FJ Dragbike Project

[fj1289]

Thanks Robert  - I started researching things and that confirms what I was seeing for this project. 

Have you seen the new TS Dash?  It is optimized to run as a dash on raspberry pi.  Also Dash Echo?  It makes your primary dash a server that can have multiple clients all Wi-Fi'd together - like additional dashes and or laptops that can make changes to the tune that are replicated thru all the clients.  You could setup the laptop for testing and have it Autotune.  Then remove it from the net during a race if rules or safety require.  And have an additional dash setup for other nice to have "gauges" - but something you wouldn't need in primary view possibly?

[RPM - Robert]

Yea, ts dash is what we are using that is the tuner studio. The dash echo is also what we will be using with the bluetooth and wifi setup, also from tunderstudio.

[fj1289]

A little racebike news - FINALLY made it out to Bandimere for test and tune on the 20th.  No runs made.  Nothing but headaches and frustration. 

This year work and home have been crazy and this was the first time I've been able to make it to a test and tune night.   

I swapped out the ECU last year after the last outing.  I could not get the ECU to log a run past first gear hard to tune without data (and is a good way to tear up engines!).  Later during some follow up testing the entire ECU became intermittent.  I decided to upgrade the ECU - but keep the same "family" and software.  I got it running on the new ECU before Christmas.  That was the last I had worked on the bike. 

First outing of the year is always the hardest - trying to find all the tools, supplies, and equipment you normally take and get it all loaded up logically.   Before loading up the bike I cranked it up - happy with how well it started, and finished loading it and everything else up.   

Found the trailer lights not working when I stopped for gas.  Great - had to replace the  wiring on the truck side last year.  Guess it's time for the trailer side this time.   Lose another hour plus at Walmart rigging up some temporary lights.   

Finally make it to the track and get things unloaded.  Warm up the bike while I get my gear ready to go thru tech inspection.  Nope.  Bike cranks well, but won't keep running.   Can't get it sorted trackside - but did use the full 4 hours trying! 

Took nearly a week to get to the bottom of it - bottom line it was a clogged fuel filter before the pump.   (I won't go into detail all the other issues getting to that point...but includes new adjustable fuel pressure regulator, manual pressure gauge to compare the the pressure sensor, a corrupt?burned out? USB adapter, trial run on yet another ECUs, etc).  As suggested, I run two filters - a course one pre-pump and a fine one post-pump.  That's the way it is supposed to be .... I misunderstood the filter ratings and put the fine filter pre-pump and the course on post-pump.   DOH!  Well that explains why the pump was a lot louder than I expected ....

Side note - trailer light issue was the truck again - eventually found a wire that had "gone black" on the inside.   Replaced it and everything works properly now.

Bottom line, finally have the bike up and running and ready to take to test and tune next Wednesday weather permitting!

[fj1289]

Thought I'd post a little about setting up the tune for the racebike.

Big picture for tuning from scratch - first do whatever it takes to get the bike to run long enough to get up to temperature. Well, first is really get everything setup properly in the ECU, check your ignition timing with a strobe, and load a "best guess" tune.  

Once you get the engine up to temp the real tuning begins.  Up to this point I am simply adjusting the entire fuel map richer or leaner by multiplying the entire map by the same factor.  Reason being is I assume if the idle area of my "best guess" is rich or lean, then the rest of the map will be too - and if not, it is easy enough to get it roughed in anyway.  Setting the idle really just involves tweaking the 9 or 12 or 16 cells in the lower left hand corner to give the smoothest and highest idle possible.   I DO NOT adjust timing for the highest idle.   I want the timing to be a few degrees lower - it will help to make a more stable idle.  One "trick" to a smooth idle is to set a column below your idle speed and add a few degrees of timing and a little more fuel - that way if the engine "bogs down" a little the added timing and fuel helps it to recover to the desired idle speed.  

Once the engine idles well the next step is to get the engine to free-rev cleanly.  This is mostly adjusting the fuel in the cells up and right of the idle region so the engine doesn't choke or bog when you try to give it some throttle.  

At this point you are ready to start some slow easy riding to datalog and start fleshing out the fuel table.   This is where the "Autotune" capabilities come into play!

I mostly use the Megasquirt family of ECU's and is what I am most familiar with and the examples I will use here.  I have also used Holley HP and have also done a little open-source tuning for a turbo Subaru.  Tuning in all cases is basically the same - USING A WIDEBAND O2 SENSOR (or exhaust gas analyzer on a dyno) - compare the AFR the engine is currently running at to the desired AFR and adjust the fuel by the same percentage as the error.  A little bit of the "magic" comes from knowing what AFR you are shooting for.  There are good generalities to follow and will yield good results.  Some "families" of engines might like a little more or less here and there, and even within the same family of engines, two "identical" engines might want a little different setting to make the last percent or two of horsepower.  But following the general rules of thumb will get you a very good running engine.  

On some dynos you can "simply" run the engine at a given throttle setting and RPM and tweak the fuel real time to maximize power and then go to the next cell.  Not so easy to do while road tuning - especially on a motorcycle!  This is where "autotune" comes into play.  There are a couple different "flavors" of autotune - those that work "real time" and those that work off of datalogs.   I actually prefer the datalog type.  Why?  Well, I originally had a laptop that would die quickly when it was unplugged.  So I learned to datalog to an SD card or small tablet and work from that.  There are also times when I notice I have some "garbage" data for some reason (usually caused by something I did or changed or forgot to undo...) and it is nice not to have that bad data tweak away from a better tune.   Also, most autotune functions will create a "lumpy" map - it is simply the nature of the beast in how the ECU interpolates between cells and how a given datalog will not evenly hit every cell in the table.  So I like to review the recommended changes and "smooth" out the peaks and holes in the table.  

The Megasquirt series ECU's use two main pieces of software - Tuner Studio to interface with the ECU and MegaLog Viewer to work with the datalogs.  The 3rd generation Megasquirt ECUs utilize internal SD card data logging.  

This is the datalog from the first "test track" session on the racebike today (viewed in MegaLog Viewer):



You select the datapoints to view on the left, see the graphs in the middle, and the selected tables from the tune on the right.

The VE Analyze button brings up this screen:



Run Analysis will then churn through your datalog and compare how the engine ran to how you wanted it to run (your desired AFR Table).



And then make the recommended changes to the fuel table you can accept or reject.  



Your fuel table should (almost always!) increase up and to the right.  As you can see the recommended table changes create some places where the numbers get smaller going up or right or make big jumps between cells.  The is where you "smooth" those areas, accept the changes, and go make another datalog.  The smoothed table loaded into the tune as seen in TunerStudio and ready to load in the ECU:




How is it "safe" to take test rides with an untuned engine?  The trick is the wideband O2 sensor and a large correction factor it is allowed to make while in "closed loop".  I have mine set to use the wideband to try to correct to the desired AFR any time it is above my idle range - and I have given it the authority to add or subtract up to 50% of the calculated fuel.   As I get the tune closer to where it should be I can decrease that authority.   There are discussions about whether or not to run "closed loop" at high power settings.   I am of the school to say "yes!" - all my "issues" when having a wideband O2 sensor give bad readings have always run the tune rich.  That's bad in that it can richen up so much that the engine will die - but it won't burn up a piston or eat a head gasket!  Most other failures I've seen while engines are running involve a fuel pump starting to fail or a fuel filter getting clogged- both of which gan lean out an engine enough to burn a piston or take out a head gasket if the ECU doesn't compensate for the rising AFR by trying to add more fuel.  

[FJ1200W]

While most of that went over my head, it is good for the update, thank you!
It's been a good year

[fj1289]

I did make it out to Bandimere again on Wednesday night.   The last Wednesday test and tune of the season.   Started with a weather delay while they dried the track - and ended an hour and half early due to weather.

Only got two runs in - but it was a successful outing.  Most importantly - the datalogger works!  The tune was not too far off - only 5% correction on the first run.

Planning on two more test sessions - Friday and Sunday this week.  Plan is to start working on the nitrous tune ups. 

[joebloggs]

I have an AEM gauge and a wide band o2 sensor I'm going to temporarily use on the FJ to set up running pods, a bit of an over kill maybe but for the sake of welding a bung in it'll be stupid not to use it.

I got the idea from the lads on the ZRXOA forum, they've been running CKV40's on their ZRX's and spent years, and hundreds of dollars on dyno time setting up the carbs getting the bikes to run cleanly when one of them went the o2 route and set their AFR's via the gauge, job done, sure a dyno can tell you what BHP you've achieved but for a road bike a safe useable tune without spending a fortune on dyno time makes shelling out on a o2 sensor and gauge a no brainer.

[fj1289]

Smart move in my opinion.  My first O2 sensor use was with my FCR flat slides, and then later on the drag bike with lectern carbs - both before I used it on the EFI system!

[fj1289]

Did some testing and troubleshooting on the race bike earlier this week - just couldn't shake the feeling something was off on the second (and last) run last time out.   

Turns out the 1st stage nitrous solenoid was not opening.  You could hear and feel the coil (and I assume plunger) operating, but the seat was not opening.   Replaced the solenoid and it is working now - with a strong enough plume it nearly snatched the solenoid out of my hand when I tested it!

Going to a private track rental on Friday - limited to 150 vehicles - so I hope to get plenty of runs to start working my way up through the nitrous settings and tweak the clutch and suspension tunes and ECU settings as needed.

Plan for Friday is first run spray a small shot (12 HP) of nitrous on the 1st stage right at the launch.

Second run add some weigh to the centrifugal clutch arms and spray same 12HP at launch and follow with 30HP on second stage about 2.5 seconds later.

If all is going well on launch - up the 1st stage to the 17HP jets (and start doing burnouts - think there will some traction issues) and up 2nd stage to 37HP jets .

Fourth run we'll up the nitrous jets again, but go from "fixed hit" on the nitrous (you just go "ON" with the nitrous solenoid) to a progressive ramp (PWM the solenoid - start maybe at 30% and progress to 100% over a second or two).  Trick here will be tuning the curve for the added fuel flow -when you PWM the nitrous solenoid, the flow rate % doesn't match the PWM duty cycle - - 30% duty cycle might be 20% nitrous flow or might be 40% nitrous flow.  And as the duty cycle increases, the nitrous flow tends to increase faster - 70% duty cycle might be closer to 100% nitrous flow.  The trick to stay out of piston destroying detonation is to remove timing and add fuel at the rate demanded by the actual nitrous flow.  As always - we'll begin rich on fuel and retarded on timing and tweak our way to the desired settings.   

Fingers crossed we've worked out all the bugs again

[fj1289]

Made it out to Bandimere this morning for a private rental event.  Basically a promoter rents the track and sells tickets.  The good thing about this - it was limited to 150 vehicles.  That meant reasonably short wait times in the staging lanes.  Even taking my time and getting out of the cow suit between rounds (it was upper 70's and sunny today!) I still got in 6 1/2 runs.  Also took time to eat lunch and pickup a couple t-shirts commemorating the 65 years at Bandimere drag strip (they are closing at the end of this season - - plan to build a new facility further away from town again).  

First run - plan to leave clutch and suspension alone and run small 12 HP shot on the first stage of nitrous.  Bike behaves and runs well!  Finally - a good starting point!
It goes a 1.73 60', 6.92 @ 102 mph 1/8 mile and 10.81 @ 126 mph in the 1/4 mile.  

Second run - plan the keep 12 HP in first stage and add 30 HP on second stage of nitrous.  These are both "fixed" hits - simply opens the nitrous solenoid full "ON".  1st stage is triggered when I let go of the clutch, and 2nd stage timed for 2.3 seconds later (just after the shift to second gear).  Again, run feels good and it goes 1.73 60', 6.76 @ 109 mph 1/8 mile and 10.40 @ 132 mph 1/4 mile.  Noticed some shifting issues to 4th and/or 5th.  Looked at the datalogs - I could see something weird, but not sue why.  Checked air pressure - it was done around 90 psi.  I upped the regulator to give 120 psi and added another .01 to the kill time for the air shift (from .08 to .09 seconds).  

Third run - swapped nitrous jets - 17 HP 1st stage and 37 HP seconds stage.  Run felt very smooth (too smooth it turns out ....) forgot to arm the nitrous system after a somewhat long wait in the staging lanes for track cleanup.  Bike went 1.90 60', 7.39 @ 97 mph 1/8 mile, and 11.47 @ 120 mph 1/4 mile.  Trying to look for the silver lining in that grey cloud - it was the best pass this combo has made on just engine...

Fourth run - do what I planned for the third run!  Run goes well!  1.66 60', 6.59 @ 110 mph 1/8 mile, 10.16 @ 137 mph 1/4 mile.  Yes!  And I've still got lots of nitrous jets left...

Fifth run - swap nitrous jets again - and change the nitrous operation from "fixed" hit to progressive (solenoids are PWM - pulse width modulated - to soften the "hit").  1st stage is 22 HP ramped up from 60% to 100% in a second, and 2nd stage is 47 HP ramped up from 60% to 100% in 2 seconds (after a 2.1 second delay ~ after the 1-2 shift).  Also did a clutch adjustment - added .010" shim under all 6 static springs, and used longer bolts to add 2 more washers (they act as weight) to the centrifugal arms.  Want to make sure the clutch doesn't start slipping when RPM drops on the shifts and am also trying to "pull down" the RPMs on launch since more power means more slip.  I also want to see how the much the clutch changes change the launch.  Run feels good!  Look at the time slip after getting back to the pits and getting the cooling fan blowing on the engine.  YES!  It went 1.62 60', 6.43 @ 113 1/8 mile, and 9.90 at 140 mph 1/4 mile!!!   MONKEY -> BACK = OFF!  (No offense Mark LOL!)  
FINALLY RAN A 9 SECOND QUARTER!!!    

So - pack up and go home?!  Hell no!  Be smart, make small changes to the tune but keep everything the same to repeat?  Hell no!  

Sixth run (this was the 1/2 run .....)  Swap nitrous jets again - this time 1st stage 30 HP and second stage 65 HP.  Extended the ramp times a little and lower the start percentages a little (will have to dig the details out of the tune files - don't remember those details now).  Line up, launches good - and DIES!  Crap!  Move out of the racing lane and immediately look down at the safety lanyard (had this happen once at the Colorado Mile) - it pulled out during the launch.  CRAP!  Insert the plug back in, restart the bike and ride to the end of the track.  CRAP!  CRAP! CRAP!  It is very close to the end of the track rental and don't know if I'll have time to line up for another run.  Luckily they have not closed the staging lanes when I get back (but I do end up being the last one for that lane!).

Seventh run (or sixth "real run) - do what I planned for the aborted run.  Back tire spins HARD on launch this time and I get SIDEWAYS quickly.  Back out and get is straightened out before going into the other lane and quickly get back into it.  Needless to say, it was not another 9.  60' is 2.09, 7.58 @ 108 1/8 mile and 11.12 @ 139 mph 1/4 mile.   It sure what happened on the launch sine the previous one left hard (until it died).  Maybe I didn't pick a good spot to line up?  Maybe the solenoid I replaced isn't suitable for PWM and didn't behave this time?  Maybe I just got luck on the previous launch?  Need to adjust suspension?  For Sunday I'll take the smart (and "safe" and lazy) way out and just go back to the last settings for the 1st stage of nitrous - and keep working on increasing the 2nd stage!

Also for Sunday I'll try going up one tooth on the countershaft sprocket.  If I get up to the speeds I'll like to get to, I'm going to need some more gear

[fj1289]

Short update from Sunday - this was as much a "social" event as it was a "track session" for me.   Did get two runs in.   First run decided to keep the 30 HP jets in first stage but start the progression at a smaller percentage and build u over a longer time.   Still spun pretty hard on the launch.  Stayed with the 65 HP shot in the second stage - just tweaked the overly rich settings down a little bit.  The 60' time was 2.144 seconds (without spinning it is usually a 1.6 or 1.7).  1/8 mile 7.315 sec @ 113.40 mph and 1/4 mile 10.735 sec @ 144.94 mph. 

Noticed a couple things on that run - bike ran a little rough on the return road; and when I looked at the datalogs the added speed had me crossing the stripe around 9700 rpm.  With the XJR rods and balanced crank I'm good with pushing it to 10500 rpm at the end, but for testing prefer to keep the rpms down a bit.  The bike was geared 18/43.  Decide to change to 19/43.  Also took a look at the plugs to see if I've nipped one (melted the electrode off).  What I find is the electrode on #2 looks a little rough maybe and has closed up the gap?!  Think for a second - how can the pistons be hitting the spark plug?  Could I have been that careless installing the plug.  Then I think - if the electrode got hot  enough to "almost" melt, then combustion pressure alone could have closed up the gap?  Go to put in new plugs one step colder - they are in the garage.  Great.  Decide the swap plugs 1 and 2 and run again to see what it does. 

Second run reduce the first stage nitrous jet to 22 HP, and up the second stage to 85 hp. 
Run goes smoothly.  60' is 1.745 sec - not surprised with the lower gearing and keeping the gentler ramp with the smaller nitrous jet.  1/8 mile 6.523 @ 117.23 mph and 1/4 mile 9.882 @ 144.63 mph! 
Very relieved to "back up" the 9.90 pass from Friday   

Had a little time this morning to check the plugs before heading to the airport for work again.  It nipped the plug in the #2 cylinder.   When I get back from this trip I'll pull the fuel injectors and get them cleaned and flowed again - make sure the #2 injector hasn't gotten gunked-up or has developed a bad spray pattern. 

[fj1289]

My understanding of clutches was expanded today.

However, I still didn't catch a lot of that. Deep. Someone obviously knows their shit.

When are you going to hit a nine already?!

More than 10 years later! 

[joebloggs]

Andy is spot on about setting up the bike – strapping the front and lowering the rear are cheap, easy, and quickly reversed.  The first couple of times I went to the track it was at stock height.  Strapping the front and lowering the back made a big difference – not only in performance but more importantly in how it felt.  Much more controllable and predictable.  With the stock engine (with carbs, 4-1, and Dyna ignition) I only raced on 1/8 mile tracks.  My best were some 6.8's at 103mph – that roughly equates to a 10.7 or 10.8 1/4 mile.  Best run I made was with the used 1314 kit (that didn't make a real good ring seal) and a ported head with oversized valves and mild cams.  Turned a couple of 10.79's at 125mph, but this was at Bandimere (Denver, CO) when the DA (density altitude) was 7600'(about a 25% horsepower loss)!  The NHRA correction factors put that at a 9.75 at sea level.  Truth be known I'd probably would had run closer to a 10.0 a sea level – don't know how well I would have launched with the additional horsepower!  

Yes, I did have a lockup clutch (single stage) from Pro Chassis Racing along with the quick access cover.  I lost the clutch and billet basket when the trailer was stolen.  Luckily I didn't have the cover in the trailer.  Nobody (besides PCR) makes a lockup for the FJ anymore.  [I'm not real inpressed with either the quality of work nor the customer support from PCR]  I looked at converting to a Busa basket and hub, but couldn't find anyone who could machine the FJ transmission shaft spline pattern in the Busa basket.  I did find someone who could do custom heavy duty billet transmission shafts and could put a Busa nose on the FJ shaft, but was very pricey.  Especially when you added in the cost of the Busa lock up and billet basket too.  I ended up biting the bullet and went big on the clutch.  Tim Hayes machined a hybrid clutch – uses FJ hub and steels with a Busa basket and frictions.  That thing is definitely metal working porn!

I'm keeping both front rotors.  One, I can probably use a little extra weight out there, and two, I don't ever want to come up short on braking power – especially if there is a short shut down area.  Also, I've tried a single caliper with a master sized for two – downright scary – no feel to it at all, "very wooden" feeling.  A lot easier to keep the extra caliper than find the right master cylinder.

Got the frame back on the wheels!  Still have to fab some spacers for the shock linkage and the swingarm bolt.  I'll be using stock Busa rearsets mounted on aluminum brackets to set them back about 4 inches and down about 2 inches.  


16" Honda CBR rim and rotors with low profile tire.  I was able to lower the fender 1.25 inches – more clearance to the lower triple and the fairing to get lower!


Flipped the lower triple clamp so I could raise the fork tubes higher and the clamps would still properly grip the wide portion of the fork tubes.  



You can see how adjusting the LinkPro changes the leverage ratio on the shock to firm or soften the rear suspension.  Adjustable dogbones then allow you to reset the rear ride height.  

Hi
Still trying to work out the best way to attach my shock etc. The 1080 spring will definitely require me to have a relay machined, as the narrower Hagon spring already contacts the arm. 
In your post you mention that the adjustment on the LinkPro allows you to harden/soften the damping, my question is which direction has what effect?
Thanks
Mel

[fj1289]

With any adjustable link or any link you swap or fab - it will only change the leverage ratio between the wheel movement and the shock travel.   

If you are going to use a different ratio linkage I would suggest leaving the stock spring on the shock (assuming it is fairly well matched to the shock damping).  If you swap a heavier spring but leave stock damping, the spring will overwhelm the shock - the linkage ratio won't effect just one aspect of shock performance - it will effect them all equally I assume...

On the Linkpro - the more you move the lower pivot point forward the stiffer the suspension gets.   The more you move the lower point to the rear the softer the suspension gets.

Tony Foale used to have some software for evaluating motorcycle suspension design and function.  Might be worth a google search. 

Good luck!   

[fj1289]

Been working on a writeup on the plan for the racebike this year - ironing out some recurring issues and what I think will solve them...

Then a couple days ago read a post on Facebook - the drag strip in Pueblo CO will be down for the season (at least the fist part of it) due to damage to the asphalt surface - looks like winter was not kind to it.  Don't know how long it will take the city to solicit proposals, dedicate funding, and get the track going again?  Past years this news wouldn't be so bad - but Bandimere drag strip just outside of Denver closed at the end of last season 

So, Bandimere was 45-60 minutes away and usually ran multiple times each week.  Pueblo is 1 hr 45 mins away and usually runs a couple times each month.  Now the closest drag strip will be Julesburg, CO about 3 hours away, and not sure how often they run.

UGH!  Definitely need to get the street bike back up so I can troubleshoot some of these issues on it too! 

Silver lining - if I get the issues sorted I'll shift focus to the 1/2 mile or 1 mile races possibly ...