lm1815 on pcb v3.0

nyabinghi · Post by **nyabinghi** » Tue Apr 01, 2008 5:24 am

I'm installing a MS2 PCB v3.0 on a c20xe engine : 60-2 wheel, with oem CAS sensor.

Because of lot of troubles with the original v3.0 on-board vr conditioner circuit I tried to made the extra mod with the lm1815 chip, in order to replace it.
I did this circuit on the porto area, following a schematic that a MS1 user have made on his v3.0 board for his C20XE (same engine as mine, with same sensor). The schematic is on picture lm1815_c20xe.jpg.
With this circuit there is a lot of tach miss because of the signal input :
On lm1815n_1.png and lm1815n_2.png pictures, I scoped the input signal at VR sensor. You can see that after the 2 missing teeth, the signal doesn't do a positive to negative zero cross as expected.
The signal fall to zero and don't cross the zero value so the lm1815 chip doesn't produce a rise on the square signal as expected.

This is strange because with the original pcb v3.0 vr circuit, the input signal do a positive to negative zero cross at this tooth (see attached picture PCB3.0_1.jpg at the next post).

Can anybody help me ? Do you think there is something wrong with the lm1815 circuit I made ?

Thanks

nyabinghi · Post by **nyabinghi** » Tue Apr 01, 2008 5:30 am

Picture of the input sensor signal with the original vr circuit :

nyabinghi · Post by **nyabinghi** » Tue Apr 01, 2008 10:37 am

Add some information on my setup

pet007 · Post by **pet007** » Tue Apr 01, 2008 11:44 am

nyabinghi wrote:Add some information on my setup

As you have wrote, the non-zero crossing after missing teeth is very strange. Maybe try to remove 330R resistor between VR in and GND. But I am not sure, if could fix it :/
Anyway, from your input VR signal is nice to see the much bigger amplitude when sensor is passing two missing teeth. And your LM1815 setup works in adaptive mode (pin 5 open). Then this bigger amplitude will cause arming treshold is on the higher level; then the following 1. normal tooth with lower amplitude is not able to reach the previous higher arming treshold => no trigger. You can try to insert aprox. 100k resistor before 0.33uF capacitor (from pin 7). This mod helps to slow down the arming speed of 0.33uF peak capacitor.
I had the same issue on C20XE engine, I couldn't rev more then aprox 2200 RPM.

nyabinghi · Post by **nyabinghi** » Tue Apr 01, 2008 2:11 pm

Thank you pet007,

I will try to remove the 330R resistor.

On your C20XE you use the lm1815 chip ? What's your setup ? MS1 ? pcb v3.0 ?
Where I was trying the on board VR conditioner I was able to rev up to 5000rpm but I had another problem : there were little tach sync loss (no squirt and no ignition) during running, at any rev.

What do you suggest for the lm1815 mode ? Why not the mode 3 with pin 5 grounded ? Have you try this ?

pet007 · Post by **pet007** » Wed Apr 02, 2008 1:01 pm

nyabinghi wrote:Thank you pet007,

I will try to remove the 330R resistor.

On your C20XE you use the lm1815 chip ? What's your setup ? MS1 ? pcb v3.0 ?
Where I was trying the on board VR conditioner I was able to rev up to 5000rpm but I had another problem : there were little tach sync loss (no squirt and no ignition) during running, at any rev.

What do you suggest for the lm1815 mode ? Why not the mode 3 with pin 5 grounded ? Have you try this ?

I am using default v3.0 PCB VR circuit and I have no problem reving up to 7000 rpm (fuel only). Only one think was necessary - I had to install a resistor 4k7 to +VR line (in serial) - because I am sharing the stock ECU.
For the LM1815 is (imho) better adaptive mode (pin 5 open), but with modification, I have already wrote. I have tested it with this mod and it was working well. When using LM1815 there is also smaller phase offset between VR zero-crossing and output IRQ signal.

My setup:
- C20XE Turbo (now upgrading to Woessener 8.5:1 pistons and H-rods, pistons bottom oil cooling)
- 315ccm Z20LET injectors
- KKK16 turbo from C20LET
- before engine upgrade: MS1 HiRes 10d2 code - fuel only, v3 PCB
- MSII Extra ready to install after engine will be back in the car

Post by **jsmcortina** » Wed Apr 02, 2008 1:06 pm

Note that the V3 VR circuit requires the VR sensor to be installed backwards compared to the LM1815 circuit.

James

pet007 · Post by **pet007** » Wed Apr 02, 2008 11:44 pm

jsmcortina wrote:Note that the V3 VR circuit requires the VR sensor to be installed backwards compared to the LM1815 circuit.

James

Yes it's true. Or I have in output from LM1815 NPN transistor inverter. This mod makes the same job.

nyabinghi · Post by **nyabinghi** » Thu Apr 03, 2008 2:52 am

pet007 wrote: I am using default v3.0 PCB VR circuit and I have no problem reving up to 7000 rpm (fuel only). Only one think was necessary - I had to install a resistor 4k7 to +VR line (in serial) - because I am sharing the stock ECU.
For the LM1815 is (imho) better adaptive mode (pin 5 open), but with modification, I have already wrote. I have tested it with this mod and it was working well. When using LM1815 there is also smaller phase offset between VR zero-crossing and output IRQ signal.

My setup:
- C20XE Turbo (now upgrading to Woessener 8.5:1 pistons and H-rods, pistons bottom oil cooling)
- 315ccm Z20LET injectors
- KKK16 turbo from C20LET
- before engine upgrade: MS1 HiRes 10d2 code - fuel only, v3 PCB
- MSII Extra ready to install after engine will be back in the car

Accordig to several post I have read on these forums, it seems that MS2 is less permissive to tooth miss than MS1.

You can see on PCB3.0_1.JPG that on my MS board there is also a phase offset between VR zero-crossing and IRQ signal. It seems that this offset is not always the same, this can make the MS2 code detect missed or added teeth and so lossing the sync.

That's why I'm trying the LM1815 mod, on the picture you can see the offset is much smaller than with the original VR circuit, and this offset is constant, that's the most important imho.

Did you have this kind of IRQ signal on your default v3.0 PCB VR circuit ? (as seen on PCB3.0_1.JPG)
Did you change the sensor gap ?

nyabinghi · Post by **nyabinghi** » Thu Apr 03, 2008 2:56 am

pet007 wrote: Yes it's true. Or I have in output from LM1815 NPN transistor inverter. This mod makes the same job.

Sorry I don't understand

nyabinghi · Post by **nyabinghi** » Thu Apr 03, 2008 2:57 am

jsmcortina wrote:Note that the V3 VR circuit requires the VR sensor to be installed backwards compared to the LM1815 circuit.

James

Thanks for the information.
This is true for all wheel installation ? tested on stock c20xe wheel and sensor ?

nyabinghi · Post by **nyabinghi** » Sat Apr 05, 2008 6:44 pm

pet007 wrote:
nyabinghi wrote:Add some information on my setup
As you have wrote, the non-zero crossing after missing teeth is very strange. Maybe try to remove 330R resistor between VR in and GND. But I am not sure, if could fix it :/
Anyway, from your input VR signal is nice to see the much bigger amplitude when sensor is passing two missing teeth. And your LM1815 setup works in adaptive mode (pin 5 open). Then this bigger amplitude will cause arming treshold is on the higher level; then the following 1. normal tooth with lower amplitude is not able to reach the previous higher arming treshold => no trigger. You can try to insert aprox. 100k resistor before 0.33uF capacitor (from pin 7). This mod helps to slow down the arming speed of 0.33uF peak capacitor.
I had the same issue on C20XE engine, I couldn't rev more then aprox 2200 RPM.

Hi,
You were all right!! I removed the 330R resistor and put a 100k resistor before the 0.33uF capacitor. It works wery well and I can rev up to 6500rpm (I didn't try more) without any sync loss.
Tank you.

I've attached a picture of the signal.

max78 · Post by **max78** » Mon Apr 07, 2008 2:36 am

I had the same problems, with the modification it works.

Thanks, maybe someone should put it into the manual.

pet007 · Post by **pet007** » Mon Apr 07, 2008 6:12 am

nyabinghi wrote: Hi,
You were all right!! I removed the 330R resistor and put a 100k resistor before the 0.33uF capacitor. It works wery well and I can rev up to 6500rpm (I didn't try more) without any sync loss.
Tank you.

I've attached a picture of the signal.

Hi,
nice to hear that good news

This problem with almost half-bigger amplitude of VR signal during missing teeth is not common for most car, but Opel VR wheels and sensors are probably exception ;)

Cheers

Tommy · Post by **Tommy** » Tue Apr 08, 2008 2:16 pm

Delete

Peter Florance · Post by **Peter Florance** » Tue Apr 08, 2008 5:57 pm

I've seen a lot of mention of the 330 ohm resistor but I've not had good luck with it installed.

joppa · Post by **joppa** » Tue May 06, 2008 8:19 am

do you c20xe guys with vr issues have action in the rpm guage in megatune?

gmt · Post by **gmt** » Thu Nov 27, 2008 12:30 pm

Hi guys, im using a c20xe too and with similar problems. Im using the same lm1815 circuit posted at http://www.msextra.com/manuals/MS_Extra ... putV22.jpg but with modifications posted here. So i have removed the 300r resistor and installed a 100k resistor before the .33uf capacitor. Now i can rev up to 6300rpm (it was 3000rpm before that modification) but not more than that. I believe there is something more i have to do.

I'm thinking about a bigger resistor value then 18k in the input, what do you think? What value do i have to test?

Peter Florance · Post by **Peter Florance** » Thu Nov 27, 2008 5:34 pm

gmt wrote:Hi guys, im using a c20xe too and with similar problems. Im using the same lm1815 circuit posted at http://www.msextra.com/manuals/MS_Extra ... putV22.jpg but with modifications posted here. So i have removed the 300r resistor and installed a 100k resistor before the .33uf capacitor. Now i can rev up to 6300rpm (it was 3000rpm before that modification) but not more than that. I believe there is something more i have to do.

I'm thinking about a bigger resistor value then 18k in the input, what do you think? What value do i have to test?

Try grounding pin 14 and see if that helps

that's how I ran it on BMW motor that logged 60-2 rpm to 9300 (unfortunate missed-shift at turn 1 @ VIR)

marconi118 · Post by **marconi118** » Sat Dec 27, 2008 7:09 am

referring to the scope pictures, how did you optained a falling edge inoutput of LM1815 (red graph) at the zero crossing of vr sigal?

My LM1815 makes a rhising edge for zero crossing. Did you inverted the output with a transistor?

For MSextra 1 do I need a falling edge right?

The question is: do I have to invert the output of LM1815 with a transistor to feed the IRQ pin of CPU with MS1Extra????

Megasquirt Support Forum (MSEXTRA)

lm1815 on pcb v3.0

lm1815 on pcb v3.0

Re: lm1815 on pcb v3.0

Re: lm1815 on pcb v3.0

Re: lm1815 on pcb v3.0