Harold Bawlzangya Racing

Author Archive

Hydraulic drum brake conversion

I had been using front only disc brakes ever since I got this cart running.  The biggest problem with the G29 is the front end is so light that the front brakes lock up too easily, thereby reducing their effectiveness.  Putting drum brakes on the rear end wasn’t an easy task.  With the Club Car rear end, bolting up drum brakes is simple, activating them is the hard part.  The G29 came with an internal rear brake and uses a single brake cable.  So that was the challenge – how to make a cart set-up for a single cable work dual cable brakes.  The easiest solution seemed to be switching from cable to hydraulic activation.

I was going to use the switch to drum rear brakes as an opportunity to ditch the 10″ rims and go back to stock 8″ rims by removing the front disk brakes.

I bought a couple of backing plates on eBay.  I noticed that once of them had locating dimples on the back (to better align the back plate with the axle) and the other was flat on the back.

So I made up this little tool to make the dimples on the back.



Here is one of the dimples I made compared to the flat one:

The next step was removing the guide plate where the cable brakes slide.

After a lot of test fitting, here are the welded up backing plates for the hydraulic wheel cylinders.

Final test fitting:

All bolted up:

While I was removing the Jake’s disc brakes, I took the opportunity to install grease fittings on the factory spindles like the Jake’s set-up.


I also had to redo the speed/odometer sensor set-up to work with the “factory” configuration.

With the Jake’s disc brakes, I had the reservoir under the seat near the forward/reverse selector.  That wasn’t going to work with the new system so I moved the reservoir under the cup holder where it is still accessible by removing one screw.

One line connection:

T-fitting bolted to the rear end:

Wheel cylinder connections:

Finally the set-up to the brake pedal:

I was working with scrap SS at this point so this was the best I could come up with to join the pedal to the master cylinder.  I have found that SS bends easier than the same thickness of steel, so I went overkill in beefing this up.

I completed this conversion in early Fall but forget how many miles I started with.  I know I hadn’t turned the mileage over, so this set-up has at least 462 miles on it.

I painted the rims black and am considering some mini moon caps to cover the lugs.

Comet 780 fine tuning

I finally got around to doing a little tuning on the 780 clutch.  My belt wasn’t going up all the way and from looking at pictures of other clutches, this was pretty common.


The picture doesn’t do it justice, but there is a good size gap between the belt and the sheaves.  That gap represents wasted movement of the clutch that doesn’t translate into any movement of the cart.

I took the cover off the clutch and pressed the spider down until I had just a slight gap between belt and sheaves.

I cut down and aluminum ring from some scrap so that it fit properly under the clip.

Now the belt goes all the way to the top.


A while back I did away with the engine set to idle at a stop.  When I was putting everything together, I took out some of the slack in the belt.  Between the tighter belt and the belt riding higher in the drive clutch, there is a noticeable difference in the speed of the cart.  My tach broke so I will have to wait and get a new speed vs. RPM to compare to the original set-up.

Back to a regular carb, this time a 24mm

It was nice having the PZ30 carb on the cart with so many more aspects to control with the jetting.  But the performance was starting to act a little erratic after 800 miles.  With the cheap Chinese manufacturing, I didn’t have high expectations on longevity.

Before going back to the Keihin that came on the GX390, I thought I would try the 24mm 3 circuit carb from VC.

Getting the epoxied, broken off low speed screw out wasnt that bad with some silver solder and a spare PZ30 jet.  Ugly, but it worked.

The flux made a mess though.  It came of easily with some carb cleaner.

The low speed screw on this 24mm carb is a different size than the Keihin/clone carbs.  Since I had the solder out, I used another jet from the PZ30 and made a “new” screw from the original.  The PZ30 jet provided a nice shoulder for the spring.


All cleaned up with the adjustable main jet.

On the cart:

It fired right up and with just a little time spent adjusting the jetting, it is running pretty well.

Last ride before Michael hits

One last ride before the rains from Michael start hitting us.

It’s October and still 90 degrees!

It is the first week of October, the leaves are falling and it is getting dark at 7:30 pm, but it is still 90 degrees in the afternoon.

With the warm weather and no rain, I have had no problems driving the cart around.

Back to original Comet clutch combination

My Chinese Vegas Carts 780 clutch had these weights and yellow springs (who knows what tension or strength they really were).

When I got my genuine Comet 780, I didn’t even try the 106 gram weights with the yellow springs it came with. I assumed it was going to be similar to the CVC clutch.

I first tried the green springs with the 106 g weights and really liked the low engagement and how much more quiet things were when going 20 mph or less due to the engine turning less RPM.  But the little 390 just didn’t have the grunt for that combination.  It definitely struggled on some of the larger hills. On flat ground it was awesome though.

Then I tried the green springs with 91 gram weights.  That was a much better set-up for the 390.  I thought I was going to keep that combination until I had another person on the cart and noticed that performance suffered on hills again. It wasn’t a big difference, but noticeable.  I considered keeping that set-up though because most of the time it is just me on the cart.  But the weather is still warm and thought I would try the yellow/106 set-up that came with the Comet.

This is what the 91 gram weights looked like after about 4 hours of use.  The color on the weights has faded.

Engaging the green springs on the weights was easy, the stiffer yellow springs takes some effort.

Back to the original set-up.

I went out for a night drive last night with the yellow/106 set-up and I am definitely keeping it.  Acceleration and hill climbing is so much better than the other two combinations I tried.  If I had a larger engine, the 106/green set-up would have worked out really well for the type of driving I do.


Misc fuel delivery

Got the new fuel pump from Amazon yesterday.  It is in a Briggs & Stratton box but made by Mikuni.  Not crazy about the housing being plastic, but I’ll see how long it lasts.  I will get a rebuild kit for the aluminum one so I have it as a back-up.

The fuel filters are full of gas again and I can see gas flowing into the top one near the tank.  All good again.

And the vent filter is doing its job.  After two rides and 22 miles, the carburetor bowl is free of debris.

Some of that debris clogged up the carb.  With a combination of carb clearner, guitar strings and compressed air, I got it all cleaned out.

The plug is looking pretty good.  Despite getting near the end of September, it is still humid and 90-92 for afternoon highs.  Hopefully sometime soon I will need to work on winter jetting.

Fuel pump not pumping that much fuel

Despite this being a genuine Mikuni:

It isn’t holding up too well, considering there are less than 75 hours on it.  This diaphragm had gas behind it.

Comet 91 gram weights

I really like the 106 gram weights with the green springs under all conditions except going up steeper hills.  The stock GX390 just doesn’t have the torque to handle the lower RPM operation of the clutch now.  Since I already have the 91 gram weights, I thought I should see how they work before possibly going back to the yellow springs.

I was only able to drive around the neighborhood once and there is definitely an improvement on the hills.  Tomorrow will be a longer drive to get a better idea on how the combination works out.

Keeping crap out

So the last few times I have driven the route where I can go 20mph+ for an extended period of time, I have had the problem of the engine dying.  I wind up letting the engine sit for a few minutes and then I have to use the choke to get it started again.  Fortunately this occurs on my way home and I’m only about one mile from the house.  When I get home, I pull the carb to find debris in the bottom of the fuel bowl.  After the first time it happened, I replaced both fuel filters.  When it happened the 2nd time, I knew the debris wasn’t coming in through the gas line.

I find that there is a vent on the side of this carb.  I added a line and installed a filter that I had used on my dirt bikes for the same purpose.  I added a filter to the over flow line just because I have several of them.   Another ride or two will tell me if this takes care of the problem.

Did a little driven clutch work

Removed a little bit of material on the CVTech driven clutch and got the sheaves a little closer together.  Not a huge difference but a lot of little things add up.



Chinese vs. Genuine Comet 780 clutch comparison

It turns out there are a couple of easy ways to tell if you just received an genuine Comet 780 clutch that was made in the U.S. or a cheap Chinese copy.

A real Comet should come in a box with some paperwork and a couple of parts:

My CVC (Chinese Vegas Carts) came wrapped in bubblewrap.

If you flip the clutch over, on the back of the fixed sheave there should be some numbers cast into the sheave.  Here is the back of the Comet:

And here is the back of the CVC:

One other thing you can do to check without touching a thing, the moveable sheave has three tabs that can be seen from the outside of the clutch.  Here is the Comet:

This is the worn out CVC moveable sheave that I replaced.  Notice no tabs on the outside edge:

If you really want to dig in and start taking things apart, the next thing you can check is what type of threads the clutch is set-up for.  If you remove the dome cover, the Comet will have standard threads, in this case 1/4-20.  The bolts holding down the pins for the roller weights are the same.  My CVC was threaded with metric M6 threads and bolts.

Also with the dome cover removed, you can take a look at the locking plates that are holding the weights in place.  Here the Comet has the locking plate folded up:

And then here is the CVC with the flat locking plate.

The Comet moveable sheave will also have some numbers cast into it.  This also shows the tabs I mentioned before:

No numbers whatsoever on the CVC moveable sheave:

If you keep going and take the weights off, there is a difference there as well.  The weights are designed differently. The CVC on the left has metal guides integrated into the arms of the weight. The Comet weight is on the right.  Plastic spacers fit into those larger openings as seen below. So you have metal-on-metal vs. plastic-on-metal.



Real Comet 780 clutch arrived

The first Comet 780 clutch I ordered from eBay arrived damaged.

But the second one showed up in great shape.

I am going to try the 106 gram weights with the green springs.  So pretty with all the different colors.

Green springs in.

Chinese comet 780 clutch

Despite buying my clutch from a reputable place, it is looking like I have a Chinese knock-off.

I noticed a week or two ago that the moveable sheave of my clutch  was wobbling quite a bit.  With the engine off, if I grabbed the sheave at the 9 and 3 o’clock position, I could rock the sheave side to side.  I ordered a new sheave from beltpalace.com along with some different weights and springs to test out.

My back is still sore and I didn’t want push it by pulling the engine out of the cart. So I dropped the swingarm instead:

The first thing I noticed was that the locking plates weren’t bent over on the bolts holding the roller arms in place.

This is how I pulled the spider off the hub.  The 2nd thing that was strange was that I had to use metric M6 bolts to thread into the spider.  Glad I still keep a container full of bolts from the Japanese dirt bikes I’ve owned over the ears.  I thought that Comet was a U.S. company and expected them to use standard threads/bolts.

And this is how sloppy the hub and moveable sheave are:



The bushings that the guide pins ride in were well worn.  The scary thing is that this clutch has less than 75 hours on it!

So finally on to getting the clutch put back together.  The hub bushing and casting around the bushing were thicker on the new sheave compared to the original:


I tried to thread the M6 bolts into the new sheave and they wouldn’t go.  Turns out the new hub is threaded for 1/4″ – 20 threads.  But it wasn’t until I tried to slide the new sheave over the hub that I that it became painfully obvious that this wasn’t going to be an easy project.  The new sheave didn’t come close to sliding over the hub, the ID of the bushing was too small.

So I sat back and tried to figure out what to do.  The evidence certainly pointed to my clutch being a cheap copy – it wore out very quickly, all the threads were metric and the lock tabs weren’t bent over.  I was feeling pretty confident that beltpalace.com was selling OEM parts or at least OEM spec parts.  The springs and weights were all bagged as Comet parts.

And if the new sheave wasn’t made by Comet, it was at least a better copy with standard threads.  I finally decided to make this sheave work.   I doubted I could find a knock-off sheave that would fit my clutch and it wouldn’t be worth it as it would wear out quickly as well.  After about an hour of working on it, I finally got the new sheave to fit over the hub.  I had to run out to the hardware store to get three 1/4″-20 bolts.  And after installing the springs and weights, I made sure to bend over the locking tabs.

I also had to take some material off the sheave so it would slide up high enough and not interfere with the spider.

Now I just need to get the clutch put back on the engine.







Dimming the running light

Not only was the running light too bright, but with it mounted so high is was almost blinding at night.

With no Radio Shacks nearby, I got this from Amazon for $6.80.

After experimenting with different resistors and different combinations, I settled on two 2K ohm resistors wired in series.

Now I have a running light:



I am going to finally cover up this opening in the electric dash:

I am going to try and use as much of my failed shroud as I can.  The first cut is always the hardest:

I am going to get the sensor and magnet mounted first.  I didn’t want to have to drill anything on the Jake’s spindles or weld anything, so I mounted the magnet to one of the bolts holding the disk in place.  I doubt that this extra weight so close to the center of the wheel will cause much of an imbalance.

I made up this bracket that will be bolted in place on the spindle:

And here is where this bracket mounts:

I made a cutout for the base of the computer:

Fits well:

I’m using a strap across the bottom two mounting bolts to hold the base in place:

I didn’t polish the face of the plate, but used some finer sandpaper to get some of the grinding marks out:

This computer goes down to centimeters.  I measured the circumference of my tire and come up with 143.5 cm.  I can only add whole numbers, so I went with 144 cm.  Opened the GPS app on my phone and drove around the neighborhood.  This should be about the most accurate test of the GPS as there aren’t any trees in the way and most of the almost one mile loop is straight.

And then I went for a drive on the paths through the woods.  The discrepancy wasn’t too bad:

Holy Crap – that is some serious run-out

A couple of weeks ago I noticed the disk brakes making some noise right at the start of a drive. It would only do it when the brakes were cold and go away after a minute.  Yesterday I figured I would finally look into why they have suddenly started making noise.  I started with the drivers side wheel in the air and when I spun the tire, I was amazed at how much the caliper was rocking back and forth.  I couldn’t believe how warped the rotor was.

I pulled the tire off and then got the rotor off.  When I checked the rotor, it wasn’t warped.  I then broke out the dial indicator.  There was no good place to mount the indicator on the front suspension, luckily the big anvil provided a solid base.

So after getting it set-up, this is what I find:



At this point I thought the hub must be warped.  I popped out the spacer that was short and traded places with the one on the other side.  When I spun the hub around again, the same spacer in a different location was still coming up low.  So now it looks like the spacer is the problem.  I pulled all 4 spacers and checked the hub. It looked good.

Taking the spacers to the table saw top, I found that the shoulders weren’t the same.

I don’t have a lathe or a mill, so I improvised with what I do have – a file and a drill press.

After several hours of taking metal off, checking, taking metal off, checking, this was what I ended up with.  I am feeling pretty good about this.  When I rebuilt two-stroke cranks, the run-out spec was .001″ for most of them.  I am really close to that here:

I am kicking myself for not catching this when I installed the brakes.  I just didn’t hear anything and definitely didn’t have any vibration or pulsating from the pedal.  But now that I look at how the caliper is mounted, I see why.

The passenger side wasn’t nearly as bad, but I spent some time and got the run-out near .001″ as well.


Well that was short lived – shroud testing and fail.

This shroud took longer to make than it stayed on the cart.  Sunday I went to the local elementary school with a laser temp gun and a 10mm socket & ratchet to test the effectiveness of the shroud I made.

The engine was somewhat warm from an earlier ride. I took the shroud off since it was cool and then drove about a mile to the school.  Driving around the parking lot was 1.03 miles.  For all testing, I tried to maintain 20 mph and 3k RPM.

I did 4 laps and pulled over and took a temp reading.  I then did another 4 laps and did another reading. The results were pretty close to each other so I felt good about that.  I simply laid the shroud on top of the engine and didn’t bolt it down.  I didn’t want to impact the temp readings while I took the minute or two to remove the 3 bolts that held the shroud on.   With the shroud on, I drove 4 more laps, removed the shroud and took a temp reading in the same spot as before.  I put the shroud back on and did another 4 laps.  I pulled the shroud and took the final reading.

The results were not what I expected and had hoped for.  With the shroud on, the engine ran about 10 degrees hotter.

The laser temp guns read a little strange when you try to use them on shiny or semi-shiny surfaces.  So this morning I thought I would try something different.  I put one of the prob from my digital thermometers that I use for smoking in between 2 of the cylinder fins and then packed it in place with aluminum foil.

This time I drove around the neighborhood.  It is a mile circle that is relatively flat, so I can maintain a constant speed fairly well.  I thought I would go 15mph instead of the 20mph from yesterday.  I thought this might even the odds between the shroud and no shroud.  I had already been out for a 9 mile drive, so the engine was already warmed up.


With the temp probe in place and the shroud off, I drove around the neighborhood and took a picture of the temperature monitor as I pulled in front of the house.  After a mile at 15 mph, this was the temperature:

Since I could see the temperature without lifting the seat, I bolted the shroud on this time and went around the neighborhood again.  Another mile at 15 mph and this was the reading with the shroud on:

I don’t think my laser temp gun was very accurate and there are certainly some flaws with this digital thermometer test.  But since both methods showed that the engine was hotter with the shroud than without, that is good enough for me.  Clearly there is plenty of cooler air flowing up and around the engine compared to the confined warmer air that the fan is blowing around and down the cylinder head.  I would guess that if you drove 5 mph the results might be different.  But who only drives 5 mph?



Hopefully the last of the stainless steel work

I finally got two projects that have been on the back burner for a while done.  With those completed, all I have are stainless steel scraps and am just about out of gas.

The first project was making a shroud to better direct the airflow from the fan around and down the cooling fins.  I spent way more time than I thought it would take on this.  It isn’t pretty, but it works really well.  I tired to make the cover fit tight in areas I didn’t want air to escape from.  With the engine running, the air from the fan will blow off leaves off the ground directly under the engine.  Forcing that hot air down, instead of just the area under the seat should help cool things off.

This project went way faster than I expected. I’ve been meaning to cover up with mangled hole where I had to make the electric body fit the gas frame.

Luckily I had one piece of stainless that was just big enough to cover it.  I tapped the 4 holes so I could just thread in 4 SS bolts.

It is kind of hard to tell, but the cut end of each bolt is sitting below the surface of the plate.  That way I can weld them together but grind off the excess.

All done:



Mother nature is making it tough to get around

Strong winds 2 nights ago did this:

And then some storms last night did this one:

Oil recycling run

I went on a drive that started as just a trip to walmart to recycle some motor oil.

It was still cool so I went to Napa to get some fuel line and then back to Home Depot to get a propane cylinder.  14 miles later I was back at home.


Oil Filter

I have been collecting parts to add an oil filter to the engine for quite some time.  Winter and then building the stainless steel swingarm/cradle this spring put this project on the back burner.  Then came the PZ30 carb.  I finally decided to commit to adding an oil filter to the GX390 engine.  It also helped that I found the pump I had been wanting on sale.

I decided that I was going to pull the oil from both the front and rear oil drains.  That way oil would feed the pump no matter if I was going up or down hill.  I also figured I would pump the oil out for oil changes, so no need for a drain.

I considered returning the oil back to the engine through the valve cover.  But the more I read about the design of these engines, I opted not to go that route.  I never found anything that identified lack of lubrication to the valve train as being a weak point for these engines.  And from what I have read, the passage(s) that drain the oil back to the lower end are the same passage(s) that feed oil to the top end.  So my biggest concern was that dumping more oil than the engineers planned for down those passage(s) would interfere with the proper lubrication of the valve train.  Finally, I already had a hole tapped in the top of the block where the governor used to be, so that sealed the deal.

I had removed and installed the engine more times than I wanted to keep track of when I was building the swingarm/cradle.  I got curious how long it actually takes me to remove the engine.  Turns out it is just about 23 minutes.  That included removing the seat back because the basket was in the way of the rear panel access opening.

With the engine on the workbench I could get started.

I also decided to remove the oil sensor while I had the side cover off.  I don’t know why I didn’t do that when I removed the governor when I first got this engine.

Plugged that opening and added the rear drain fitting:

Front oil drain fitting:

And finally the fitting for the oil return:

I wanted to use 3/8″ aluminum fuel line to plumb most of this system.  I thought the aluminum lines would help dissipate the heat as the oil moved through the system.  I was really impressed with myself that I nailed the bends on my first try.

But as I sat back and looked at it, I talked myself out of using the compression fittings and aluminum line.  I figured that even if I tried my best to support the lines, the continuous vibrations would eventually cause a leak in one of the fittings.

So I switched out the compression fittings to barbed fittings for rubber lines.  At this point I was still planning on having lines run from both the front and back drain plugs.  I used some clear tubing and connected the front and back to a T fitting.  I experimented a little with tilting the engine up and and down to see how far I could tilt the engine before I got air in the line.  It turns out that I could tilt the engine up to a 24 degree angle and still pull oil through the fitting.

24 degrees is right at the point that oil is still in the tube:

I didn’t think I would be able to tilt it that far and still pull oil, so I made the final revision to the system to just use the front drain fitting and plug the rear drain.  Pulling the oil from just the front will simplify things and will be one less place for a leak to occur.

This is the pump I will be using.  It uses about 3 amps, is rated for continuous duty and is self priming up to about 3 feet.  I got lucky and caught this on sale. It wasn’t that much more than the cheap, noisy pumps on eBay and Amazon.

I wanted to mount it below the oil level so it would never have to dry start it.  There was really only one place for me to put the pump and pull that off.  Luckily I still had a few pieces of scrap stainless that I was able to weld up a simple mount on the front corner of the engine cradle/swingarm.

Side view:

Top view.  The engine is about as far forward as it will ever need to be, but just to be on the safe side, I positioned the pump so I could still slide the engine forward 1.5 – 2 inches.

It was a lot easier to decide where to mount the oil filter.  I welded on a bracket to the frame here:

And that puts the filter next to the engine with enough clearance to remove and install it with no problems:

Turns out I think my concern about dumping too much oil through the valve cover was valid. This pump moves quite a bit of oil:


The last piece I needed was a way to control when the pump runs.  I currently have this timer set to run the pump for 15 seconds every 10 minutes. It is easy enough to change if that combination doesn’t work out.

With the motor off and the seat down, I can’t hear the pump running.  I think I may hook up an indicator light somewhere so I can keep an eye on when it is running.

PZ30 carb dialed in

I thought I had the original PZ30 carb working pretty well.  I wanted to experiment with a few things so I spent $23 and got another one from Amazon.  There were a few areas where this new carb looked to be better made.  At least this manufacturer (right) is making an attempt to look like an original:

I put in the new carb and it was in fact better.  I was able to use a smaller pilot jet and my idle is now a steady 950 RPM.  Plug looks good as well.

I quickly bought another one before Amazon ran out of the ones made by that particular manufacturer.

First night ride

I was only able to stay out for an hour as a big storm was coming in.  One thing I forgot to bring were some clear glasses.  With no windshield, I got a few bugs in my eyes.  I bought 2 LED light strips that I was going to try to incorporate on each side of the bumper.  But I decided to take the easy way out and put one on the back of the basket.  I verified that I have that LED strip wired up as a running light.  It is brighter than I would like.  Down the road I may try to find one that is dimmer.