the other is pulling fractionally harder.

 

If you get the settings correct, you should be able to setup a non-sound chip to be the same as the sound chip, thus saving money. BUT this depends on whether there are delays to the motor drive whilst the chip plays out sounds before applying power (common in commercial ESU sound projects. I am less sure how it applies in a Zimo chip, though the Zimo manual describes a CV to set this value for steam projects implying it is customisable).

 

One advantage of two Zimo chips is an over-abundance of function outputs, including Servos. So, scope for all the instrument illuminated, an animated driver and second man one of whom smokes from time to time, etc..

 

 

- Nigel

[Welly]

What I suggest is to use 2 decoders with their track inputs connected in parallel, one decoder will be the sound decoder and the other one be beefy enough to drive your all your traction motors. They can have the same address CV and common CVs. One would need to disconnect the traction motor decoder in order to re-blow the sound decoder but that is very unlikely.

[JohninAustria]

I want to power it using one of the Zimo sound decoders as discussed on the Digitrains thread here, but the current consumption is too high. The larger scale Zimo chips don't have as much memory, so that's them ruled out.

 

Dear Pugsley,

 

The early versions of the MX690 were produced with a smaller memory chip, but now they use the same larger memory as in the MX64x range. All the current ZIMO sound decoders have the same size memory (32 Mbit). See this page on the ZIMO website for details (see the row marked Sound-Speicher). You should be able to fit any sound project into the MX690 large scale decoder, as one designed to fit in the 00/H0 scale decoders (MX640, MX642 or MX643). I hope this information will help you avoid the complication of using 2 decoders

[Pugsley]

John, thanks, that is excellent news! That saves an awful lot of trouble.

 

Thanks to everyone else who replied, I particularly liked the idea of the animated secondman having a fag!

[locoworks]

i assume the 3 motors in the bogie will be wired in series?? this gives you a 9 volt bogie?? you may need to be careful and limit the top speed CV. it is PWM so it still gives the full 12 volts to the motor output but the 'off' time will help prevent over powering the motors.

[Suzie]

6x 3V = 18V, pretty much the same as the track voltage for larger scales, so you should be OK with them all in series on a large scale decoder.

 

 

 

 

 

[melmerby]

6x 3V = 18V, pretty much the same as the track voltage for larger scales, so you should be OK with them all in series on a large scale decoder.

 

Just a thought:

What would the loco decoder do with the multiple BEMFs when using several motors like this?

What would the overall controllability be?

 

Maybe somebody has tried it before and found it works OK.

 

Keith

[nightstar.train]

A possible idea is to not use the decoder to provide the power. Not sure if it's possible as i'm fairly new to DCC, but can you get a chip to drive a servo in a speed curve manner? If so you could use a model aircraft/boating speed controller to drive the motors, and have that driven by the DCC chip servo output. Speed controllers can provide huge amperages (like 40+, depends on the type) so no worries for that, just as long as your track can supply it, and they draw very little current from the controller.

[Pugsley]

6x 3V = 18V, pretty much the same as the track voltage for larger scales, so you should be OK with them all in series on a large scale decoder.

That's what I thought, thanks for confirming that.

 

Just a thought:

What would the loco decoder do with the multiple BEMFs when using several motors like this?

What would the overall controllability be?

 

Maybe somebody has tried it before and found it works OK.

I'm not sure, is the honest answer. They are identical motors, so if they're all in series the decoder may see them as one large motor? I guess I might end up switching the feedback function off if it doesn't work.

 

A possible idea is to not use the decoder to provide the power. Not sure if it's possible as i'm fairly new to DCC, but can you get a chip to drive a servo in a speed curve manner? If so you could use a model aircraft/boating speed controller to drive the motors, and have that driven by the DCC chip servo output. Speed controllers can provide huge amperages (like 40+, depends on the type) so no worries for that, just as long as your track can supply it, and they draw very little current from the controller.

That's an interesting idea, but I don't think it will work, unless you can control a servo using pulsed current?

[Edwin_m]

If there is no suitable decoder that can source enough current, may be worth using a small decoder and an amplifier circuit? BEMF certainly wouldn't work though!

[Nigelcliffe]

 

That's an interesting idea, but I don't think it will work, unless you can control a servo using pulsed current?

 

 

Zimo decoders (again) have the solution with servo outputs. All decoders, except the not-yet-in-shops little MX621.

Servo output has option of position proportional to speed setting. With either one servo (full forward to full reverse) or two servos (one for speed, one for direction). I think its meant for opening regulator and controlling reversing lever on live-steam locos.

 

 

- Nigel

[Pugsley]

Thanks for the responses. The servo control part sounds interesting, but luckily is no longer required. It seems that the large scale Zimo decoders now have as much memory as the smaller ones, so I can use one of those instead.

[mikeg]

Pugsey, just a though, how often will the loco be likely to reach the stall current? I have used an OO sound chip in a O gauge 08 and it runs well, I also have a loco with two axle hung motors both 12v and wired in series with a TCS T1 chip and the loco runs slowly and the top speed is very realistic as the motors are only getting about 8/9v each so run quite slow the stall current for each motor is 1.5A but have not been able to check if two in series is 3A the decoder is 1.8 with 2.5 max and does not get hot when running, so I think you could use the smaller zimo decoder but have a cutout unit for just under the decoders limit, I think lenz do a cutout thats 1.5A max but you can get them from Maplins, not sure if they work on AC as ment for DC controllers.

 

hope this helps

 

mike g

[Pugsley]

I'll admit that it is unlikely that it'll ever reach the stall current, however, if I ever got the chance, I would load it up just to see what it could do!

 

I think I'll just go for the large scale Zimo decoder, now that I know that they can be programmed with exactly the same sound and settings as the smaller scale ones.

[Edwin_m]

I'd expect the stall current of several motors in series to be the same as that for a single motor, maybe less if they are not getting their full rated voltage. Not sure about using cutouts either, as I think the decoder electronics may fry more quickly than the cutout operates.

[melmerby]

the stall current for each motor is 1.5A but have not been able to check if two in series is 3A

hope this helps

 

mike g

 

Current does not increase when devices are in series, it is more likely to decrease. Consider the motors to be varying resistances (a bit of a simplification I Know!) If you connect two in series the resistance increases and the current falls. Motors when operating are much more dynamic than that but the general rule should still apply.

 

Keith

[Crosland]

I'd expect the stall current of several motors in series to be the same as that for a single motor, maybe less if they are not getting their full rated voltage. Not sure about using cutouts either, as I think the decoder electronics may fry more quickly than the cutout operates.

Stall current is simply aplied voltage divided by armature resistance.

 

Assuming equal motors then then the voltage across each motor is divided by the number of motors in series and the stall current reduces by the same ratio.

 

Andrew Crosland