Nerd's Webster build

[werowance]

and when it comes to starting it, seems you need to be able to spin it over fast before it will start to pickup and take off.

 

[awake]

and when it comes to starting it, seems you need to be able to spin it over fast before it will start to pickup and take off.

My Webster experience has always puzzled me - I have found that my cordless drill, on its low speed setting, will NEVER get the Webster to start; I have to use the high-speed setting on the drill. Meanwhile, though, if the carb is dialed in right, I can start it by hand, holding onto the rim of the flywheel and yanking it over. I wouldn't think the result would be very fast, surely not even as fast as the low-speed setting on the drill - ???

 

[werowance]

its probably just mine not dialed in good. i use high speed on my dewalt drill and the rubber disk against the fly wheel.

 

[awake]

But as I said, slow-speed drill won't start mine either - much to my surprise. I'm still puzzled over this one ....

 

[Nerd1000]

Quick update: the engine is mechanically done, aside from some repairs needed (more on that later).

So to get it running I need fuel and spark. I've made a start on a simple carb but not much opportunity for machine work lately. So I've focussed on the other issue:

This horrible dog's breakfast is my proof of concept for an electronic ignition. The heart of the arrangement is a coilpack from a Honda Jazz (aka Honda Fit) that I bought on eBay for the princely sum of $23. These are what's known as a 'smart' coil, meaning that all the electronics for switching the coil and making a spark are in the coilpack itself. To operate the coilpack needs 12V power, a ground to the chassis and a +5V command signal: the coil will charge for as long as the +5V signal is applied, then trigger the spark when it's cut off. In my case the command signal is coming from an Arduino nano, which in turn is monitoring a hall effect sensor. At the moment it just pretends to be a 'Kettering' system by applying +5V whenever it senses the magnet. I had lots of issues with interference from the spark resetting the Arduino, so as you can see I've gone to some lengths to shield things and added a large capacitor between the power rails to try to filter some of the noise. The biggest improvment came from ensuring a good connection between the plug lead and the coil, the sparks jumping there were making a lot more noise than the plug itself.

Of course once I got this working I couldn't help but want to try it, so I sprayed some starter fluid into the engine and spun it over with a drill. After some mucking about it did fire a couple of times... And a bit after I noticed the valve timing was out by 180 degrees. It seems that the retaining compound holding the crank throw onto the shaft has sheared from the force of the ignition.

So I now need a better solution. I was thinking I might knurl the end of the shaft and press the throw on, with retaining compound as well. This seemed to work for the crankpin anyway. Does anyone have a better suggestion?


Finally, given the ongoing discussion of piston rings I thought this was interesting. Here's the cylinder bore after some 'running in' under drill power. I was quite interested to see how how the honing has changed into an almost polished surface, you can just see some fine grooves in it which I hope are the 'valleys' of the correct plateau finish.

 

[awake]

Duplicate post - can't delete it, so editing out the text.

 

[awake]

I'm impressed that you were able to suppress the noise sufficiently to get the breadboard to work. Well done!

FWIW, I TIG welded the crank throw to the shaft (and the crank pin to the throw) - nothing crazy, just a small, quick weld to lock them together without introducing significant distortion. I can't imagine I didn't get some distortion, but it was minimal and not something that I can detected either by eye or by the functioning of the engine. (I've never put a dial indicator on it, since it seems to work just fine - "if it ain't broke ....")

Also FWIW, I used keys to secure the flywheel, timing gear, and ignition cam to the shaft. The rotational position of the flywheel doesn't matter, of course, but for the other two, this meant trusting that I had worked out the correct positions and machined the parts correctly to get the timing right. I did make it so that I can adjust the ignition timing, so that just needs to be close, but the valve timing needs to be REALLY close from what I can tell. I don't know how close I got to the optimal timing, but I did get a result that runs well, so I'm happy with that. Keying the shaft to the throw would also be a possibility, but again introduces another point at which the design and the machining really has to be right.

 

[Nerd1000]

Yes I'm also considering just reassembling with more Loctite and then drilling to install a Dutch key.

 

[Nerd1000]

I decided to go with the Dutch key option. Crank was reassembled with more Loctite, then after it set I drilled a 3mm hole at the join between the parts. I then cut down a dowel pin to fit and loctited it in. All this at the price of a 3mm drill, which I accidentally ran into the vice jaws ☹

 

[Nerd1000]

Well, it runs!


Apologies for the shakey video.

I had to give up on the Arduino ignition. It worked for a a few sparks but I think the ongoing interference actually damaged the Arduino even though it was able to operate without glitches. Also I totally bungled the soldering when trying to build a more permanent board...

As a plan B I adapted the arrangement seen here: Simple Transistor Ignition System to use the hall effect sensor I already fitted, and to deliver 5V control pulses to the smart coil rather than directly driving an ignition coil. It seems to work well.

I also built a simple carburettor, which deserves its own post when I get the time.

 

[minh-thanh]

Congratulations !

 

[awake]

Congratulations!

 

[werowance]

congratulations. seems to run smoothe.

 

[ZebDog]

Another Webster run congrats

 

[CFLBob]

Congratulations! Now what's your next engine?

 

[Nerd1000]

I think I'll get the Webster set up properly, then I've got a lot of other projects to finish (e.g. a Hercus 9" model C lathe I've been rebuilding) before attempting another engine.

My plan for the Webster is to set up a little generator for it to drive (I like the idea of charging my phone with it just for fun). But first it needs some housekeeping, e.g. a case for the ignition electronics and (first priority) a muffler. My partner has declared the open exhaust too loud to tolerate.

I spent today repairing the flywheel, I made some adjustments to the valve timing but neglected to put the second roll pin in with the consequence that the thinnest part of the flywheel (where the gear presses in) exploded during a test run. Thankfully it wasn't a complete write-off, I was able to replace the damaged section with a press fitted steel plug that accommodates the gear. Clearly someone had to be that dumb and today it was me

 

[Nerd1000]

Has anyone else had problems with the roll pins working loose on the Webster? This has happened twice now, thankfully no damage the second time despite both pins escaping at the exact same moment!

I measured the pins and they seem to be the correct size for the holes (about .132 for a 1/8" hole, though there is a lot of variation from pin to pin). Wondering if the pins I bought are no good, or if I should find some extra way of retaining them. Alternatively I could try to replace them with bolts, I suppose.

 

[CFLBob]

I have not seen that issue.

 

[awake]

Has anyone else had problems with the roll pins working loose on the Webster? This has happened twice now, thankfully no damage the second time despite both pins escaping at the exact same moment!

I measured the pins and they seem to be the correct size for the holes (about .132 for a 1/8" hole, though there is a lot of variation from pin to pin). Wondering if the pins I bought are no good, or if I should find some extra way of retaining them. Alternatively I could try to replace them with bolts, I suppose.

I did not use roll pins in my design, but did have a similar problem. I used a key for the cam gear, flywheel, and ignition cam. The cam gear and ignition cam each have a set screw that bears on the key, and they have remained secure. The flywheel has two set screws, one on each side, through the central hub and bearing on the key; these had to go in at an angle because I couldn't drill 90° into the hub with the rim in the way. I don't know if it is because of the inertia of the flywheel, or because the setscrews are at an angle, but at first the flywheel would work loose after running for only a short time. I finally put some medium-strength thread locker on the set screws, and that seems to have solved the problem.

 

[werowance]

i have not had that issue either however i did put locktight on them as i drove them in.