KTM Owners Forum banner

1 - 17 of 17 Posts

·
Registered
Joined
·
1,204 Posts
Discussion Starter #1
Just wondering if anyone has experience of this on the 990 as I have some questions.
The manual says nothing about it, so I guess KTM say not to do it.. but of course the internet says to do it, and I have been know to tinker with things. :giggle:
 

·
Registered
Joined
·
499 Posts
was looking for the air bleed screws whilst the cover was off, couldn't recognise anything.
having said that: when I took the left cover off the airbox there was access to the hoseclamps fitting the throttle bodies to the heads, so I would presume the other side will have access to some other things?
 

·
Registered
Joined
·
1,204 Posts
Discussion Starter #3
I have an open air box (Motohooligan) so I can see and access the adjuster easily. I'm not sure what the "air bleed screws" are that you're referring to. There is just one screw to adjust the balance and I think it is close to the throttle wheel/linkage, maybe on the left side.

It seems you only need access to the MAP sensors, pull the hoses off and link them to your gauge, or leave them connected and instead use TuneECU to read the live values from your MAP sensors. I'd rather take actual reading from an independent gauge and adjust it that way if needed.
But that raises the question of which method is best, having the throttles actually balanced equally or having the ECU think they are balanced as there will no doubt be a difference due to sensor accuracy.
 

·
Registered
Joined
·
499 Posts
to my understanding:
  • first balance the throttle bodies (the primary flies in our case) via gauges.
  • (re)set TPS.
-use the air bleed screws to adjust the mixture independently at idle.
Note: the air bleed screws do NOT affect any throttle linkages, they only let more or less air pass the primary fly valve without affecting the TPS.

tricking the ECU they are balanced may give an uneasy running engine, and can potentially make 1 piston do more work than the other and that can affect performance and lifespan of some parts.
 

·
Registered
Joined
·
1,204 Posts
Discussion Starter #5
Balancing by external means could be seen as "proper balance", especially as I would use a homemade setup of hoses with oil to show the level. It would be balanced more accurately this way than using 2 separate gauges.

As far as I am aware the MAP sensors are not calibrated units so they will almost always have a different reading even if the pressure is actually the same. The ECU must not be too sensitive to this I guess.

I don't want to adjust the TPS as I have read that doing that may offset the custom map I have (in a power commander). Anyway, as long as the TPS has not moved since it was setup at the factory there should be no need to adjust this.
I can't see how the TPS could be effected by adjust the balance, it would still be correct reporting the position of the throttle.
 

·
Registered
Joined
·
499 Posts
I'll try to explain what I mean, and no doubt you'll know most of this already:
all the sensor values are hard values. when you (re)set the sensor values in the ECU it takes into account the offset between actual and ideal and uses this in its calculations when the engine is running.

similarly for the TPS (if and when it has moved), for example: ideally the TPS voltage at throttle closed (i.e idle) is somewhere at 0,5V~0.6V for the ECU (full range = 85degrees = 2.8V). either by TPS sensor disturbance or remounting, in the ECU it shows 0.8V at full close. when you (re)set the TPS the ECU will take the actual voltage 0.8V and use the offset 0.2~0.3V to get the ideal 0.5~0.6V.

in the instance where the secondary throttle body has changed position relative to the primary body (the one with the TPS sensor) the ECU only sees the primary, not the secondary body, and responds to the primary. that could mean the secondary is more or less open relative to the primary and that gives uneven engine running.
so get the everything working nicely, the 2 must be balanced first, the the TPS (re)set.

on the SMT there are 2 absolute pressure sensors one on each inlet. I don't know how much this is influencing the ECU output values but I would say it is smart enough to recognise a substantial pressure difference between the intakes and give an error message when outside the hysteresis band. perhaps in the TuneECU software you can read the pressure sensors realtime and adjust accordingly without the need of gauges/manometers?
 

·
Registered
Joined
·
1,204 Posts
Discussion Starter #7
I'm going to need to think about this a little more but I'm getting there, as in I have read what you wrote 4 or 5 times and it's starting to make more sense.

on the SMT there are 2 absolute pressure sensors one on each inlet. I don't know how much this is influencing the ECU output values but I would say it is smart enough to recognise a substantial pressure difference between the intakes and give an error message when outside the hysteresis band. perhaps in the TuneECU software you can read the pressure sensors realtime and adjust accordingly without the need of gauges/manometers?
Some people have used (or tried to use) the sensor outputs in TuneECU but the update is so slow that it's hard to adjust it accurately (apparently).

Also some people have set the balance using a gauge only to find out that what looks balanced on the gauge reports as out on the TuneECU readout - which do you trust... well I know which I trust (the external gauges) but my question is which is better. Being correctly balanced but looking unbalanced to the ECU, or being unbalanced but looking balanced to the ECU?

These map sensors are crucial to the fuelling map. I wonder if the ECU looks at the difference at idle and uses that as an offset to normalises for any difference in reading... or maybe it just has different fuelling for each cylinder... so maybe questions... my head hurts now..o_O
 

·
Registered
Joined
·
499 Posts
I see what you mean, which doe you trust....
I would same as you trust the gauges and that's the key here: use the gauges to balance the primary flies at full closed ( think that means they are actually fully closed, not just a pubic hair crack open, somebody will need to confirm that).
at that point the TPS and then the fine-tuning with the air bleed screws if they are there (has to be, I havent seen one without).

the map sensor would ultimately be looking at the TPS as that is the major input, the rest are adjustment factors in the "optimum-fuel-injection-duration-and timing-taking-into-account-the-lambda-feedback" calculations. air temp and pressure are all correction factors in that calculation, and possibly the inlet pressure sensors work together with the lambda sensors to get the fuelling etc optimal.

the Ducati's I worked on only had a single air temp and airbox pressure sensor with 2 lambda as feedback. so basically 1 set for both cylinders. the kabooms have dedicated input and feedback sensors per cylinder, so much more accurate but also more intricate.
 

·
Registered
Joined
·
1,204 Posts
Discussion Starter #9
The lambda sensors are only used for part throttle opening and below a certain revs (say 6,000 rpm), like for when just cruising.
Once you go WOT (or above 6,000 rpm) the ECU switches to open loop control, in which case it will run a power map rather than an emission map. The main controls for this are rpm, tps, air temperature, manifold air pressure, ambient air pressure and finally coolant temperature.
Mine actually doesn't have lambda sensors as they have been physically removed from the exhaust pipe and mapped out in the power commander.

So if I adjust the TPS I could be shifting my entire map a few percent up or down.
 

·
Registered
Joined
·
499 Posts
ay, but then that would be accounted for in the ECU calculation already me thinks.
TPS is the mains, the rest you mention are correction factors.
if the bike was setup at that time it will have been with that TPS value, provided nobody changed anything or tinkered with the sensor.
a power commander (or other piggy-back unit) is like a mind-control unit: you use the ECU for its processing power, hijack the inputs to go to the power commander and force the power commander outputs over the ECU output. and thus you can also fool the ECU the lambda's are still there so no errors.

if you hook up TuneECU and read the actual TPS voltage, its at that actual position of the primary flies (so again fully closed). at the time of setting up the power commander, the TPS voltage may have been slightly different. so yes, resetting the TPS will bring things back to what they were.
for example: at setup time it was 0.5V, and now over time because the flies and linkages have bedded themselves in a bit the current TPS is 0.47V. you may (or may not) have noticed the engine is idling/running a bit rough.

doing the TPS reset will take the offset (as described earlier) into account and the bike should run like it did fresh from the dyno.

another reason to do throttle body synchronisation and a TPS reset is due to linkage wear between the 2 bodies.
besides that, resetting TPS does no harm provided all else is still set properly! no point in TPS reset if one of the other sensors is clogged up or fowled.
 

·
Registered
Joined
·
499 Posts
getting back on this, I looked through the SMT manual again on the throttle body topic (chapter 41). there they use the official KTM diagnosis unit, but as it is a voltage you should be able to make it with a multimeter with the ignition on:

the value should read between 0,56V and 0,64V. if it is outside hese readings the TPS should be adjusted accordingly. open and close the throttle by twistgrip and check again, adjust if necessary.
and yes: the throttle valve should be FULLY closed.
 

·
Registered
Joined
·
1,204 Posts
Discussion Starter #12
Do I need Tune ECU for any of this?
Like to ensure the throttle valve id fully closed, or does that mean making sure the engine is warm.

I'm gonna let this one simmer for a few months now.
Once it warms up (and dries up) I'll pull the 990 out of hibernation and then start investigating, as well as properly reading the manual.
I guess as long as I record all values before I adjust anything I should be fine, as I can always revert back if I upset something.
 

·
Registered
Joined
·
499 Posts
for measuring the 0.57~0.64V you don't I suspect;
just need ignition on, and a multimeter with some prongs (like opened up safety pins).

the nice thing about TuneECU as I read it is it can do all this stuff.
 

·
Registered
Joined
·
1,204 Posts
Discussion Starter #14
I thought I'd update this thread as I have now had TuneECU connected and done some more reading.

TuneECU is no good for balancing the throttle bodies. It can read back the static values just fine which is good for checking that they are all within spec, but when the engine is running it can't keep up. Thinking about this afterwards it makes sense. The value from the MAP sensors will be a pulse, and only the peak value (or peak vacuum) is the required measurement. The bike ECU obviously samples the crap out of that to ensure it catches this peak value, or it has a peak voltage filter to capture it... but the value shown by TuneECU only updates about once a second, and not even both at the same time. This means the voltage readings for both MAP sensors varies wildly and are rarely ever a similar value, so trash info. It's a shame TuneECU can access the MAP value that the ECU has interpreted it to be but it doesn't seem to.
Am I doing something wrong?

Anyway, I have bought 10 meters of 4mm ID clear aquarium air hose (for £3.99 delivered) and with that I will make an oil filled manometer so I can balance them properly. I did think about making up a peak voltage circuit so I could read from the MAP sensors directly but tbh I'd rather not mess around with them. Making up a manometer sounds like more fun and will be useful for other bikes.

One thing I have done which has improved the bike immensely was to cut a small slot in the throttle cam with a dremel cutting wheel. This has made the initial throttle opening much smoother which has made the bike smoother at low speed but also better when coming off the throttle into corners. It's a massive improvement and I think all of the 990's should have been like this from the factory.



TPS - I checked mine and initailly the reading was odd and I thought would need adjusting but it didn't.
The closed position was reading as 2% (0.8 volts) and the fully open was 95% (3.72 volts).

Firstly the closed position is only to be measure with the engine warm as 2% is the "choke" setting, and sure enough it dropped to 0% once it had was run and the revs dropped to normal idle.

The fully open throttle not being at 100% is probably intentional. From what I read on superduke.net each bike has it's own setting but it ranges from 95-98% as the max yours can have. Setting it to 100% results in staling issues. After reading this I am happy and will not be adjusting it. Adjusting the balance will not affect the TPS btw. The TPS is read from the cable driven throttle front body and the adjustment is on the rear linked one, so adjustments made to the rear are not seen by the ECU.
 

·
Registered
Joined
·
499 Posts
The TPS is read from the cable driven throttle front body and the adjustment is on the rear linked one, so adjustments made to the rear are not seen by the ECU.
which is why you start with the that body and then use the balancing to get the rear one sorted I suppose.

good to know you found out TuneECU can't be used for dynamic balancing, as you describe it the measuring resolution to the PC is too low (bu the ecu is fast enough to react).
 

·
Registered
Joined
·
1,204 Posts
Discussion Starter #16
good to know you found out TuneECU can't be used for dynamic balancing, as you describe it the measuring resolution to the PC is too low (bu the ecu is fast enough to react).
Well, it's not that the resolution is too low, it's the software or com link refresh that is too slow... but that isn't it either. I could live with one update a second as long as everything was updated at once every second but it isn't. It seems like it reads one value, updates the software and then reads the next.. and so on.
The result is that the MAP sensors aren't read together at the same place on the vacuum pulse so the values reported back will be almost always be different, which is useless for comparing them.

TuneECU can still be used to check that they operate and are within spec. The manual says they should all read the seem when at atmosphere, and if not to replace the one that doesn't.. Well it's not going to happen to have them all reading the same as they are not precise instruments and they are not adjustable, but I suspect as long as they are all within 0.1 volt they are close enough.

Mine showed pressures of 1014 hPa on the diagnostic screen, and the external sensor showed as 1004 hPa in the text field.
The voltages were - 3.65v & 3.61v for the MAP's and 3.62v for the external sensor.

You could also link all three together with T's in the hoses and pull a vacuum on them with a syringe to check they all respond equally (watching on TuneECU) but it's not really needed unless you are chasing a fault.
 

·
Registered
Joined
·
499 Posts
but would be cool to dee the ecu reacting to it!
agree, a few mV off should be within the acceptable % error (or hysteresis)...
don't forget there's some voltage drop over all the contacts which can account for the differences between analogue and digital measurements. as you've got it should be all good then!
 
  • Like
Reactions: TrouserSnake
1 - 17 of 17 Posts
Top