I hope this is in the correct section if not could a moderator please move it to the correct section...

Basically I have a small project I am working on, that I require help with as I need to control the speed of a motor, which I done electronically before, but never one that is pulling 67A off a 12V supply...

I have designed and built the lure machine, ready for dragging a dummy bunny to aid in the training and maintain the fitness level of a bird of prey, but could be used for dog fitness etc...

But I need to slow the speed down electronically, at one point it was clocked at 5,600 RPM before we fitted the pulleys so speed may have been increased further, I am happy with the speed and torque except for when I wish to load the spool in which case it's running too fast and following a snag in the line resulted in the spool flying dangerously fast in my direction...

So I was trying to work out a way to slow the motor down, without having to use a mechanical solution, as that would over complicate the project...

The motor I am using is a Starter motor (part number is 25092D - M35G) - basic specification of motor is 25092D M35G-12V-0.8Kw AMBASSADOR PETROL STARTER MOTOR 1.5 HM PLUS, that is running off a 12V Motor Bike Battery to make it portable, as the unit will be used in the field...

So it's a series wound dc motor, so am I right in thinking that I could use a MOSFET to control the motor speed using a voltage level shifter on the gate. As I believe a MOSFET is not concerned about the size of current passing through it...

This is what I am thinking of:- NTMFS4846: Power MOSFET 30V 100A 3.4 mOhm Single N-Channel SO-8FL

As I have never done any projects before involving a MOSFET before, or very high currents I need as much help and advice as possible as I don't think I should treat it the same as I would a transistor?

Would I also need to incorporate overload and short circuit protection around it, or will it be ok as it is?

Also would I be safer to use more than one MOSFET with the source and drain connected together?

I wish to work on the design so that I can fully understand how to safely control high current applications, I still have the facilities to build and test the unit here, which I prefere instead of buying in something purpose built...

So any help on this would be greatly appreciated...

 

First, how fast do you want the dummy prey to move?

From this information, we can calculate the necessary speed of the motor. I recommend that you mechanically slow down the spool, first. Then, electronically control the speed further. To do this, you will need a large pulley on the spool drive. And if you can make the motor pulley smaller, it would help also - although from your pictures it appears to be pretty small already.

It works like this. The smaller pulley has a smaller circumference - call is Cs. If the big pulley has a circumference of Cb, then the motor speed will be reduced by the ratio of the two - Cs/Cb.

Then, you can drive the motor electrically to a higher speed - which will make the motor happy. I would recommend using PWM to control the motor's speed rather than by direct DC. This will allow you much better low speed control. Unfortunately, I have not driven a motor with such high currents.

Your frame, being already built, does limit your options for gearing the motor down. But it looks like some improvement is possible.

 

0.8kW @ 12V = 67A. However, the starting current may well be several times that; so a 100A FET seems inadequate. Putting FETs in parallel would be an option for PWM control, but a respective low-value resistor in series with each source might be needed to balance the FET currents. Like djsfantasi I have no experience of driving high-current motors. You might find relevant help on a robot forum.

 

IMO a starter motor is a bad choice of motors, it is a series wound motor and operates essentially in a run away condition, this is why you should never run it without a load of some kind.
A shunt wound or P.M. motor would be a better choice. For that application I don't see the requirement for that kind of power?
Max.

 

I did something like that several years ago. The power was about the same or a bit higher. It was 12V treadmill motor that was supposed to be in excess of 1.5 HP. I used an LM3524 to generate a pwm signal and an LT1158 to drive multiple mosfets in parallel. I just used a half-bridge configuration (i.e., top and bottom mosfets) instead of mosfets plus diodes.

Here are some pictures of the board. For this control, I just used a 4 oz PCB. For a more powerful unit, I used 25 mil copper strips. Mosfets were soldered directly to the PCB or strips. Gate drive and resistors (one for each mosfet) were done with a modified Manhattan method. I used Teflon tubing where I needed insulation.





John

 

I agree, You won't be happy with the starter motor. It's just wrong for this use.

 

I agree, You won't be happy with the starter motor. It's just wrong for this use.

That is odd...

The old Ford longshaft starter motor is still made in India specifically for a very similar purpose, training race dogs. When I was looking for a rebuilt Ford starter, I found the ones made in India were cheaper and well made. Been using one since about 1995 to launch model sailplanes with no problems. BTW, I use the 6-volt, two coil version at 12V. I also use 2-0 welding cable to a marine battery for power. That controller used the 25-mil copper and 5-paralleled mosfets top and bottom.

Here is a link to the brand I got 20 years ago. I didn't see the 6-V version right off. It may still be made: http://www.injoy-1.com/motor.htm

Amazon has a 6-V version, for comparison, which I have no experience with: http://www.amazon.com/Tractor-1952-1964-Mounting-B5c-11002a-C3nf-11001/dp/B005UDYNQ4

At such power ratings, they are not continuous use, but baiting is not a continuous use application. The less powerful device I mentioned would probably stand up to longer term use. In my application, it was used only to retrieve the launch line from the winch (longshaft Ford motor). The retriever is reversible, but that functionality is not utilized..

John

 

I think I would have looked at a automotive/truck winch type motor.
Max.

 

I think I would have looked at a automotive/truck winch type motor.
Max.

Why? What advantage would that give?

According to this source,

While series DC motors consume more power than PMDC motors to produce their electromagnetic field, they’re powerful, efficient at high speeds and produce the most torque; hence, they are used for heavy duty winches.

In my winch application, a series motor is preferred. I believe the same holds true for luring. There is virtually no visible difference in the construction of auto/truck winch motors (say, Warn) and these series wound starter motors.

Just so there is no confusion, the first device I mentioned and show pictures of was for a PM motor. That is used only for retrieving the launch cord; although, some people will use a PM motor for launching smaller models. The motor used for launching is the series wound starter. Both controllers were the same design, except the controller for the winch was a little more robust and had more mosfets.

For luring, a PM motor might work, but I see nothing wrong with using the series wound motor since the OP already has it.

John

 

Apparently there are now smaller starter motors made which are P.M. so the series consideration does not apply, and from what I have seen truck winches currently being sold appear to be P.M. also.
A P.M. motor is by definition a shunt field motor, and as long as when a series motor is used, it is understood that it should always be used under load, especially when used at full rated voltage.
Max.

 

Thank you everyone for your more than helpful comments...

DJSFANTASI, I can alter the pulley ratio slightly as I used torque lock fittings so can should I need to play around with the ratio, but as she is moving now is an ideal speed for a fully trained bird of prey,

Alec T I agree with your comments, can we do that with FET's?

MaxHeadRoom, I looked at all the designs out there and all use starter motors due mainly to there ease and availability at short notice should things go wrong, it's just a quick drive to the nearest scrap yard...

Thank you JPanHalt, not heard of the Manhattan method, will have to look that up. So basically you believe it is possible to vary the speed of the motor?

Thanks guys, I seem to have stirred up some debate over the type of motor I have used, sorry...

Basically I have the industry standard, as the speed at recall is not the problem the faster the better, but when initially loading the spool she is dangerously too fast I had a slight snag and the drum I was loading off was launched straight at my head from a distance of 5 foot, thankfully it missed me...

So following on from your idea's and part numbers I just now need to research the Manhattan method and workout the schematic...

Would I slow the speed down by varying the PWM, I presume the LM3524 has that option, also pushbutton start where could I implement that?

 

The Manhattan method basically refers to building up, off the board instead of flat on the board. I didn't mean to use a confusing term. Here is a closeup of the winch (Ford starter motor) controller:


Notice that the gate drives and resistors are off the board. The earlier pictures I posted were for a less powerful control. This is the one for which I used 25 mil copper for the main power conductors.

I used to have a picture of the completed winch, but can't find it now. The major difference in construction was that the drum for line was attached directly to the shaft of the starter. That is why the Ford long-shaft is specified. There was very little drag in the system, which had both good and bad effects. The bad part was that backlash could happen and cause a tangle of the line. That was solved using a Hall effect sensor (actually two) to detect direction of rotation. A short pulse was automatically applied to stop backlash. Your system with the belt drive may have enough drag to not suffer backlash.

John

 

can we do that with FET's?

If you mean parallel them then yes; jpanhalt's done it .

 

I did have some backlash, hence the reason the drum came flying at me head height, but I was just manually pulsing the motor by dabbing the cable onto the starter motor... not exactly very advisable, but was very keen to see if she would actually work...

So where do I start, I am about to download the datasheets for the components you have specified and start from there...

But as I have never used any FET's before I can foresee me pestering you for advise along the way...

Is there any do's and don'ts with FET's that I need to be aware of, obviously I am going to need one hell of a heat sink, got that sorted I have a large slab of aluminium left over from a previous project, where by it was used to aid in calibrating some sensors...

I presume the parts you specified are still readily available and for a reasonable price...

 

I did not post the schematics earlier, because they are almost 20 years old, the components are probably dated by now, and this was my first foray back into hobby electronics after many years of absence. There are things I would do differently today. But, I have attached a schematic for later version of the control, as it uses a nice chip, the TPIC 2101. I am not sure it is still available, but it allowed speed/power control by a potentiometer and had soft start that could also be easily set. Disregard all of the logic stuff above. I eventually used a PIC 12F509 for a lot of that.

I picked up my original winch on my trip to Cleveland this morning. This was the very first model, but should give you an idea of the size ("winch2"). It has a rebuilt Ford long-shaft starter. The power module was removed awhile back when I started using "winch3" (discussed below). We realized quite early that variable speed on the winch was useless for launching model airplanes. We simply pulse the peddle that drives the starter solenoid. You may come to the same conclusion. That decision allowed us to use a simple drag brake as shown in winch3. This winch has the brand new Ford long-shaft starter from the luring company linked to above. So long as there is tension on the line, the V-belt is lifted off the pulley by the fairlead/bale and there is no braking. Once the line relaxes, it stops quite nicely with no backlash.

My efforts were then concentrated on the line retriever that I posted pictures of above.

If you decide to make a variable speed lure, you will need a mosfet gate driver. That is what the LT1158 does. There are many similar drivers on the market. You will also need a way to generate the PWM control signal. That is done by the other chips I have mentioned, but I would suggest the TPIC2101, if it is still available, or take the next step and use a microcontroller.

Good luck.

John
Winch 2

Winch 3

View attachment TPIC controlSMD_as built.sch

 

Here is one similar, but not what I wanted... his runs at 43MPH...

http://www.lure-machines.co.uk/

I wish to move away from a controller, as I do have the Pic Kit 2, but not used it in 4 years... well I believe it is the Pic Kit 2 I have as I did upgrade it...

Could we use a NE555, or is that too unstable?

What FET would you use?

Ps I am unable to view schematic, what software do you use to view...

 

I have seen the 555 used for generating a PWM. Some designs claim a 5% to 95% range. There are lots of options, and as such, I doubt one can find the best one for all conditions. Whatever works is my advice.

John

 

I put a 555 circuit together for testing a bi-directional Mosfet PWM worked well, I can post the diagram, If I can find it
Max.

 

http://www.ebay.co.uk/itm/370630423624

http://www.ebay.co.uk/itm/LT1158CN-...al_Components_Supplies_ET&hash=item25789fc6e5

Should I grab them at that price?

 

The pwm control chip is reasonably priced, should I go with that to Base the first building blocks on and possibly get it up and running and fully tested first?

Is there a better mosfet driver we can use, not many of those available, is there anything else we could use £10 for a single chip is a tad high as we may blow it on first run...