AEM WIDEBAND 30-4900 WITH M4 (M2R) VOLTAGE DIVIDER

[AdamW]

The AEM analog ground (brown wire) should be connected to Motec LA-(pin 32), This wire/pin should not connect to chassis ground at all.
Note your M4 will need to have the advanced options enabled to allow you to set up a custom Lambda calibration.

[chick0]

Thanks adam and david....

Made the voltage divider. Converted 4v to 1v i did....

Using multimeter im getting 1v after the divider and also 1v right on pin 31 = +la input.

When i went to the AD conversion tables i wrote down the exact conversions... issue is this, the motec is reading a lower lambda although the voltage tested at the pin is the correcr voltage...

Example:

14.4 afr = .71v, i should be seeing 1 lambda and even wrote the a/d conversion correct to .71 volts. However the motec shows lambda as .83.

Anyone advise why its showing a lower lambda value even though the ecu is recieving the correct voltage input

[AdamW]

Can you post your map here, I'll take a look

[chick0]

Heres my map adam,

so i converted 0-4 to 0-1v using a voltage divider: 3.9k and 1.3k

my ad table reads between .7 lambda and 1.2 lambda

calculation: .7 lambda = sensor voltage (.3v) x 1024 = 308 (307.2) rounded up my values

1.2 lambda = 1.1v x 1024 = 1127


example of error: at 14.4 afr lambda should be just under 1.00 motec reads .83, voltage at pin 32 reads .71 which going by my ad table should also read just under 1

done it at another point 13 afr .57 v reading at pin 32, however motec lambda on screen shows .77, ad table is calculated to show just under .9

the motec is receiving the correct signal but not converting my values correctly

[AdamW]

i havent had a chance to look at your map yet but a quick glance at your math I think where you are going wrong is you are assuming the motec 1024a/d bits are working over a 0-1V input. Im pretty sure the last time I messed with one I remember it was 0-1.2V. So, a 0V input would read 0a/d bits, a 1.2V input should read 1024a/d bits.

Your AEM device connected through a 4:1 divider should give roughly the following:
0.58Lambda = 0.125V = 106.7A/D bits
1.23Lambda = 1.125V = 960A/D bits.

Of course your resistors probabaly have a 5% tolerance and the motec A/D table only increaments in 0.05Lambda so you will have to tweak things a little to get them to match.

[chick0]

Ill check it now adam... resistors are %1, 1/4 w metalfilm ones. And both measured to the third dec. .000 spot on, wanted to make sure i was cal. Correctly... but ill check what you havejust advised and get back to you

[chick0]

Thanks adam all is working perfect now!

Just recieved a lambda error now and then, unfortunately the m4 does not state what the error is in relation to, only that its an error. Think it happened when i was playing with it, trying to get it to hit a target afr and it didnt hit the target in time or couldnt

[Rude Engineering]

Hmmm, I would not expect to need a voltage divider as the input can measure 5V. I would use the following cal without the divider:

A/D Lambda
3 0.50
67 0.55
130 0.60
193 0.65
257 0.70
320 0.75
384 0.80
447 0.85
510 0.90
574 0.95
637 1.00
700 1.05
764 1.10
827 1.15
891 1.20
954 1.25
1017 1.30

According to the help the input value is sensor voltage / 5 x 1024.

[SprinterTRD]

Some of the help is incorrect, probably a copy and paste error from the other inputs.

The lambda input can only handle around 1.75V

[AdamW]

I agree with sprinter here, the help file is wrong. I'm pretty sure the last time I tested mine with a simulator a 1.2V input gave me the 1024 A/D value. I remember I could put a little more voltage in than that and the A/D value would continue to climb but I don't think it went as far as 1.75V, I think it topped out somewhere about 1.3V.