small steam engine for a model boat.

[optima21]

It would be good if the electrodes could be fed with an Ac signal. That would stop the electrolysis.

never thought of that or came across it with my earlier googling, but has been said before there is an expert for most things on here, and this could be your specialist subject. After your post I started googling again and there is a LM1830 chip available for about £10, so with a a different power supply (higher voltage) it would be a possibility to use. If I have to remove the boiler for inspection once a year, changing electrodes shouldnt be an issue, but only time will tell and its always good to have a plan B to fall back on.

http://pdf.datasheetcatalog.com/datasheet/nationalsemiconductor/DS005700.PDF

http://www.circuitstoday.com/liquid-level-indicator

I also made a safety valve .filler plug for the boiler last last week. I've decided that I like to have fluted edges rather than knurling to grip things and here Im milling the flutes (using a ball nosed mill) around the top lip of the safety valve body after turning it on the lathe.

and the machining the inside of the valve, using a D bit to cut the seat for the ball and threading it to fit the spring adjuster.

and making the spring adjuster, its one of those fiddly little jobs, its threaded 5mm o/d and had 4 slots cut along its length to let the steam out.

and the finished parts of the safety valve which will operate at 25psi.

 

[Kayos]

you're not the only one, its a steep learning curve for me too.

completely off topic, but I forgot to post this on my last post.

I recently saw a film from the mid 1960's about how they were making computers for the Apollo space craft. now I know what Im doing isn't rocket science but when I saw this and you see how intricate and fiddly some parts of the early computers were to make its a bit of an eye opener. the first 10 minutes explain how you find you way in space, the original sat nav, then it gets onto the computer bits.

I watched a series on how they restored a guidance computer, was interesting to see how they were made in the first place

 

[Carl Wilson]

Excellent work with the safety valve, inspiring. I think the level indicator with the ic is a good idea.

You might be able to get one of those chips for nothing. A lot of makers, National Semiconductors included, used to do free sample components. This was/is so that development engineers could trial them.

It would be worth a try. I used to do that a few years ago when building prototype electronics. Sounds impressive but it really wasn't, just very simple stuff.

 

[optima21]

Excellent work with the safety valve, inspiring. I think the level indicator with the ic is a good idea.

I think I need to go down the route of using the LM1830 chip, as it seems to be the correct answer to the electrolysis issue, and the circuitry seems to be fairly simple. I'll be looking at using a 12V step up converter (from 3.7v) to power the chip as it only takes a few milliamps, and and probably use the output from that to drive the pump motor directly from the 3.7v (ish) from the lipo batttery, so I'll be keeping the amount of energy wasted due to the voltage conversion to a minimum. I did start looking at 12V N20 size geared motors but that will be a last resort to use as some of the motors have a 1 amp stall current which is far too high. on the 3v motor/gearboxes Ive got, it looks like the 50rpm and 150rpm gearboxes are identical, but the windings in the motors are different as the 50rpm motor's winding has 5 ohm resistance and the 150rpm motor has a 32 ohm resistance. the 100rpm motor/gearbox has a different gearbox and the resistance of the windings in 8 ohms.

when I said this wasn't rocket science, for me its on a par with getting apollo 13 back home when they had a limited amount of power available and had to make the most of it.

Ive put an offer of £7 for one of the chips on ebay and this is a link to the step up converter Ive ordered off ebay. how can they make them so cheaply and small?

https://www.ebay.co.uk/itm/262990281027

 

[Carl Wilson]

You might find this useful. Interesting circuit which could help you I think.

 

[optima21]

one thing that has really been confusing me is getting the gas burner to work, at one stage I found it impossible to light, and started to think that it could be because the butane and air wern't mixing correctly because the 2.5mm gas jet I made was too big and would need to get a smaller jet to increase the gas velocity to better mixing. I didnt fancy my chances of drilling a 0.15mm hole so I bought one (no.3 gas jet)

https://www.ebay.co.uk/itm/163841096373

unfortunatley the body of the jet was too big and I dont have 1BA threading equipment, so it ended up being machined so it was a bit smaller. (compared to the original one which is a no.5 jet which is 0.2mm

and drilling air holes in the burner holder

and I still couldnt get the burner to light, so I had a rethink and wondered where I was going wrong, even thinking if it was a dodgy can of butane lighter fuel. I had a look my primus propane stove and looked at the ratio of area air intake (buy the burner jet) to the area of holes in the burner and it indicated that on my burner I had aprox 6 times as much air intake, so the reason it wouldnt light was because it was too lean.

so I got a gas jet I could light, and you'd think that would be problem sorted, but no, the velocity of the air/gas mixture was too fast and only the end of the burner furthest away from the burner would light. this was with tubular burners of different designs

I also thought a flat burner with holes on the top would work so here I am drilling 90 holes between 0.8mm and 1.6mm in 0.5mm stainless steel.

and after drilling and deburring the holes. I'll cut to the shape of the burner if I decide to use it later.

then I started thinking about needing to reduce the speed of the gas in the burner, I guess a bit like the difference between a standard lens and gas lens for tig welding. so I case up with a selection of "diffusers" to play with (I have no idea what they should be called) to be fitted between the gas jet and burner holes/slots). the three on the left are machined from brass bar. then we have the top of a sintered bronze air silencer and then some wire wool. the three I mde were pretty useless but the air sliencer and wire wool gave the desired results. the wire wool works well, but I decided that for long term performance the silencer may be a better option as holding the wire wool in tube may be difficult.

so here I am drilling the 8 off 1.0mm dia holes on 5mm pcd on my milling machine using co-ordinate milling

and there are websites out there that will work it out for you if you havnt got a dro that can do it

https://littlemachineshop.com/mobile/bolt_circle.php

Ive ordered a couple of these silencers to pay with and it Im really lucky once I cut then from the brass fitting they will be the right diameter for what I want without any extra machining.

https://www.ebay.co.uk/itm/112379829498

 

[optima21]

and here are some pics of the trial burners and diffusers sorry about them being blurry but making take isnt too flame retardardant so I had to be quick. also the air:butane ratios arn't consistant but this is just to indicate how subtle differences can make a difference and Im guessing why people shouldnt be playing with gas, with the first picture being far too rich (I think).

1) flat burner using air silencer

2) flat burner using no diffuser

3) flat burner using wire wool

4) tube burner using no diffuser ( this needed 5 of the 6 air holes covering up with tape to work)

5) tube burner with air silencer (this needed all 6 air holes to work)

6) tube burner with wire wool diffuser

so thats another part of this project closer to completion, but as you can see its just one issue after another, but where is the challenge in doing easy things. Methylated spirit burners are far simpler in comparison.

 

[waddycall]

Maybe you have too many holes in your burner for the volume of gas you’re burning. If you choke it up with a diffuser inside the burner your also richening it up as the airflow will suffer more than the gas flow. I’d be tempted to try a tube with a series of single holes along it. Do you know how much heat input you need?

 

[waddycall]

You could also try some fine mesh on the underside of the plate. That would also prevent light back to the injector.

 

[optima21]

Maybe you have too many holes in your burner for the volume of gas you’re burning. If you choke it up with a diffuser inside the burner your also richening it up as the airflow will suffer more than the gas flow. I’d be tempted to try a tube with a series of single holes along it. Do you know how much heat input you need?

not alot of heat is required, my gas tank will hold about 20grams of butane and at a rough guess that may last 30 to 45 minutes. Im learning most about making parts of this project as I go along. my only real experience of playing with burners and gas is using a Sievert blowtorch. Im finding all a bit of a steep learning curve as there is very little information easily available and some parts of this seem to be a bit of a black art. Im just posting some of the success and failures as I go along.

 

[waddycall]

I guess if the injector is a size you’d find in a similar sized boiler the gas rate should be near enough. That’s where we’d start with a new burner and fiddle around from there. Our atmospheric burners had mesh under the top plate to help even out the gas flow and prevent light back to the injector.
Amazing work you’re doing. It’s got me looking at lathes again! I've been regretting selling mine.

 

[optima21]

It would be good if the electrodes could be fed with an Ac signal. That would stop the electrolysis.

well this is one of those twists in this project and is better because of it. I got one of the LM1830 Fluid level detector chips, and while it seemed to work I couldnt get any sort useable output from it, but after a bit more googling found posts like this one, using logic gates.

https://diageram.blogspot.com/2013/12/simple-fluid-level-sensor.html

with this circuit diagram

the main thing that got me interested was that it said it could work down to 3V which meant no voltage converters and less wasted energy. It was the first time that I've used a logic chip (I have tried with transistors before), and it has capacitors in it, which confuse me as they seem to work in wonderful and mysterious ways.

and here's my version of the above circuit

with a few tweaks due to available components and issues with noise in the signal, and I needed to dig out my old xp netbook that had a cheap usb oscilloscope for that wont work with windows 10.

so the first part of the circuit (left hand side of the diagram above) is a oscillator which is the AC, and used the first logic gate , the green line is 0V so its a proper square wave AC signal

then the signal gets fed to one of the electrodes of the boiler

and if the water level is low this is the signal received but the buffer/driver which uses the second logic gate and is the middle of the circuit above.

and if the water level in the boiler is making contact with the electrodes

I did have trouble with noise but by changing one capacitor for a much higher value in the buffer/driver part) I ended up with an output signal like this, which is pretty much a straight line

this was then used to drive a mosfet like my original circuit diagram on the previous page.

so once again thanks for your comment Carl, I now have a circuit which works well from a boiler with cold water to boiling water in it (havnt tried it with pressurized steam and hopefully it will work). it also works directly from the 3.7v lipo battery, so no voltage converters to waste power. the mosfet drive transistor drops only 50mv with the motor offload and when drawing 0.12A drops 0.575v across it, which is much better than a bipolar transistor which I would have been using before working on this part of the project.

so this is my prototype circuit (well its really the second one as it is built from the circuit diagram above just to prove it worked and I had drawn it correctly). I just need to make a circuit board for it. for that I'll looking at using this type of prototyping board (when they arrive in the post)

it works, does what I want it to, but cant really explain why it does work and that is as technical as Im wanting to go with the voodoo and black magic that is electronics, think I didnt do too badly with it though.

 

[tom2207]

really enjoying watching this , could you occasionally put something like a pound coin in your pics , just for scale , your work is so fine it just looks normal , till you remember the scale your working at .

 

[Carl Wilson]

That's interesting. I wonder why the LM1830 didn't work. A logic ic is ingenious here because it doesn't care about signal level (amplitude) so much, if it's above a certain threshold it's a 1, if not it's a zero.

The capacitors on the probes are I think to stop any Dc component getting through. The two diodes are a rectifier of sorts, like you'd get in the demodulation stage of an am radio. The capacitor after the diode smooths the output to a steady Dc. That's why putting in a bigger one made the signal into a straight line and the smaller one had classic half wave ripple.

I'm enjoying your posts immensely. I love a project like this, bit of mechanical and a bit of electronic engineering too.

 

[Carl Wilson]

The burner stuff is interesting too. In gas turbine theory I vaguely remember there being a maximum velocity for the flame front at which the flame literally blows itself out. That would explain why when the gas flow is too quick it won't burn. By the time it has got to the end of the burner above its slowed down enough.

Another way to promote mixing is to induce swirl.

 

[optima21]

just a small update to say Ive got the circuit board for controlling the boiler water level sorted and here is the circuit diagram again (this time scanned in) also shown is a layout diagram and the circuit tracks from below (which is a reversed image). just in case it is of interest to anyone out there, a float switch is an easier option, but in my circumstances was not an option

shown below are the tracks on the finished circuit board, together with a blank circuit board and a veroboard (which I would have used previously). I didnt use the veroboad this time as I thought I would have wasted alot of space getting the tracks to fit, as you make the tracks as you go along with the other one and end up with a more compact layout.

I made a couple of tracks from wire for the positive and negative rails

the rest of the tracks were made by bending the leads of components as they were fitted, and there isn't much of the circuit board that isn't used.

and the component side, its was a bit fiddly to do but I was lucky that I drew the layout on the third attempt and checked it three times, and luckily it worked. the circuit board is less than 30mm x 30mm so ended up being pretty compact.

now I need to get back on with redoing some of the mechanical parts, but where there is a will there is a way as the expression goes.

 

[Carl Wilson]

That is brilliant, a really nice compact layout but still repairable.

 

[optima21]

time for another update, but first another picture of the completed circuit board which has me pleased with how compact it ended up with the component density.

of course there is no need for an electronic controller for a water pump if the water pump doesnt work reliably, as I was suffering from pressure variations from a steady 50 psi and then it would drop to 10 psi, which is no good if you're boiler is running at 25psi, as it wont fill. It was guesswork if it would work or not, so after exmaining it closely, the seats of the check valves were leaking badly, and looking into it further Im thinking that the balls I was using were rubbish quality and not round, and after a bit of looking end up getting some silcone nitride balls, as they are harder, rounder, lighter compared to steel and stainless steel ones. I'd never heard of them before

https://www.ebay.co.uk/itm/274308698688

so I bought some and tried to see if I could get better check valve seats buy using different ways of making them, I did get improvements but I still couldent get a perfect seal. so then I started looking at using O rings, and eventually had success with them.Shown below are some of my sample seals, and to get the O ring to seat properly it was necessary to cut a grove to hold then at the bottom of a 3mm hole, so thats there too with one of the 3/32 balls and a 1.5mm x 1mm wall o ring. and Im glad they are available in Viton and the seals at the steamy end of the engine would be just past the operating temperature on nitrile o rings, the standard nitrile having an operating temperature of 120C and the viton being over 200C

this meant that I needed to make a new body for the pump which was almost identical in appearance to the previous one except that I decided to put the sealing O ring on the pump ram, the ram was also extended so that there wouldn't be any bubbles in pump when it was working. and the extended end made it easier to fit the O ring over the ram onto the groove. ohh and the O ring is 1mm ID and the ram is 3mm od. and I was wondering what size to make the groove and then found that there are tables for the grove sizes for O rings, to allow for compression of the o rings like this one

https://www.sealanddesign.com/page/o-ring-groove-design-metric

so with basically a new pump it was time to have another play, and that was a disappointment as with the 100rpm motor I could only get a water pressure of 8psi. I was expecting better, so thought I'd try the 150rpm motor before having a rethink. Im glad I did as that proved to be the answer. Im thinking it could be because the slower speed and low capacity of the pump allowed the water to get past the balls in the valves when they changed direction.

so now for some proper testing. I was wondering how to test the pump under pressure and was going to connect it to the model boiler and fill that with compressed air to get the pressure, then after a period of time would empty the boiler and measure how much water there was..... I was just about to start making the fitting and thought why not prick a hole in the tube and let the water leak out. and it worked perfectly and was such a simple solution. the puddle of water on the plate was easy enough to measure with a syringe.

so now for some stats, the boiler has a capacity of 20cc and will have a working pressure of 25psi, and I was wanting a maximum current consumption of 150mA.

at 35 to 40psi output pressure the pump will deliver 25cc of water at a current consumption of 75 to 100 mA over a 5 minute period. I'll consider that to be a success, as I can increase the output my making a new scotch link to allow 10mm stroke of the pump as it is currently only 6mm. free running the pump will deliver 33cc of water at 65 to 75ma, and I managed to get 100psi out of the pump before the hose blew off.

shown below are the banjo bolts, I still had trouble sealing the O rings so used one that was recessed into the end of the banjo bolt..... the thread on the bolt is M4x0.5mm for scale, and the O rings are 1.7mm ID, 0.5mm wall thickeness and 2.7mm OD, they were bought for something else but were perfect for this (its the banjo bolt in the middle)

 

[Shox Dr]

Oooo. Do I spy a 25mm dia 60psi guage .

 

[optima21]

is it really a month since Ive posted on this thread, well the boiler was steamed to 25psi on my cooker, but wasn't happy with the safety valve, and it would appear that they're not as simple as they may first appear, and Im still trying to make one that functions as I want, and that will be a separate post on here.

so then I decided to make a new check (one way) valve for the water feed to the boiler, and I'd decided to use O rings to seal the balls in the valves as it was possible to get perfect seal. This was because I couldn't get a perfect seal using a metal to metal seal, but trying to lap the seals show how rough brasso is compared to 14000 grit diamond paste. this check valve needed to have a recess machined in it to accept the O ring. I machined the body, fitted an o ring in the groove and tested it with compressed air and got a perfect seal. due to the size of these parts the only way I can remove the o ring is with a pin, which then damages the seal so a new one has to be used. I then silver soldered the body together, pickled it to remove the oxide coating and then ultrasonic cleaned it, fitted a new seal and tried it again. it leaked like a sieve, so Im thinking that it must have distorted with the silver soldering, not much, but enough, so the o ring seat was refaced (the thread in M4x0.5mm for scale)

and machining the hexagon on the top for the check valve

and the disassembled check valve on the drawing for it.

and once again I had a perfectly working check valve. so after this I decided to recheck the seal of the O ring held inside the boiler feed pump body and that had the same issue, which was soon fixed. So the lesson learnt is that in future I need to do final machining after silver soldering and not before.

and then on to some proper testing, here is the boiler being fired up on my cooker, and I think I can say that the silicone coated wire (for the water level electrodes) and silicone tubing for the water feed and pressure gauge are heat resistant. the water is in a beaker so I can measure the water consumption, with module for powering the pump next to the pressure gauge, and the water pump just below it.

I decided to run the boiler for 30 minutes to see how it performed, firstly vented to atmosphere and then again at 25psi pressure. Under pressure the pump was feeding the water into the boiler too fast and the pressure would drop down to 10psi, so I decided to try again with the 50rpm motor. This worked much better and the pressure dropped to a minimum of 20psi, even when the pump was running for over 30 seconds. I was planning on using a water preheater, but it may not be necessary now. When I ran the boiler for 30 minutes vented to atmosphere, it used approx 22ml of water from the beaker and the pump was running for 9 min 40 seconds. when I ran the boiler for 30 minutes at 25psi, it used approx 19ml of water from the beaker, and the pump ran for 8 min 20 seconds. when the boiler got to pressure, it was 14 minutes before any additional water was required, and when the boiler was shut down, there was only 6cc of water left in it (plus the water in the tubes at the bottom), so I'm thinking that boiling water under pressure has a consistency of foam as it bubbles up. I think I'll be fine, just need to be careful with the heating of the boiler. so the next part will be working on the burner and gas supply again