Ion Sense

Hardware inside and outside of the ECU
User avatar
kb1gtt
contributor
contributor
Posts: 3407
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: Ion Sense

Post by kb1gtt » Sun Dec 23, 2018 11:21 pm

Per the below, a 1mS to 2mS spark duration is normal. It looks like this spark is about 2mS at idle, which might be a bit high. It also appears we are seeing detonation pressures, as we occasionally see the combustion indicator trigger 2X in one combustion. This engine may have some problems, which are great to look at. This engine may be showing us thing the original engineers didn't get to see.
http://crypton.co.za/Tto%20know/Ignitio ... 0time.html
Welcome to the friendlier side of internet crazy :)

User avatar
kb1gtt
contributor
contributor
Posts: 3407
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: Ion Sense

Post by kb1gtt » Mon Dec 24, 2018 1:16 am

Some things to ponder. The low part of the yellow signal was 4.16V, and the higher part of the flat section of the yellow trace was 4.75V. There are occasional very short low spikes, but generally the signal a delta of about 0.6V with a 4.5V offset. If we look to get this signal into rusEFI, how should we capture it? I would think we would need a sensing resistor, then sense the current that flows through the resistor. Perhaps this MAX471 chip.
https://www.ebay.com/itm/1Pcs-MAX471-Vo ... ctupt=true

Oh, also the flame test shows the goo in the pencil coil burns instead of melts.
Welcome to the friendlier side of internet crazy :)

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Tue Dec 25, 2018 9:34 pm

Got Oleg - a racing driver instructor who lives nearby - to drive the Saab a bit. I think Oleg was just a bit over-qualified for the job?

Both videos follow the same script - full throttle run & 50% throttle run. Second one is maybe a bit better quality



https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

User avatar
kb1gtt
contributor
contributor
Posts: 3407
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: Ion Sense

Post by kb1gtt » Tue Dec 25, 2018 11:59 pm

Should you do the diesel blend to induce knock?
Welcome to the friendlier side of internet crazy :)

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Wed Dec 26, 2018 1:08 am

kb1gtt wrote:
Tue Dec 25, 2018 11:59 pm
Should you do the diesel blend to induce knock?
The videos above are Saab - I think there is no point adding diesel into Saab's fuel tank since stock ECU would adjust too quickly. I plan to put diesel into red Miata running under rusEfi control hopefully this weekend.

Also looks like CDM is affecting the ion sense signal. Jared, can you get the schematics of CDM in terms of how is it affecting ion signal?

https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

User avatar
kb1gtt
contributor
contributor
Posts: 3407
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: Ion Sense

Post by kb1gtt » Wed Dec 26, 2018 11:02 am

I can try to schematic-ize the CDM for this signal. That was the yellow wire correct?

Also FYI, it's long hours for me at work until the end of the year. I'll try to get it in before the weekend, but I'm most likely to attempt it during the weekend.
Welcome to the friendlier side of internet crazy :)

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Wed Dec 26, 2018 1:50 pm

Its the corner yellow wire. Says I1 on case
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Sun Dec 30, 2018 4:07 am



https://sourceforge.net/projects/videoc ... ranscoder/ worked not great but gave me command line commands to rotates videos

Pieces joined together using https://www.openshot.org/ - this looks like nice softwar but video rotate did not work :(
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

User avatar
kb1gtt
contributor
contributor
Posts: 3407
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: Ion Sense

Post by kb1gtt » Sun Dec 30, 2018 11:14 am

Do you have a snippet with the OEM pull up what ever that is? Perhaps just a series of pictures, OEM, no pullup, 1k ohm, etc. Pictures can be easier to look at then pausing the video. The video is great as it shows lots of great stuff relative to time, but when comparing the differences, it's nice to jump from one to the other very rapidly with the eyes.
Welcome to the friendlier side of internet crazy :)

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Sun Dec 30, 2018 1:54 pm

kb1gtt wrote:
Sun Dec 30, 2018 11:14 am
Do you have a snippet with the OEM pull up what ever that is? Perhaps just a series of pictures, OEM, no pullup, 1k ohm, etc. Pictures can be easier to look at then pausing the video. The video is great as it shows lots of great stuff relative to time, but when comparing the differences, it's nice to jump from one to the other very rapidly with the eyes.
Audio says and video shows OEM module is connected in the beginning. Without audio narration it would be impossible for me to know which picture shows which conditions, hopefully there are options to work with the video. Maybe leave comments right on youtube with timestamps regions for each pull-up.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

theflyingdutchp
Posts: 91
Joined: Wed May 18, 2016 11:03 pm

Re: Ion Sense

Post by theflyingdutchp » Tue Jan 01, 2019 1:24 am

Russian, can you connect a scope channel for the knock detection output signal of the CDM module? Also, I may have missed it, but what is the blue trace? Correct me if I'm wrong but its still wired as: red ignition demand #1 (CDM -> coil), green combustion detection #1&#3 (CDM -> ECU), and yellow coil-feedback #1 (coil -> CDM).

Also, maybe for inducing knock you can purposefully tune the engine badly by reflashing the Trionic ECU? Obviously not for too long, but it might be nice to see what a good solid knock response looks like.

It may also be useful to log data from the SAAB ECU (such as ignition advance & knock count) once you get logging from the oscilloscope going. Perhaps you could tap another, user-controlled, system on each datastream to establish a common time reference IE: turning on the headlights, cranking the starter, etc.

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Tue Jan 01, 2019 1:33 am

And we have knock and all kinds of success!



And in text form https://github.com/rusefi/rusefi_docume ... ing_rusEfi
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

theflyingdutchp
Posts: 91
Joined: Wed May 18, 2016 11:03 pm

Re: Ion Sense

Post by theflyingdutchp » Wed Jan 02, 2019 6:18 am

russian wrote:
Tue Jan 01, 2019 1:33 am
.
Awesome! Nice job

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Sun Jan 06, 2019 4:19 am

Today we have a joint cross-Atlantic effort to gather some ion sense wareforms while knocking engine - see https://github.com/rusefi/rusefi_docume ... _rigol.zip for data, there is a readme file inside with details.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

XXLO
Posts: 5
Joined: Sun Feb 10, 2019 7:28 pm
Soldering skill: yes
Coding skill?: yes

Re: Ion Sense

Post by XXLO » Sun Feb 10, 2019 8:02 pm

Hallo.
I found this side by seachring "ion sensing" films on youtube.
Verry interesting. Thank you for your videos.

I have uploaded the CDM/COP wireing from Opel engine Z20NET. (same Trionic 8 that SAAB use) There is old analog and new digtial CDM wiring. The only difference i see, is the color of the ion sensing feedback cable (coil->CDM) of coil 2+3.

You listed up which cars are use ion sensing. There is also a Mecerdes Benz V12 engine. M137 - producend since 1998.

Have you seen this funny guy at youtube?
(he has a couple of ion sensing videos)
He build a closed loop basend on ion sensing. He can autotune the igniton timing to hit the optimal peak preassure point. For his rotary engine 52° after top dead center. I have read in the saab papers, that the optimal peak preassure point for pistion engines are at 10-15°.

Is in the combustion signal (CDM->ECU) the peak preassure point included and visible?
Has the saab a autotune to reach optimal peak preasure point?
(like the youtube guy)
Attachments
CDM COP wiring.jpg
CDM COP wiring.jpg (120.46 KiB) Viewed 604 times

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Sun Feb 10, 2019 8:42 pm

I have some info at https://rusefi.com/wiki/index.php?title ... b_9-3_2005

Good to know about v12! See my v12 wet dreams at viewtopic.php?f=3&t=1504

I've seen one of his videos but that's a good point that I should poke him. Just left a couple of comments on his videos :)

At the moment the ion project is a bit on hold, need to spend some time on other projects but definitely planning to get back into this in a few months!
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

kenzone
Posts: 1
Joined: Mon Mar 12, 2018 2:54 pm
Soldering skill: yes
Coding skill?: yes

Re: Ion Sense

Post by kenzone » Tue Feb 12, 2019 7:08 pm

@russian May I ask whether you managed to test the fitment of the Demio COPs on the miata valve cover?
Also what are the chances the Demio COPs' outputs play nice with the Saab CDM?

Another youtube user who appears to have implemented Ion Sensing successfully is Nathan Coleman.
Their more recent video showcases a filtered clean pressure signal.

Notes from most recent video:
"Modified Ion testing circuit, passing the full wave signal limited to +/-5V. Full wave allows the low pass filter to kill the ignition noise to zero with no DC offset, as you can see when I adjust the filter it disappears entirely. The ion current flow is positive, spark current is negative. Bias voltage is now 180V, up from 90V, seeing a maximum of 300uA ion current, fluctuating between 50-100 at idle. In the video I'm adjusting the sensing resistor to give the strongest signal without hitting the 5V limit. Spark current peaks at approx 70mA. I thought it would be more, good news as now I don't need as many zener diodes to dissipate the energy. Will probably use 3x50V 5W instead of 12x15V 5W. Almost ready to finish the processing circuit to collect the peak current per cycle to feed into the ecu for datalogging."



User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Tue Feb 12, 2019 11:28 pm

kenzone wrote:
Tue Feb 12, 2019 7:08 pm
@russian May I ask whether you managed to test the fitment of the Demio COPs on the miata valve cover?
Also what are the chances the Demio COPs' outputs play nice with the Saab CDM?
Lack of time and priority for this is the only reason why I have not tried :( I would say very little chance of me trying this in February or March, some chance in April+.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Tue Feb 12, 2019 11:34 pm

kenzone wrote:
Tue Feb 12, 2019 7:08 pm

Another youtube user who appears to have implemented Ion Sensing successfully is Nathan Coleman.
Their more recent video showcases a filtered clean pressure signal.
Thank you for the link! Just poked asking for updates :)
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

XXLO
Posts: 5
Joined: Sun Feb 10, 2019 7:28 pm
Soldering skill: yes
Coding skill?: yes

Re: Ion Sense

Post by XXLO » Wed Feb 20, 2019 8:52 am

Here is an other paper of ion sense information. (i dont see it in the first post of this thread)
http://www.not2fast.com/ignition/SAEPeakfind.pdf

There is written:
"Empirical studies [7], under steady stateconditions in a dynamometer, indicate that this isapproximately true but that the model has difficultiesdescribing low load phenomena, even if data is collectedunder ideal external conditions and then averaged. Thiscomplication is due to the fact that the post flame peakessentially vanishes if the load is less than 20% MBT2(see Fig. 3). It is worth mentioning, in this context, thatnormal driving on the highway is close to the low loadsituation."
"An ion current signal sampled during a single cycle on boardthe SAAB car while driving on the highway (solid line). The enginespeed was 1500 RPM and the load was approximately 25% of MBT.There are several wide peaks in the signal and the post flame phase isnot visible. The pressure trace is shown for comparison (dashed line)."


i think this the reason why you cant see peak preassure point in the oszilloskop videos.

To low load in idle and rev up while car is standing still.

User avatar
russian
Site Admin
Posts: 9522
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Soldering skill: yes
Coding skill?: yes
Contact:

Re: Ion Sense

Post by russian » Tue Apr 23, 2019 3:20 am

Bosch (7330) F5KE0 Industrial Combustion Sensor $8.75 & FREE Shipping on Amazon for Measures Ionic current to determine combustion efficiency "CATERPILLAR CAT 3600 SERIES" - they use ion sensing in large diesels natural gas engines?

https://www.cat.com/en_US/products/new/ ... 57059.html Lube Oil System - Refill 187.0 gal (US)
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

XXLO
Posts: 5
Joined: Sun Feb 10, 2019 7:28 pm
Soldering skill: yes
Coding skill?: yes

Re: Ion Sense

Post by XXLO » Thu Aug 15, 2019 7:19 am

XXLO wrote:
Sun Feb 10, 2019 8:02 pm
There is also a Mecerdes Benz V12 engine. M137 - producend since 1998.
-> this car has a DC voltage ignition.
German info: http://www.fahrzeug-elektrik.de/Ecce.htm
German bock: https://www.amazon.co.uk/Wechselspannun ... 3183389126

Post Reply