I can haz auto tune!
 

Turned up today, however, because of college (and then karate) it's to late for me to fit it now, but shall fit it tomorrow and give it a blast and post back what its like!

Looking forward to the results. :sbike:

Personally, I try to avoid rolling around with other blokes, especially in my PJ's


:moon:

Ok then, I stand corrected.

Still dancing with another bloke, and in your PJ's.....:rotf[1]:

I do shotokan!

Only been doing it for a few weeks so im still only an ungraded white belt :P

But its fun and only picked up my Gi yesterday!

On the auto tune front...

It took me a couple of hours to install.
screwed in the lambda sensor which by the way, is stupidly long and gets fiddly to fit between the cylinder.

Fitting the rest was a doddle apart from finding a power source. They suggest using the tail light, but I didn't want to start removing cowling to get to the wiring. However, there are under the seat, 2 unused plugs, and one has a power supply that comes on with the ignition, so I cut that wire and used that.

Also plugged the AIS up, which is a shame cause I kinda liked the backfiring but nevermind...

What it doesn't state in the manual, is that when you plug the autotune into the PCV, you need to put the 2 supplied plugs into the empty plug sockets as the auto tune wont work otherwise!

Spent ages trying to work out why I wasn't getting a reading on my laptop from the sensor, and that was why. Soon as I plugged in the empty plug sockets, voila, I was getting a LIVE AFR reading on my laptop... pretty darn awesome.

AFR was set at 13.7 to 13.2, though recommended for performance is between 12.7 and 13.3 I believe, or something like that

Took the bike for a quick ride in the pouring rain... and well, it performs admirably!
Even with the custom map, there was still a small amount of surging. However, with the autotune connected...


None!



It rode like I was riding silky smooth, no jerking no lurching... that thing is incredible.

Just need to take it for a good ride, connect up my laptop again and accept the trim settings which will then allow it to get it even more spot on.

Possibly turn down the AFR too, need more powah!!

Nice to think though that when I add the stage 1 filter, it'll automatically adjust!

Nice one.

Once you have complied your custom fuel map please can you send me a copy so I can have a look at it.:errrr:

I'm not sure if it actually changes the map itself Kev, rather than adjust the fuelling to match the set AFR table.

Without the autotune, the PCV map adjusts the yamaha fuel map to whatever values are inputted by a percentage adding or removing fuel at that RPM and throttle opening.

Whereas with the autotune, it runs on the Air/fuel ratio set in the AFR table, like 13.7 and uses the wideband sensor to pick up the burn ratio on the exhaust, and then adjusts to suit.

You can see it working on the PCV software, if you open the throttle, the AFR will be all over the show as its adjusting. but once you hold the throttle steady, the AFR will read ~13.

Least I think thats how it works anyway Kev

so if I send you my map map, I believe id I'd just be sending you the original map that I put on there myself and not a corrected one by the autotune.

Correct me if Im wrong however.

Just looked up... I stand corrected!

Will send you in a few days time!

What a ball ache its being...

I loaded a map onto the PCV that best suited what I've got and it ran ok.
Now the auto tune just seems to be making it worse!

So I may just reset the fuel map and start from the beginning.

What afr have you added to the target table, can you copy and paste the table to the forum so i can have a look. You can only put a target afr in the high throttle openings on the later Xt's and just leave the rest on the table with zero's.

The surging on the 07-09 models will just confuse the autotune and make a dodgy map correction, also you must block off your AIS system as this will show an overly lean afr at part throttle as well.

Simon.

I won't post anything just yet. Simply because I need to have something worthwhile posting!

Currently using 13.4 for low rpm/medium throttle openings and 12.8 for high rpm and WOT.

AIS is blocked off. Used electrical insulation tape over the airbox outlet and then refitted the pipe with the clamp.

How do you know about only being able to use the AFR in high throttle openings? Doesn't that kind of defeat the point of having the autotune?

You can't put the target values in the lower throttle openings because of the surging on the later bikes, it will just confuse the autotune. When Dynojet eventually release the new 02 controller you'll be able to add the values then.

I know because I've been dynoing bikes for the last six years including Billy McConnells Motec equipped Supersport bikes.

Simon

I can vouch for Simon, he know what he is talking about.

The XTX & XTR 07 to present models have been a real challenge. I was also unaware that the Auto tune battled to correct the A/F ratio below 4000rpm.

Really?

Below 4000rpm you say? That an approximate figure or you know that for sure?

If thats the case then ill reset the AFR figures to 4000rpm and leave them blank and try it like that.

That quote was from Dyno Jet Europe, when you buy a PCIII for a 07 XTX or XTR onwards the instructions say to map the bike above 4000rpm. I personly had problems with my 07 fueling jection mapping below 3500rpm. Have a look at the A/F ratio you will see what I mean.

http://i201.photobucket.com/albums/a...ore7200RPM.jpg

Ok then, I'll be honest, I'm lost!

I always thought that the idea of fitting a Power commander, fuel mods etc, was to get rid of the low end surging and to improve the fueling over the base map.

Now your saying that its not mapped that low? so how do you get rid of the surgin?!

All the maps that I've downloaded to my PC5 all had inputted data from 1750rpm right the way to the red line. But if they say they don't map below 4000rpm, how could that be?

Adding a fuelling device does help with the surging, the open loop circuit is to lean & shows improvements when extra fuel is added. The closed loop is were we have no control over, believe me I have tried every trick in the book to try to take control over the closed loop & so far no one in the world has been able to do it yet or show us how to do it. Dyno Jet claim to have a controller soon to be released, until then we are in the dark.

We are not alone, I was contacted this past week by a Yamaha dealer to have a look at a XJ6S for them, which has been recently released in Australia with the same characteristic's as the XT660X/R surging. Yamaha Australia have said to the customer that's the way they are, it is a design characteristic.

This was one of the very reasons why I back dated my 07 ECU with O2 sensor to a 2005 ECU with no O2 sensor, I wanted full mapping control over the fuel injection, especially with my 700 conversion. http://www.xt660.com/showthread.php?t=8478

I have spent many hours on the dyno trying to map the 2500 to 3500 rpm range of my 07 XTX, there was an improvement over the standard mapping using the Power Commander, all I am saying is it was not perfect. Dyno Jet Europe stated not to remap under 4000rpm not me.

Hopefully once Dyno Jet release the O2 controller we will have better control over the closed loop circuit.

You guys need to be very careful here.

A lot of factors are not taken into consideration here and you need to understand them before trying to chase a problem that might not be there.

1st, Kev, regarding your dyno report with the bike running rich and lean before settling down from a A/F ratio point of view.

Are you using Dynojet's Tuninglink to map the bike or doing it manually? Manually is not recommended and you will see why.

One of the things I have learned over time is to ignore what you see on the power run as posted by the dyno report. Reason being is that the bike when mapped with tuninglink it is under load. Quite a lot actually, plus it is normally done in 4th gear. The motor very slowly accelerates from say 2000 RPM to the limiter under these conditions and the fuel is mapped accordingly. That is similar to real life conditions.

When you do a power run, you hook up 4th, max out the throttle and hold on. No load is applied to the rear wheel. The bike picks up RPM much quicker and it's fuel demand and airflow is very different to what will happen under normal riding conditions. Under quick, no load acceleration runs, the fuel map is always chasing the RPM range and is always lacking behind to some degree. So what you see on the dyno run will hardly ever happen in real life.

So, it looks wrong on the power run, but it's not. Do you understand what I am saying here. So don't go and try to fix that rich and lean condition, because all you will do is fix it to read right on the dyno, but mess it up for when you ride. And yes, you might even pick up Horsepower and torque, but again, you are chasing conditions that are dyno related and not real life. Dial in a lot of load on the dyno and do a power run again and see what I am talking about here. The A/F ratio will be different under load than without it.

Also, a lot of guys forget that the accelerator pump is active when they build the map with tuninglink. When the computer builds a map, the accelerator pump does not really matter because the run takes so long, the pump stopped adding fuel long before the tuning software does it's bit. When you dyno the bike for power and no load, the accelerator pump is active and it can mess with the A/F readings, causing people to bark up the wrong tree.

Now the same thing will happen when you are using the Auto tuner. The bike never accelerates the same when you ride it. Some days you pull off at a steep hill and the next time at an uphill. That requires different amounts of fuel, plus people don�t stay at a fixed RPM and throttle input for a long time to allow the tuner to learn and adjust IT just takes a stab at it and it messes things up. Rider input is never the same, so the tuner tries to compensate, but gets it wrong because human input gets in the way.

There are a few things one can do, but it still does not fix everything on the XT.

On superbikes, I build a fuel map for each gear, allowing the auto tuner to map the fuel map for each gear. This is great because in the lower gears where the bike accelerates quicker it will eventually build a special fuel map that is different to the fuel map at the same RPM and TPS settings in taller gears.

Also, do not give the Auto tuner a lot of flexibility and if possible do not let it map lower PRM ranges. There are too many variables at low RPM to allow the tuner to do it�s tuning correctly.

I said this before, do not allow the auto tuner to make massive changes. The less it is allowed to change the better. It comes back to having the bike tuned correctly on a dyno 1st, where conditions, load and throttle input is controlled and then only let the Auto tuner do fine tuning.

People believe that the Auto tuner means that you don't need to go to the dyno. WRONG!!!.. Very wrong. It causes all kinds of problems if it does not have a good map to start of with.

Also, even after tuning it on the dyno and riding it around for a week, it still needs someone who understand what is happening to decide which values suggested by the Auto tuner to accept and which to ignore. This is not a magic bullet for tuning problems.

In some cases it will cause a lot more problems than what it fixes.

Don�t get me wrong, it�s a great device, but you have to programme and use it correctly.

Most of the problems with the newer XT seem to happen on a light throttle opening and very little load so mapping the bike in steady state in Tuning Link won't highlight the problems because the dyno is always holding the bike back unlike the real world where we have hills and wind and big trucks to slipstream.

When I was using Autotune on the 5 race bikes I found that it was best not to allow it to make AFR changes bigger than 5% or below 40% throttle and to always save a copy of the original map as it did go tits up quite regularly, if you accidently accepted any wrong fuelling changes it tended to snowball from there and the map would end up being completely wrong because the power commander would try to interpolate between a correct figure in 1 cell with an incorrect figure in the next and then try to correct it the next time which made matters worse.

I won't get into an argument on how to map bikes on the dyno as people have there own ideas and I know mine is the correct way to do it LOL

Well said Simon

Hi Simon.

Why do you say that the Tuninglink on the dyno cannot simulate very little load at very small throttle openings? You can easily adjust the PID inputs to simulate almost any type of acceleration and amount of load applied.

I agree with you to not allow Auto tune to tune at the lower RPM ranges and to restrict how much it will adjust by.

Time after time Auto tune gets those settings wrong with riders complaining that with time it just gets worse. The dyno on the other hand seems to be getting it right. All of them claim the low end with small throttle openings are so much smoother, more responsive and throttle response is improved.

I agree with you that it can simulate any conditions but not all, on the Dyno it's virtually impossible to match the on/off small throttle and variable road speed conditions that most people encounter with day to day riding.

You can improve the surging problem as Kev says by chucking a load of fuel in low down but it only masks the problem.

There is a problem with the XT part throttle fuelling that can't be cured even with a power commander and a skilled dyno operator, if there wasn't Dynojet wouldn't waste time developing an 02 controller for it in the first place.

Richard and Martin at Dynojet UK have spent a lot of time on the dyno trying to cure this problem and failed and they've both got degrees in some sort of Electrical engineering and they train all the Dyno operators in the UK so should know what they are doing.

Dynojet haven't exactly rushed to bring this new controller on to the market as they don't have the problem over there in the US.

Dynojet we are waiting.

Freez I have been using Tune Link for a while now.

I also agree with Simon by adding fuel low down, masks the problem for the 07 onwards XTX/R's.

Come on Dyno jet we need your help.

If I remember rightly one of the responsible men of the tech department of DynoJet USA joined our forum a little while ago. I need to dig out and search his username, perhaps one of you can pm him asking for more details.

I was only half right, we do have not one but 2 DynoJet geezers on the forum, one from the states and one from UK here are their usernames:

Dynojettech is the american one and europeanbob is the UK one.

Feel free to contact them by pm Kev and see what they have to say.

Thanks

I contacted Dynojettech the day he joined the form via PM, he did not return my PM. You know I contacted Dynojet a week or 2 back & there reply was no one knows what is going on with regatds to the O2 controller.
:hippy2:

Great! What hopes do we have? After all why joining this forum in their respective work position and not even attempt to provide any assistance?

Like I said, I don't have anything against the Auto tuner, I recommend it to people but it is something you need to be careful with as users can cause a lot of problems for themselves if they let the unit go wild. I just don't advise people to use the Auto tuner and think it eliminates a good dyno tune to start with.

The other thing I maybe need to understand is what is the 07 XT's are doing and what are people complaining about? Still a lot of surging?

Another thing we need to try and see if we can figure this out is to see if we can hook up the PC5 to detect what gear the XT is in. I think that is another big problem why the Auto tuner is not getting the mapping right. The bike will have different amounts of load, depending in what gear the rider is in and that in turn has an effect on the amount of fuel required or the way the bike breathes.

Another interesting thing I read the other day. I was reading up on some engine development work they did on F1 and on the Pro race bikes. One of the tests showed cylinder pressure changes at a fixed throttle input and RPM.

I was shocked to see how the pressure varied between strokes. You would imagine that since throttle input and RPM is constant, pressure inside the combustion chamber should also be. Not so.

Things just get worse with a single cylinder motor, especially at lower RPM where breathing is even more erratic.

If pressure changes from stroke to stroke, then it means the amount of air flowing into the motor is also not constant and that in turn means fuel supply has to be adapted at a very rapid rate.

If I map low end, small throttle openings on the dyno with tuning link it takes 5 to 20 seconds easy to tune just one of the hundreds of RPM/Throttle settings and that is with constant load and a fixed throttle input in a controlled environment.

Auto tune uses the same probe, so no wonder it never gets the low end right, because things never remain constant for 5 to 20 seconds to allow the unit to sort the fuel out.

You will need a much faster lambda sensor and method to adjust the fuel supply that what currently is available to sort that problem out while you quickly pop through the gears.

Ok then...

In laymens terms then please...

How... can I use the autotune efficiently, to create a good map?

i.e. recommended settings, RPM, throttle openings.

As for visiting a dyno... as far as I'm aware, and according to the dynojet website, I do not have an approved dynojet testing centre anywhere near here! Hence why I bought the autotune. Not forgetting aswell, dynojet states that the autotune is good for users who aren't able to use a dyno which is the main reason I bought it.

I was on the phone last night till after 1100pm to an engineer in Greece, he deals directly with Dyno Jet, with the development & testing of their products & has done alot of work with the Auto Tune. He agrees with Freez that you should get a base map made on a dyno then use the Auto Tune from there if you are having problems.

He did say (as Freez mentioned) you need to be very care full with the Auto Tune, one can mess the mapping up very easily.

I also thought about PCV and auto tune but according to this thread (from people that obviously have experience with it) the auto tune doesn't do what dynojet is saying it can do so i don't see any reason to buy one.

The Autotune does do what Dynojet designed it to do, it's just the later XT's have a problem that can't be fixed just with a fuelling change. Put the Autotune on the early bikes without the lambda sensor and it will work perfectly, you just have to understand how it works and have a good base map to start with.
Any of the custom maps done by Kev or myself or any other competent dyno centre will be a good starting point just make sure the map was done for the modifications you have.

So if i understand you correct the 2007 models and the (hopefully coming soon) O2 sensor controller will work fine with auto tune? meaning no professional tuning needed?

We hope so, until tested we are in the dark.

I want another XT this next year. I am also waiting for the O2 controller results, otherwise I will be fitting a 00 or 01 ECU to my new 2010 XTX, with the wiring loop from a 2005 XT660 like I have now on my 07 XTX.

My suggestion if you do not have access to a dyno is to get a map that is made for another XT that runs similar mods to your own.

Then, 1st thing is to set the warm up time on the unit to start only after say for example 5 min. This prevents the unit from tuning while the bike is cold and the choke might be active, adding more fuel than it should. Work out how long your bike takes on average to warm up and then set the warm-up time to exceed that. If you idle your bike for 5 min before you ride, then you need a much longer time.

Then keep one very important thing in mind. It takes time for the Auto tune to sample and adjust, so the more constant throttle input and RPM you can give it the better. Just to build the initial map and get a good base line. This will involve you getting out there and riding around at for example 10% throttle and fix RPM for at least 10 seconds, before trying the next. Remember, take it easy, give it time to do its thing. Small smooth changes is key here.

Also, do not use massive values to allow the Auto tune to make big changes. The smaller the values it is allowed to work with the better. Small like in max 5 or so. Do not give it 20, 30 or 40 limits.

If the smaller then 20% throttle opening, at lower than 4000 RPM, feel good, then don�t let the unit tune there. Leave it alone, or use like 2 to 3 values as a max.

The Auto tune is a well worth investment, so please do not get me wrong here. It just has to be used correctly, like anything else. It has limitations, so be aware of them and work around it. Biggest limitation is sample and correction time, so don�t; expect the unit to work correctly if you give it like 0.1 second to sample a setting.

Fuel injection bikes these days have sensors to compensate for air pressure and temperature, but they are never perfect as with everything electronic. There is always some sort of tolerance.

So, this is where the Auto tuner comes in very handy. It will make the necessary corrections while you ride to keep the ratio at the required amount.

For examples on superbikes, not so much on the XT. The superbikes use ram air, and that is impossible to set up correctly on the dyno. Ram air pressure changes with speed, so again, impossible to simulate indoors as it constantly changes. So, this is where the Auto tune really works well. It will make the corrections to ensure the EFI compensates correctly for ram air.

Air temperature is another. I can feel that my personally XT runs crisper and cleaner some days than others, purely because of air temp. With the auto tune, it will make those small corrections and ensure the bikes electronics compensate correctly, ensuring that crisp performance all the time. But, it does not need to make 10 or 20 % corrections to do that. Small is the key here. The only bikes that might need large correction values are the superbikes that need to correct for RAM air. Assuming you start with a good base map, so don�t give these values to the tuner to play with on the XT.

Then, look at the suggested values the tuner recommend and if you see that some places have very erratic values, for example -20 at 100% throttle at 3000 RPM, but 3250 RPM is +8, that�s bad. It should flow into each other, not make big jumps in value.

The more you ride the less you need to allow the Auto tune to correct and the better your map will become with time.

Some good advice Freez.:toothy10: