Page 9 of 12

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Thu Jun 27, 2019 1:36 am
by mck1117
Also, the resistor in the connector is something like 30-300 ohms, laser calibrated in the factory. It requires both the resistor in the connector, and the 61.9 on the board to operate.

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Thu Jun 27, 2019 9:01 am
by Реконструктор
Thank you very much, guys! :)

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Tue Jul 02, 2019 2:36 pm
by JRD McLAREN
..What's wrong with this board...
.. it lost SPI comunication, when +12V for heater is connected ...
:roll: :roll:

Edited: .. on frankenstein ...

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Tue Jul 02, 2019 9:26 pm
by kb1gtt
????

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Tue Jul 02, 2019 9:30 pm
by JRD McLAREN
..I will post more details tomorrow..

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Fri Jul 05, 2019 6:36 pm
by JRD McLAREN
Ok ... let's move ...

This is my current testing setup ...
DSCN4860.JPG
DSCN4860.JPG (1.91 MiB) Viewed 7875 times
Orange, Brown, Red and yellow wires are for SPI
Single Green is for heater ..
Blue and (another) Green is power supply for CJ125 Board
Grey and Black are UR and UA outputs.

Single White connected to power board is for heater +12V

When no LSU is connected CJ125 get response to console ( cj125_calibrate )
When no +12V power supply for LSU, we got response from CJ125

Code: Select all

2019-07-05 20_28: EngineState: confirmation_cj125_calibrate:15
2019-07-05 20_28: EngineState: cj125: Check ident=0x60 diag=0x3F init1=0x89 init2=0x10
2019-07-05 20_28: EngineState: cj125: Starting calibration...
2019-07-05 20_28: EngineState: confirmation_cj125_calibrate:15
2019-07-05 20_28: EngineState: cj125: Cannot start calibration. Please restart the board and make sure that your sensor is not heating
...... or ....
2019-07-05 20_34: EngineState: confirmation_cj125_calibrate:15
2019-07-05 20_34: EngineState: cj125: Check ident=0x60 diag=0x3F init1=0x89 init2=0x10
2019-07-05 20_34: EngineState: cj125: Starting calibration...
2019-07-05 20_34: EngineState: cj125: Done! Saving calibration data (7338 1219).

When all is connected, have no successfully response.

Code: Select all

2019-07-05 20_30: EngineState: confirmation_cj125_calibrate:15
2019-07-05 20_30: EngineState: cj125: Check ident=0xF8 diag=0xFF init1=0xFF init2=0xFF
2019-07-05 20_30: EngineState: cj125: Error! Wrong ident! Cannot communicate with CJ125!
2019-07-05 20_30: EngineState: cj125: Error! Cannot set init registers! Cannot communicate with CJ125!
2019-07-05 20_30: EngineState: cj125: Starting calibration...
2019-07-05 20_30: EngineState: cj125: Calibration error (init1=0xFF)! Failed!
It does not matter, if I have STM Discovery + CJ125 only ..
or if I have connected Frankenstein board also ...

.. Just for record ...

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Fri Jul 05, 2019 6:37 pm
by JRD McLAREN
...I will try to connect SPI as andrejka have it on the prometheus board ... :roll:

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Fri Jul 05, 2019 6:44 pm
by Pet007
Hi Guys, (zdravim JRD McLarena do Blavy ;)
please, how are you calculating Lambda from IP current ? I am absolutely confused from CJ125 PDF and LSU4.x technical sheets.
For ex. when you take table from page 7 of 4.9 technical sheet (lambda-value vs IP curent) and comapre it with this equation (code taken from Arduino lambda shield project), there are two different results of lambda :( So where is the thrue ? Thanks for any hints.

Code: Select all

LAMBDA_IP = 1000 * (CJ125_UA -1.5) / (61.9 * 17); /* V=17 * or V=8 -result in [mA]
k = 0.2095/2.54;     // 20.95 % O2 ~ 2.54mA of IP
LAMBDA_O2 = Lambda_IP * k;    
LAMBDA_VALUE = (LAMBDA_O2 / 3 + 1) / (1 - 4.77 * LAMBDA_O2); //Calculate Lambda value acc. to BOSCH LSU 4.9 Technical Product Information Y 258 E00 015e.
PS. (and what about rich area V=8 compared with V=17 - lean area - whean I want to measure from aprox. 10.0 AFR with good accuracy)

PS:JRDMcLAREN: Ted sem na tebe zrovna vzpominal, delal sem ecu na atmo C20XE v Kadetu pro kluky z Teplic, nabrzdilo se 200PS....chteli ale 250PS na vackach 276/276 a Dbilas sani (tzn. este bez 4-klapky)
EDIT: add missing Lambda_O2 equation

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Fri Jul 05, 2019 9:24 pm
by JRD McLAREN
Ahoj ...

Look at this: https://github.com/rusefi/rusefi/blob/master/hardware/CJ125_board/bosch_lsu49.pdf
Screenshot_20190705_231940.png
Screenshot_20190705_231940.png (71.6 KiB) Viewed 7863 times
Ja tomu celemu chapem tak, ze chipaku bosch povies, ktory "amplifer factor" chces pouzit...
a on potom na zaklade toho "upravi" vystupny signal ... teda nastavujes si tym "rozlisenie" ...
PS:JRDMcLAREN: Ted sem na tebe zrovna vzpominal, delal sem ecu na atmo C20XE v Kadetu pro kluky z Teplic, nabrzdilo se 200PS....chteli ale 250PS na vackach 276/276 a Dbilas sani (tzn. este bez 4-klapky)
..bez upravenej hlavy mozu na dalsie kone rovno zabudnut ... :mrgreen:

EDITED:
sorry for language ...

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Sat Jul 06, 2019 6:22 am
by Pet007
Thanks, ohh I've missed the last page of 4.9 technical sheet. So, there is an equation:

Code: Select all

 Vcj125 = Ua [V] = 1.5 +(61.9/1000*v)*Ip[mA] - where v= 8 (for reach area or 17 for lambda > 1)
 Ip[mA] =  (Ua[V] - 1.5) * (1000/(61.9*v))
and then use linear aprox. from IP vs Lambda table (attached) ?
lsu4.9_bosch_table_lambda.png
lsu4.9_bosch_table_lambda.png (8.29 KiB) Viewed 7841 times
Anyway, proto version of my CJ125WBO controler in 14p7 style (and case - due damaged original PCB) with PICMCU and MAX7219 LED driver:

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Tue Jul 09, 2019 10:40 pm
by 960
Are there any gauge in TS to show AFR from CJ125, or how do you do that?

I have a CJ125 up and running, but no way to show it at a gauge.

Btw, I spotted this today while making a tuningfile for a Ford Focus.
It has a lot of CJ125 settings.

ols_32on32_x0UjFkZUgJ.png
ols_32on32_x0UjFkZUgJ.png (437.08 KiB) Viewed 7795 times

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Wed Jul 10, 2019 5:53 am
by mck1117
If you have the cj125 enabled, the normal AFR gauge will show the measured AFR.

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Wed Jul 10, 2019 1:44 pm
by 960
Ok, didnt know. Haven't had any reaction at that gauge yet.

I'll have another look.

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Thu Jul 11, 2019 6:03 am
by mck1117
In that case, it probably isn't working properly. Can you set the debug mode to CJ125, and share a datalog?

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Fri Jul 19, 2019 6:11 pm
by 960
I can do that next time I test it in car.

This is the gauges showing something at bench(No sensor connected).
javaw_DbRXHfQ8RF.png
javaw_DbRXHfQ8RF.png (550.11 KiB) Viewed 7685 times

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Sun Jul 21, 2019 4:22 am
by 960
This manual has some interesting CJ125 info:

https://www.bosch-motorsport.com/media/downloads/ecu_ms_4_sport_turbo.pdf

From page 97-

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Mon Aug 12, 2019 11:54 pm
by Ahmad
Hi friends,
Has anyone worked with L9780 ic for AFR measurement?
Any suggestion?

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Tue Aug 13, 2019 9:40 am
by kb1gtt
I have not, and I don't think anyone here has. It seems to be an interesting chip, very similar to the CJ125. What attracts you tho this instead of the CJ125? My 30 second review of the datasheet seems to indicate very similar features. Are there differences that of of interest?

Below is a picture of the block diagram for the L9780 as well as a link to the datasheet.
L9780_Block_Diagram.png
L9780_Block_Diagram.png (70.69 KiB) Viewed 7557 times
https://datasheet.octopart.com/L9780-STMicroelectronics-datasheet-43533507.pdf

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Tue Aug 13, 2019 2:17 pm
by Ahmad
I make a Lambda display with CJ125 three years ago.
In my new project (ECU for 8 Cylinder lean-burn Generator) I dont like retry intuition secrets of Bosch! :geek:
Also, L9780 has internal ADC and is newer than CJ125.

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Tue Aug 13, 2019 8:33 pm
by mck1117
Internal ADC is interesting - it's also not bosch, so documentation is far easier to find, and it's far easier to actually buy one (Mouser has them in stock). And they're cheaper: $6.55 in single quantities.

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Sat Aug 17, 2019 10:48 am
by JRD McLAREN
960 wrote:
Fri Jul 19, 2019 6:11 pm
I can do that next time I test it in car.

This is the gauges showing something at bench(No sensor connected).
You have any progress .. ??

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Sun Aug 25, 2019 11:51 pm
by 960
JRD McLAREN wrote:
Sat Aug 17, 2019 10:48 am
960 wrote:
Fri Jul 19, 2019 6:11 pm
I can do that next time I test it in car.

This is the gauges showing something at bench(No sensor connected).
You have any progress .. ??
No,

I gave up on the whole rusefi project.. :-)

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Mon Aug 26, 2019 6:16 pm
by JRD McLAREN
Wow ... :shock:

So,I have no working CJ125 board with frankenstein .. .. (also)

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Tue Aug 27, 2019 11:53 am
by kb1gtt
I do not understand your comment. I think you have the cj125 board and frankienstien but you do not know how to use them together. Is that a correct understanding?

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Tue Aug 27, 2019 3:25 pm
by JRD McLAREN
..so ...

Have working SPI communication ...
Have working Heater PWM ... (it heat the sensor) (and FET also ... no fun :evil: - wrong design maybe??)

Have correct output UA and UR voltages measured with DMM ... but not with ST / rusefi

My discovery board have added backup battery.

I don't know "how to connect analog outputs to discovery board correctly" ...
.. to measure "right voltages" from CJ125 chip ...

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Wed Aug 28, 2019 2:59 am
by kb1gtt
What are the correct UA and UR voltages? I think that's 14V-ish. It sounds like you need to make a resistor divider? Similar to the 12V resistor divider on the analog circuits of Frankenso.

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Wed Aug 28, 2019 4:05 am
by Abricos
Maybe it's helping ...

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Fri Aug 30, 2019 11:03 pm
by Ahmad
I found LQFP version of CJ125 in a local store with low price :D
It is marked 30615 (or 30522).
This package has separate VCC/GND for Signal path and better noise rejection.
Any comment on schematic will be appreciated.

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Mon Sep 02, 2019 5:03 pm
by Fantomiaso
kb1gtt wrote:
Wed Aug 28, 2019 2:59 am
What are the correct UA and UR voltages? I think that's 14V-ish. It sounds like you need to make a resistor divider? Similar to the 12V resistor divider on the analog circuits of Frankenso.
Having troubles with my project so just checked values today on my board.
For lambda = 1 Ua is 1.5V at all UB voltages, maximum voltage is near 4V
For new LSU 4,9 sensor at operating point Ri is near 300 Ohm, so according to CJ125 datasheet UR should be near 0.75V. For sensor at room temperature near 2,5V.

I'm trying to make CJ125 work with 32f407 chip and no success - zero WBO ADC readouts, no changes in PWM heater pin (100Hz with near 1/5 duty cycle). First of all seems pulldown resistor needed at SO pin. Here is how it looks with internal STM32 pulldown:
photo_2019-09-02_20-08-43.jpg
photo_2019-09-02_20-08-43.jpg (126.74 KiB) Viewed 7221 times
Third (blue) channel is SO. As you can see, pulse tail has transition process. With HiZ STM input it takes about 5-7 times longer, influencing first bits of SO transmition if they are not logic "1". With 2k2 external pulldown front/tail of SO looks good.
So... After dealing with SO tail I checked all currents and voltages on sensor and CJ125. IP, VM, 450mV displacement, Ua and Ur looks valid, . Then I checked if UA ADC pin work: I use external voltage source, pin configured as BARO input and it works good.Then I configured it back as CJ125 UA, lifted CJ125 pin to the air and feed ADC input with regulated voltage source. No reaction both in self-stimulation, external stimulation and stopped mode. UA and UR both changes if I power lambda heater with external power source according to sensor temperature. Maybe I dont know something about tuner studio settings?
Here is my TS project: https://rusefi.com/forum/download/file.php?id=5086

Any suggestions?

Re: rusEfi own wide band controller board add-on using CJ125

Posted: Mon Sep 02, 2019 8:33 pm
by JRD McLAREN
Another "reverse engineering" of my second LSU controller.

No DIAHG and DIAHD are used, no divider for UA and UR.
DSCN4878.JPG
DSCN4878.JPG (90.04 KiB) Viewed 7214 times