Peak and hold injector driver board

[myk777]

Here's post I made with some scope pics of the P+H circuit in action, thought it might be of interest here.

http://www.msefi.com/viewtopic.php?f=29&t=32048

Mike

[undesiredshoe]

Ok so im building this board but im confused on how to connect it up. Im not using megasquirt but this board should hopefully work with my car. How my injectors are set up is theres a common 12v which is a red wire for all the injectors and then different color wires coming out of the injectors to the ecu. I believe the different colored wires are what the ecu uses to trigger the injectors. Im just confused on how i would go about connecting up the wires. If anyone could help that would be awesome.

[racingmini_mtl]

First, which board are you talking about: the internal one (V1.1A) or one of the external one (V2.0 or V2.1)? Second, do you have a wiring schematic and can you post it? And how many injectors?

From what you're saying, you would leave the red wire alone and the different color wires would then be connected to the p&h board. The p&h board will need to be connected to the car ground and to the ECU (and possibly to 12V depending on the board). How this is done depends on which board you have.

Jean

[undesiredshoe]

Im going to be ordering the 1.1A board. My car is a 4 cylinder, Nissan ka24e. The four red wires are 12V and those wires are the ones i cut and placed the resistor pack inbetween. Each injector then has one red wire going in and a different color wire coming out which I assume grounds to the ecu and is what the ecu uses to trigger the injectors. I hope you get what im trying to say.

[undesiredshoe]

Ok so i found this diagram on your site and its helping me make sense of more stuff now.



I would connect the different colored wires coming off the injectors to the top of the board and then at the bottom of the board, connect the other half of that wire to the same wire going to the ecu?

[racingmini_mtl]

Not exactly. The V1.1A board doesn't connect in between the ECU and the injectors in the way you mean. The 4 connections at the bottom of the picture above need to have 5V pulse when the injectors are injecting. So it needs to be connected somewhere inside your ECU. (I can't exactly say where and how in your case) The signal outside the ECU is a grounding signal which will not work with this board.

If you want to just replace a resistor pack, you need the V2.0 board (for 4 injectors or the V2.1 board for 8 injectors). With the V2.0 board, you will connect 4 wires to the injectors and 4 wires to the ECU (like you would the resistor pack). In addition, you will need to connect 4 wires to ground for the injector grounds (which is no longer done at the ECU) and a 12V and ground wire to power the board.

The V2.0 fit in it's own box independent from the ECU. You can look at my post from Feb. 10 (on page 10) on this thread to have an idea of what it looks like and also at this. This is what the finished box looks like (without the connector):


If you're interested in getting a board, you can contact me by PM or by email (boards at jbperf.com).

Jean

[undesiredshoe]

Dang, anyone want to buy the components needed to build the 1.1a board?

[racingmini_mtl]

The components to build the 1.1a board are almost all used for the V2.0 also. You'll need a few additional things for the V2.0. The BOM for the V2.0 is here.

Jean

[myk777]

I've been doing a bunch of injector testing and thought I would post this up, someone might find it of interest. What I have is a before and after scope shot of injector current (100mv=.90A) using a 2ms pulsewidth (approximately my idle PW) both at 13.2 volts.

Before is the P+H circuit as drawn, as you can see the chip never transitions into the "hold" mode with my setup.

After is the same P+H circuit with the resistor in the time-out circuit changed from 39k to 15k Ohms, as you can see in the picture this forces the chip to transition into the "hold" mode at around 1.8ms.

web01.jpg

All this adds up to a more consistent closing time of the injector because now all my injection events end from the "hold" phase, where before some were from the "almost peak" phase and some were from the "hold" phase depending on how long the PW was or how high the battery voltage was. It didn't make a huge difference but it was a fun exercise nonetheless.

It was also suggested to change the current sense resistor to a larger value (I would have needed ~.15 Ohms resistor) this should in theory also force the chip to transition to the hold mode sooner, but it has the down side of lowering the hold voltage substantially.

I checked this modification from 8v-16.6v input and it seems to be a good solution since my injector opening time worst case at 8v is about 1.3ms.

Anyone see any reason why I shouldn't do this?

[Poolside]

Way to get in there and make it the way you want it. :thumbsup

As long as the injector is fully open, it doesn't matter when the transition to 'hold current' occurs. And if your confident of the 1.3ms worst-case opening time, you're good.

Aside: Hey 777, is that a USB/PC scope you're using? What brand and model? Does the software allow the notations?

[racingmini_mtl]

As mentioned above, there should be no problem since you've established that the injector will be fully open even at a battery voltage of 8V.

Jean

[myk777]

Thanks for the review, I couldn't think of a reason why it wouldn't work, but it's always good to have more people think about it.

The scope is a USB/PC unit model is DSO-2150 made by Hantek 60MHz, they are sold by the dozens on ebay, pretty reasonable too. It's my first one so I really can't compare it to anything but it seems to work good and it easy to use. I did the notations in powerpoint, I'm not sure if the scope program has that functionality or not, I'm just so used to ppt at work that's what I always use.

[racingmini_mtl]

I will have complete kits available soon for the V1.1 board. More details are available here. Depending on what the interest is for this, I'll continue to make this available or not. If there isn't a good response, I just won't renew the component stock once I run out.

This is only for the V1.1 board. I'm looking at other options for the V2.0 and V2.1 boards.

Jean

[linuxman51]

anyone else have a problem with the V1.1A board? It doesn't quite match the assembly guide on the website and I'm having a couple odd issues (hopefully just something I've missed). Here's what I've got:
Output 1 is always grounded. output 2 works fine (its tied into output 1 on the injector driver on the board), so whatever is causing it to ground out is after the input on the board (and also isolated from inj output 2 since it doesn't seem to affect it).

Output 3 works fine, output 4 does nothing at all.
Again, both 3 and 4 are tied into the second output on the board.
On output 1 I have replaced the fet (no change) and the IC (again, no change). Nothing appears to burn, nothing smells funny, and nothing gets hot.

I'm wondering if I've got the 5v hooked up to the wrong lead: there are 6 pins instead of the indicated 5 on the website (however a 5v is silk screened, and thats where I have 5v running into the circuit).

Any ideas?

[linuxman51]

ok i have corrected the issue with #4.without a chip in the socket on #1 it is still connected to ground.

[Poolside]

Nothing appears to burn, nothing smells funny, and nothing gets hot.

That's a good deal right there. Don't let the smoke out.

- Jim

[racingmini_mtl]

As mentioned on the V1.1 section on the web page, the board has 6 pads with the 5th pad being a ground (which was not there on the V1.0) and the 6th pad being the 5V supply (as you assumed). The 5th pad is for the low current ground (associated with the 5V supply) which is used to power the LM1949 chips. This has to be connected to the ground on the MS board and if it is not present you will not have a functional board.

I definitely need to update the web page to reflect this because I agree that it can be confusing with just a small note in writing instead of the picture showing the connections. I apologize for any problem this has caused.

If that was not your problem let me know and we'll work this out.

Jean

[linuxman51]

This thing is kicking me hard. with the power off, I can't find continuity between anything and ground (Well, not direct continuity. I can pull a couple K ohm between various pins and ground). I can't help but think i'm just missing something small and stupid except the other three circuits work fine (wire had pulled out on the #4 input and was only intermittently connecting the circuit.. that was only moderately embarrassing)

[linuxman51]

As mentioned on the V1.1 section on the web page, the board has 6 pads with the 5th pad being a ground (which was not there on the V1.0) and the 6th pad being the 5V supply (as you assumed). The 5th pad is for the low current ground (associated with the 5V supply) which is used to power the LM1949 chips. This has to be connected to the ground on the MS board and if it is not present you will not have a functional board.

I definitely need to update the web page to reflect this because I agree that it can be confusing with just a small note in writing instead of the picture showing the connections. I apologize for any problem this has caused.

If that was not your problem let me know and we'll work this out.

Jean

I will try this when I get home. Why with that not connected tho would the other three work? (just curious. Circuits do weird things at times)

[racingmini_mtl]

It may have been grounded partially through the LM1949 to the high current ground which is not how it's suppose to work but it may be sufficient to get thing partially functional depending on the chip itself and the associated components. Hopefully nothing was hurt by this.

Jean