Speeduino is now on Github Sponsors (Rather than Patreon): https://github.com/sponsors/noisymime
For any other general hardware. Can be OEM parts (Eg Coils, injectors etc) or anything more specific.
By Jansun
#25707
Hi all,

I was wondering if it is possible to calculate the time between injection start and the actual wideband reading.

I would say there is a time difference and it should be a constant that dependents on rpm, engine volume and maybe backpressure or so. Someone here who knows anything about this?
User avatar
By PSIG
#25715
Jansun wrote:
Fri Apr 20, 2018 1:21 pm
I was wondering if it is possible to calculate the time between injection start and the actual wideband reading.
I can't tell you how to calculate it on paper for any point in operation; but I find it as best I can for several locations and then have a rough idea what the delay is for the remainder of the range. As the processor and TS readouts are not all instantaneous for all data; I look at the output traces in MLV and determine the relative delay of the various data traces relative to the time, including O2 (Part1). Then find the delay from the beginning of an identifiable injection point event (perhaps AE) to the start of the responding readout of the event by O2 (Part2) relative to the timeline. Delay = Part2 - Part1

Having said that, Part1 is relatively small and varies from system-to-system and even at different data rates, and simply using Part2 (actual sensor delay) seems to generally get me in the ballpark for usable information. Hope that helps some, and I'm always looking for better ways to do it.

David

PS: Some WBO2 controllers can be programmed for certain advanced functions, such as the Innovate controllers that can be set for slower reporting but internally average the Lambda between reports. This is very useful to dampen a jumpy O2 gauge. The average is not a concern (a good thing), but be aware some controllers can have or be set for different reporting speeds, affecting the 'delay'. Take that into account for any controller, and especially if yours is one of the programmable types or set for that function, but it should still show in the data delay timeline if you don't know what it is.
By Jansun
#25771
Thanks,

The reason i want to know is a bit different then maybe expected, but its for a datalogger. Still original ecu, so cant use TS and MLV properly.. I rebuild my bike and now i want to log the o2 sensor, reletively to TPS and RPM. So I think, if i have a stable situation between this three reading, i can make a afr table out of it. But.. the only way to get it to work properly is to have the right delay at each reading.

I noticed the sample time and its pretty long; about 80ms (1/12 of a sec, got the Innovate MTX-L set) So to be in a stable situation i need to wait first for 80ms and then calculate an average of the afr over a certain time, other sensors have to be stable in that time (stays the complete time in one cell of the table). How shorter this time the more results i get, but the less i'm in a stable situation, i have to blend out map, wall wetting etc).

So to get back to the delay, besides from calculating an average I could add a min and max. Then, if i see a big difference between the min and max i now i'm wrong with or the delay, or sample time(not afr sample time, total sample time. Two values tweaking is harder then one, so i was hoping to get this one out.

What are common time delays, to put in as a start?
User avatar
By PSIG
#25787
I am not following. Delays can be any value, based on many, many factors. I can't help much here, as I do not know your purpose. Why do you care what the original tune's AFRs are? Is it tuned exactly how you want it? If so, why change the ECM? If I were trying to build a start-up tune for A-N fueling using a running engine and data-logging, I would probably just use a throttle stop to hold a certain TPS, and log the injector pulse width as it accelerates across the RPM. Set Speeduino to provide the same PW at the same TPS and RPM points (refer to calculations in the Wiki and code). Repeat for different TPS hold points, perhaps ±10, 40, 70 and 100%. I might also capture spark advance. Blend the tables, start it up, and begin tuning for the fuel you're using and what you want the engine to do. I do not know if that is a good suggestion or bad, as I don't know what your purpose is. :?

David
L TWIN ducati vr signal need help

As above a tooth log would be useful. Which condi[…]

Modified 3 cylinder

You would still need to remove the intake setup of[…]

In my experience the problem only usually occurs[…]

3rz-fe turbo. help with iac valve

READ THE WIKI. https://wiki.speeduino.com/en/conf[…]

Still can't find what you're looking for?