Lister Engine Forum

Alternative fuels => Waste Vegetable Oil => Topic started by: dax021 on June 07, 2018, 02:20:53 PM

Title: WI
Post by: dax021 on June 07, 2018, 02:20:53 PM
I have finally got a batch of WVO ready for consumption in my SR2, but am reluctant to try it without WI.  I still haven't got my head around how to install WI, but remember reading a post where someone modified an oil bath filter to feed water into the intake.  To me this would be the easiest solution, but am concerned about whether it will work and also whether the filter will still work as an air filter, not having the oil bath.  I know Ed and Glort will be able to advise.


Title: Re: WI
Post by: EdDee on June 07, 2018, 03:55:23 PM
Hi Peter,

Glort's the man for WVO advice.... but.... in the interim, fuel it up, run it, if it looks like its starting to ail, a squirty bottle and a bit of water puffed down the snout while up to running temp to break any carbon free....

But let's wait for Glort to give a real answer!!

Title: Re: WI
Post by: dax021 on June 07, 2018, 08:07:35 PM
Thanks Ed, will see what Glort has to say.  Also hoping the OP of the oil bath WI will jump in
Title: Re: WI
Post by: dax021 on June 08, 2018, 10:33:07 AM
Hi Glort, thanks for the info. 

Engine is in a shed, permanantly mounted running a 4Kw alternator @1500RPM.  I think the displacement is 1100cc, twin obviously.  I use it for standby power for when my solar can't cope or when welding, running a compressor, etc.  I had it serviced when I bought it and have since put on just over 200hrs without any problems.  I guess on average I run it about 5-10hrs a month.  I read your (or was it Ed's) comments about feeding a drip into the intake and watching the exhaust, but haven't quite figured out how, being a twin and the fact that the exhaust goes through the shed wall and then underground.  Maybe i'm overthinking the whole thing and should just run the WVO and squirt a litre or so of water down the intake once a month.

Thanks for the advice.
Title: Re: WI
Post by: BruceM on June 08, 2018, 05:44:03 PM
Mr X's had a very simple and effective system for WI that I liked. He used a shallow covered pan (I'm thinking Tupperware or similar) located just barely below the intake manifold, with a small tube and flattened brass end into the intake manifold.  No chance of disaster in failure, stops when engine stops.  The shallow pan with water just below the intake is to allow the minimal intake vacuum to do the job and provide relatively uniform rate of delivery. A needle valve or crimped brass tube in that small silicone (RC fuel line) line could restrict the water consumption rate.

The addition of methyl alcohol to the water to prevent freezing in winter was also discussed.  Glort reported methyl alcohol improves diesel power so that's a plus.

The additions I was thinking of making were to add a shut off valve controlled by engine temperature and a magnet-hall effect sensor to sense throttle position.  I'd keep WI off until engine was up to temperature and load was better than half load.  I've also pondered adding a float valve to the pan, making it quite small, and feeding the float valve from a larger gravity feed tank mounted on the wall.  The tank should last 8 hrs of run time, since I do a check over and oil service at that interval.  If the very small pan tank has an overflow tube that runs to a catch pail, a float valve fail will be of no concern. 

 I have some relatively low power very small 12V solenoids for low pressure that I could use. A snap switch could be use for head temperature sensing, it could feed the throttle arm magnet/hall effect or magnet-reed switch for a combined signal. A power transistor or DC solid state switch could then operate the solenoid.

My Listeroid does much of it's service in air compression so AC power isn't always present.

Glort, let me know if you think this is overkill for a remote start, unattended engine; I already have the controller shut down the engine if has no electrical or compressor load for 10 minutes.  I assumed that WI on a cold (just started) or unloaded engine would be a bad idea.  Getting rid of the throttle position sensing would make it much simpler, though it is certainly not too tough.

Title: Re: WI
Post by: BruceM on June 09, 2018, 08:13:57 AM
Thanks for the helpful advice, Glort.  It's much appreciated.

A low power, 12V dispensing pump does have some merit- same controls as a solenoid valve and perhaps about the same current.  The supply could then be anywhere, and no float valve simplifies things too.

The smallest one I can find uses 300 ma at 12V. No doubt the water flow is excessive but I suppose I could run it off a cmos 555 timer can cut the duty cycle to 1 second in 10 or so.  I'll still have to rig a bypass for excess flow.

Maybe I'll find a better suited pump on ebay tomorrow AM.
Title: Re: WI
Post by: BruceM on June 09, 2018, 04:11:40 PM
Good ideas for thought. I can see why petrol injectors would be appealing as a means of atomizing and metering for your oil burner design. 

I think small droplets are OK for WI in the intake manifold so a brass tube into manifold with venturi effect seems appealing; even with a micro DC pump it could be set up with a tee-bypass and crimped brass tube insert flow restriction so that water head is insufficient without intake manifold venturi/vacuum.  Silicone tubing would make that readily visible for adjustment.

I don't have AC power when doing a lot of air compressing; I like to unstrap the generator head as it does affect my fuel consumption.  So I'd like to keep my power use to the minimum.  I do have a little 40W panel and 12V battery (hand me down's from my no-alternator car that have lost too much capacity) in the engine shed, but I like to keep running power draw to a bare minimum. 

I'll take a look today at floor (pump) vs wall mount tanks (gravity) and see what seems easiest. 

Title: Re: WI
Post by: BruceM on June 10, 2018, 12:20:13 AM
I checked out my engine room for water injection possibilities.  I think I'll try a covered shallow pan just below the intake manifold method developed by Mr X first.  There's a nice spot on the wall that I can add some brackets and a plastic storage bin as the feed tank.  I found the small solenoid valve in a parts box in the engine room, and even a needle valve so this could be a very cheap to build setup.  I've got to look at my thermal snap disc collection next.

edit- sorry, wrong video link here, will re- upload and update the link next.
Here's the on-topic video I meant to list:
Title: Re: WI
Post by: 32 coupe on June 10, 2018, 12:36:08 AM

Nice video, as always !


Title: Re: WI
Post by: BruceM on June 10, 2018, 01:44:00 AM
Thanks Gary, Alas, I uploaded the wrong video- it was one showing my Listerioid setup but not the one I had intended for this thread. 

Because of the ease of implementing Mr X's intake venturi feed method as I discussed, I'm rounding up parts  to give it a try.
Title: Re: WI
Post by: BruceM on June 10, 2018, 05:00:00 AM
It's Mr X's method as he reported here; a shallow pan with water level just below the intake manifold, feeding a pipe in the intake manifold. Intake suction pulses pull the water up the last inch in via venturi.  He flattened the end of the tube to get reduced flow and finer droplets.  I'll report on how it goes. I'll just be adding the solenoid valve and head temperature switch. 

Title: Re: WI
Post by: ajaffa1 on June 10, 2018, 08:47:56 AM
Very much looking forward to this, I love to learn new things from people with experience. More pics and videos PLEASE

Title: Re: WI
Post by: dax021 on June 10, 2018, 10:52:56 AM
Yes please, also looking forward some pictures.  I presume with a twin, one would just T the water line, one branch to each intake?
Title: Re: WI
Post by: BruceM on June 10, 2018, 03:48:59 PM
You nailed it, Glort.  The CS intake vacuum is so limited that flow rate will vary with height above the water, and even a few inches will change the flow rate.  Mr X already proved this approach works using a shallow pan to reduce the variation, so I will pursue it further. The flow rate will also vary with restriction of air filter. I'm not sure what Mr X used, I have a paper filter and he may have used the stock oil bath.

That is the advantage of Glort's preferred metered, pressurized system; a consistent flow rate.  But I'd like to see how this intake venturi/vacuum method works out.  As you pointed out, Glort, a fairly wide range of WI is OK, and once set up if flow increases by 50% because the air filter is getting dirty, it should be fine. A reduced WI rate as supply is used up is also not a show stopper as long as the variation is not more than 75% for a useful volume of water.

Your rather brilliant and original idea of using an inverted bottle to maintain a constant level would completely eliminate the water level variance ... A two liter soda bottle inverted in a small container might be an easy to service WI supply with a totally constant water level.  A float valve filled small container would also do that and allow for a larger water supply.

I'll know more today after I test the intake vacuum and venturi flow rate variance. I want to try different tube diameters in the intake manifold to see if that affects suction. 

Glort or any other experienced WI users:  Could you provide a guesstimate of a range of water flow that you think would be appropriate for my Listeroid CS 6/1?  The minimum effective water flow rate to reduce carbon buildup using dino diesel would be particularly helpful. 

Dax021- yes, a tee should do a twin.  I'll try to collect vacuum and flow data today if the wind doesn't do me in. I'm still at the thinking and collecting data stage.  I will update as I progress.

Title: Re: WI
Post by: BruceM on June 10, 2018, 07:35:33 PM
I measured the suction on the running CS 6/1 intake manifold at only 35 mm or 1 3/8 inches of water.   So the range of water that can be used without a flood on stopping is very shallow.  For the emitter in the intake manifold I used 3/16 OD brass tube with the end cut at 45 degrees facing away from the flow. (

I see why Mr X flattened the end of his tube to restrict flow-  with the open tube, as you raise the water level (in my case raising a 10 ml syringe body attached to a telescoping stand), the flow is fairly brisk (>1 ml per second) until the water gets below the flow level and then it just stops. 

This tray feed method could certainly be used effectively as a batch WI process- get the engine warmed up and then fill up your tray; it is essentially gravity flow until the water level gets down to the stopping level of 35mm below the highest point of the manifold emitter. If you fill your tray above 35mm, you risk an intake flood if the engine should shut down.  The WI flow rate should vary linearly with water level.   

This is not what i had in mind for my use since my generator shed is quite remote and I'm more interested in something automatic that just needs water every 8 hrs of engine run time, with my usual runs of 40-60 minutes.
I'll need to add some complexity; a constant water level to be regulated via float valve or the upside down bottle method Glort suggested.  I can now fiddle on my bench with the manifold orifices  and needle valve to get the desired flow rate with a head of about 20mm of water.

If I use a 4 ml/minute flow,  8 hours of run time would use about  2 liters of water.  Is that sufficient? Would pulses of higher flow be better for dislodging deposits?


Title: Re: WI
Post by: BruceM on June 10, 2018, 09:02:14 PM
Thanks for the help with the flow rate, Glort.  I can carry 3 gallons (11 liters) of water in a 5 gallon pail with lid per 8 hr oiling service.  In winter I'll add methanol to avoid freezing.  That would allow up to 23 ml/minute water injection.  I can use a 5 gallon pail with lid for a cheap gravity feed storage vessel.  The float valve tank can be as small as will fit the $9 float valve for 1/4 OD plastic tube supply I found on Amazon. 

Air to pump water via bubbles is something I have read about.  One deep well pumping company sold windmills with an air compressor attached to the blades. It that pumped deep water wells via rising air bubbles for stock moving parts in the deep well.

I may play around a little with your bubble idea.  A small overflowing tray to be the constant water level tank inside a larger storage tank, with water pumped to the tray via air bubbles or micro water pump... not a bad idea since it eliminates the float valve.  Mostly I need to scope out the limited space in my engine room to see what will fit and will make filling the water tank easy.

Title: Re: WI
Post by: mike90045 on June 11, 2018, 01:27:32 AM
What's used in windshield washer antifreeze - is it methanol ?  Besides the color dye, anything else in it?
Title: Re: WI
Post by: BruceM on June 11, 2018, 02:05:50 AM
I suspect some detergent and ammonia plus the dye and methanol.  I'm not sure I'd want to use that for water injection.  I have 5 gallons of methanol I bought at the auto race track that I can use up.

I did find a cheap up to 100ml/minute volume dosing pump that draws only 80ma of 12v and could work for someone that wants to do a direct pumped WI on a CS or other small engine.  For most if driving from a small switching 12V supply off the generator mains, that should suffice for drawing up from a covered bucket. You could even get fancy and switch the AC supply through a snap disk epoxied to the head.

I'm still undecided. For gravity flow I did find decent small float valve for 1/4 OD plastic line.

Title: Re: WI
Post by: BruceM on June 11, 2018, 04:08:17 PM
After thinking and sleeping on it, I'm going with Glort's suggested pumped WI. I've ordered the cheap dosing pump listed above.

I do like things as simple as possible, and I find gravity quite reliable and cheap,  but in this case the pump draws less power than a solenoid valve, and I can just sit a 5 gallon bucket with tubing through the lid in the corner and feed the dosing pump with that.  I'll add some new Basic code to my existing PICaxe 40x2 engine controller so that it waits 10 minutes after load is detected before turning on the pump; so no new head temperature sensor required.  Likewise, stopping pumping with no load and on any condition caused shutdown is another couple lines of basic since the sensor inputs and associated code are already there.

These peristalsis type pumps can run dry with no problems and are self priming, so no float switch is required to stop a run dry situation. I'll can adjust flow rate via voltage dropping resistor and/or duty cycle as well as flow restriction. Voltage reduction is ideal since current will also be reduced and the motor will last longer.  I'll test the pump as soon as I get it; perhaps Thursday.

Title: Re: WI
Post by: BruceM on June 11, 2018, 11:33:28 PM
Glad you've had some experience with this type of pump.  I ordered some slightly smaller silicone tubing to use with it; that should reduce flow rate and roller pressure as well; on similar pumps  flow rates are specified for several different sizes of tubing which gave me that idea. 

I should get the pump late Wednesday.

Title: Re: WI
Post by: BruceM on June 12, 2018, 02:52:00 AM
I had to search widely to find a pneumatic diaphram pump with the right diaphram material for handling biodiesel. The stuff destroyed my hand crank barrel pump and my air drill powered vane pump in just a few days.

The China made motorized ball valves have changed my automation world; I can get for under $50. what used to cost $350 to $500. The ones with the super cap that close on power off, thus acting like a super low power solenoid valve are especially handy.   

But by far the best thing China has done for me is Zennioptical.  I have MS related vision problems- difficulty moving eyes and focussing which resulted in ferocious headaches shortly after putting my glasses on in the morning.  By experimenting with under-correction to various degrees, and finding just the right correction for various tasks and distances, I do WAY better.  This took buying a trial lens set, trial frame set and owning a lot of glasses.  By sticking to under $10 bargain frames including RX lenses, this was possible.  It also made me realize that the US opticians had been screwing me my whole life.  The Zenni bargain frames and lenses are better quality, by far, than the $350 glasses I had been getting.

Title: Re: WI
Post by: BruceM on June 13, 2018, 11:37:26 PM
Got the cheap peristalsis pump and tested it.  It's junk.  Wouldn't pump at all as delivered, the bare motor shaft is friction drive on the nylon rollers- a bit of oil on the shaft meant it was just slipping and not pumping. Only the tubing pressure on the rollers pressing against the motor shaft?? Not an reliable design for remote use.  I cleaned it and limbered up the tubing and then it worked- but it draws 250ma of 12V, not 80ma.  Specifications from China are rarely meaningful.  At 6V it still draws 230 ma but barely moves.  The motor is overloaded for this application and draws way more current than needed for the task.  It is certainly never going to last the 6000 hours service life specified.

I've ordered a different pump, $15, with alleged gear drive, 6V, 30ma. 

Title: Re: WI
Post by: BruceM on June 14, 2018, 03:27:35 AM
I could live with the 250ma, but the pump I got is just too cheesy in design.  The motor is over loaded, irregular in sound, warm. I don't see it lasting the rated 6000 hours.  It did put out about 100 ml/minute at nearly no lift as specified.

A low pressure drive of a water vessel is a good idea, as is using a small bubbling of compressed air to pump water up via tubing into the intake manifold, but I'm still hoping for an acceptable quality electric pump.  Otherwise, I'll reconsider.  I'll get the new pump on Friday.

Compressed air is a bit annoying at times- things like regulators, pilot valves, and check valves all leak and that is "normal".  For example a low pressure regulator will occasionally not seat well and slowly leak a lot of air over half a day.  No matter the brand. I do have a variable amount of leakage the entire time the Listeroid is running, since the controller needs air for shut down control and control of the air compressor.  The pilot valve for the compressor leaks a little (again, it varies) and the air compressor unloaders which must be pressurized if the air compressor is belted but not being used. 
Title: Re: WI
Post by: veggie on June 15, 2018, 11:55:32 PM

A bit late to this thread. I just saw your video. (Nice system by the way  ;)  )
Have you considered these small 12 volt pumps to push the water into your misting nozzle?
They are typically used for small fertilizer sprayers. (

A lot of people use them for water injection here in the great white north.
No gravity required and they can be activated by manual switch for temporary dosing,  a timer, a temp switch, or whatever you like.
Control flow with a needle valve in the discharge line.

Title: Re: WI
Post by: veggie on June 16, 2018, 12:13:17 AM
....aslo, if you want to get really nerdy you can add a 12 vdc pulse width modulator for flow/speed control.
I use one of these pumps to push warm thin WVO through my 5 mic filters and I added this little unit to control the flow rate.
Just dial the speed to the flow that you want. Very inexpensive. (


Title: Re: WI
Post by: BruceM on June 16, 2018, 12:46:22 AM
Thanks, Vege.  That's WAY more pump and power use than needed for my desired trickle of 20 ml/minute.  When doing only air compressing, I have 12V from battery, charging via 40W PV.  So realistically 2.5 amps and perhaps 1 amp on a cloudy day, and running off battery when dark. Right now, my run time current draw it tiny, perhaps 40ma total. 

I got the Yosoo brand peristalsis pump today.

The rollers ARE gear driven, it works fine without fussing with the tubing, but it draws 250 ma at 6V.  Not terrible but a far cry from the listing of 30ma.  They are lugging the motor pretty badly with an inadequate gear ratio. Somewhat disappointing.

I spent the afternoon updating my Picaxe engine controller software for the new 40x2 part, which is really a PIC 8 bit 18F MCU.  It allows software reassignment of all input/output pins so now I've got loads of spares.  The new parts are upwardly pin compatible, just some minor software mods needed. 

I do appreciate the child- oriented simplicity of the Picaxe as I become more impaired; never any mysterious serial programming problems, dirt cheap, great tech support.  I've used the Arduino where I needed interrupts, and time critical real time bit banging and RMS AC voltage calculations for my inverter, but alas, their programmers are error prone and flaky, there were serious bugs in their math library and undocumented features in the compiler relative to interrupt routines setting flags.  Basic problems that reminded me why products cobbled together with Open Source software from multiple sources are not my preference when I have serious work to do.  I'll also add that for 5V operation, the PIC 18F series uses roughly half the power of the comparable AVR parts, even when adjusting clock for equivalent MIPS. 

Title: Re: WI
Post by: BruceM on June 16, 2018, 06:17:44 AM
I did some fiddling with the Yasoo pump.  It must be the type you used, Glort, as you can oil it and it does run a bit smoother.  The other roller-friction on the bare motor shaft type just spins the motor shaft without moving the rollers if you oil it. 

It can't be slowed down by voltage much- won't start or run below 5V...and still has way more flow than needed for my 6/1.  I tried some smaller thin wall silicone tubing but it didn't work at all.  Ends up the wall thickness is critical so that the roller closes the tube completely.  Another slightly smaller tube with the wall a bit thicker was too thick and would stalled pump.  The tubing when squeezed in the space between the rollers and the cavity wall has to be JUST right.

I'm in no hurry so will look some more.  What I want is 20 ml/min at 36 inches of head. Car washer pumps are gross overkill for a Listeroid WI, I think.  I have an old MB one from my parts car I can try on reduced voltage to see how it goes.

I did experiment with compressed air bubble lift.  Alas, it requires a substantial depth of water to work so isn't practical for a bucket lift.

Still pondering the best solution for my needs.

Title: Re: WI
Post by: dieselspanner on June 16, 2018, 08:54:20 AM
Back in the day I had a VW camper which had the screen washers driven by compressed air from the spare tyre, so....

How about a system using a tyre as the air reservoir (the pressure won't vary much over 10 or 15 litres of dosing), using a needle valve to control the flow with the fuel rack opening and closing a valve?

Engine stopped, no flow, fuel rack open, off she goes. Should you be worried about the motor stalling with the rack wide open, mount the inlet as a drip over a 'sump' with a drain fitted with a valve held closed by the vacuum in the inlet manifold. As Glort said, worst that can happen is you run out of air or water.

Simple, cheap, no electrics at all and there's probably enough bits in the 'it'll come in handy pile' to knock up a test rig for free.

Cheers Stef

Title: Re: WI
Post by: BruceM on June 16, 2018, 04:42:50 PM
I had the same idea about a funnel; I may play with it today.  There is an endless low pressure air source available without sipping on my 500 gallon compressed air tank... that's the exhaust.

RC model airplane folks use the exhaust pulses tapped from the muffler to run a diaphram pump for pumping small amounts of smoke fluid into the preheat/muffler.  A servo actuated smoker valve controls the output flow.

Another pump off of crankcase pressurize pumps fuel in aerobatic models to insure uniform fuel feed. Some use the simpler system of pressurizing the fuel tank via muffler tap.  They also had tiny check valves, as I remembered.

Alas, the RC diaphram pump is for very high rpm engines, it wouldn't be suitable for WI on on the CS engine. 

Pressurizing the water storage tank just enough to allow the intake vacuum to finish the job is one way to go.  A 3-5 gallon vessel that can handle 1.2 psi yet be cheap and easily filled is a tough nut.

The CS exhaust pressure is likely more than sufficient to pressurize this tank.

Pity the all plastic peristalsis pumps aren't readily modified.  With a tiny geared down motor they would be sweet.

Pulsed 12V to the 6V pump would preserve some torque, though at the price of motor heating.  Problem is all the available pumps I can find so far are 10-1000x the needed flow; too far to adjust by PWM, really.

I still have gravity. Even when I get to where can't lift a 3 gallon bucket I could always feed the supply tank with a little submersible pump from a bucket on the floor. 

Our first rain in months today. A good day to look at giant scale RC gadgets for fuel/smoke pumping to see if anything might be suitable.

Title: Re: WI
Post by: BruceM on June 16, 2018, 06:50:01 PM
Sweet little exhaust-diaphram fuel pump, Glort!
No penalty for flow restriction on these, so it should do the 20 ml/minute rate.  Water/meth should be no trouble either. 
Title: Re: WI
Post by: BruceM on June 16, 2018, 07:04:32 PM
Alas, those pulse diaphram pumps are run from pulsing vacuum- hooked up to the gas engine intake manifold I'd guess.

The CS intake manifold vacuum isn't likely strong enough to operate one. 
Title: Re: WI
Post by: BruceM on June 17, 2018, 02:47:06 AM
Is is likely to be able to draw up water 36 inches?  Sure looks like a sweet solution- didn't realize they existed for regular engines...I'd only used them for RC engines when I was a young man. 
Title: Re: WI
Post by: BruceM on June 17, 2018, 05:27:21 AM
I'll have get one and test it, it might be just the ticket.  Thanks for the great idea, Glort!
Title: Re: WI
Post by: ajaffa1 on June 17, 2018, 08:44:06 AM
Good thinking guys, my ride on lawn mower uses one of those as a fuel pump, it ran faultlessly for 6 years until the engine head gasket failed and the crankcase pressure increased. Bloody thing vented hot sump all up my leg. Replaced the head gasket and flushed out the diaphragm pump, runs like new.

Not sure how much pressure there would be in a Lister crankcase because they already incorporate a spring steel diaphragm to release the pressure. If it works you could fit two one for WI and one to pump fuel rather than gravity feeding.

Very interesting.

Title: Re: WI
Post by: ajaffa1 on June 17, 2018, 10:58:37 PM
Most small 2 stroke have a small diaphragm pump built into the bottom of the carburetor. They have a lift pipe and a return to the fuel tank. They shift a lot of fuel, as you will know if you have ever had the return pipe break or blow off.

Only trouble with a four stroke with this sort of WI is that you are going to get two injection pulses for every combustion stroke, doubt it would matter but you might want to experiment with a needle valve for correct dosing.

Title: Re: WI
Post by: BruceM on June 18, 2018, 04:16:15 PM
I've got a pulse pump on order; too interesting to not try out.  Bob raised an interesting point; with 2 diaphram actions per crankcase per intake gulp, you do get dribble into the bottom of the intake manifold unless you implement a little shallow dish to catch it.   That's the advantage of Mr X's vacuum fed approach- all water is entered into the peak intake airstream.

 I also ordered a kit of RO tubing, float valve and shut off valve for $13. so that I could do a gravity fed constant water level version of the intake manifold vacuum fed WI.  The diaphram pump could also feed a small constant level tank- by just having an overflow port to drain back to the supply bucket.  All quite cheap to implement so worth some dabbling to find a nice setup for CS WI.
Title: Re: WI
Post by: BruceM on June 21, 2018, 04:12:24 AM
I got the vacuum pulse type diaphram fuel pump today and tested it hooked to the 6/1 crankcase.  It works nicely but has a total lift of only about 13 inches of water (zero flow); intake and outlet combined, it matters not where the pump is between them.  At 10 inches it pumps a steady flow way more than needed, flow starts falling off rapidly from there. These pumps are direct single diaphram types so pumping lift can be no more than the vacuum in the crankcase.  I'm sure some CS engines with less hours may have better crankcase vacuum, but that's all I've got.  It's lower than typical gas lawnmower engines because the crankcase volume of air is so large relative to piston displacement volume, I think.

So I decided that gravity is pretty reliable and cheap and I've had enough pump fiddling. The float valve and plumbing kit came along with the diaphram pump. I've got the two tanks made up with fittings, float valve and plumbing mounted.  Tomorrow I'll do some flow rate adjustment testing and then will add my little solenoid valve for Picaxe control.  I sleeved down the 3/16 ID brass tube in intake manifold to 1/16 ID for starters. I'll have the top of the regulated water level in the float filled tank just below the brass intake manifold inlet blow the manifold, so I can see the static water level in the supply tubing. The intake manifold will pull it up the last 1.5 inches.  I used clear Sterilite storage containers (that I had on hand) for the two water tanks so monitoring them will be easy. The 5 gallon supply bin is intended to be filled to 3 gallons and low enough that I can fill it easily.  The big lid will allow me to fill it from a 5 gallon bucket.

I'll do photos and/or video once I get the bugs worked out,

Title: Re: WI
Post by: EdDee on June 21, 2018, 10:29:07 AM
Hi Bruce,

I am using a gravity/float bowl type setup....has worked flawlessly for me in the past... (I didn't want additional pumps and things that could go wrong....)

I have a 3mm copper pipe that sits center of airflow in the air intake, coaxial to the intake manifold. The copper pipe stops slightly inside the intake port on the head (1/2" or so).... There is no constriction at the end of the pipe, free flowing...

The float bowl can be height adjusted so the level of the float bowl is around 4 to 6" below the level of the intake center (This will vary according to the length of water feed pipe you use, your air filter, and of course the air intake plumbing... also, be careful to prevent multiple dips in the water feed pipe, cumulative head is a bugger on a low pressure system) .... Adjust this height to change water flow. As soon as you restrict the pipe and put a nozzle on it, it gets prone to blocking... There is little difference in effect from mist to droplet, but droplets do work better.

With the above arrangement, there is no chance of flooding the engine, as intake draft is required to induce water...set and forget.... Oh yes, one small addition was a small valve on the water line to allow for water shutoff, seldom used, only when the tube was withdrawn to prevent spillage ultimately...

In the beginning, I shut the water off before shutting down and turned it on after warm up, ultimately, I ended up simply leaving the valve open, there was no ill effect on my system either...

Title: Re: WI
Post by: BruceM on June 21, 2018, 03:43:17 PM
Ed, Thanks for the info on your successful gravity-float valve-vacuum WI setup.  I'm intrigued that you had about 5 inches of water vacuum; I'm guessing the engine in question isn't a CS type? Perhaps you were thinking 4 cm which is just what I measured.  I measured my 6/1 Listeroid running intake vacuum at about 1.75" in the middle of the intake manifold and airstream.  Your coaxial tube projecting into the head inlet (good idea and easy to implement ) might be that much more effective but it intuitively seems unlikely to me. I'd really like to know if this 4-6" of water vacuum at inlet is on a CS type engine!

I did note that my engine didn't complain a bit about modest rate WI when cold, but I didn't try starting cold with water. Like you I like things as simple as practical. 

Title: Re: WI
Post by: BruceM on June 21, 2018, 05:56:53 PM
All these fuel diaphram pumps are vacuum pulse driven. You'd have to change the design to run off pressure- the spring would have to be on the opposite side of the diaphram, where presently there are check valves, and the diaphram material would have to be stiffer to allow higher head. 

Didn't sleep much last night so I'm not sure how much progress I can make today.
Title: Re: WI
Post by: BruceM on June 22, 2018, 12:05:18 AM
I got it working and the flow rate set.  I ended up removing the flow restriction at the intake manifold, back to 3/16 ID brass tube as shown before in this thread.  With 1/16 inch ID at the intake manifold, I was only getting 1 ml/minute.  I kept removing the sleeved tubing one at a time until they were all gone.  My tiny pneumatic solenoid valve is pretty restrictive for water, cuts the flow in half at any giving water level height, but it's free so I'll use it for now.  The long 1/4 OD (RO type) tubing did restrict flow also compared to my prior test with a short loop of larger ID tubing. For a larger than 20 ml/min flow, I'd suggest larger tubing.

I now have water level marks for about 10 and 20 ml/minute flow rates; the float arm is adjustable and I just change that for the desired flow rate.  Photos when it cools down.  It was tedious to keep fiddling and using a 60 ml syringe body for the water supply to see water consumption- timing 2ml drop. 

I'd like to add an inline filter to catch any crud in the water- any suggestions for a small inline fuel filter that won't clog when used for water and water/methanol?

Title: Re: WI
Post by: EdDee on June 22, 2018, 01:32:03 PM
Hi Bruce,

In reply to a previous question, it IS a CS style roid, 12Hp at 1000rpm, now running at 9-10hp at 750rpm...

The inlet manifold is 1 1/2" or so galv pipe, leading to a large truck type air cleaner with little input restriction... 4.5" of water is what it pulls on the pipe, but, my error, that is to the bottom of the float bowl, the water level is about 2" above the bottom, so around 2.5" actual pull.... my apologies!

Title: Re: WI
Post by: BruceM on June 22, 2018, 03:20:03 PM
Thanks, Ed.  That's consistent with my measurements given your increased RPM. You were right on the mark with adjusting water level to achieve the desired flow rate, thank you.

I'm not thrilled with the float valve I'm using.  Instead of a sharp cutoff I get dribbling, then dripping, with half an inch of rise in water level beyond dribbling to a full shut off.  About 3/4" difference total between initial and later running water level.  That's more variation than I want as it represents a doubling in flow rate.  Not a show stopper, but annoying.  I'm going look into it further.  For now I'll adjust so it starts at 25 ml/min and tapers down to 15.

Title: Re: WI
Post by: EdDee on June 22, 2018, 05:10:12 PM
Hey Bruce...

Dont stress re the dribble... thats a progressive thing with the float valve scenario...

Run it, see where it averages and set level accordingly....

Title: Re: WI
Post by: BruceM on June 22, 2018, 05:32:25 PM
Thanks Ed.  I am going to add an overflow pipe to the float tank- so that float valve failure doesn't flood the floor and intake manifold. 

It's right on the wall, so a 1/2" tube through the wall is a snap.

I've written to tech support at CDI- they have bought up all the float valve companies in the US so hopefully can make a good suggestion as to what might work best for my low pressure feed, low flow rate application. 

The consolidation of small companies in the US is alarming.  I've seen this in electronics as well-  6 good filter companies all bought out by the same company, product lines slashed, prices more than doubled with lead times of 4 months, no stocking of the formerly stocked products.  The purchasing company is awful, no decent engineers, no good tech support, a bunch of bullshit artists.
Title: Re: WI
Post by: dieselspanner on June 22, 2018, 05:56:02 PM
Another idea occurred whilst plodding around the lanes on me old tractor this afternoon..........

Most of you guys are running 'Roids, presumably with taper roller bearings and no oil pump. how's about mounting a small piston pump to drive the WI in the same spot?

Given how little power it would take running one off the exhuast or inlet cam follower might be a goer too..........

Title: Re: WI
Post by: BruceM on June 22, 2018, 11:15:39 PM
Here's my gravity fed, float valve/tank - intake manifold vacuum water injection setup.  My next step is a software update for 10 minute delay after running to allow for warm up, and turning it off whenever when no-load or shutdown condition is detected. I also have to add a  Mosfet to my interface board  for the solenoid valve control, and run wire to and properly mount the solenoid valve.
Title: Re: WI
Post by: BruceM on June 23, 2018, 11:36:19 PM
A question for our more experienced WI users:

My Listeroid 6/1 engine is automatically shut down when running with no load for more than 10 minutes. Would it be a better or worse to do WI during no load time.  Either way is just as easy.  I note that at 20 ml/min WI there is no noticeable change in running at cold idle in 9OF temperatures. 

The next is- should I wait for a 10 minute warm up (winter AM starts at 15F) or just turn on the water after starting.  Again, either is just as easy, the question is what's best.  Normal run times are about 1-1.5 hrs, with occasional 3-4 hour runs (woodworking or pneumatic string trimmer). 

Title: Re: WI
Post by: BruceM on June 24, 2018, 02:17:48 AM
Thanks, Glort.  I'll wait for 10 minutes before turning on WI, and will only turn it off just before shut down.

Title: Re: WI
Post by: BruceM on June 24, 2018, 04:44:18 PM
One more question:
My well water is moderately hard (magnesium, calcium) but very good quality drinking water. Is that going to be acceptable for WI? 
Title: Re: WI
Post by: BruceM on June 25, 2018, 01:11:33 AM
Thanks, Glort.

There is a white deposit in the teakettle after some months- it's moderately hard water.  I will probably run it for a while then pull the head and check things out visually. If the water flashes to steam I'm hopeful the minerals will be blown out with the exhaust.

Collecting rain water isn't very satisfying here.  We used to get 15 inches a year but now it's less. The bulk is in August thunderstorms (deluge).  We can go 6 months without measurable rain. I do have gutters so it's possible...but bugs are drawn to water like crazy here since it's so damn dry.  Single digit humidity and around 95F with winds all the last month. 

I'm just debugging the software update now- it didn't turn on the WI solenoid as it should have in 10 minutes with a load.  No signal at the processor pin so clearly a software goof. Shoot the software guy.

Title: Re: WI
Post by: BruceM on June 25, 2018, 04:25:18 AM
No fog here.  I could do a solar still but that's more work than it's worth for WI.

I found the WI software update problem-  there was none, I had forgotten that I had pulled down that output pin to ground since it was previously an unused input.  The new hardware and software is working fine now.  I went with a 7 minute with load warm up, and water off when there is no load.