Donations towards server fund so far this month.

 
£0.00 / £100.00 per month
Page:
Home > MS Code Discussions > MPi injection characteristic data

fastcarl

User Avatar

6962 Posts
Member #: 507
Fastest A Series Mini in the World

leeds/wakefield.

carefull or i'll fire you as my no 1 driver,

WWW.FORCE-RACING.CO.UK PLEASE CLICK HERE


turbo hogster

1641 Posts
Member #: 178
Post Whore

stowmaket suffolk

me to carl way over my head i think i will stick with carbs its far more civilised rather than being marsian let alone russian!!!! lol

always looking for them bigger bunches of bannanas


turbo hogster

1641 Posts
Member #: 178
Post Whore

stowmaket suffolk

woh way above my head but i supose threadings are from the afr gauge

always looking for them bigger bunches of bannanas


TurboDave16V
Forum Mod

10979 Posts
Member #: 17
***16***

SouthPark, Colorado

No, they're traces of the crank VR and injectors... :)

On 17th Nov, 2014 Tom Fenton said:
Sorry to say My Herpes are no better


Ready to feel Ancient ??? This is 26 years old as of 2022 https://youtu.be/YQQokcoOzeY



Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

Sorry to drag up an old thread, but I'm considering injection timing, again.

I struggling with calculating the time from the injection pulse to when it gets into the cylinder.

I think that the consensus of opinion is that the data from TDs trial shows that there is a single injection pulse per port per cycle. Hence that pulse must be timed so that half the fuel goes in the outer cylinder, the other half sits in the port until the inner cylinder inlet valves open.

At idle, with a very short injection pulse, the pulse starts just after TDC. However, the outer cylinder inlet valve does not close until at least 200 degrees after that.

Hence, the time in the port must be equivalent to approx. 180 degrees of crank rotation.

At higher revs, the pulse seems to spread around TDC, still taking half a rev to get to the cylinder.

Agree?

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


jbelanger

1267 Posts
Member #: 831
Post Whore

Montreal, Canada

It's very difficult to see from the data but if the patent is any indication, I think that the fuel is injected on the overlap of the inner cylinder closing and the outer cylinder opening. That would also be part of the reason why you want minimal exhaust overlap with the cam on an MPi.

Jean

http://www.jbperf.com/


Bat

User Avatar

4559 Posts
Member #: 786
Post Whore

Bermingum

Hi,
That was my understanding of it too :)
Cheers,
Gavin :)

VEMs Authorised Installer / Re-seller. K head kits now available!

WB/EGT gauges. Click here for customers write-up

Visit www.doyouneedabrain.co.uk

My Mini build diary


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland


On 5th of Jan, 2008 at 09:47pm jbelanger said:
It's very difficult to see from the data but if the patent is any indication, I think that the fuel is injected on the overlap of the inner cylinder closing and the outer cylinder opening. That would also be part of the reason why you want minimal exhaust overlap with the cam on an MPi.

Jean


That is how i always thought it worked until we saw Dave's data.

Given that the injection pulse is at TDC on cylinder no. 1, it is far too late for that to be the case.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


jbelanger

1267 Posts
Member #: 831
Post Whore

Montreal, Canada

Could the fuel travel time be about 300 degrees instead of 180?

http://www.jbperf.com/


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

I've done some calcs, based on Marcel's method:

http://www.starchak.ca/efi/speed.htm

He estimated that it took 2.5ms for the air to travel 75mm in the inlet port.

I think it's more like 125mm for an MPi manifold, so let's say 4.1 ms.

He also calculated that at 3700 rpm, there are 22.2 degrees per ms.

So crank degrees from injector to to open valve is 4.1*22.2 = 91 degrees.

So, at 3700rpm the pulse must start well before 91 degrees BTDC to get any fuel in the inner cylinder.

And from Dave's trace it does, so no problem.

EDIT: apologies for numerous edits, but go the numbers a bit wrong.

I think that I've just proved that the original theory as stated by Jean & Bat is correct - LOL.

Edited by Paul S on 6th Jan, 2008.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

However, if you do the calcs for 1000 rpm you get the same answer, 91 degrees.

But if you reduce the VE for someting reasonable you get a larger angle. 167 degrees at 50% VE.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


Bat

User Avatar

4559 Posts
Member #: 786
Post Whore

Bermingum

Hi,
That's a worry, as surely VE will increase when the boost goes in?
Cheers,
Gavin :)

VEMs Authorised Installer / Re-seller. K head kits now available!

WB/EGT gauges. Click here for customers write-up

Visit www.doyouneedabrain.co.uk

My Mini build diary


alpa

520 Posts
Member #: 2093
Post Whore

Grenoble, France

Paul, do you assume the air speed depends on the VE ?
I think it depends on the pressure gradient (intake/cylinder) and the air VOLUME (which is constant as air fills all the room). The impact of the pressure gradient is difficult to estimate, it's a dynamic behaviour.

Edit: to me the problem is simple. The average air speed depends only on the RPM because the volume to transport is rougly the same. The instant air speed is oscillating and reflects intake/cyl pressure ratio oscillations which are hard to estimate.

My understanding is that we have to use the average air speed to estimate the transport delay.

Edited by alpa on 7th Jan, 2008.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm


Dylan8660

3 Posts
Member #: 9781
Junior Member

Ireland

I built a four injector manifold to run on the standard mpi ecu, it wasn’t entirely successful but thanks to the graph in this thread I had reason as to why that might be. I’m now developing a manifold in which the injector angles aren’t as extreme that will hopefully provide better running. But to cut a long story short, which can be found here: http://www.theminiforum.co.uk/forums/topic...r-mpi-manifold/ . During my trials I came up with this theory about fuel injecting Siamese ports using four injectors.
(see attachment)
Obviously we’re looking at only one of the two inlet tracts here, but what this suggests is the potential duty cycle can be doubled to 50% as the two opposing injection pulses can overlap to each other’s fixed point.
The reason for posting it here is that I certainly don’t have the knowledge, tools or talent to take this any further even if it’s possible, but there are obviously very talented people on here that could tell me if it’s workable or might find the theory of some use.

Slight change to the diagram as it may not of been too clear.


Attachments:

Edited by Dylan8660 on 18th Mar, 2012.


Rod S

User Avatar

5988 Posts
Member #: 2024
Formally Retired

Rural Suffolk

Dylan,

Although this is added onto a very old thread dragged up, I'll add a "current" point of view.

As you rightly say in your TMF post, the two MPI injectors are indeed driven by a 4 channel MEMS with the outputs paired up.

However, so far as we know, this is because they needed 4 channels to alter the timing of the first/second injector pulse to compensate for the siamese "port robbing" issue and meet emissions targets for the final years of the (classic) Mini production, and wanted to do it with parts readily available.

The limit of MPI is the ECU can't easily be reprogrammed for higher outputs (note I said can't easily Colin).....

So if you look to replicate what Rover did with a modern programmable ECU, you can do it with two injectors, four or six if staged, but you simply need an ECU that can vary the injection timing between the inner and outer cylinders.

MS2 (siamese code) or MS3 (with outputs doubled up) are the preferred on this forum for a 5 port head.

If you are just looking to double the injection capacity on an MPI setup, well I don't think you can as you can't (easily) re-programme the inputs.

Schrödinger's cat - so which one am I ???


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

I'd had a lot of success with mid-pulse timing and firing two/three injectors per port all at the same time. Timed so that the AFRs are equal in the inner and outer cylinders.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


Dylan8660

3 Posts
Member #: 9781
Junior Member

Ireland

Thanks for the replies, I'm sorry for dragging up an old topic but considering its title and content I didn't feel it inappropriate. Although it could be looked at as being staged I feel the reference points for injector events could give a little more finesse of control. I’m not just thinking about simply doubling or tripling up but rather if the use of four injectors could open possibilities that two don’t offer.

Another diagram hopefully offering better clarity of thought.


Attachments:


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

OK, so how do you account for wall wetting ???

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


jbelanger

1267 Posts
Member #: 831
Post Whore

Montreal, Canada

You're better off using 4 injectors in a staged setup with the current dual pulse or single pulse modes of the siamese code rather than what you propose.

This way you can size the primary injectors for optimal low load behaviour with the secondary injectors for maximum power instead of having to use 4 identical injectors which will compromise low load behaviour.

And as Paul mentions, you'll have to control the individual pulse widths and/or timing as a function of load and RPM to take into account wall wetting.

Jean

http://www.jbperf.com/


Rod S

User Avatar

5988 Posts
Member #: 2024
Formally Retired

Rural Suffolk

On 18th Mar, 2012 Dylan8660 said:
Another diagram hopefully offering better clarity of thought.


Dylan,

What you show in that plot is what the MS2-Extra (siamese code) was designed by Jean to do, with the ability to vary the timing points over the whole RPM/load range by setting start/mid/end pulse values at any given load/RPM value.

Whether you do it with two injectors, double them up but still acting as two, or stage them from four or six, isn't really the issue.

Staging is best IMO but, to achieve the spread you desire in the plots you show is quite straightforward with the siamese code.

What I'm not quite sure of from your first post is whether you are attempting to do this with the Rover MPI MEMS CPU which you first mentioned.

MEMS is "fixed" code (unless you re-programme at machine code level) so can't be altered like an MS2/3 which is just done from a laptop.

Hope that makes sense.

Schrödinger's cat - so which one am I ???


Dylan8660

3 Posts
Member #: 9781
Junior Member

Ireland

Thanks Rod, I'm not looking to use the std rover ecu as I'm aware of the re-mapping issues. I was just wondering if it was possible and you've given me a few pointers if I decide to take it onto the next level. If I do I'd want it to be as plug and play with the rover sensors, alarm and immobiliser as possible and from what I've read so far there may problems with that, particualy the tp and flywheel sensors.

Home > MS Code Discussions > MPi injection characteristic data
Users viewing this thread: none. (+ 1 Guests) <- Prev  
To post messages you must be logged in!
Username: Password:
Page: