I've been searching for about 30 minutes and can't find an answer or similar problem to what I am experiencing..

Just bought an AT110 (auto tune pro) for my powervision. I hooked it up to my bike and did two rather long tuning runs to dial in the VE tables..

The second run showed 1% average change on 176 cells....

I felt good with that, as I covered all my normal riding habits and a few aggressive runs through the gears.

I decided to try a tuning run, to adjust the timing. I set MAX VE to 0, and MAX Spark to 10.0. I got the hits screen, and had a few warnings. They dealt with warm up stuff, I wasn't concerned as the bike had cooled down since the last VE tune run.

I took off on my run, and it lasted about 20 minutes.

I checked the PV and the hits screen showed no hits what so ever... it was all green, and the "MAP:L" was flashing....

I checked and there were no stored codes in the ECM...

Not sure what caused that. Did something get disconnected between the tuner/AT110/ECM?

Done for the day. Gonna try it again tomorrow, just trying to figure out if it's something I did, need to do, to get a successful timing AT run....

Thoughts?

 

From the manual:
MAP:L The map sensor is reading too low (usually deceleration). Map <20 KPA.

What did the log show for MAP readings ?

I would also make sure the plugs are secure.
.

 

I didn't hit the "data log" button on the bottom of the green screen. It's my understanding you need to do that to be able to retrieve the log. I just leave it on the green screen, put it in a bag on the seat, and at the end of the tuning run, look to see how many cells I managed to turn red. I also look to see percentage of change... I was shocked to see all cells blank and "MAP:L" flashing...

I knew that MAP:L means low map sensor reading, that's why I checked for codes. I figured if there was a problem with the MAP sensor on the bike, it would throw a code.... none. That led me to believe it was something I selected (or didn't select) on the PV or the wiring came loose..

I was hoping someone would have had it happen and say, "Oh yeah, you forgot to XXXXX."

I'll check the plugs, but it was running great, and I did 2 successful VE tuning runs, 30 min each, about an hour before I did the timing run..

I guess tomorrow I'll just start from scratch and hope I just set it up wrong or something quirky happened...

Thanks for the suggestions...

 

Yes, you have to hit Datalog to record the run for viewing later, but there is a wealth of information in the logs.

The reason I asked about them, if it was showing map readings of 40, 50, etc. it would indicate it everything was plugged in properly, but the PV was just ignoring them.

Remember, the message I posted above is only for the default settings (Min Map of 20)

There is a setting to adjust that (Jamie from FM has advised in the past users change that to 30 so as not to screw up decel areas).

Is it possible you accidentally changed the Min Map setting when you were adjusting for Spark Learn?
.

 

I thought we'd done this already but it was a discussion over on HTT.

I would maybe start by asking Roy at Dynojet if the spark tuning in the autotune process works at all or should be avoided before you start chasing ghosts and de tuning your tables unnecessarily.

The process seems to do some averaging rubbish that includes the data from the knock retard removal, and not just the actual knock events ...or at the very least that was what it seemed to be doing if you try it with AT basic and narrowband. Maybe the AT100 is different but I can't see how.

If Roy says it is good, then go with that advice.

 

Thanks for the replies.....

I originally did have min MAP @ 30kpa for the VE runs, but changed it back to min 20kpa for the timing run..

I've read the timing auto tune only removes timing, based on knock events. I will call Fuel Moto on monday with a few questions on the whole timing auto tune thing, or maybe I'll just check in with Roy @ DynoJet. I have several questions having done it now for the first time.

Here is what I found out playing with it this morning, if my conclusions are in wrong, feel free to correct me.

The "MAP:L" just intermittently flashes at idle and no-load use of the throttle (just revving the engine in neutral). When I switch back to an auto tune for VE tables, the intermittently flashing "MAP:L" did not keep cells from being filled on the hit screen in VE mode.... seems to be normal circumstance and the PV was working properly every time I switched back to VE tuning... This should rule out a problem with the wiring, PV, or actual MAP sensor (still no codes)....

I've never done an auto tune for timing before, because I've read it should only be done with wide band O2 sensors. A completely green hits screen, void of any cells filled in was new to me. I am going to assume that a "Hit" isn't cell checked, but rather "hits" are actual measured knock events in a given cell.

I suspect that the total lack of "hits" was because there were no recorded knock events... does this sound reasonable?

Anyways, I'll call someone Monday and get an answer... I'll post up what I find out...

FWIW... the Auto Tune Pro, when tuning VE tables, hit far more cells than I ever did using AT basic with the OEM O2 sensors. I guess that should be expected because of the larger range the wide band O2 sensors can read. I forgot to check the % of change after the 1st AT run, but after the second AT run, the average cell change was 1%....

The bike, with just the two VE table sessions, runs noticeably better...!


Additional FWIW: I actually took a powervision class from DynoJet a couple years ago. They spent most of the time on marketing (here's what we have and what it does) and then using the PV on a dyno... a lot of it was over my head, as I was brand new to the PV & EFI tuning. Plus, it was a class for people in the business of tuning bikes, they let me go to see how I liked it. I learned a bunch, but a lot of it was over my head...

I suspect DynoJet (or anyone with good tuning credentials) could probably make some good money putting together a tuning class for owners wanting to learn about tuning. Just a class covering some basic tuning theory and protocols. Then actual step by step for AT basic & AT pro... the videos I've seen cover basics, but always leave me with several questions...

 

Sorry, I forgot to say about that bit...
To autotune (anything) you have to be in the datalog screen - that is when it starts measuring and crunching new numbers.

Thing is, the hits screen only shows you learned VE cells but tells you nothing at all about spark tuning, if you happen to have tried to enable that. [One indication of a half thought out idea if you ask me but never mind]

When you exit the datalog screen only then does it tell you anything about how many cells have been learnt.

If VE learns the same value that is still a learnt cell, and you get the average and max etc as you have seen before. Learned spark timing cells on the other hand, is always zero unless there was either a knock event, or the knock control is still slowly removing the applied knock retard. If you had no knock, you get 0 learned cells?!? [again kind of back to front and not useful if you ask me - if there is no knock in the cell you were in then there was no knock ...er, that means we have learnt something, no?]

As for the status messages flashing on the datalog screen, that's just what PV seems to do ...I'm kind of too busy riding to notice what they may or may not be trying to say

 

Thanks for your response....... I think you confirmed what I suspected..

If I may ask for clarification. Are you saying that while auto tuning spark, a blank hits screen will only be blank if I have no knock events, or it will always be blank and the only way to tell if anything has been learned is after leaving the data log screen....?

I am beginning to wonder what, if anything, a timing auto tune really does.... unless your tune is grossly advanced....

Thanks for the help in understanding this timing tune stuff...

 

The hits screen only shows VE hits. If you are trying to lear spark but not VE, then it will always be blank. I could not find a "Mode" while in Datalog that shows anything to do with spark/knock.

Exiting Datalog mode to where it tells you the average change in VE etc is the only place it will tell you anything about spark/knock - if there were no learned cells, then there was no knock detected during that run. The only way to see how much and where anything was learned is to use the WinPV software to compare a before and after map.

Spark tuning using the Delphi knock system only ever retards, it never adds advance.

The problem between Delphi and Power Vision is that the Delphi knock strategy is to retard timing in the event of a knock event ...and to keep it retarded for a period of time even though you may no longer be at the rpm/load that is subject to too much advance. If there is no further knock detected then it gradually removes the amount of retard back to 0 and normal spark table operation.

Now if the Powervision spark tuning ONLY took its learning from actual knock events then it may be useful, but it can not or does not differentiate between actual knock activity, and Delphi's timing management after a knock event. That is just wrong.

 

There is simply not enough good information in the Harley forums on tuning spark tables for a DIY type to do it ...other than, if you get knock - retard timing in the table around the rpm/kpa that it happened.

Which spark table you you start with is a bit of a joke... have a look at 40 base maps across all the different tuning devices and you will find 40 different tables ...sorry, make that 80 - I forgot about front and rear