View unanswered posts | View active topics It is currently Tue Oct 22, 2019 7:18 am



Reply to topic  [ 85 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6 ... 9  Next
New wideband controller ALM compared to Innovate LM-2 
Author Message
TO220 - Visibile

Joined: Tue Aug 09, 2011 10:25 am
Posts: 10
Thanks Fred. That makes a lot more sense :)

I know the guys want to make a splash for their new product. But, they need to realize that almost ALL of the existing wide-band controllers out there already use the Bosch CJ1xx chips :) If you really want to compete with the speed of the Innovate controllers, you need to work directly with the sensors, not the interface chip-sets. It's not THAT hard :)


Thu Aug 11, 2011 3:44 am
Profile
TO220 - Visibile

Joined: Tue Aug 09, 2011 10:25 am
Posts: 10
jharvey,

An interesting use of near-instantaneous AFR response time is...

If one knows the exact AFR, on at least, a per revolution basis, couldn't one just use that signal to directly control the injector timings ? i.e. set a target AFR table for specific rpm/load conditions and use a straight software PID to control the injectors ? No fuel/trim maps, etc. It would automatically deal with fuel rail pressure issues, intake/exhaust mods, cam changes, etc.

It would be universal to almost any engine.

Fun idea :)


Thu Aug 11, 2011 6:30 am
Profile
Moderator
User avatar

Joined: Tue Jan 15, 2008 2:31 pm
Posts: 15340
Location: Home sweet home!
Fun idea until you get a misfire and it reads lean when it's rich! :-) Anything could be dealt with, but you'd still need a tune at higher RPM and you'd still need a starting point to get the engine running in the first place. It could be useful as a verification technique, though, I guess.

What I didn't like was the coming in as "diyhansen" and then switching to the company name. I openly allow vendors of relevant products to promote them here in this section. There was no need to do it in an underhanded way like this.

_________________
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!


Thu Aug 11, 2011 11:09 am
Profile WWW
1N4001 - Signed up
User avatar

Joined: Tue Jun 10, 2008 5:17 pm
Posts: 1607
dr.mike wrote:
An interesting use of near-instantaneous AFR response time is...

Fun idea :)
Fun yes, but is there a sensor that has the possibility of doing it? I don't know of one, but I'm also not an expert on the matter either. I only vaguely know the consumer stuff.


Thu Aug 11, 2011 11:17 am
Profile
TO220 - Visibile

Joined: Tue Aug 09, 2011 10:25 am
Posts: 10
The current Bosch LSU4.2 is fast enough. The LSU4.9 is slightly slower, due to its sealed design, IIRC. It's the interface circuits that slow them down. I am pretty sure that I could drive one to about 1kHz ( 1000 samples/sec ) with the right design.


Fri Aug 12, 2011 8:49 am
Profile
Wideband Wizard

Joined: Tue Jun 10, 2008 2:53 am
Posts: 241
Location: Toronto Canada
The 4.9 and the 4.2 are almost identical, the 4.9 is meant to run in diesel and the 4.2 is meant for gasoline.

The 4.9 has a smaller heater element, less powerful and less thermal mass, from this I assume that diesel tend to burn hotter and there is greater variation in temperature WRT to gasoline, I think this is true.

4.9 has smaller holes on the tip for gas to enter the diffusion chamber, from this I assume that diesels either produce more exhaust gas or the exhaust gases have higher velocity, not sure about the first but the 2nd is true generally.

Personally, my opinion is that 4.2 is slightly superior because of the larger openings for gas to enter the sensor.

My guess is that european/asian brand lambda controllers will use the 4.9 and north american brands will use the 4.2, the choice is only due to the local availability of the 4.2 vs 4.9.

The problem with O2 lambda sensors, atleast the bosch ones is that they are very slow, to get the thing to start responding to a large gradient change in the exhaust gas takes 50ms to 75ms. This is for a brand new sensor. As the sensor ages, the accuracy drops by a very little bit but the response time gets much slower.

Because of this, you can rely on the accuracy of a lambda sensor regardless of age (usually), but you can not depend on consistent response times as the sensor ages.

Innovate stuff is fast, but their design depends on the response of the lambda sensor to some change, because the response time of a lambda sensor takes a huge hit as the sensor ages, sensors have a very short lifespan until their response time is too slow to work with the innovate controllers. Although the sensors are too slow to work with the innovate controllers, their accuracy is still perfectly good and you can use the sensor on other non innovate controllers without problems. This is the reason why people keep getting the famous "E8", error 8, even on sensors with low mileage on them.

_________________
www.14Point7.com


Tue Aug 16, 2011 11:45 pm
Profile WWW
Moderator
User avatar

Joined: Tue Jan 15, 2008 2:31 pm
Posts: 15340
Location: Home sweet home!
Awesome post, Alan! :-)

_________________
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!


Wed Aug 17, 2011 1:11 am
Profile WWW
Wideband Wizard

Joined: Tue Jun 10, 2008 2:53 am
Posts: 241
Location: Toronto Canada
To clarify what I mean by "to get the thing to start responding to a large gradient change in the exhaust gas takes 50ms to 75ms"

I have a chamber filled with 0.8 lambda gas, the chamber is not sealed but the chamber has a long tube to the outside, this creates to little bit of back pressure to make sure the 0.8 lambda gas does not escape and stops free air from coming into the chamber. I use a 24v solenoid switch to switch a tank of compressed free air @ 40 PSI, into the chamber. I use the power signal of the solenoid as a trigger to my scope. Once the scope is triggered, I plot the "feedback" signal of the lambda sensor (it is the black wire on the sensor), on most lambda controller designs the "feeback" is closed loop controlled to maintain 0.45v above the virtual ground (it is the yellow wire on the sensor).

If the gradient of the exhaust is large enough and applied fast enough, you should be able to see a small blip on the feedback line, the blip represents the control error. The magnitude and duration of the blip is roughly equal to the quality of the closed loop control, that is if the blip is large and long, the controller is not very good.

Anyways, I can see a small blip @ around 75ms after the solenoid is switched. My opinion is the blip represents the response of the sensor to the change in gas. My solenoid probably takes about 10ms to fully energize, and perhaps it takes 10ms for the compressed air @ 40 psi to make its way to the sensor, so my guess is that it take 50ms for the sensor to actually respond.

When Dr.mike is throwing around numbers like 5ms and 1 khz, I can not relate to that figure at all.

_________________
www.14Point7.com


Wed Aug 17, 2011 1:41 am
Profile WWW
Moderator
User avatar

Joined: Tue Jan 15, 2008 2:31 pm
Posts: 15340
Location: Home sweet home!
Have you done this experiment with multiple controllers? If so I think a thread is in order! I trust your opinion on all things wideband. I trust it for a reason, you've always been the first to criticise your own stuff and very objective about all things.

Have you considered switching the solenoid with higher voltage P&H style? What about further instrumentation to cut some of the guessing out of the solenoid and air movement? Perhaps colour in the air and a camera watching it at a high frame rate? or an LED and receiver and measure the intensity received? Or something like that?

_________________
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!


Wed Aug 17, 2011 2:33 am
Profile WWW
Wideband Wizard

Joined: Tue Jun 10, 2008 2:53 am
Posts: 241
Location: Toronto Canada
I have done this with my own stuff and other controllers as well. I did not bother to refine my testing jig, as I am quite convinced that regardless of controller design the bosch lambda sensor is not capable of resolving individual cylinder AFR at the collector, that is the holy grail. With that out of reach, resolving response time further is of no tangible benefit.

My stuff is not as fast as innovate, but faster than most other controllers. I am fine with that. If individual cylinder metering was even remotely possible; I would shank a mofo get the crown for fastest response time.

Edit: When I saw the "150ms faster than the LM2" I thought; holy sh*t someone has invented time travel, because the lm2 is ~100ms response time, guessing that based on my testing with the the LC1, so the ALM controller is reading AFRs from the future.

_________________
www.14Point7.com


Wed Aug 17, 2011 3:44 am
Profile WWW
Display posts from previous:  Sort by  
Reply to topic   [ 85 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6 ... 9  Next

Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group
Designed by ST Software for PTF. ColorizeIt.