95 replies to this topic

[finch661]

 

in regards to tuning, if you went with a wide band lambda sensor, could you make the tuning more automatic (system constantly adjusts the fuel/air ratio throughout the rev band to 14:1?)?  

 

I'm sure its possible providing the algorithm takes account of varying engine states - e.g. when accelerating, the mixture is deliberately rich and decelerating its deliberately lean as well as at idle and while coasting. You'd probably only apply real-time adjustment while cruising. The most difficult part would be understanding how the engine responds to changes in fuel metering. You can easily induce wild oscillations if the algorithm is too "aggressive" or be ineffective if too "mild".

 

My first preference for "auto-tune" is based on statistical measurement of the air/fuel ratio (limited to lean/rich with a narrowband sensor) in each RPM/Load cell. Providing you take enough samples, you can confidently determine if a cell is rich or lean then make a small discrete adjustment for that cell only. You exclude cells that cover idle and coasting, etc. 

 

that makes sense, still learning about this. i have been tempted by an injection system for a while, was planning on using spi parts. 

 

love this engineering/programming project!

[inim_repooc]

Something that you may find useful is that the A-series makes more power with bigger droplets of fuel, with less fine atomisation - something injection can work against if im right in thinking? would it be possible to get this injector to atomise less finely? (not sure how the spi injects, but I know this is what the main advantage of carbs was vs the 90's lucus custom injection systems - it could cost around 12hp top end vs a webber carb with the same intake) 

 

just food for thought! 

[miniBrain]

Hi, I think you've explained something that I was wondering about. The SPi TBI injector squirts onto the throttle butterfly where I noticed the fuel condenses and runs off into the intake throat in larger droplets. In fact, the butterfly is always wet. Until you mentioned it, I didn't think this was an optimum arrangement - it just seemed wrong to "ruin" the fine spray from the injector. But now you've mentioned that large droplets benefit the engine, it seems that this TBI is actually designed with this in mind. I could be completely wrong about this and will need to research further. Food for thought indeed!

[inim_repooc]

Hi, I think you've explained something that I was wondering about. The SPi TBI injector squirts onto the throttle butterfly where I noticed the fuel condenses and runs off into the intake throat in larger droplets. In fact, the butterfly is always wet. Until you mentioned it, I didn't think this was an optimum arrangement - it just seemed wrong to "ruin" the fine spray from the injector. But now you've mentioned that large droplets benefit the engine, it seems that this TBI is actually designed with this in mind. I could be completely wrong about this and will need to research further. Food for thought indeed!

Aha, well at least that's something!   going from what you say about the injector set-up makes me think that more research is definitely needed, I diddnt know that the TBI was setup like this - would be interesting to find out more! 

[finch661]

in regards to the spi (and this set up), the timing is still set through the dizzy isnt it?

[coopdog]

Amazing work

Goodluck with it

[miniBrain]

in regards to the spi (and this set up), the timing is still set through the dizzy isnt it?

 

Yeah, SPi's were all conventional dizzy timed - this setup is too but will be fitting coil-pack & programmed ignition later. 

 

EDIT: SPi's are timed via the ECU as per spiguy's post below

Edited by miniBrain, 20 November 2015 - 05:30 PM.

[miniBrain]

Amazing work

Goodluck with it

 

Cheers

[spiguy]

in regards to the spi (and this set up), the timing is still set through the dizzy isnt it?

 

Hmm, not really. The timing isn't set by adjusting the position of the dizzy like on a conventional, non ECU controlled setup. The ECU alters the timing by altering when the low tension side of the coil is switched. The duration that the rotor arm is adjacent to each plug outlet on the cap is sufficiently long to allow the spark to be delivered to each plug outlet over the full range of timing advance and retard as controlled by the ECU.

 

In other words if you went out and rotated the dizzy a bit, there may be a point where you could create the situation that the rotor arm was either just past the plug outlet, or not quite at the outlet, when the ECU has adjusted the timing to maximum advance or maximum retard. Conversely, a slight rotation of the dizzy either way won't make any difference to the ignition timing, as long as you don't rotate it too far.

[miniBrain]

Thanks for the clarification spiguy   I've corrected my post accordingly.

[Ethel]

There'll be two factors with droplet size:

More surface area with small droplets means more fuel can vapourise before it's trapped in the cylinder, taking up more space that would otherwise be filled with air.

Bigger droplets will have more inertia, that will have an effect on fuel distribution between the Siamese ports.

[finch661]

forgot to ask if you saw any improvement on fuel consumption?

[miniBrain]

forgot to ask if you saw any improvement on fuel consumption?

 

Not really measured it accurately yet - its looking like about 30-35 mpg. But that's stop-start and I tend to drive "enthusiastically" 

[inim_repooc]

Just been thinking about this project today and was wondering if you have any updates on whats happening? 

 

 

*Edit - been looking through your website, looks like you are now planning a kit around April?

Edited by inim_repooc, 10 February 2016 - 05:41 PM.

[miniBrain]

Just been thinking about this project today and was wondering if you have any updates on whats happening? 

 

 

*Edit - been looking through your website, looks like you are now planning a kit around April?

 

Hi,

 

I've added auto AFR correction and been testing this. Its running fine and I 'll get something up on the website soon. Next step is to test the programmed ignition  - I've just bought a coil pack and connectors so hopefully will have this running in the next week or so.

 

I want to provide it as a kit at some point but things are taking longer than I thought. I had planned to fit the ECU to a Golf GTI this winter where I could test it to the limit but other stuff keeps getting in the way!