Replacing Mashima motors

[sir douglas]

Put a voltmeter on the wires from the controller when it is turned down to zero to see if it isn't completely turning the power off, even though the dial shows that it should be.

ive got an anologue multimeter but im not sure how to use it or read it but the needle moves in the DC settings with the controller at zero

 

i video for example, it starts with motor running with controller power (a bit of red tape on the shaft to show that its rotating, switch to neutral and it stops, switching it back on it starts up, then i turn to zero and knock it against the end of its rotation to show it at zero, again show it with knob at zero motor still running, switch to nuetral to stop it

 

so i think it is as cpman said

 

Portescap motor will start to run on tens of millivolts i.e. tenths of a volt so I suspect that your controller does not give zero volts out when the dial is fully anticlockwise which is confirmed by your statement that the motor only stops when you set the direction switch to its off middle position.

Mike

 

would that mean replacing the potentiometer? and sorry for sidetracking the thread

[cpman46]

I'm not sure replacing the pot will have any benefit but cannot say for certain. It would be useful to see the circuit diagram for your controller (I take it it is a home constructed one and not a commercial offering re-packaged). Once the circuit is known it may be possible to suggest a modification to ensure 0 volts out when the speed control is at zero.

Regards

Mike

[sir douglas]

it is a Hornby r965, with only the input and out put modified to make it handheld http://www.rmweb.co.uk/community/index.php?/topic/133015-handheld-from-a-Hornby-r965/

[laurenceb]

Hornby locos will move with the ( Hornby train set) controler set to zero. Found this out when working on a friends railway when a Charlie parked in a siding out of the way joined in uninvited

[rodent279]

To get 7 volts at the motor with 12v input and 500mA (.5A) current draw needs a 10 Ohm resistor. 500mA is the quoted stall current at 7.2 volts. The voltage drop across 10 ohms will be less at lower current draw, which will happen at lower voltages. So at .1A the drop will be 1.0 volt, but the applied controller voltage will also be lower. So the suggestion that the first 3 volts are "removed" doesn't quite give the correct picture.

Alternatively, a 6.8v Zener diode with a suitable series resistor will shunt any voltage in excess of 6.8v away from the motor.

Assuming a 500mA load @ 6.8v, the resistor needs to drop 12-6.8=5.2v. From Ohm's Law, R=V/I = 5.2v/0.5A = 10.4 Ohms. The power dissipated in the resistor will be VxI = 5.2v X 0.5A = 2.6W. Choose a 3 or even a 5W to be on the safe side.

 

Edit:-Ha ha, that's not going to work is it? It won't work when you reverse the polarity! Stupid boy! That's what happens why you try posting late at night when you're tired.

Edited May 23, 2018 by rodent279

[JohnGi]

Alternatively, a 6.8v Zener diode with a suitable series resistor will shunt any voltage in excess of 6.8v away from the motor.

Assuming a 500mA load @ 6.8v, the resistor needs to drop 12-6.8=5.2v. From Ohm's Law, R=V/I = 5.2v/0.5A = 10.4 Ohms. The power dissipated in the resistor will be VxI = 5.2v X 0.5A = 2.6W. Choose a 3 or even a 5W to be on the safe side.

 

Edit:-Ha ha, that's not going to work is it? It won't work when you reverse the polarity! Stupid boy! That's what happens why you try posting late at night when you're tired.

 

Try two opposed zeners in series, one will drop the zener voltage, the other will drop the normal diode forward voltage (approx 0.6V).

[rodent279]

Try two opposed zeners in series, one will drop the zener voltage, the other will drop the normal diode forward voltage (approx 0.6V).

Hmm. That could work..... I'll have to do some experimentation!

[AndyID]

Hmm. That could work..... I'll have to do some experimentation!

 

Don't waste your money on expensive power zeners. Just use a string of cheap 1N4001 (or equiv) diodes in series to "lose" volts. Each diode will drop about one volt. For bi-directional operation put another string of diodes in parallel "pointing" in the opposite direction.

 

Don't put a resistor in series with the motor. It's only going to encourage "jack-rabbit" starts like the old rheostat controllers.

[johnarcher]

Don't waste your money on expensive power zeners. Just use a string of cheap 1N4001 (or equiv) diodes in series to "lose" volts. Each diode will drop about one volt. For bi-directional operation put another string of diodes in parallel "pointing" in the opposite direction.

 

Don't put a resistor in series with the motor. It's only going to encourage "jack-rabbit" starts like the old rheostat controllers.

Thanks for that last comment, I tried the two little 10mm square motors I have without the resistor (see my post #133), and they do start more smoothly.

Unprotected of course, but as I'd never move anything much above a mild trundle I wonder if that really matters? 

 

I will try your diodes idea (though I don't really understand such things) perhaps for safety.

 

Would a diode other than a 4001 drop more than 1 volt - that would be simpler than a string of 4 or so?

Edited May 26, 2018 by johnarcher

[AndyID]

Thanks for that last comment, I tried the two little 10mm square motors I have without the resistor (see my post #133), and they do start more smoothly.

Unprotected of course, but as I'd never move anything much above a mild trundle I wonder if that really matters? 

 

I will try your diodes idea (though I don't really understand such things) perhaps for safety.

 

Would a diode other than a 4001 drop more than 1 volt - that would be simpler than a string of 4 or so?

Alas, no. Silicon diodes that can handle the required current have a similar forward voltage drop. It's likely to be a bit less than one volt, but not much.

[rodent279]

Most small diodes have a forward voltage drop of about 0.7v. Larger diodes, capable of handling hundreds of amps, drop several volts when forward biased.

Edited May 29, 2018 by rodent279

[AndyID]

Most small diodes have a forward voltage drop of about 0.7v. Larger diodes, capable of handling hundreds of amps, drop several volts when forward biased.

Yes, they start to conduct at 0.7 volts, but the voltage increases as the current increases.

[johnarcher]

Just a quick note to say I received a couple of the 15mm square motors today, and they look very promising.

I have only run them 'in-hand', but they seem to start very smoothly and run slowly, even without feedback.

[bertiedog]

The reason the motors run on "zero" is the controllers' maker has used a voltage regulator IC, which can not control it's output to zero as there would be no reference voltage to allow the circuit to work. Putting diodes in series is the best answer, but you may need a new reversing switch wired in after them and before the track. A back to back pair will work without a new switch.

You may need several diodes in series if you use coreless motors.

 

A Resistor could be used, they do not affect starting, but they heat up, unlike diodes. 

 

An older trick was to put a six to twelve-volt filament lamp in series, they exhibit high resistance at low voltage, but as the voltage increases the current makes the filament glow and the resistance fails just what is needed. Usually, 12-volt car interior bulbs will work, or side lights, LED car lamps do not work.

Stephen.

 

Stephen.

[AndyID]

A Resistor could be used, they do not affect starting, but they heat up, unlike diodes. 

 

Excuse me, but that's complete nonsense. A resistor does affect starting and diodes heat up too.

[bertiedog]

A resistor is neutral it merely delays the increase in voltage causing the motor to start. As soon as the starting voltage is enough the motor moves. The only difference would be the degree of rotation of the control. For a particular current to flow at this voltage the current has to pass through the resistor and heat results. As the resistor has a neutral and basically linear response there is no jump or glitch as the voltage increases.

With a diode the voltage lost it set by the loss through a forward biased junction, and is constant, producing heat like a resistor, but it can stand it better.  Adding either does not cause bad starting or cure it.

[AndyID]

A resistor is neutral it merely delays the increase in voltage causing the motor to start. As soon as the starting voltage is enough the motor moves. The only difference would be the degree of rotation of the control. For a particular current to flow at this voltage the current has to pass through the resistor and heat results. As the resistor has a neutral and basically linear response there is no jump or glitch as the voltage increases.

With a diode the voltage lost it set by the loss through a forward biased junction, and is constant, producing heat like a resistor, but it can stand it better.  Adding either does not cause bad starting or cure it.

 

Sorry, but a resistor is definitely is not "neutral". You need to understand that the speed of a perfect DC motor (one with no losses) would be determined only by the voltage applied to the motor. A perfect motor's back EMF would be exactly the same as the applied voltage and the motor's speed would be linear with respect to the applied voltage, regardless of the load on the motor.

 

Unfortunately motors are not perfect and they do have resistive losses (the motor's internal resistance) and that means the back EMF will be less than the applied voltage. It also means that when the applied voltage is increasing the back EMF will be zero until the motor starts to turn. It also means that as the load on the motor increases the motor slows down. The motor has to do more work so it draws more current and that results in a greater drop in voltage across the motor's internal resistance, a greater difference between the applied voltage and the back EMF with a slowing down and potential stalling of the motor.

 

Adding external resistance only makes things worse. It makes a motor even less "perfect". It will slow down more as the load increases and it will suddenly accelerate as it starts under load (the well-known "jack-rabbit" start problem exhibited by simple rheostat controllers).

 

First you said diodes don't heat up, now you are saying they do heat up? Yes, they absolutely do heat up. The power dissipated is simply the voltage drop across the diode times the current passing through the diode.

 

Diodes are a much better method of dropping the voltage for the simple reason that they are not resistors (although they are not perfect and they are slightly resistive). The reason they are much better is because, unlike a resistor, the voltage drop across a diode is approximately constant regardless of the motor's current demand. According to Ohm, the voltage drop across a resistor is directly proportional to the motor's current demand.

[leopardml2341]

An older trick was to put a six to twelve-volt filament lamp in series, they exhibit high resistance at low voltage, but as the voltage increases the current makes the filament glow and the resistance fails......

 

Stephen.

T'other way around chap!

 

Cold filament = low resistance

Hot filament = high resistance

 

Often used as current limiters in simple electronic controller circuits.

 

When the controller is short circuited, current flows through the low resistance cold filament which then heats up and glows, becomes high resistance, limits current and gives visual indication to boot

[AndyID]

Cold filament = low resistance

Hot filament = high resistance

 

Often used as current limiters in simple electronic controller circuits.

 

When the controller is short circuited, current flows through the low resistance cold filament which then heats up and glows, becomes high resistance, limits current and gives visual indication to boot

 

Yes, filament bulbs are very effective positive temperature coefficient (PTC) thermistors. In theory it should be possible to use a negative temperature coefficient (NTC) thermistor in series with the motor that would effectively cancel out the motor's internal resistance. The motor would then run at constant speed even when the load increased or decreased. Unfortunately I've never been able to find a NTC thermistor with the right characteristics to make it work

[mmagliaro]

I bought a handful of the 10mm x 10mm x 15mm long square-body Minebea motors from eBay and did some testing on them.

I have not tried on in a locomotive yet.

 

Starting voltage is 0.55v (free running).  By 1v, they make enough torque that it is rather difficult to stop the shaft by gripping it with fingers, so I expect these would make plenty of power for a model loco.   By 3v, I could not stop the shaft by gripping in my fingers no matter how tightly I could squeeze.

 

0.5v  15 mA

6v  26 mA

9v  68 mA

12v  202 mA   warm.

 

Given the big jump in current about 9v, I would follow the suggested "6-9" volt rating on these and not push them above 9v, so use some in-line zeners to limit it to 9v.

 

I did some brutal "stall testing" on it.  Gripping the shaft with pliers, a full stall for 30 seconds at 6 volts pushes the current to 325mA.   The case got very warm, but the motor did not fail and afterwards, it still ran as well as before (would still start at 0.5v, and still drew the same current as before).  I repeated this test at 9 volts for 30 seconds.  The current was a positively brutal 515 mA.  The case got too hot to touch, but the motor still survived and still ran as well as before. 

 

I disassembled a few.  The armatures are definitely 6 pole.  They all run whisper quiet and they start up so slowly that you can count the revolutions of the shaft by eye.  These are typical qualities of a high end Faulhaber or Maxon coreless, so I was impressed.  I took them apart because I wanted to see if it was possible to convert them into dual-shaft motors by bracing the armature and pressing the shaft partially through it.  The short answer is "yes", but it requires attention to detail when reassembling.  Mark the case and be sure to put it back together exactly the same way, and to be very careful with the finger brushes.  It took me 3 tries of disassembling, pressing the shaft through, and reassembling before I was able to do this and still have the motor run as well as it did before I took it apart.

 

Although in all fairness I have not tested one in a loco, the motors are very smooth, powerful, and quiet.   And they withstood full stalls and high heat without self-destructing, so I think they could be used in place of a Mashima.   They can run slower than a Mashima 1015 or 1220 without cogging, and they have more power for their size than a Mashima.  HOWEVER, they are not built as robustly.  The Mashimas have heavy bearings and tough brushes in them.  I doubt these Minebea would tolerate abuse, heavy shaft loads, or stalls/overloading as well as a Mashima can.  

 

Given the precipitous climb in current between 9v and 12v, I suspect that over 9, we start hitting mechanical limits, possibly the bearings, so I would limit the voltage to 9.  A couple of back-to-back 3.3v zener diodes inline with the motor would do the trick there.  I would test with an ammeter when running in a loco to see if the current climbs over, say, 150 mA.  If it does, I would limit the voltage further with higher voltage zeners.  It will limit the top speed of the loco, but most model locos run way too fast as 12v anyway.

 

Sorry I do not have a set of photos of all this.  But I've played around with these and wanted to dash off what I had found.

 

Edited June 5, 2020 by mmagliaro

[Barclay]

This is very interesting - I have some of the 15mm square Minebea motors which are 12v and they seem to be smooth and slow revving. Most of my stuff is small however so I've been tempted by the 10mm ones but put off by the 9v rating. 

 

For a complete electronics numpty could you explain how this diode set up should be assembled? 

 

By the way you say they may not be as robust as a Mashima but you seem to have abused yours pretty well and they still work!

 

Many thanks

[Michael Edge]

The 15mm ones are fine although they do run a bit warm, I'm not so sure about the 10mm ones though. I've had some success with them but some failures as well.

[Captain Kernow]

I can't remember now, but was the Mashima 1224 the longest in the '12' series, or was there a longer one, please?

 

[polybear]

5 hours ago, Captain Kernow said:

I can't remember now, but was the Mashima 1224 the longest in the '12' series, or was there a longer one, please?

 

 

When I saw your question I thought there may be a 1226 - but a google search finds nothing.  So I guess 1224 is the longest.  There is a 1426 though.

HTH

[Podhunter]

I came across Taff Vale Models at RailWells in recent years.