Have fun and save money by making your own fuel mixer
A schematic illustrating the new mixer that Ed made for his Ottawa (Figure 2 in the text).
I picked up a 4 HP 1922 Ottawa ES 4282 about two weeks after our fall show one year. It was headed for the scrap iron dumpster, as it was left over from the tag sale and no one wanted it. The piston was stuck, the water hopper had freeze cracks and most parts were missing. I thought that I would practice the art of removing a stuck piston and use the Ottawa ES for a yard ornament.
Well, after two years, I got the piston out and after another year got the freeze cracks welded. So, I decided to have fun trying to get it to run while minimizing expenses and made my own mixer.
Figures 1, 2 and 3 show what I think the original mixer looked like as well as the one I made. The simpler the project, the better it comes out for me.
On the original, I was glad to see a simple fuel nozzle coming into one side of the venturi and the needle valve on the other side. The original mixer had a more elaborate air flow design for good engine performance under load and at high RPM, but I plan to run it at no load and at the slowest RPM I can get. So I expect my simpler mixer may work even better for my purposes.
I have two 5-gallon buckets of scrap brass, and in going through them I found a 1-inch ball valve left over from some well work. This seemed ideal as the port on the engine head was 1-inch, and this valve came apart to allow the ball to be installed in it. It also had provisions for the valve stem where the packing box had a 3/8-inch pipe thread that I could use for the needle valve. This worked great as all I needed to do was install a 3/8-inch NP to 1/8-inch NP bushing and a 10-24 rod coupling for the needle. I would have preferred a 10-32 for the finer adjustment, but I think the 10-24 will work with the large knob.
To form the needle, I just chucked it up in the drill press and patiently used a file to form the needle. For the fuel nozzle, I used the tube from an old ballpoint pin after cleaning the old ink out. I took a lot of care to try to get the fuel nozzle and the needle to line up. I did this by putting a drill bit with some bushings snug in the bore where the 3/8-inch bushing went, and using the drill press and drill press vise, I marked the center of the hole for the fuel nozzle. Then, using a drill size for the nozzle, made the hole for the fuel nozzle.
For the fuel nozzle and needle valve to work, the low pressure formed by the air flow through the venturi is needed. I wanted to fill the valve with lead, but do not have the machinist skills or good machines to bore the venturi. Instead, I elected to cast the venturi in place. The trick was to use a stick of sidewalk chalk (1-inch in diameter and 4 inches long) that kids use to write on the sidewalk (Crayola 51-8020). The seats in the ball valve were 13/16-inch in diameter, so I used the chalk for the large diameter of the venturi. I chucked the chalk gently in the lathe and used some coarse sandpaper and a coarse file to take the diameter of the chalk initially down to 13/16-inch, fit checking using the removable ball seat. The snug fit here worked well in keeping the chalk stable while pouring. Next, I installed the chalk in the valve/mixer and carefully marked the locations of the needle and fuel nozzle. The fuel nozzle needs to be centered in the small diameter of the venturi. Then, I went back to the lathe and formed the narrow diameter of the venturi to 5/8-inch, a SWAG.
Since there was a snug fit on the chalk in the ball seats, I filed some vents so the molten lead and air could flow for the pour.
I haven’t done any foundry work, so I decided to use the valve body itself as the caldron to melt the lead directly into the valve. I did this by mounting the valve on a piece of scrap metal to hold it vertical, placing it in my grill and placing coals of charcoal around it. This worked very well, as you can see in Fig. 5.
One of the issues is that the chalk will try to float in the molten lead. So I fashioned a length of perf strap to hold it down. This worked well. One thing I learned is that the bottom end has to be well sealed to not let the molten lead run out. I used a pipe cap with two sheets of aluminum foil to help seal the threads. I filled the void in the pipe cap with chunks of chalk, but I think dry sand would have worked better. The pipe cap did unscrew without much difficulty. One problem I encountered is that the lead started to run out between the 10-24 rod coupling and the bushing. I had packed this void with Teflon tape, but it did not seal adequately. I should have put a washer and nut on the screw used in place of the needle valve. I moved the charcoal to the other side and placed a damp rag on the bushing to cool that area until I was finished with the pour.
I wanted to be able to unscrew the removable ball seat, so I put Teflon tape on the threads and coated the ball seat with carbon black from a candle, much like what is done when pouring babbitt bearings. I used old wheel weights from a garage for the lead and melted them directly over the chalk with a propane torch.
After chipping out the chalk with a small screw driver, I used a wood carving Dremel tool to remove the lead that filled the vents and any extraneous deposits. The last bits of chalk were cleaned off with sandpaper. I had to make some alignment adjustments to the fuel nozzle and then solder the nozzle in place.
As of this writing, I still have to make or find the exhaust lever and exhaust lever bracket ... anybody have them? I also need to make an ignition system and skids, so I have not had a chance to actually run the mixer yet. If anyone has comments or something to share, send them to firstname.lastname@example.org.
Contact Ed Stoller at email@example.com.