Please help me with my MSDIYPNP build (knock with KS4)

[Bean Bandit]

Hello,

I hope to get some help regarding what to put on my MSDIYPNP board.

Short background: I had an MS2 v3.0 board, sold it and now upgraded to the MSDIYPNP for a neater look and sequential ability.

I have a 4A-GE 16V engine and am in the process of some upgrades, currently the engine is stock but new head with different cams, ITBs, full sequential setup are ready to go on soonish. I'd like to run the MSDIYPNP on my stock engine first to come up with a decent map and facilitate further troubleshooting.
This means in particular I want to:
- first run the stock dizzy and will upgrade to COP (sequential or wasted spark - depending what coils I can source locally) with Nippondenso 24/1 CAS
- first batch fire the injectors and later run sequential
- wire in my phormula KS-4 knock monitor and use MS for logging and intervening

Therefor I purchased the MSDIYPNP and the sequential mod.
Now so far I soldered in the following parts:
- MAP sensor
- Relay 1 & 2 for operating an electric fan and TVIS (which will later be redundant due to the ITBs)
- LM1815 for second VR signal (G)
- all the parts that need to be there regardless of the configuration

Now to the questions:
- I assume I can already install the sequential mod board and just use INJ1&2 to begin with (as I plan to wire the new injectors in through the DB15 connector to avoid modifying the original harness)?
- can I prepare all the necessary connections for the two VR sensors (NE & G) and COPs (will also be wired in through DB15 for same reasons as above) and just leave out the jumpers, as long as I'll be running the stock ignition on normal trigger?
- how do I go about connecting my KS-4? The device has two outputs (trigger and sensor data) - the trigger cable will pull to ground so this is what I will connect to the 'Knock In' - do I still need to build the whole circuit?
- regarding the data logging of the KS-4 (0-5V signal), is it possible to modify one of the general purpose inputs, do I have to make a custom circuit in the Proto area or am I way off and can simply connect it to some pin and which I can then choose to be logged accordingly?

I think/hope I haven't forgotten anything

Thanks in advance for the doing the lengthy read and any input/advice/help

Best regards Bean

EDITED Title to hopefully receive an input eventually

[Bean Bandit]

addition to the knock device:
I just found out that the trigger is pulling to ground when knock is occurring and stays there until the alarm is reset on the unit so this will be less useable then I thought... I assume the data (0-5V) signal is a bit tricky to use with the current settings of the MS, as even when set to going high I'm not sure if it will trigger the MS all the time as a slight knock wont send the signal up to 5V... any ideas on how to resolve this are welcome too

[Bean Bandit]

75 views an no opinions? I'm surely not the brightest when it comes to electronics but don't think my questions are too tricky to answer by those who are working with MS for years...

[Matt Cramer]

addition to the knock device:
I just found out that the trigger is pulling to ground when knock is occurring and stays there until the alarm is reset on the unit so this will be less useable then I thought... I assume the data (0-5V) signal is a bit tricky to use with the current settings of the MS, as even when set to going high I'm not sure if it will trigger the MS all the time as a slight knock wont send the signal up to 5V... any ideas on how to resolve this are welcome too

I think this could actually be wired straight to an ADC input.

[Matt Cramer]

Hello,

Now to the questions:
- I assume I can already install the sequential mod board and just use INJ1&2 to begin with (as I plan to wire the new injectors in through the DB15 connector to avoid modifying the original harness)?

Yes, it won't complain about unused I/O.

- can I prepare all the necessary connections for the two VR sensors (NE & G) and COPs (will also be wired in through DB15 for same reasons as above) and just leave out the jumpers, as long as I'll be running the stock ignition on normal trigger?

Yes. Same as above.

[Bean Bandit]

addition to the knock device:
I just found out that the trigger is pulling to ground when knock is occurring and stays there until the alarm is reset on the unit so this will be less useable then I thought... I assume the data (0-5V) signal is a bit tricky to use with the current settings of the MS, as even when set to going high I'm not sure if it will trigger the MS all the time as a slight knock wont send the signal up to 5V... any ideas on how to resolve this are welcome too

I think this could actually be wired straight to an ADC input.

will try that, I assume there's no need to build the knock circuit on the board then.
To have MS use the knock signal (I'll hopefully get in through ADC, as suggested) and automatically retard ignition I guess some reprogramming will be necessary.

Thank you very much for your help.

Best regards Bean

[Matt Cramer]

No, the standard knock sensing code should work fine as is.

[Bean Bandit]

No, the standard knock sensing code should work fine as is.

will it trigger even if the signal doesn't go all the way to 5v (i.e light knock with only 4v and not trigger with normal trace <2v) - as I couldn't see a setting to enter from what voltage it will trigger (just going low or high)?

sorry for all the questions but getting anxious to finish the MS, so I can focus on the remaining to-do list

[Matt Cramer]

I thought you said the knock box was sending a 5 volt signal that was pulled to ground when it heard knock?

[Bean Bandit]

I thought you said the knock box was sending a 5 volt signal that was pulled to ground when it heard knock?

addition to the knock device:
I just found out that the trigger is pulling to ground when knock is occurring and stays there until the alarm is reset on the unit so this will be less useable then I thought... I assume the data (0-5V) signal is a bit tricky to use with the current settings of the MS, as even when set to going high I'm not sure if it will trigger the MS all the time as a slight knock wont send the signal up to 5V... any ideas on how to resolve this are welcome too

the alarm trigger does but unfortunately stays there... - just to be sure the knock circuit on the DIY board is for conditioning the signal only, right?

EDIT: here's a pic what the signal from data will look like:

the scale is hard to read but it's 0.5v increments (so knock was a spike just above 4.5v)

[Matt Cramer]

You could probably feed that signal into the DIYPNP knock conditioner; it's a straight up threshold detector.

[Bean Bandit]

You could probably feed that signal into the DIYPNP knock conditioner; it's a straight up threshold detector.

this is getting a bit too technical for me again - how does adjusting the pot in the circuit change what MS will see in terms of sensitivity. I.e. when I set it to sensitive will it trigger early with a slight voltage rise and vice versa? if this is the case this could be indeed an option - the only downside crossing my mind then is, how can I log that signal at the same time? would it work to simply split it and feed it to both knock circuit and ADC input, or am I asking for trouble here due to interferences?

As mentioned electronic circuitry isn't one of my strong points, planing to change that in the future

[Matt Cramer]

You could probably feed that signal into the DIYPNP knock conditioner; it's a straight up threshold detector.

this is getting a bit too technical for me again - how does adjusting the pot in the circuit change what MS will see in terms of sensitivity. I.e. when I set it to sensitive will it trigger early with a slight voltage rise and vice versa?

Exactly. It looks for a particular voltage, and if your signal goes over it, it triggers. The pot adjusts the voltage.

if this is the case this could be indeed an option - the only downside crossing my mind then is, how can I log that signal at the same time?

You can't, at least not easily.

[Bean Bandit]

hmm...

so currently I'm looking at 5 options I guess.

1. use the 0-5v data signal for the knock circuit and have no log
2. install a switch to choose between knock circuit and log
3. double the signal, with lots of help from the community as I have no idea how to do it
4. use the 0-5v data signal for log and somehow make the trigger signal work as a fail-safe in combination with the knock circuit
5. bring in another thought (there is another signal available): the raw signal from the knock sensor - and use that for the knock circuit and use the data one to log

- my knowledge about the knock circuit is very limited - could I make the trigger signal work and have the MS simply retard timing overall by a few degrees and therefor creating something like a limp-mode until I reset the alarm, which will stop the trigger from being pulled to ground and hopefully have the MS return to normal operation?

- when you say not easily, does that mean supper hard involving a huge circuit and signal conditioners or not easy for a noob like I seem to be?

- I assume the knock circuit being less sophisticated than the unit from phormula interpreting the raw sensor signal, would it still be recommended to use the unit that way?
as the MS documentation say it's more to protect from a bad tank than anything else... (I won't rely on the circuit to tune the engine but around here some days get very hot ~49°C and that's when I want to be on the safe side)

open for suggestions

EDIT: added more accurate information

[Matt Cramer]

Do you have a plot of that graph from the Phormula with a time scale on it clearly readable? Depending on how short the voltage spike is, it may not be possible to log it at all.

[Bean Bandit]

Do you have a plot of that graph from the Phormula with a time scale on it clearly readable? Depending on how short the voltage spike is, it may not be possible to log it at all.

unfortunately not but the data signal rate can be set from 5-100 samples/second. I have no idea where the limit of MS is...