Pocketbike Forum banner

1 - 18 of 18 Posts

Registered
Joined
3 Posts
Discussion Starter #1
Hello. New to this forum.
A few months ago I started working on this 39cc "pocket dirt bike" that had been lying in backyard under a tree for a few years. A few years back my friend got this bike and we worked on it and rode it back then, but then it broke up and we abandoned it.. but now I wanted to get some fun tinkering so I decided to revive it back from the dead and see how much power I can get out of it. As a background I have played with twostroke moped engines for almost 10 years now, I have a lot of experience in porting cylinders and making exhausts etc. So I found this forum when I was googling what the internet has to offer for these types of bikes. I thought to share my experiences with it. For a grown man it's a pretty ridiculous machine to ride but still quite fun... at first I considered it a complete child's thing but actually it's surprisingly capable and enjoyable to ride.

Here you can see the bike about 2 months back when I first got it running. Seat and foot pegs not attached here. I have made the exhaust pipe about 8 years ago to a different bike and it was modified for this bike about 5 years ago... The front fork tubes are very rusted and other tube is completely jammed so there is no front suspension.

156974



When I worked on this last time 5 years ago, I found that the stock ignition system is quite poor. The spark was weak when starting and also I suspected that the ignition will hit a limit at around 11000rpm, because even though I had made a tuned exhaust and ported the cylinder, I couldn't get it to rev past 11000rpm. So now the first thing I did I made a battery-operated CDI ignition box which creates the spark primary voltage from battery, so I can get full power spark at low revs too and also the ignition is programmable. The system is my own design and uses a microprocessor similar to an Arduino board. I can fully adjust ignition timing curve and guaranteed no rev limiter, tested at over 15 000rpm in other engines.

156975



The stock ignition coil is still retained, I use that to give the CDI trigger pulse. It is done by using the original kill switch wire from the stock coil. The kill switch wire is attached to the primary ignition coil, and gives an AC signal which can be used for triggering a CDI (with proper filtration).

Originally I did not disassemble the engine because I wanted to do this just for laughs and wanted to see if it would run well after just upgrading the ignition and some basic carb cleaning. However I found that the flywheel keyway was messed up and it caused the flywheel to slip losing ignition timing. So I needed to fix the keyway and taper by remachining them in a lathe, so I had to disassemble it. Then I found that the bearings were rusted frozen and the crank was actually slipping in the bearing races. Overall the bike was in very bad shape from lying outside for many years.

156976



I still didn't want to buy new parts other than absolutely necessary, so I just threw all parts in vinegar for rust removal and then ultrasonic cleaner. The bearings spun perfectly well after that and I assembled the engine again. I even reconditioned the old crank seals by swapping the rusted internal springs from some seals that I had lying around...
I glued the crankshaft to the bearing races with thread locker to help prevent slipping.

156977



Then I made a new reed valve block by casting it from aluminium and machining with my lathe and dremel. I have recently learned the metal casting process, it's a great way to make small engine parts in your own garage and very well doable after some reading and learning from youtube metal casters.
The reed valve is a minarelli scooter reed valve from aliexpress, cheapest I could find. But it is still a ton better than the original reed valve which has very poor flow path.

156979



Installed reed block. I also upgraded the stock carburetor to a chinese 19mm dellorto copy. The stock carburetor is a piece of garbage. It can work in a completely stock engine but as soon as you start modifying you will run into problems. You cannot find larger mainjets for the stock carb. First I tried to modify the stock mainjet to an adapter which enabled to use 5mm dellorto mainjets. But then I ran into a problem that the float bowl ran dry because the engine started to drink so much fuel after these modifications. The needle valve and petcock is badly restricted in the stock carb. Immediately after switching to the dellorto copy carb the engine gained huge power and driveability (not because of the larger size, but because of better fuel delivery at part throttle and adjustability).

156980



The stock clutch is in a bad need for some stiffer springs. I modified some piaggio scooter tuning springs to fit in it (have to bend the spring end 90* and elongate the mounting pin in a lathe slightly). Also drilled the shoes. With this modification the clutch engages at a whopping 9000rpm and the bike starts to have some respectable acceleration.


156982



I have also ported the cylinder but not much. Exhaust port is widened and raised and I ground a boost transfer port but the timings were still quite conservative. At this point the bike ran reasonably well and the power is good, I estimate it accelerates to 45km/h (28mph) in about 5-6 seconds with a 85kg rider (187 lbs) and could easily smoke the back tire flat under a minute (doing few second burnouts is a good and fun power test but I want to preserve the tire). I don't have any numbers about stock performance, but I think it has easily double or triple the stock power.

But despite all these modifications it still refuses to rev past 11000rpm or so. When driving it is limited to about 10500 and 45km/h is the top speed with current gearing. It could go faster with taller gearing but I don't want to do it because it will hurt the acceleration. I know that it is possible to get a lot more revs out of it but need to find out what is restricting it.

Latest modification is to raise the cylinder by 1mm with a spacer plate and change the cylinder head.
This brought the timings to 190 exhaust and 133 transfer.

156983


I tested it with the spacer plate and it only brought the power down and no help at top revs!
Power loss was due to the combustion chamber volume going up. Then I knew that I need to make a new cylinder head.
I machined the stock cylinder head off from the cylinder. Then I cast some scrap aluminium into a tin can and machined a new cylinder head out of it.

Here is the aluminium block that I cast and the piece of stock cylinder head that I sawed off the cylinder...

156984



New head machined. Now I can get a proper combustion chamber and smaller 10mm thread spark plug as a further optimization.
The volume is very small. I want to experiment with a ridiculously high compression ratio and see what is the limit. I calculated that with current ~1mm squish gap the comp ratio should be between 16 and 17:1. The stock piston has a relatively tall dome and influences the volume a lot. The squish gap is still large, but I have found out that the crank bearings are so worn out that the gap will almost close at high rpm. The piston had already touched the stock head 馃榾 It seems that it is necessary to opt for some new bearings if I want to continue tuning.

156981



I have made a series of videos about it. I added english subtitles for first 4 videos, will try to add for the rest too.


My goal is to tune it so that the peak power is around 12000rpm and it can rev to 14-15000 when driving. I want to get a ridiculous acceleration and sound appearance. The 45km/h top speed is already scary but I want to improve that too.

Let me know if you found this interesting.
 

Registered
Joined
3 Posts
Discussion Starter #2
Finished the cylinder head and bolted it on.

157026



It was hard to accurately measure the volume but I believe it is around 2.5cc with the piston dome in account, and with 0.8mm squish when measuring. But as installed the squish is now 1mm because I wanted to test the fitment and starting first before fine tuning it. So the compression ratio should be around 16:1 I think, with the 0.8mm squish it would be 17:1...

When starting it, there is now much more resistance from compression and it also kicks back. I dropped the ignition advance to only 5 degrees when starting, that reduced the kickback. That already tells me the burn speed is at another level compared to the stock head. Then I dropped the advance at powerband from 32 to 20 degrees.
There is a huge difference in power. The head has now restored power that was lost when the compression was dropped by the cylinder spacer. Compared to stock cylinder height, there is noticeably more oomph. Basically the port timings and combustion chamber should be now in order..except I still have room to tighten the squish gap but I have to replace the crank bearings first. There seems to be no detonation but I have run it pretty rich and only short rides.

Max rpm is still pretty much the same. Compared to stock head and cylinder height, there is now maybe 300rpm more. If you wonder how I can determine rpm values, I do it by calculating the engine noise frequency from video playback on computer with an analyzer program(audacity). I think it has topped at 11400rpm when cold. When it gets hot it starts to misfire at top and cannot get over 11k any more. I think it might be due to running rich. I changed a bit smaller main jet and it improved the misfiring, but might be still too rich.
My list to try to improve the rpm has still some things to try:
-carbon reed petals
-smaller main jet
-less ignition advance at high rpm
-more thorough porting on cylinder (reshape transfers)

If these cannot bring it over 12000, I think I need to start cutting the cases or crankshaft :unsure:
I have a theory that the crankshaft rotation blocks the reed inflow at high rpm. At low rpm the intake charge has sufficient time to "hop" between the crank cutouts as it rotates, but as rpm increases the crank webs have rotated in front of the intake before the flow has properly started. I thought I could try to grind "side passages" to the cases or grind the crank webs so there is more space to flow over the crank. Although it starts to get a bit too serious and time consuming (like the cylinder head and cdi box wasn't already)..

Then I accidentally found out the reason for the seemingly low top speed. The clutch is actually slipping at speed. During a test ride one of the clutch springs snapped. When I was experimenting to find a way to install the extremely stiff clutch springs, I had ground one clutch spring a bit thinner from the hook end, so it snapped. Just before it snapped(probably the spring giving way and clutch gaining grip), I felt the bike accelerated easily past the previous 28mph even though the rpm did not change. (didn't have gps on so no new top speed, and I hit the brakes because I was frightened as the front wobbles at that speed). So I have to install a bit less stiff clutch springs.
 

Super Moderator
Joined
2,355 Posts
Consider a Kevlar burst shield for the flywheel 馃榿

Consider methanol race fuel or N2O to overcome the reed inducted motors poor flow path 馃挭
 

Registered
Joined
74 Posts
Really nice project you got here.
Would you share your work on the cdi unit?
Would be great to build my own cdi unit.

Keep it up
 

Registered
Joined
1,598 Posts
Finished the cylinder head and bolted it on.

View attachment 157026


It was hard to accurately measure the volume but I believe it is around 2.5cc with the piston dome in account, and with 0.8mm squish when measuring. But as installed the squish is now 1mm because I wanted to test the fitment and starting first before fine tuning it. So the compression ratio should be around 16:1 I think, with the 0.8mm squish it would be 17:1...

When starting it, there is now much more resistance from compression and it also kicks back. I dropped the ignition advance to only 5 degrees when starting, that reduced the kickback. That already tells me the burn speed is at another level compared to the stock head. Then I dropped the advance at powerband from 32 to 20 degrees.
There is a huge difference in power. The head has now restored power that was lost when the compression was dropped by the cylinder spacer. Compared to stock cylinder height, there is noticeably more oomph. Basically the port timings and combustion chamber should be now in order..except I still have room to tighten the squish gap but I have to replace the crank bearings first. There seems to be no detonation but I have run it pretty rich and only short rides.

Max rpm is still pretty much the same. Compared to stock head and cylinder height, there is now maybe 300rpm more. If you wonder how I can determine rpm values, I do it by calculating the engine noise frequency from video playback on computer with an analyzer program(audacity). I think it has topped at 11400rpm when cold. When it gets hot it starts to misfire at top and cannot get over 11k any more. I think it might be due to running rich. I changed a bit smaller main jet and it improved the misfiring, but might be still too rich.
My list to try to improve the rpm has still some things to try:
-carbon reed petals
-smaller main jet
-less ignition advance at high rpm
-more thorough porting on cylinder (reshape transfers)

If these cannot bring it over 12000, I think I need to start cutting the cases or crankshaft :unsure:
I have a theory that the crankshaft rotation blocks the reed inflow at high rpm. At low rpm the intake charge has sufficient time to "hop" between the crank cutouts as it rotates, but as rpm increases the crank webs have rotated in front of the intake before the flow has properly started. I thought I could try to grind "side passages" to the cases or grind the crank webs so there is more space to flow over the crank. Although it starts to get a bit too serious and time consuming (like the cylinder head and cdi box wasn't already)..

Then I accidentally found out the reason for the seemingly low top speed. The clutch is actually slipping at speed. During a test ride one of the clutch springs snapped. When I was experimenting to find a way to install the extremely stiff clutch springs, I had ground one clutch spring a bit thinner from the hook end, so it snapped. Just before it snapped(probably the spring giving way and clutch gaining grip), I felt the bike accelerated easily past the previous 28mph even though the rpm did not change. (didn't have gps on so no new top speed, and I hit the brakes because I was frightened as the front wobbles at that speed). So I have to install a bit less stiff clutch springs.
Looks like you could use a billet case instead. They are cag cases cnc made to be more designed like a proper reed port engine with the reed intake to be closer to the top rather then the back. Here is one on ebay in Europe. I think it would be a fun build. There was a member here that has been working with one but he hasn't been progressing much with it. I think you could really benefit from it.

 

Registered
Blata quadard. 2 xhinese copy quads. 3 lucky 7's. 1 lucky 7 trike. 1 cag.
Joined
32 Posts
Superb work. Make it look easy mate, not seen anyone casting their own parts before, impressive I'll be keeping a close eye on this thread. Great to see someone really pushing these little engines again. Seems like a lot of the OG builders have gone quite on here lately so great to see new blood taking up the mantle. I've just started building a new engine myself, I'm a good few notches behind you on the knowledge and skill scale but I'm learning. They are very forgiving engines to work with. If I'm honest I couldn't give a toss about max rpm etc. The track we race on is so small and tight its all about acceleration rather than top end. Nice work on the clutch I need to do something similar myself.
Just a suggestion get yourself a 15mm pumper carb set up mate. They've worked brilliantly for me and for on the fly or continues tuning which you will likely be doing they are perfect. Can get the complete set up for less than 拢20. Keep going bud, great work.

Max
 

Registered
Blata quadard. 2 xhinese copy quads. 3 lucky 7's. 1 lucky 7 trike. 1 cag.
Joined
32 Posts
Looks like you could use a billet case instead. They are cag cases cnc made to be more designed like a proper reed port engine with the reed intake to be closer to the top rather then the back. Here is one on ebay in Europe. I think it would be a fun build. There was a member here that has been working with one but he hasn't been progressing much with it. I think you could really benefit from it.

Love the look of these, blown the budget this month but will be keeping my eye on one of those. Can these be used in conjuction with the BBK and FCC crank kits off ebay, had good success with one in the blata build, thinking this may be the next step. Very interesting. Think it was EPR who was building one, hopefully he may see this and chime in.

Max
 

Registered
Joined
74 Posts
Love the look of these, blown the budget this month but will be keeping my eye on one of those. Can these be used in conjuction with the BBK and FCC crank kits off ebay, had good success with one in the blata build, thinking this may be the next step. Very interesting. Think it was EPR who was building one, hopefully he may see this and chime in.

Max
According to the seller's website it's possible to install 33mm or 35mm stroke crankshaft without modifying.


 

Registered
Blata quadard. 2 xhinese copy quads. 3 lucky 7's. 1 lucky 7 trike. 1 cag.
Joined
32 Posts
According to the seller's website it's possible to install 33mm or 35mm stroke crankshaft without modifying.


Mate what a website, has loads of performance parts, spares etc. Is it in Italian, im kind of navigating around. My question would be items such as rhe bbk? Are they actually of a higher quality than the ebay chinese stuff or is it all the same?
A great source of spares though mate, thanks for the heads up.
I'm going to put together a shopping list for payday, lol.
 

Registered
Joined
74 Posts
Mate what a website, has loads of performance parts, spares etc. Is it in Italian, im kind of navigating around. My question would be items such as rhe bbk? Are they actually of a higher quality than the ebay chinese stuff or is it all the same?
A great source of spares though mate, thanks for the heads up.
I'm going to put together a shopping list for payday, lol.
Yes very interesting performance parts on that website.
I have no idea if it's better quality.
Their full circle crank is 33mm stroke.
So that crank, billet case, rubber intake for 18/21mm dellorto carburetor would probably be very nice I think.
 

Registered
Blata quadard. 2 xhinese copy quads. 3 lucky 7's. 1 lucky 7 trike. 1 cag.
Joined
32 Posts
Really good website. Yeah be a killer combo. Im pretty sure looking at it most of the performance stuff is the same as ebay so you can find it cheaper on there if your prepared to order direct and wait. But loads of other good bits like quality crank bearings, FCC and this little beauty!
Not seen this before and 70cc can that be right? I see one of these going on the Christmas list and straight into the offroad miniquad. Its even got the reed plate on the top like the polini series 2. Thats quite interesting. For 200quid brand new. I'm going to have to get one and find out. Top work mate.

Max
158053
o
 

Registered
Joined
74 Posts
Really good website. Yeah be a killer combo. Im pretty sure looking at it most of the performance stuff is the same as ebay so you can find it cheaper on there if your prepared to order direct and wait. But loads of other good bits like quality crank bearings, FCC and this little beauty!
Not seen this before and 70cc can that be right? I see one of these going on the Christmas list and straight into the offroad miniquad. Its even got the reed plate on the top like the polini series 2. Thats quite interesting. For 200quid brand new. I'm going to have to get one and find out. Top work mate.

Max View attachment 158053 o
I'm thinking they used that 39.7mm stroke crankshaft and a air-cooled scooter cylinder. Bore is 47mm s氓 it would be 鈮68cc




 

Super Moderator
Joined
2,355 Posts
Before you buy all of those parts to make the CNC billet case work!

And try to jam the billet aluminum case on to the Cagllari Daytona and realize that the carburetor is about where the seat is supposed to be and realize that the exhaust won't fit.

and not being able to purchase a clutch that can handle the power and constantly disassembling the engine to replace the wrist pin bearing .

You can get a used BZM or Polini engine and build off of a proven power plant!!!

The CAG engine is fun to build off of but when if starts costing $$$ to get the power that other engines have stock is it honestly worth it ???
 

Registered
Joined
1,598 Posts
So that 70cc engine is meant for a scooter. That case was not mean with a pocket bike in mind. I would want a cag case where the intake reed port is on the top like that case but more flat like a polini series 1 case then the carb and intake manifold might fit with an exhaust. I would love to cast my own cag case to but i don't have the right tools to do so. The cag case is so inefficient it bothers me.
 

Super Moderator
Joined
2,355 Posts
If someone wants to do the design work I will print the crankcase out for test fitting then we just have to send it to China or something for it to be cnc'd

It would be a lot of work but in the end it would definitely be worth it.


Notice in the picture of that way overpriced engine you can see the crankcase halves don't fit together properly this is the problem that other members have reported that have purchased those casings the other thing that members have reported is the bearing wells are not deep enough meaning the crankcase needs additional machine work after you get it to make it functional.
 

Registered
Joined
412 Posts
So that 70cc engine is meant for a scooter. That case was not mean with a pocket bike in mind. I would want a cag case where the intake reed port is on the top like that case but more flat like a polini series 1 case then the carb and intake manifold might fit with an exhaust. I would love to cast my own cag case to but i don't have the right tools to do so. The cag case is so inefficient it bothers me.
This case is made for cag bike. It has the same fittment as the standard cag engine, you will only need an exhaust like on the elite, what is under the engine.
The biggest crank has 39,7mm stroke and every cylinder for minarelli engine fit on the crankcase. But it isn't worth, buy a polini 6.2 engine with a 5 port cylinder and you got the same power. And it's easyer to fit the polini into the cag than get a exhaust that fit into the cag frame. Also it's at the end of the day cheaper to buy a complete euro bike and you know that you buy quality.
 

Registered
Joined
3 Posts
Discussion Starter #17 (Edited)
Good discussion here. Glad that my project brings some interest.

Buying new cases is definitely not an option because I want to do this on low budget and humor basis :D
The more power I can get out of it with original or diy parts the funnier it is.
I agree that the stock cases are poor and the reed valve flow path is probably a bad restriction. I have thought about casting new cases, I think it would be quite possible to do it and machine them with just my lathe and drill press. The cases are so small and simple. It's just a lot of work and time consuming, I have too much other projects too. But I'm definitely thinking about it...
Other possibility is to try to figure out a new creative way to attach the reed valve block on top of the cases close to the cylinder but that's not very easy to do either.
But first I'm going to try "blueprinting" the cases as it is mentioned quite a lot on this forum. My view is that it is not going to help much.
Haven't done much for the bike lately. I tried a smaller main jet and retarded ignition to 15 degrees but it still cannot rev past 11500. Next I'm going to buy carbon reed petals and see what they do.
Here is my newest video, still without english subtitles though. But at 2:24 and 16:40 you can see a little driving and burnout. It is super fun to burn the rear tire with it, but there is so little rubber on it that I don't want to do it for long, otherwise I would need to replace it all the time...


I also noticed that the cylinder plating is a bit worn at some spots. I guess the chinese did not use a proper nicasil coating for it. It looks more like a chrome coating to me. Due to this issue, I wanted to test what would happen if I removed the lower piston ring. This will reduce friction and preserve the cylinder a bit longer. I was expecting that reduced compression and leak from transferports through the piston ring groove would cause a noticeable power loss, but I didn't feel any difference in power. Maybe even the opposite, I could almost swear that it had actually more compression when I tried to start it with only the top piston ring, but I don't have any measurements. But I'm pretty sure it doesn't wear the cylinder as fast now.
 

Registered
Blata quadard. 2 xhinese copy quads. 3 lucky 7's. 1 lucky 7 trike. 1 cag.
Joined
32 Posts
Sorry, didn't mean to derail things, just hadn't seen that engine before and thought it was interesting. Thsnks for the info chrisi, good to know.
I'm all about the budget philosophy, that's kind of the whole point with these things, the chinese stuff anyway, cheap fun. And trying to make power cheaply.
If by blueprinting you mean smoothing and flowing the inlet through the case and the head I did that faintly for the first time to my last build that engine is very strong. I'm doing the same to the next one, this is a totally stock case and head and piston, so we will see how it runs. I would say its definitely worth doing. Every little helps and it all adds up.
Yours seems to pull nicely, acceleration is good. I've played around with the rocket keys, advancing seemed fatser to me but the track we race on is very tight with only two shortish straights so its all acceleration, the advance seemed to help with that. I'm no expert that's just my take.
 
1 - 18 of 18 Posts
Top