Members Login
Username 
 
Password 
    Remember Me  
Post Info TOPIC: MAD port jumper selectable map


Guru

Status: Offline
Posts: 963
Date:
MAD port jumper selectable map


The MAD port plug on the bike wire harness has a ‘connector’ or dummy plug on the end of it to keep it clean and dry. Or so I thought. The shop manual states there are two possible connectors that may be installed depending on ECU model number. One type is just a plug, The other has a jumper wire in it. The manual lists the following:

ECU 21175-1087 open connector
ECU 21175-1089 open connector
ECU 21175-1090 Short circuit connector
ECU 21175-1091 Short Circuit connector
ECU 21175-1092 Short Circuit connector

The short is between the COV 3 input and ground. When I was tracing out the COV 3 code for the MAD I ran across something very strange. The ECU I'm working on is a 21175-1089 which says no jumper installed, yet the code looks for a jumper at power up and loads different map values for 2nd and 3rd gears depending on whether the jumper is installed or not.

Not only that, there is a value in NVRAM that determines whether the change_the_map_if_COV3_jumpered feature is enabled or not. It basically makes it look like no jumper present even if the jumper is installed. My test ECU was jumper disabled at the factory but it turns out this NVRAM value can be changed by the MAP port.

I found that the COV2 (Cylinder) input actually has a 5th setting that allows you to modify this Jumper enable value.

A quick glance at the maps shows little or no difference between the jumpered and unjumpered maps for 3rd gear. The 2nd gear map, however, has a 10% increase across the entire upper 3/4s of the data.

Again the usual disclaimers apply. I still don't know how map data like this translates into actual changes in injector timing or horsepower. But I thought it was an interesting find.

I guess I will have to decide whether or not to modify the MAD design to add a 5th position to the COV 2 knob. Luckily I built it with a 6 position rotary switch so I just need to add another resistor and move the locking ring.

I'll add the above mentioned gear maps to the Enginuity map viewer definiton next time I update it.

-- Edited by RidgeRacer at 05:47, 2006-12-07

__________________


Member

Status: Offline
Posts: 13
Date:

If you were able to provide me with a prototype, I can talk to Marc at Factory Pro and see if we can get a bike on the dyno there so we can play with the settings and get some more concrete answers for you. What would be really great would be if we could re-flash the NVRAM, which would let us "try out" the mappings from various different boxes for different markets to see which one does what, and why. Factory Pro dynos have a very fast IR four-gas analyzer, which tells you a very great deal about what is going on inside the combustion chamber (even a few degrees of spark timing change will show quite clearly on the four-gas data stream).

I get down that way about once every month or two, and I need to do a few dyno runs with my personal bike anyway since the motor is so far from stock (how else are you going to tune a unique configuration to really get the most out of it without a dyno? Even shooting for a set lambda value is not optimal -- the best AFR generally varies with the torque curve (better torque = leaner mixture for best power)), and things like cam timing and combustion chamber shape alter what the "best" value is at a given throttle opening and engine speed. Heck, changing your ignition timing will read as a change in lambda if your only tool is a wideband O2 sensor! So off to Factory I go. Ping me via email and I'll get with Marc and see if we can make that happen.

__________________


Guru

Status: Offline
Posts: 963
Date:

AWade wrote:

I What would be really great would be if we could re-flash the NVRAM, which would let us "try out" the mappings from various different boxes for different markets to see which one does what, and why.



Just to make sure we are on the same page here....

I only covered the NVRAM briefly in the other thread, What is the MAD, let me go into a little more detail here. The CPU has several internal FLASH banks that hold the runtime code and air/fuel map data. While these can be FLASHED at the factory they are not designed to be modified in situ. The biggest limiting factor is that they can only be bank erased. To write a single byte you must first erase the whole smash. Obviously its kind of hard to write an onboard FLASH routine that as step one erases itself.

So for the ECU to be able to remember anything after the key is turned off, like diagnostic error codes, it needs some kind of external non-volatile memory (NVRAM). The zx-12 ECU has a small, 512 byte, serial eeprom. While technically a FLASH chip it allows you to Randomly Access any single byte of Memory and read/write it. In other words it acts like a RAM chip with the one important exception that it doesn't lose its data when you remove power. This is what I refer to when I say NVRAM.

The NVRAM is external to the CPU and can not be FLASH'd like the CPU. It can only be accessed by CPU software. Of course using the BDM port I could download a small piece of code into CPU RAM that would change the NVRAM contents and then execute it. But to me it seems more trouble than its worth.

I have dissassembled the software enough to know exactly what number will be written to what NVRAM location for every setting of the MAD box. What I don't know is exactly how much offset to the injector time the the MAD values provide. 5%, 10%, more? That information is also in the code, I just haven't tracked it down yet.

Also I think I spoke to some head guy at Factory Pro last year when I was trying to track down the plugs and figured I might as well start with Teka and Yosh since I knew they must know where to get them. (the zx-12 and suzuki gixxer plugs are the same) I think if they were really interested they have would have already done something. I gave them enough info over the phone and posted on the net that they could reprogram a Teka and use it on a ZX-12. I've been half waiting for them to release a new Teka that also works with a -12.

But I would be more than willing to talk to them again and send them my prototype, as long as they sent it back with a plug attached smile

But, back to what you said above about re-flashing the NVRAM with 'mappings from different boxes'. By boxes I assume you mean ECUs and mappings mean different air/fuel maps, not various MAD settings?


 



-- Edited by RidgeRacer at 16:58, 2007-04-14

__________________


Member

Status: Offline
Posts: 13
Date:

Yes, my bad on confusing the terms for the non-MCU flash memory and the MCU's onboard flash memory. I'm not a total genius, but I have a pretty decent understanding of MCU architecture. ;) And yes, I was meaning running different OE mappings on the dyno, not trying different trim table settings via MAD. All you would really be able to do with dyno-ing various MAD settings would be to establish the resolution of the edits that could be made, as well as the max increase or decrease in fueling.

I'd be particularly interested to see what changes were made between different boxes in terms of spark timing. This is an oft-overlooked area of tuning, and it's dangerous to play around with unless you know what you are doing; furthermore, it's really only possible to get a clear idea of the impact of smaller changes made to ignition timing by using four-gas analysis.

I'd have to assume that the market share for a now-obsolete bike is not sufficient for an aftermarket supplier to have done the R&D work. Whereas the Teka SFI and the various Yosh boxes will work on any fuel-injected sport motorcycle made by Suzuki going back almost 10 years, if the MAD only works on ZX-12R bikes there's no longer any real reason to spend time and energy developing it from a commercial sense. Since it's obviously pretty easy to add the functionality now that someone (you) has worked out the various input lines and how they operate, it would only require the additional programming to get the Teka to work with the ZX-12R; however, it costs money to hire a programmer...

I will say that if you've had a look inside a Teka SFI, and you have some programming experience for embedded systems, you will very likely be able to write the additional code for the ZX-12R and make a dual-make Teka SFI from a Suzuki-only unit without much trouble. Then again, you could get the same functionality with your small box with knobs, switches and some resistors. The Suzukis cannot, AFAIK, make changes to fueling "on the fly" with nothing more than switches and resistors (i.e. the original Yosh box), but from your explanation, it appears you CAN get this with your box and the ZX-12R. I'd say there's nothing to be gained from building something more complex that has the same functionality. I think the only real reason Factory Pro developed the Teka SFI is the ability to change fueling in real-time, having a greater range and finer steps than the original Yosh box did, and the ability to change idle setting per cylinder.

__________________


Veteran Member

Status: Offline
Posts: 80
Date:

does anyone know for sure that the yoshi box can't make changes 'on the fly'?

i'm suspicious that Yoshi say to turn the bike off while making the changes is because of non-technical reasons (maybe just covering their arses/asses) ... teka use this as a selling point for their unit but as far as I can see there's nothing special that the teka can do (apart from having more settings) that the yoshi box can't - I mean they're both only fiddling with the same interface pins on the ecu

I'm wondering whether the teka sends a very short pulse on the 'write' line whereas the yoshi unit recommends holding it low for several seconds ...?

also teka warn that with suzukis, there's a hiccup when the settings are changed - length of the hiccup is probably related to the length of time that the write line is held low???

anyone able to shed some light here?

cheers,
MarkW



__________________


Veteran Member

Status: Offline
Posts: 68
Date:

From what I can recall from 10 years ago, no the Yosh box does not allow on-the-fly changes. The Teka SFI will allow on-the-fly adjustments, except for when you're adjusting the idle mixture on all cylinders simultaneously. I haven't noticed any hiccup during writes. The only anomaly I've experienced is the idle settings not being respected while adjusting the main settings.

I'm learning the Yosh EMPro stuff now - what a revolutionary change from the old Yosh box.

__________________


Guru

Status: Offline
Posts: 2338
Date:

tcchin wrote:

 no the Yosh box does not allow on-the-fly changes.


Actually the yoshbox does allow on-the-fly changes - the only problem is that the individual cylinders must be "burned" each one at a time. The L/M/H can be adjusted while running the engine and burned (write) after adjusting the mixture.




__________________

When asking a question, you can also consider posting it to facebook:

http://www.facebook.com/ecueditorcom

Page 1 of 1  sorted by
 
Quick Reply

Please log in to post quick replies.

Tweet this page Post to Digg Post to Del.icio.us


Create your own FREE Forum
Report Abuse
Powered by ActiveBoard