MST3K VE vs MSnS-E WB AFR

[kyle242gt]

How's that for alphabet soup? :rofl:

Does MST3K interpret the MSnS-E target AFR as the EGO switchpoint or does it use stoich? (IE, the voltage entered in MST3K's setup screen)

My MST3K-derived VE's are almost equal to the VE table I'm running, and it's tuned to 16:1, so it makes me think that MST3K is using a variable EGO switchpoint (IE 16:1 at cruise, 13:1 at WOT).

Just want to make sure before I start making changes to my VE table.

Thanks for any help!

[Neil Peart]

you just did that to make the k-jet boys head hurt!!

:P

hey, where are the P-shop files foo!

:-P

[adrianpike]

How's that for alphabet soup? :rofl:

Does MST3K interpret the MSnS-E target AFR as the EGO switchpoint or does it use stoich? (IE, the voltage entered in MST3K's setup screen)

My MST3K-derived VE's are almost equal to the VE table I'm running, and it's tuned to 16:1, so it makes me think that MST3K is using a variable EGO switchpoint (IE 16:1 at cruise, 13:1 at WOT).

Just want to make sure before I start making changes to my VE table.

Thanks for any help!

Well, before your experience, I would have probably voted for it using the voltage in the setup screen. I thought that MST3K was pre-msns-e.
Probably best to email the MST3K developer : BiggRanger at direcway dot com

[Matt Dupuis]

Unless he's changed something to allow use for target AFR, which I doubt, MST3K uses the voltage in the setup screen. AFAIK, the target AFR voltage isn't output into the datalog file, and the last time I played with MST3K, the AFR section didn't work at all that way. Instead it was meant for modifying the 14.7:1 AFR target voltage into different AFRs for a given point of the map.

You pretty much need to know where you want the mixture rich and where you want it lean, and when using MST3K to interpret different target voltages, you've got to manually transfer over the points that correlate to each target AFR.

I.E., when I want a mixture of 13.0:1 in a given area of the VE table, I must set my target voltage to 2.00 and then filter the datalog. Then once it filters and interperets the information, I have to then pick the VE values that are in the data points where I want an AFR of 13.0:1. If you are really picky about your target AFR table, you may need to use MST3K 6 or 7 times in this fashion before you have your table fully fleshed out. Keeping the number of different target VE values small (i.e., 12.0, 12.5, 13.0, 14.7, 16.0) limits the amount of times you need to filter a set of datalogs.

MST3K is really meant for a narrowband O2 sensor... It doesn't work so hot with a wideband. It'll WORK, but not well. Someone should really come up with another tuning tool... I'll be installing my narrowband sensor during the initial phase of tuning my car, I think.

[JayPSC]

Yeah this very true, with the new target AFRs in extra there's no reason why it can't write its own VE table to an extent. It knows how much it's having to richen/lean the mixture to get to the target AFR, be handy if it wrote it to a seperate VE table on the fly.

Jay

[kyle242gt]

I sent the author an email, we shall see what comes of it.

Matt, I get what you're saying, and I expected as such, but the logs don't bear it out. Very odd.

I just looked at two DLs together, and damn if the MST3K values with the switchpoint set to stoich aren't spot on, particularly in places where I real-time tuned them to match the target AFR. :e-shrug:


When I scale in MST3K, the numbers are off.

Target AFR / MST Error (MSnS - MST3K)
16.0 / 4 to 6
15.75 / 3 to4
15.5 / 0 to 2
15.0 / -1 to 1
14.5 / -3
13.5 / -9 to -12

For example, I'm running 87VE and think it's 13.5, but MST3K when scaled says 99VE, 9pts too rich; Set to 56VE = 16:1, MST3K scales to 51, 5pts too lean.

The unscaled (but filtered and calculated in MST3K) VE's are almost all within a point or two, regardless of AFR.

Fuel for the fire. It sure looks like it's somehow picking up the right crossover point and doesn't need to be scaled... that's for NB weenies.

[Matt Dupuis]

Yeah this very true, with the new target AFRs in extra there's no reason why it can't write its own VE table to an extent. It knows how much it's having to richen/lean the mixture to get to the target AFR, be handy if it wrote it to a seperate VE table on the fly.

Jay

Target AFR is being used by 90% of the people incorrectly, IMHO. I believe the proper way to do it is to tune the VE table to get the desired AFR, and keep tuning it until you're happy. Then you can enter the determined "ideal" AFRs into the VE table so that MS can monitor and fine tune it until you're running at your previously determined "ideal" tune.

What people tend to do is to set the AFR table first and drive around twiddling with W and Q until their O2 sensor is only correcting by a few percent. It works, but I disagree with it, much like I disagree with other common methods of tuning that also "work".

It's important to get the VE table set properly first, because during acceleration you have no O2 sensor correction and it suddenly switches from whatever EGOcorr you're at to 100%. You could be driving around on a lean VE table, with the AFR table richening things up, and then when you step on the gas the EGO correction drops to 100% and you get a lean stumble. This will frustrate the hell out of you when you're trying to get the acceleration settings right!

If the datalog output had a column for target voltage or delta voltage (Vtarget - Vcurrent), it would be able to tell MST3K what it needed, and you wouldn't need to fart around with any target voltage settings or AFR corrections in MST3K. It would essentially be reading the target voltage from your AFR map, and looking for those crossover points instead of the one you input in the settings screen.

Edit - Kyle, I'm not sure I understand your post fully, but after reading it a few times I think you're saying that MST3K can figure out the target AFR, and looks for those crossover points? Hmm... that I kind of doubt, but I haven't looked at the datalogs lately and I certainly haven't looked at them too much since they were changed from the original "classic" style. Maybe there's something in there that I'm unfamiliar with? Are you communicating directly between MST3K and MS, or are you collecting datalogs and exporting your VE tables from Megatune?

[JayPSC]

Completly agree on this, I haven't touched my AFR targets yet, gonna get my VE table tied down first & add AFR targets later. I've just started to re-tune my car with CFIs & WB02 & OMG what a difference the WB makes, well worth the investment. Luckily I have a friend who's pretty clued up on this so I just tell him what AFRs I want & pretty much leave him to it.
You are correct on the Vtarget - Vcurrent datalog, would make things 100x times easier for MSextra users. Worth mentioning on msefi.com maybe?

Jay

[kyle242gt]

What people tend to do is to set the AFR table first and drive around twiddling with W and Q until their O2 sensor is only correcting by a few percent. It works, but I disagree with it, much like I disagree with other common methods of tuning that also "work".

It's important to get the VE table set properly first, because during acceleration you have no O2 sensor correction and it suddenly switches from whatever EGOcorr you're at to 100%. You could be driving around on a lean VE table, with the AFR table richening things up, and then when you step on the gas the EGO correction drops to 100% and you get a lean stumble. This will frustrate the hell out of you when you're trying to get the acceleration settings right!I'm pretty sure it keeps a rolling EGO%, and doesn't reset on accel?

Speaking of which, care to share your accel setup? Mine is currently off completely for DLing, but IIRC you are THE Accel Wizard :-P

Not to get into a pissing contest (damn glad to see you posting again, BTW! :cheers: ) but my map is living proof that realtime AFR target adjustments work handily. Five MB of datalogs, and using MST only on 40-100KPA (16:1) MST is suggesting no more than one or two points of adjustment.

Where it doesn't work well is areas that you tend to zip through rapidly, low rpm, high load, etc.

Edit - Kyle, I'm not sure I understand your post fully, but after reading it a few times I think you're saying that MST3K can figure out the target AFR, and looks for those crossover points? Hmm... that I kind of doubt, but I haven't looked at the datalogs lately and I certainly haven't looked at them too much since they were changed from the original "classic" style. Maybe there's something in there that I'm unfamiliar with? Are you communicating directly between MST3K and MS, or are you collecting datalogs and exporting your VE tables from Megatune?
Man, I have no trouble believing it took a few reads; it took a few writes to get it even that coherent. The DLs do not have a column for delta, target, anything. I am confused as hell :e-shrug: Kyle,
Right now it's all stoich (or your ego switchpoint), there is a map modification tool to adjust the bands across your preset MAP values to a specific AFR. But all MST3K is doing is finding crossover points for your EGO switchpoint.
Darren Clark
...so... I'll do what I was supposed to do: use MST3K multiple times for multiple AFR's.

HEY :omg: I wonder if the error has something to do with the way MST calculates the differential between AFRs; IE, instead of using 0-5V for the Tech Edge WBlin, it's using the old DIY-EB/TE SVout table... there is a table to enter AFR factors (voltage = AFR).

I'll take a look, just to satisfy my curiousity.

[Matt Dupuis]

I'm pretty sure it keeps a rolling EGO%, and doesn't reset on accel?

I'm pretty sure it USED to switch right back to 100%, but maybe it's changed. I'll have a look tonight.

Speaking of which, care to share your accel setup? Mine is currently off completely for DLing, but IIRC you are THE Accel Wizard :-P

Nope. Mine's brand new, not even set up for the 5.0L on #357 injectors that's freshly installed in my car. Your accel settings are completely unique for your engine/injector combo, so I'm afraid I can't help you...

Not to get into a pissing contest but my map is living proof that realtime AFR target adjustments work handily. Five MB of datalogs, and using MST only on 40-100KPA (16:1) MST is suggesting no more than one or two points of adjustment.

(snip)

HEY :omg: I wonder if the error has something to do with the way MST calculates the differential between AFRs; IE, instead of using 0-5V for the Tech Edge WBlin, it's using the old DIY-EB/TE SVout table... there is a table to enter AFR factors (voltage = AFR).

I'll take a look, just to satisfy my curiousity.

Entirely possible you're onto something there, but I don't think MST3K cares about what type of sensor, only the voltage. There is the Wideband settings section in Settings, but I think that's for directly connecting MST3K with MS. Also, I thought you were using MST3K with target AFRs across the whole map, meaning, it was deciphering where you asked it to be rich and where you asked it to be lean, but based on what you've said above, I understand that you're only using it for one AFR at a time? If so, maybe it is something to do with a voltage difference...

[sgtrauksauff]

If you're using MST3K to run your engine, I'd be very very careful.

You might end up being stuck watching horrible movies, and Tom Servo and Crow might pull some practical jokes on you. And watch out for Pearl, if she finds you tuning engines instead of watching bad movies, she'll have Brain Guy do something to you....


:)


(I had to say something, nobody else mentioned the REAL MST3K in the posts that I've read)

--sarge

[Matt Dupuis]

Actually someone called him on that on msefi.com, and he admitted that's why it's named the way it's named.

[Magnum TE]

MST3k was a funny show, it was all about the Mitchal epp and Gamera :-P