Using electromagnets to control inlet Timing on steam engine.

Thread Starter

kgkfr

Joined Feb 24, 2014
17
Hi, I need some help with a project I am working on. I am building a six cylinder radial steam engine, work on the engine is quite far along, but nowhere near finished. I have finished the crankcase, pistons, connecting rods, all machined from solid, and cleaned up the cylinders from seamless stainless tube. I am trying to copy an Idea I saw on You Tube for a steam engine made by Dan Gelbart, this is the link
http://www.youtube.com/watch?v=ccjTMQwKWNs
it has mechanically opened valves but uses an electromagnet to keep the inlet valve open for a variable length time to be able to control the speed an power of the engine, if you watch the video he gives a very good explanation of how it is done.
I have designed a similar system with valves built into the piston crown, operated by the connecting rod. The inlet valve will open mechanically at TDC but I want to keep it open using an electromagnet, I have designed it all but I know very little about electronics so could use some advice as to how to go about it.
I have purchased six12 volt dc 5 wattt magnets and a 600 pulse per rev optical shaft encoder and have been told I can use MOSFET to switch the magnets on and off. What I really need is to learn how to get the information from the shaft encoder to operate the magnets and how to vary the time they are energised possibly using a potentiometer.
I plan to test the engine on compressed air and if it is a success then to build a steam generator and condenser.
All advice gratefully accepted, I will say again I have a good mechanical understanding but electronics is a new field to me.
Thanks

 

crutschow

Joined Mar 14, 2008
34,280
One problem I see is that each of the six values needs a different operating (start) time, however you want all of the delay times the same length. The easiest way to make 6 delay times identical but adjustable,is to use some type of digital circuit.

One approach would be to have six digital counters with a programmable delay. A thumbwheel switch (2 or 3 digits for example) could be used to select the delay to be programmed into each counter. Each counter's start time would then be triggered by the appropriate signal from the shaft encoder.

This could likely also be done with a single microprocessor but that would require programming of the device. The program itself should be relatively simple but learning to write the program obviously has a significantly learning curve.

Do you have an idea what the minimum to maximum delay times you want?

Edit: If you wanted to use rotary control of the delay instead of thumbwheel switches, then a panel rotary encoder and an up/down counter could be used to generate the desired digital delay value.
 
Last edited:

Thread Starter

kgkfr

Joined Feb 24, 2014
17
I have done some calculations for the delay times, I would think they would be ranging from 20° minimum to 135° maximum of the crank rotation. I do not want to run at very high RPM I would think 1500 maximum but hopefully between 500 and 750 for any practical use. A rotary control would be better I think, as the idea is to fit it in a lawnmower eventually, if it ever gets that far.
This a link for the encoder on ebay, I only ordered it yesterday, so it will be about 3 weeks for delivery.
http://www.ebay.co.uk/itm/281232555011?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

Encoder 600P/R Incremental Rotary Encoder
DC5-24V voltage power supply
Speed: more than 5000 r / min
Shaft: 6mm/0.23" Dia, 13mm/0.51" Length
Size: 39mm/1.53" Dia, 36.5mm/1.43" Length
Output: AB 2phase output rectangular orthogonal pulse circuit, the output for the NPN open collector output type.

If the spec of this one is not good enough then I will purchase a suitable model, this one was really not expensive.
 

crutschow

Joined Mar 14, 2008
34,280
So if I understand correctly, the value will be mechanically opened and then held open by an electromagnetic for a specific (selectable) number of pulses from the incremental encoder. For 600 pulses per revolution that would be about 33 pulses minimum to 225 pulses maximum delay. Does that sound correct?
 

Thread Starter

kgkfr

Joined Feb 24, 2014
17
Yes that is correct, the encoder is dual channel, they mention up to 2400 pulses per rev in the advert this what is written on the site
"600 p/r (Single-phase 600 pulses /R,Two phase 4 frequency doubling to 2400 pulses).
With small size, light weight, easy to install, cost-effective features.
For measuring goods angle, acceleration, length measurements and the rotational speed
Apply to a variety of displacement measurement intelligent control,measuring height of civilian body scale and so on
Connection:
Red = A phase, Green = B phase, White =Vcc power +, Black = power -
Notice: AB 2phase output must not be directly connected with VCC, otherwise, will burn the output triode.
Specifications:
DC5-24V wide voltage power supply
Speed: more than 5000 r / min
Shaft: 6mm/0.23" Dia, 13mm/0.51" Length
Size: 39mm/1.53" Dia, 36.5mm/1.43" Length
Output: AB 2phase output rectangular orthogonal pulse circuit, the output for the NPN open collector output type."

Perhaps this mean more to you than me, its all new to me, there will I hope be some sort of specification sheet with it when it arrives.
 

alfacliff

Joined Dec 13, 2013
2,458
electromagnets are slow to release. it takes time for the magnetism in the core and the armature to fade. industrial stuff using electromagnets that must release quickly use a short pulse of reverse current to demagnetise quicker.
 

crutschow

Joined Mar 14, 2008
34,280
600 pulses/rev means the resolution is 0.6 degree of one crank revolution. I would think that is adequate for your needs. Or do you need the higher resolution that the 2400 pulses/rev would bring. 2400 pulses would complicate the circuit somewhat, increasing the number of parts by about a third.
 

GopherT

Joined Nov 23, 2012
8,009
electromagnets are slow to release. it takes time for the magnetism in the core and the armature to fade. industrial stuff using electromagnets that must release quickly use a short pulse of reverse current to demagnetise quicker.
Fast electromagnets use ferrite cores instead of iron. If the high saturation of iron is required, then use insulated iron powder like a Micrometals.com core.
 

alfacliff

Joined Dec 13, 2013
2,458
powdered ferite might do for the core, but what about the valve? ferites are a bit fragile for shock loading. banging open and closed might break them.
 

Thread Starter

kgkfr

Joined Feb 24, 2014
17
Regarding the pulses per minute of the encoder, no I would have thought 600 was enough to do the job, it just because it mentions the 2400 pulses in the INFO on there site, I was thinking more steps better control, but making it simple is the main thing.
I don't think there is any thing to be gained by using an electronic method over a mechanical system, accept the complication of making a variable inlet system that was operated mechanically, I could control speed and power using a pressure regulator, but it would not be as efficient as variable inlet timing. I just saw how Dan Gelbart had done it when I was researching the project and it thought that it was a very good idea. The valve system is 75% mechanical, exhaust fully mechanical, inlet is opened mechanically just the dwell time before closing.
The inlet valve consists of an aluminium outer sleeve with a stainless steel piston and attached to the end is a short length of steel rod for the Magnet to pull on, the steel does not have to touch the magnet, a gap of a couple of thousands of an inch would be OK so I could use insulated Iron powder instead of steel rod, I have mixed resin and this powder previously and it makes a very tough machinable substance with good magnetic properties, this would possibly help the magnet to release more quickly and possibly a pulse of reverse current would help If this can be built into the electonics.
This is a link for the electromagnets I purchased from ebay they are the 30mm diameter ones
http://www.ebay.co.uk/itm/271345245...eName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649
 

shortbus

Joined Sep 30, 2009
10,045
Instead of an electromagnet to 'hold' the valve open, the valve could have a mechanical latch or 'dog' to hold it open. Then a electrical solenoid would be used to release the latch at the required time. The latch would be spring loaded to grip the valve rod when it opens, then the solenoid would release it when the correct time comes.

Electromagnets are not good a holding something small in size, like the end of a valve.
 

Alec_t

Joined Sep 17, 2013
14,280
Your shaft encoder is the incremental type, not the absolute type, so you will need some way of determining when a piston is at TDC to trigger the electronic interval measurement. Perhaps an opto-switch actuated when the valve opens?
 
Top