Link CAN lambda and methanol

[NathanFreke]

I have a thunder and am using the 2 x onboard lambda channels and then have 2 x can lambda plumbed in as well so have 1 lambda per cylinder. 

I am using methanol and have the issue where the LSU sensors only go to .67 lambda, whereas I need to be targeting .65. 

Im sure I’m not the only one who has experienced the issue of this, so wondered what / if any solution there is?

i guess using the NTK sensors would work, but is there any way this can be sorted by link?

[Simon]

Its a deliberate limit from our end as the sensor accuracy is very poor if we try read any richer.

Its only an issue for the guys running methanol which in the grand scheme for us is a relatively small market.

 

[NathanFreke]

So no solution possible?

This is one of your sponsored cars.. would be cool if something is possible, don’t really want to get a plex unit instead

[Tim Hardisty]

LSU 4.9 specifications state a lower limit of 0.65 - this is not an issue with Link; just the limit of that sensor. PLX may supposedly provide an output with a richer lambda reading for the same sensor but, as Simon stated, it is not a value I would want to rely on.

NTK LZA09 ("L2H2") is possibly more accurate in that lambda range, but that needs a completely different hardware interface.

[NathanFreke]

Yes it would need to be used with NTK sensors, not LSU. 

I didn’t know what the implications were to make an NTK work with the link, but it sounds like it’s fairly complex

[Tim Hardisty]

1 minute ago, NathanFreke said:

Yes it would need to be used with NTK sensors, not LSU. 

I didn’t know what the implications were to make an NTK work with the link, but it sounds like it’s fairly complex

Completely different hardware interface.

LSU 4.9 is a great choice for the vast majority of applications, so undoubtedly the right choice from a commercial point of view. The NTK is twice the price and - technically - obsolete!

[NathanFreke]

Apart from for methanol

[Tim Hardisty]

5 minutes ago, NathanFreke said:

Apart from for methanol

Do you really need lambda less than about 0.7, even on pure Ethanol (or methanol)?

[ClintBHP]

You would not want to be doing any active manipulation of fuel on the floor measurement of the sensor, but you could measure the difference between 0.71 and 0.68 as its a percentage you will get a good idea of the % of fuel to remove to get to 0.65

[Tim Hardisty]

Sorry to be dumb/inexperienced...why do you need/want .65 lambda? That's SOOO rich!

[ClintBHP]

10 minutes ago, Tim Hardisty said:

Sorry to be dumb/inexperienced...why do you need/want .65 lambda? That's SOOO rich!

You can run Meth down at 0.55 Lambda in drag applications.

[Tim Hardisty]

2 minutes ago, ClintBHP said:

You can run Meth down at 0.55 Lambda in drag applications.

OK...guessing that's for cooling reasons? So closed-loop lambda control isn't really that relevant and you'd run open loop and just chuck loads of fuel in?

[ClintBHP]

8 minutes ago, Tim Hardisty said:

OK...guessing that's for cooling reasons? So closed-loop lambda control isn't really that relevant and you'd run open loop and just chuck loads of fuel in?

CLL isn't very good on boost so most people would be open anyway.

In practice rich mixture has little downside except fuel economy, so you would run as rich as possible spark permitting and with big capacitive discharge systems you can go low.

[Tim Hardisty]

CLL should be very good, on boost, for most boosted applications. If not, there's an ecu problem that I don't have the (Link) experience to comment on.

But, taking your point at face value, the "inability" of an LSU 4.9 to measure very rich lambdas is  therefore a bit of a moot point? The point of a wideband is to control lambda outside of stoich. running, to the tuners wish/desire, and you're saying it's not relevant for this application? Although excess fuel might cause bore washing: on a full-out drag car, rebuilt every few runs, who cares; but the majority of Link customers are not in that category I imagine!

 

[AbbeyMS]

My 2 pence worth , 0.65 on Meths is super rich but the difference between 0.8 and 0.65 lambda is safety the power difference will be very small from what I have found.

We we was running a methanol drag car we was running NTK sensors but was only using the lambda's as a reference we looked upon the EGT's to get an better idea on how the car was running.

Running CLL on full throttle on the Link is something that I would like to do but we dont have the ability to change the closed loop control from low load to high load like Syvecs. I.e have a much smaller adjustment parameter when on high load parameters and it would be good to have the ability to change the CLL % to a different figure from +(add) to -(minus) again like Syvecs.