Project TaoTao : ATM50-A1 Four-Stroke Experiences & Upgrades

[190mech]

Great vid as always Brent!Kinda makes me want to get a 4 stroke,,NAH!!!

[90GTVert]

Thanks. lol You'd be too bored building 4T exhausts... although it would be nice to see a GY6 50 pipe that was seriously designed instead of just a large pipe and loud can.

[190mech]

I read a comment from an expert tuner that the stock GY6 50cc muffler gave the most power of any exhausts...

[90GTVert]

I know the stock pipe on the 2T surprised me. If it worked to 96cc (as far as I tried it) then maybe the 4T pipe works well for big bores? I've got an eBay pipe here, so someday we'll see what happens on the TaoTao.


I wanted to do a compression test, just so I have numbers to compare if I ever need them. I also wanted to see how my gauges work with the 10mm adapter. I didn't have a 10mm adapter for the Snap-On compression tester, so I watched eBay and found one for $30 shipped. I also wanted to see how a cheaper gauge worked. There have been questions of which gauges work and don't work that are affordable on multiple occasions. I picked up an Actron compression tester for close to the same price that I paid for just the adapter for the Snap-On.



The top hose is the Actron, bottom is Snap-On. I used teflon tape where the Actron adapter screws together, hoping to ensure a good seal.




I started with the Actron gauge and measured 165psi.



Then I switched to the Snap-On tester and it said 183psi.



The Actron adapter is easier to work with and will fit the Snap-On, so I tried it with the Snap-On and got the same reading as with the Snap-On adapter.

Kinda shows why compression psi figures are tough to use for performance builds and why I switched over to doing actual compression figures for my 2T builds. Sometimes 180 would work on a 50cc but on the 90cc I would be getting detonation at 150psi. Using 7:1 corrected on the 2Ts I've had reliable results. I'm planning on getting some good cc and calculation based compression figures for the various stages of this 4T along the way.


I messed up the floorboard a bit while checking compression. I was picking up the corner a little to reach the head easier and I've cracked the plastic by the bolt. It would have made more sense to unbolt it, but I'm used to bending panels a bit on my other scooters. The more flexible floorboard on those hasn't given me much trouble. As with a lot of things on the TaoTao, the floorboard is not at the same quality/thickness as my Ventos and Keeway (all QJ manufactured). Still, can't really blame the part for it since I'm the one bending it.




I also noticed while removing the seat at one point that I've already cracked the mounts in the seat bucket. I have done this to both of my Tritons, but over time. This is at 200 miles or less. Maybe this is what the weight limit is all about? Everything else on the scooter seems to deal with me pretty well, but the seat bucket would probably appreciate if I were under the 225lb stated limit.








While speaking of panels, here's what the TaoTao actually looks like most of the time. Not pretty, but I can work on most of what I need to by removing the seat bucket (4 nuts) at most.




I did another oil change at 164 miles total, 101 miles since the previous oil change. Since I measure oil in and out of the engine, it's easy to tell how significant any oil loss is. I add 750ml and I drained out maybe 740ml. The mark just above the oil is 750 and I made a mark above that at 800ml for reference. 10ml over 101 miles is more than I'd like, but the seal is not moving much oil. Even at a 500 mile interval it should be safely around 700ml at that rate. BlowinSmoke suggested using an o-ring just behind the seal. He said perhaps that would redirect enough oil away from the sealing area to eliminate the issue. I may try that at some point.



The drain plug magnet is still doing it's job. I've got about the same debris at 101 miles now as I did at 45 miles earlier. Not much of anything on the screen side again.

[aeroxbud]

That is a lot off difference between those gauges. Have you ever had the snap-on one calibrated?
Your scooter is much easier to work on than my Aerox. Just to get to the carb I have to take off the rear seat pad, both rear panels, locking mechanism and seat bucket. That's eight bolts and a lot of screws! I have done it so many times now, I have it down to an art

[90GTVert]

If it's been calibrated, it's been years. It's a hand-me-down from my father who was a mechanic most of his life.

Gotta have easy access for a project. I was worried about that when I chose the scoot, thinking maybe I would have been better off with a larger frame model. I couldn't do anything the to SunL or Venus without removing the engine because the frames are so close. This one's setup nice so once the panels are out of the way the valve cover is easy to get to and the CDI is right in front of that. You could do most avg jobs with the seat out and front of seat panel off.

[2strokd]

2strokd likes that

[90GTVert]

Seems like forever to me (5 days lol) since I updated the thread. I wanted to get a few things together and get a vid done so all the stuff is here at once.


You can check out this update in video form and watch the runs in HD...

www.youtube.com/watch?v=K5j8gHg5kMc


Since last time, I started by seeing what I could get out of the carburetor and CVT with just basic tuning. I made a bunch of WOT runs about 2 miles each way and wrote down max speed each way, RPM at those speeds, air'fuel ratio, and max CHT. A simple trial and error method that anyone can do, really only needing a speedometer and a place to run. The more gauges the better though. These are the results of the test passes.

Avg Speed is average of top speed each way. Avg RPM is average of RPM at top speed each way.

Main Jet	Max CHT	Avg Speed	Avg RPM	Air/Fuel Ratio
72	361°F	33MPH	6,820	13.5 - 15
75	352°F	36.45MPH	7,265	11.5 - 13
78	341°F	34MPH	6,995	10 - 11.5
79	338°F	33.6MPH	6,965	10 - 11.5
80	337°F	33.35MPH	6,950	10 - 11.5
82	331°F	32.85MPH	6,865	10 - 11
85	292°F	28.5MPH	6,675	10

Even though the wideband had been showing rich readings, I wasn't sure it was accurate. I actually did the runs from 75-82 another time just to be sure results were the same. 85 was very obviously too rich. It felt dead compared to the rest and the part throttle sputter was terrible. I went with the 75 main when I got similar results on a second set of runs.

Here's what the spark plug looked like after 150 miles of running rich at 10-11.5:1 air/fuel ratios most of the time (pre-tuning).



I re-adjusted the idle mixture screw. Then I shimmed the needle up to richen it. I had been pretty certain the sputter that many say is a lean stock issue was really a rich condition (sputters tend to be and the wideband hits 10:1 when it happens), but I just had to be certain. Sure enough, it got worse. I ordered a needle from Parts For Scooters for $3. It may have worked, but I didn't like how thin it was where it passes through the slide and how much it could move. Not really sure it matters, since the needle shouldn't pull all the way out anyway.





I left the stock needle and called the carb done so I could move on to CVT tuning. I started with finding the Dr Pulley sliders that would work best. I've compared rollers and sliders in the past, and I've always liked sliders. Might as well cut to the chase. I did take a run with 5g rollers first, to see what the difference would be in 5g rollers vs sliders. That surprised me. The sliders raised RPM substantially vs rollers of the same weight. I've tried them side by side in my Minarelli engines multiple times and they always produce similar RPM. Not the case for the GY6 50 variator. Here's a look at speed and RPM with various weights.

Avg RPM is average of RPM at max speed each way. Avg Speed is average of top speed each way.

Weight	Max RPM	Avg RPM	Avg Speed
5g Rollers (Stock)	7,650	7,000	32.75MPH
5g Sliders	8,220	7,690	34.45MPH
4.75g Sliders	8,380	7,800	35.45MPH
4.5g Sliders	8,450	7,910	34.85MPH
4.25g Sliders	8,670	8,110	34.8MPH

4.75g sliders gave the best all around performance for me. 4.5g sliders were a little quicker accelerating, but I think revs would be a little too high for best cruising/speed results. The 5g sliders would probably be a better choice if the scooter were a little faster, but the TaoTao is lacking the power to pull me very fast. I went with the 4.75g sliders.

I was surprised to see the stock rollers looking as good as they did when removed. I started noticing flat spots after the first ride, but it seems they've worn in pretty round.





The stock bushings were getting a little worn, so I replaced those with Dr Pulley bushings as well.




Then I worked on the clutch springs. I found that 1,500RPM springs gave me the latest engagement before the rest of the CVT started shifting.



I went for a 34.9 mile ride and speeds increased just a little. There was a 1-2MPH gain in average WOT cruise speed, now 33-34MPH @ around 7,8-7,900RPM on average. Max speed went up nearly 1MPH to 38 @ 8,100RPM. I can easily go faster on slight declines or with a little tailwind while just cruising than I can tucking down in "honest" conditions. CHT and oil temps didn't show any big changes vs stock. Max CHT was 378F. The highest oil temp I recall seeing was about 230F.

The big improvements were in acceleration and gas mileage. Gas mileage went up to 95MPG vs just under 82MPG with the stock setup. Not only is the main jet smaller, but with the improved CVT tune I could stay at lower throttle in town. I didn't need to pin it every time I tried to move. 0-30MPH average time was 18.71 seconds vs the stock average of 26.88 seconds. A huge improvement.

Overall, I'd say the basic tuning was a big success. I already had the stuff, but a jet kit is about $15 from Parts For Scooters , a set or two of sliders at $20 each from there or elsewhere, and the Hoca clutch springs are about $10 a set. That's $45 to maybe $100 depending just how wide of a selection of stuff you want. Sounds kinda pricey, but it feels amazing vs stock and additional tuning supplies usually come in handy (at least for me).



The next thing I tried was removing the restrictions from the stock airbox. I kinda hate calling them restrictors in one sense, because with these scooters we hear "restrictors" and automatically think something is holding us back. Removal is very easy. Take the screws out and pop the cover off of the airbox. You'll find this under the lid.



It can be pulled up and out with ease, leaving a much larger inlet.

 

There is still some restriction though. There's a cap on the end that can be removed, or the whole inelt hose can be unclamped. I opted to remove the entire hose.









Reinstall the lid and that part's finished. Once I started the scooter, it wouldn't go more than a few feet WOT before the wideband pegged at 16.0 (leanest) and the engine surged. I ended up going from a 75 main jet to a 90 main jet, a 20% increase, to get it running right.

I went for a 31.9 mile ride that night to try it out. I really couldn't feel a difference, but I was hoping maybe speed would increase or something. WOT cruise speed/RPM and top speed/RPM stayed basically the same as before.  38.2MPH max tucked, .2MPH faster than before. Tucked speeds are kinda odd now though. I try to keep them fair, but power is so low on the stock scoot that anything being different can change speeds easily. CHT was a little cooler at 267F. I think that may be more from the few degree cooler night vs other runs. Oil temp was again around 230F. Fuel economy dropped a little, to 91.6MPG. Not sure that tells anything either, because I had a headwind to fight for a bit and it's such a small percentage difference that minor things could have changed that.

0-30MPH acceleration averaged out at 18.58 seconds. It was 18.71 previously.

My opinion is that removing the restrictions essentially did nothing other than require a larger main jet. Even the part throttle richness was still intact. I did opt to leave the restrictions out though. Mostly for later on if airflow becomes more important.


Last but not least, I swapped on a Keihin PD19J carburetor in place of the stocker (also labeled PD19J, but definitely not a Keihin). It's not hard to see casting and quality differences in the two carbs.









The difference that most interested me at that point was the adjustable needle.



I tuned the Keihin and it liked the same 90 main jet that the stock carb used. It had a 33 pilot and I ended up with the mixture screw about 1 turn out. I set the needle in the leanest position to finally see a change in part throttle mixture and get rid of most sputtering. Even on the leanest setting, air/fuel ratios at part throttle were most commonly 11-13:1.

The Keihin's accelerator pump seems to deliver more fuel than stock. If I pump the throttle it will peg the wideband at 10:1 and with a few more pumps the engine may die.

I went for a 32.6 mile ride with the new carb and again got about the same speed, RPM, CHT, and oil temps as before. The two big changes were 0-30MPH acceleration and fuel economy. Mileage increased to 98.2MPG, likely due to better part throttle settings. 0-30MPH acceleration was almost 7 tenths of a second quicker. It didn't seem as if the scooter made more power, but rather RPM picked up quicker off the line without the excess fuel causing a sputter and stumble.

I was pleased with the carb, but I think similar results could be had with simply an adjustable needle in a stock un-sealed carb. That's not to say I think the stock carb is the same quality, but it could likely work the same at least for a while. I'm sticking with the Keihin on Project TaoTao.


After seeing 98MPG, and being so close, I just had to hit 100MPG. I don't recall getting 100MPG on anything. I went out another night and rode using part throttle to cruise 25-30MPH and tried to be gentle with the throttle. I avoided WOT for the most part, other than for safety concerns with faster moving traffic. 43.9 miles of riding later, I filled up with 0.399 gallons of gas which made 110MPG! I don't really plan to ride that way, but at least I finally got into 3 digit mileage at least once.

[dude]

Sorry if I missed it what is the overall gear ratio? And does it have 3.50x10 tires

[190mech]

Another great vid Brent!!!

[90GTVert]

It does have 3.50x10 tires and I do not know the gear ratio yet. I'll get there eventually. I know I post all the stats at the beginning of my other projects, but I haven't had it all apart yet to count teeth on gears and check compression and such.

[flashback30]

great video Brent

[aeroxbud]

That is quite impressive gains from just tuning the CV. Just goes to show what a little bit of setup can do.

[stepthrutuner]

Great stuff. Nobody does it better.

[90GTVert]

Little bit of an update...

My plan for the next batch of updates was to try out a high performance coil and two different CDIs. All of them are from ScrappyDogScooters, because the prices are good and I figured it's a source that should be around so if the stuff is decent we'll know where to get it. All of it came from their eBay page, cheaper than offered on their site.






I was starting out with the coil. It's the "red hot" version. They sell two "high performance" coils for the 139QMB. One is orange and labeled as follows :

"PERFORMANCE IGNITION COIL
High energy ignition coil boosts that weak spark into a real shocker. This will improve starting, acceleration, and fuel mileage. This is plug and play for GY6 scooter atv 50, 125, 150 engines in 152QMI, 152QMJ, 157QMI, 157QMJ, 139QMB, 172MM, CF250, and all other 4strokes. As on all performance parts, no guarantee, as these parts will not fix a problem you may have elsewhere. Fits all Chinese 4stroke engines as well as your budget. $19.95"

Then there's this one, with this description :

"Red hot high energy GY6 ignition coil
This GY6 performance ignition coil is similar to our orange coil, but hotter! These fit all the fourstroke scooter/atv engines like the 139qmb, 152qmi, 157qmj, 172mm, 50cc, 125cc, 150cc, 250cc. Simple one bolt installation. Tank, Jonway, Honda Helix, NST, JMstar, etc. $24.95"

So this is the big dog.



Honestly my first impression pulling it out of the packaging is that it felt really cheap. The stock coil, wire, and cap seemed better made to me, but what really matters is on the inside I guess.

Before I went out and tried to ride with it, I wanted to see how if it gave different readings or if I could perhaps even see if it would make a stronger spark. I hooked up the stock setup (with NGK cap replaced at PDI) to an HEI tester. It's a tester that has a very large gap that will show you if something's weak. I forgot to take a pic, so here's a link to one for now. I did take video of the tests, and that'll be included in the next video update. The stock plug did as I expected and provided a snapping spark on each revolution.

Then I switched to the "red hot" coil and repeated the same test, same ground location. It provides what at first appeared to be an intermittent spark. I looked at the video later for a better view than I really got in person. Then I saw that it was making little baby sparks and then a big spark and then baby sparks and a big spark. The big spark looked to be more intense than what the stock coil was putting out, but it only happened every 2-3-4 revolutions. I found this to be less than impressive.

I went on to cut each end off of the plug wire and re-secure it to the cap and coil and to make sure the cap was gripping the plug well. No improvement. I tried swapping the stock wire and NGK cap on just in case, and still no better. I even had some times when I got nothing but baby sparks.

While I had the coils stripped down, I did take ohm readings to see if there was a difference between the two. I have only a very very basic understanding of electronic components so unfortunately the numbers don't necessarily say a lot to me, but they are below. I know that a coil uses a primary and secondary coil with different amount of windings to up voltage, but I don't really know if reading the numbers says a lot. My thought would be that if you have a larger spread, like the stock coil, since a higher ratio of secondary turns vs primary turns is supposed to provide a greater increase in voltage, you would get a spark more capable of jumping a big gap with the stock coils specs because the lower resistance indicates less turns and higher indicates more. Is that how it really works? You tell me. Maybe they mean nothing because of wire thickness or material conductivity. What makes sense in my head with electrical components rarely seems to be true. I know just from car stuff and seeing different rated primary resistances that lower primary resistance is said to be harder on the chain before the coil. Not sure that the same thing even applies with this system.

	Primary	Secondary
Stock	0.3Ω	3,218Ω
Red Hot	0.6Ω	2,702Ω

I also hooked each coil up to a standard NGK plug with a standard gap. In that case, I can't tell a difference between the two sparks. I would think that while I won't be gapping my plugs to 1/2", the HEI tester results should tell me the "red hot" will struggle more under real conditions no matter what it looks like in regular atmospheric pressure in clean air.

I opted to leave the red hot alone and not bother riding with it. Again, I'm not an electronics whiz, so if you guys and gals can tell me my logic is all wrong and why I should give it a shot, please do.


Moving on, I put the stock coil back and went to test a different CDI. I noticed by the end of a ride (at night the first time) that RPM had dropped maybe 200 or so on average at cruise speeds. By the time it was happening I did have a little wind to deal with and just thought perhaps it was related.

I went out today while it's hot to do a few test passes. Again RPM seemed a little low, but again I had some wind so I wasn't totally convinced of anything at first. By the end of 18 miles RPM had clearly fallen farther from where I initially tuned the CVT to work. I was around 6900RPM IIRC at times.... where normally I average around 7800RPM WOT cruise. It was hot today, around 90 and sunny so the road and everything else is hotter too (maybe I should have put the scrappy coil in the sun?). I normally ride at night for any real distance, so I thought maybe I'm just now seeing how bad heat is in the CVT. I took a look at the stick-on CVT temp gauge when I got back and it was up to the 171F mark.



I've never even seen the thermometer register anything before on my night rides. Still, I've always thought CVT cooling is a good thing, so I popped the cover off to see what I could do easily. Unfortunately they make the newer covers more complex, so it's not as easy to just take a big chunk out.





I had a 3 3/8" bi-metal hole saw that fit the job fairly well, so I used it and then filed a bit to smooth out sharp edges.



I took another ride in the heat of the day and when I returned the temp sticker was barely "lighting up" 149F. A small victory... but still RPM was low when riding, even from the start of the ride.



As of now, the whole CVT is apart to be cleaned and re-greased and what not. I have a feeling I'll end up needing to swap the contra spring. My Hoca clutch springs don't seem to be affected much at all. Even when everything else is dropping revs I can get over 4,000RPM, closer to 5K, before it starts to wanna engage when I'm stopped. I guess if the cleaning doesn't work I could install lighter sliders, but I don't see that as the actual issue.



While it's apart I changed the oil since I'm nearing 500 miles (482). I also decided to swap the crank seal again. It hasn't really been a big issue, but I'm tired of looking at oil in the CVT so I wanted to try something. Blowin_Smoke PMed me an idea a while back. He mentioned trying an o-ring behind the seal to possibly divert some oil from the actual sealing surface. Figured it's worth a shot.