Old/new compressor

I started discussing this in my villiers thread, so thought I'd start a new one... It's probably not really going anywhere for a little while, but I'll be doing little bits and pieces here and there so I might as well just start it now and keep it updated as and when anything happens.

I've got a couple of compressors with various parts missing or broken so between these and some other parts not doing anything I should have enough to get a decent size unit together. One thing I've wanted for a while is an engine powered compressor so I'm not tied to the mains and seeing as a suitable working electric motor is one thing I haven't got that's what this will be.

Parts I have from a 1947 Dunlop ST5:
Receiver
Pressure switch
Charge cooler
Non-return supply valve

Parts I have from a 1959 unknown model:
Receiver
Sherry & son parallel twin compressor pump unit

Other stuff
Various pipes, connectors etc from both units
A couple of pressure regulators and dryers/oilers
Another old car alternator

Power source
Tecumseh HH100 10hp petrol engine with electric start (I'm not sure what year it is, I'll find out sooner or later but it's not hugely old) This is the only 'medium-modern' Tecumseh engine range I consider to be any good - the 'consumer' type, as fitted to many cheapy mowers, are usually terrible but these industrial/commercial units are pretty reliable.

The plan:
So, the idea is to use both receivers mounted on a wheeled frame and linked together to give me more standby capacity, and the engine to drive it. This engine is massively overpowered for the job, but it means I can run it slower then a smaller engine so it'll be quieter and it should last longer as while it'll have a reasonable load it'll be under less stress. Another important consideration is it's the only available electric start horizontal crank engine I have.

This is where the electronics mentioned come into play... I want this compressor to operate in the same sort of fashion as a 'normal' electric powered one - i.e. run up to pressure and then shut down, then when the pressure falls to the preset level to start up again. Most petrol powered ones people make up have a direct drive where the pump is working all the time and requires manual intervention to monitor the pressure and start/shut down the engine (a pain) - or just a relief valve (bad practice in my opinion).

I intend to use the pressure switch in it's original capacity - open/close at the appropriate pressure levels... Shutting down the engine when it reaches high pressure is simple, it's just an electronic changeover to operate as a kill switch. The starting is a different matter - when the pressure falls I need it to turn on the ignition (open the kill switch), operate the starter motor, then release the motor once the engine has fired up but leave the ignition on.

The HH100 does have it's own alternator, but it's just a simple single phase job that uses an external regulator/rectifier unit so is no good to me for this purpose - hence the car alternator again. A car unit will not only be better for keeping the battery charged but has the added bonus of the battery light - this I can use as a sense point so that when the bulb goes out (engine up to speed) the starter motor is de-energised and this can be arranged so it doesn't interact with or affect the operation of the kill switch.

I'll put up some 'before' photos soon-ish...

So here we go again - questions/comments to the usual address

I think that your biggest problem is going to be, as you say, getting the engine to restart automatically. You will need a battery, starter motor and heavy duty switch. An inertia type starter might help as the ignition can be left on all the time. Your theory is sound but may not be simple to achieve. I think you might have some weight problems as well. Good luck and let's see some photo's of work in progress.

DanBoy wrote:I think that your biggest problem is going to be, as you say, getting the engine to restart automatically. You will need a battery, starter motor and heavy duty switch. An inertia type starter might help as the ignition can be left on all the time. Your theory is sound but may not be simple to achieve. I think you might have some weight problems as well. Good luck and let's see some photo's of work in progress.

The auto-start is the only real 'problem' as such. The engine has it's own inertia starter so that's taken care of, I've got a couple of car batteries so the intention was always to use one of these. For the h/d switch, I'm not sure which one you mean - the starter will be switched by a standard start solenoid. The pressure switch is designed to control the original electric motor and as I'm only going to be using it just above signal level it'll last forever (well, you know what I mean).

Simplicity is relative - it's not quite as simple as pulling a rope but it shouldn't be as complicated as this:

Which is a morse code auto keyer with resistive paddles.... I made that as the requisite project for my ham radio exam.

As for the weight, I'm expecting it to finish off somewhere roughly around the 250kg mark all-in. Might be a shade less... The frame/trolley will be getting a 50mm tow hitch so I can move it around with one of the garden tractors if nothing else.

I'll probably get some photos of some of the parts up later today, if the rain holds off and I can sneak a few minutes out to do it.

I'm watching with interest. It all makes perfect sense to me.

I've always liked auto-start machines, as a kid I always wondered how Lister did it with their Start-o-Matic lighting plants (now of course I know, but way back then it must've been revolutionary).

Will be interesting to see how straight forward it all ends up being. I'm told my Homelite generating set has electric start capabilities on the 28v DC side, wonder if I'll ever get to try it out???

It does sound like an interesting project, like you say the engine starting will be the problem, here's a thought probably a stupid one LOL what about the engine running all the time with a clutch to operate the the compressor using a solnoid/miro switch and a lever.
Good luck and keep us updated on your progress,

Stu.

My first thought was to use a constant run engine, with a pressure controlled throttle link and a centrifugal clutch. This way, when the pressure drops it would open the throttle and speed up which would operate the clutch.

There are a couple of problems with this though - the way I'm probably going to be using it the engine would be sat on tickover for fairly long periods which isn't too good. There's also the fuel to consider - I think a bigger engine running less will use less fuel than a smaller one running more...

I might end up doing this on a different unit used in a different way powered by something like a suffolk, but that's a 'maybe some point later' thought.

There's plenty of cheapo remote central locking/alarm kits for cars on ebay with a remote start option, can you hijack the gubbins out of one of those?

I thought I'd just throw something into the pot and see what you thought of it, sometimes it's best to stick with your first idea, if it doesn't work out nothing lost only time and you could re-design it.

Stu.

Chillitt wrote:There's plenty of cheapo remote central locking/alarm kits for cars on ebay with a remote start option, can you hijack the gubbins out of one of those?

Well, I could, but that would mean spending money and my 'parting with money' allergy is really playing up at the moment... Most of those units would also have a dedicated proprietary IC rather than discrete components to keep the cost down so just using a portion of it could cause problems with unintended interaction. That would also make it harder to repair if the need arose. Plus, I have all the parts to do it the way I'm thinking of.

stationary stu wrote:I thought I'd just throw something into the pot and see what you thought of it, sometimes it's best to stick with your first idea, if it doesn't work out nothing lost only time and you could re-design it.

Stu.

No problem, I welcome ideas and comments - even criticism if it's constructive - sometimes an obvious solution that's right in front of you is hard to see...

Right then, here's what I'm working with...

Tecumseh HH100


I do have everything for this engine, cowl and all! it's partially stripped so it's easier to move about and because it needs a clean. (Yes, it's sat on a set of cast iron wheels, not sure what they're going on yet...)

The compressor pump it'll be driving


Data plate from the sherry's receiver


Data plate from the Dunlop


A Dunlop info plate


And finally the pressure switch


So, all I have left to do is give it all a wipe off and wire it up

Paul im watching this topic with interest .... be good to see how sort of auto starting on clutch arrangement , bit of head scratching involved i think ....

matt

The pressure switch taps off between the pump and the non-return valve and has a decompressor - this means the motor never starts against any resistance from the pump. Just so long as I can get the engine firing up quickly enough it shouldn't need a clutch or anything else.

All that said, I think a centrifugal clutch might be sensible, just as a precaution if nothing else. It might never be needed but I'd rather have one and not need it than not have one and wish I did... I have a few spare so it wouldn't hurt.

Only one way to find out if this would work, kept me busy for half hour anyway



hope you don't mind Paul couldn't resist the challenge

blackvanman wrote:Only one way to find out if this would work, kept me busy for half hour anyway

hope you don't mind Paul couldn't resist the challenge

I don't mind at all - in fact it's very similar to my 'mk1' idea with a couple of differences... The engine ignition is a magneto, so instead of switching it 'on' by supplying 12v, you switch it 'off' by grounding it, so you need to put in a relay to flip it. I take it the connection to the alternator you've labelled as 'EX' is the battery/indicator light? If so, I did think about using a relay on that until I remembered how they work! The battery light bulb is used a current limiter for the excitation current for the field coils - putting a relay coil in place of this wouldn't give the right effect to the coils and putting one in series with a bulb would mean it wouldn't have enough current to operate the relay.

The pressure switch is double pole, single throw, both poles closed at low pressure and open at high and always electrically isolated from each other. I plan to use one pole for the ignition and the other pole for the start circuit.

With this in mind I'll roughly describe the circuit as I can't be bothered to draw it!

Ignition circuit -- at low pressure the switch is closed and I need the ignition wire o/c. Use pressure switch to supply current to operate a normally closed relay and lift the ignition wire. At high pressure the pressure switch opens, allows the relay to fall closed, grounding the ignition wire and stopping the engine.

Start circuit -- Tap off between the battery light and alternator and connect to transistor. When the pressure is low and the engine is not running the battery light will light - use the transistor circuit to operate the start solenoid. When the engine fires and goes up to running speed the battery light will extinguish, removing the signal from the transistor circuit and stopping the starter. I've got a basic circuit but it's very much based on guesstimates of current requirements/component characteristics and needs a certain amount of 'refining'.

There will be a few more components involved in the start circuit, but that's an overview... The next design step is to get the engine and alternator mounted on 'something' so I can take a load of measurements running and not, do a bit of maths, look up transistor data sheets, do a bit more maths, then prototype and test it.

Should be fun!

Of course, you could've used an optical sensor circuit on the starter, which would be fine until the bulb blows on the ignition light

My caravan split charge relay works from a line tap (scotchlok connector) from the ignition light, that doesn't kick in until the engine is running I tested it, as I wondered if it would run my battery down if I was sat in the car with ignition on but engine off (which I wouldn't do, as I believe it's not good for old fashioned coil ignition systems) so would you really need a transistor circuit?

I've got a MOSFET on one of my circuits, they could probably handle things even better. I'm also about to rip apart an old microwave (for the large capacitor & diode), dunno if there's likely to be anything you'd need in there? Last one I dismantled had some relays, but it was a 24v switching circuit (very old digital model) I kept most of the important bits anyway. I've got masses of resistors as well, quite a few 5 watt, loads of 1/2 watt, think I've got a bag of 1 watt as well.

I expect you've got most of what you need, but I've got bags of used components & shed loads of new resistors, loads of old Jiffy bags as well, sure I could stump up the £"whatever it costs these days" to pop it in the post. Can't have the MOSFET though, it's still in use (& cost me £3.99 )

All sounds tickety boo...... Just a matter of assembling the parts (he says casually) Let us know if it all works or what needs modifying.

nutgone wrote:Of course, you could've used an optical sensor circuit on the starter, which would be fine until the bulb blows on the ignition light

My caravan split charge relay works from a line tap (scotchlok connector) from the ignition light, that doesn't kick in until the engine is running I tested it, as I wondered if it would run my battery down if I was sat in the car with ignition on but engine off (which I wouldn't do, as I believe it's not good for old fashioned coil ignition systems) so would you really need a transistor circuit?

I agree it's not good to leave the ignition on with the engine not running, if the points are closed (or at least, if the coil is in circuit) it can damage the coil.

'Theoretically' the battery light lead shouldn't really operate a relay correctly, but I am yet to take the real-world measurements I need to figure it out properly. I haven't finalised the wiring design by a long shot, so as I get on with it I'll update.

Thanks for the offer of the parts too - my 'stuff' boxes aren't here any more so the available parts are much depleted, although I should have most or all of what I need. If I find myself wanting I'll let you know

It's all sounding good and ready to go, only question I'll ask, you've probably covered it and I can't see it, the starter kicks it (energized) how will it turn it's self off once the engine is running? sorry if you've covered this.

Stu.

No probs Stu, I don't mind clarifying or re-explaining any point.

There is going to be a car alternator on it as well. As in a car, there will be a battery/ignition light - when you start the car and the engine is running (more importantly, the alternator is up to speed) the battery/ignition light goes out.

This is what I will be using as a signal source for the starter. The pressure switch turns on the ignition, battery light lights, starter operates, engine (hopefully!) starts, battery light goes out, starter disengages.

Once it is up to high pressure, the pressure switch turns off the ignition and the engine stops. Use the air, pressure falls, switch clicks over, turns on ignition etc. etc. the cycle continues until you turn off the master switch.

Hope that's clearer, for you and also for anyone else who may have been too quiet to ask the same thing!

Paul

OK try this, had a quick word with my dad, turns out I had drawn up a very simular circuit to what he used to build for his commercial standby gen sets, tried and tested:
slight mods here


personaly I wouldn't go down the transister route, you would still need a relay to switch the starter, or a bloody big transister

Right I'm with you now Paul and that makes sense.

Stu.

Andy, I hope the following explanation doesn't sound like I'm having a go! - But there are a couple of things on your circuit that make it not work...

The pressure switch I have is two poles switched as a gang, on at low pressure and off at high - due to the design of the switch there is built-in hysteresis, the low-pressure turn on occurs at around 80psi and the high turn off is around 110psi (these pressure figures are dependent on where you set them...)

The point between the alternator and bulb you have tapped off for the relay is effectively at 0v when the light is lit because it grounds through the alternator field coils, rising when the alternator starts to generate and turns the bulb off. The way you have the other side of the relay coil to ground means if it were to work at all, it would be the opposite of what is required.

I hope there isn't a problem with the multiple posts like this - I'm just breaking up the subjects...

There are a few (what I think are) good reasons I want to go solid state with this:

The first is current - the battery light acts as a current limiter for the excitation voltage to the field coils, allowing maybe 0.2A to flow. If you were to put a relay coil in parallel with the bulb, the amount of current flowing through the field coils would be much higher (about 1A or more) and with that amount flowing through the field coils they wouldn't last very long.

There is also voltage, I measured the indicator connector on my car when it was running earlier and it was only up to about 6volts, but the bulb was off. If you put half the rated voltage through a 'normal' automotive relay it will flutter and get very hot. One of those relays would have no chance of running the starter motor (70A from spec sheet) either so you'd still need to double up and have a starter solenoid.

"So just run the solenoid in place of the relay" - that'll just make the current situation worse, a 'normal' 12v starter solenoid pulls around 5A to operate (or so I seem to remember from when I measured it a while ago) and I really don't fancy drawing that much through the field coils.

I can run a starter solenoid straight off something like a 2N3055 or similar though.

So, here's a basic circuit diagram...



As I said in my previous post, the tap-off point for the sensing circuit is effectively at 0v, grounded through the alternator field coils. To operate the starter, I need something that will 'flip' this, sending a positive signal to the starter solenoid - then when the alternator is generating power, to remove that signal and allow the motor to stop.

I only really want this to sense whether there is a voltage or not, if I can get away with it not really drawing any current I think it'll be better.

Keep the comments coming - it makes me think about things!

So Paul, would you say my split charge relay on my car isn't such a good design???

Reason I ask is mine seems to have packed up, it needs to be able to switch a maximum of around 20amps, probably more like 15 in real terms (actually, a quick calculation reveals my caravan fridge is 100w, which makes it somewhere around the 7-9amp range, depending on voltage & heating element resistance, the battery charge side is hardly likely to reach this, I would guess). Inside the box I currently have is just 2 relays, connected in parallel, soldered to a small PCB.

The relays in there at the moment are pretty small, no wonder they failed. It works, as described earlier, with a Skotchloc connector from the alternator "battery light" wire.

I know you can now get these fancy "self switching" relays, which sense voltage, but the wiring's all there for mine now. I have a spare 2n3055 transistor somewhere, as well as loads of other components, what do you reckon would be a better system?

Or should I just buy another split charge relay (a better one this time) & connect it into my existing wiring?

I only bought that one as it came with most of the cable to complete the job, I've no idea what's happened to it, it was always dodgy on the battery charge side, but last time I used it I got nothing on either output (all fuses were fine, but I had to replace at least one of them IIRC).

I thought about diodes, like they use on boats, but these drop the voltage slightly, which isn't preferable (not for battery charging anyway), also they would need to be pretty big diodes to handle the current. I'm guessing whatever I build will need relays, which I may have, or may be able to lay my hands on. But if I could do it without I would prefer it.