Dead time for low impedance injectors with resistors?

[dpuma8]

Hi everyone,  I have RC engineering 550cc low impedance injectors with resistors wired in and I am having a tough time getting consistent AFRs after using Auto Tune.  I will tune to get a 13.5 AFR at idle but then the next time I will see 12 or 15 as my AFR.  From what I am reading, my injector dead time amount could be affecting this. 

 

I emailed RC Engineering and they sent me this document in an email.  I asked if these values were for low impedance with resistors and he said they were.  For Megasquirt, I am entering .263 for 13 volts @ 40 psi of fuel, correct?  It seems to me that these values are too low and should be near .900, right?  Any insight?

Edited July 25, 2017 by dpuma8

[Chickenman]

Edit: And the below quote is apparently " sketchy " . I didn't read far enough in the thread to see the replies to several E-mails back and forth. I've used the " Strikeout " command to indicate that this is all questionable information from RC engineering.  

 

Originally Posted by RC Eng

Hello Ryan,

The lag times for the low impedance 550s (PL9-550) is 1.1 ms at 13.5 volts. 


Thank you for your interest in RC Engineering. Please feel free to call or e-mail us with any further questions you may have. 



RC Engineering, Inc. 
20807 Higgins Ct.
Torrance, Ca 90501
T (310) 320-2277
F (310) 782-1346
rceng.com

 

Edited July 26, 2017 by Chickenman

[Chickenman]

^ Ummm... yeah, I'd say it's a bit low.

 

A bit of a Tip for you. Don't let Auto-Tune or VEAL analyse the VE Cells around idle. It doesn't do a good job.; Open VEAL and highlight a 3 x 3 block of Cells around your Idle speed and Kpa. Right click and Choose " Lock Cells " . 

 

Edit: Also limit EGO control to a minimum RPM of at least 200 RPM above your idle speed. Especailly if you are running a performance cam. Adjust idle VE Cells manually using the above method. . 

 

When you save the Tune, use the " save As " Option and give the Tune a Unique name like "  550cc Tune Mod #1.msq, 550cc Tune Mod #2.msq, DFWI Tune.msq ( inside joke ).  When TS closes it saves your tune as " CurrentTune.msq.... but it only saves one file. CurrentTune gets overwritten each time you make a change and close TS. So you can get lost in a real hurry.

Edited July 26, 2017 by Chickenman

[Chickenman]

Mind you, this chart says the RC 550cc Low-z injectors ( 3 ohm ) are 0.59 ms at 13 v.  Whos' right?? 

 

http://www.turbochargerkits.com.au/efi-tech/fuel-injector-dwell-times/

 

Edit: Ok, I've found a second site that says .59ms at 13 volts. Injector Rehab. These guys I would trust more than some unknown site.

 

http://injector-rehab.com/shop/lag.html

Edited July 26, 2017 by Chickenman

[Chickenman]

BTW, I believe resistor boxes add more Lag Time than using PW Current limiting as the the injector receives only 6 volts during the Peak and Hold cycle. Mean while a PW current limiting receives a full 12 v ( 12 v being a measured set point ) during Peak pull up and them the Hold function is Pulse Width Modulated. 

 

What does that mean? I'd try somewhere between .700ms to .900 ms. See where your PW falls at idle. 

[Chickenman]

Apparently I didn't read far enough in the Thread that contained the 1.1ms information from RC engineering.  On page two the E-mails from the Customer ( Ryan ) were continued and he got back this response. 

 

 

Originally Posted by RC Eng.

Hello Ryan,

We suggest using the values provided by AEM. 

Thank you for your interest in RC Engineering. Please feel free to call or e-mail us with any further questions you may have. 

RC Engineering, Inc.
20807 Higgins Ct.
Torrance, Ca 90501
T (310) 320-2277
F (310) 782-1346
rceng.com

 

So apparently RC Engineering doesn't have a clue what their Latencies are. Or the receptionist is answering Tech questions.....

 

BTW, AEM recommends .59 ms at 13 volts. 

 

Here's a link to the original thread:. make sure you read page two... Oy vay 

 

http://www.3si.org/forum/f1/rc-550cc-lag-time-429664/

[seattlejester]

For my 440cc low impedance supra injectors with a resistor box I have it set to 1.150ms for dead time.

 

Is that chart correct for having the resistor inline on a low impedance setup? Adding the resistor would raise the impedance and require more dead time, of course the resistance of the resistor would play a factor as well.

[Chickenman]

^ That's a very good question. And I have no freakin' idea.... nor can I find any reliable information on that. I've been looking for a long time for the answer to that exact question ... sigh.

 

But .263 ms with resistors? I don't think so. 

[spiff]

First I would go to the ego controller and choose "disable" to make sure it's not veal or closed loop thats affecting your idle afr's. Even with it disabled tunerstudio will read af but autotune or any other form of closed loop will not interfere with anything.

 

As far as dead time goes I can't help you but I just used the default tunerstudio 0.9ms on my Z20E (low ohm injectors with resistor pack) and had zero issues. 

[Namor]

So, I have spent a lot of time reading about deadtime and trying to figure out what to set mine to (DSM 450cc low imp injectors with resistors.) I first tried measuring deadtime using a graduated cylinder and test mode in tunerstudio.  I got very accurate results, but couldn't come up with a good way to test at various voltages because to use the megasquirt drivers, I couldn't have the engine running.  So I was able to come up with a number, but it was only for 11.9 volts, which doesn't help a lot for a running engine.  

 

After some more reading, I came across something so simple that I felt like an idiot for not thinking of it.  Ensure you are setup for batch injection and then cycle between 1 and 2 squirts per cycle.  If your deadtime is accurate, your AFRs shouldn't change.  If your deadtime is too long, when you go to 2 squirts per cycle you are going to be injecting more fuel than the computer thinks and you will go rich.  If your deadtime is too short, in 2 squirts per cycle you will go lean.  Now I have a pretty good number for both ~11.9 volts and ~13.9 volts and between the two I can set up a slope.  

[dpuma8]

Darn, so no solid answer.  I will try higher dead time numbers but my AFR problem happens at all RPMs, not just at idle.  This makes me wonder if anyone here in this forum actually has a solid tune.

Edited July 27, 2017 by dpuma8

[Chickenman]

Can you post up your .msq so we can have a look at it? 

 

At least by using the dropping resistors you are halfway ahead of the game.  Megasquirt and Tuner Studio PW Current limiting seems to be hit and miss from MS forum questions and articles. . 

 

There was a discussion on Guild Of EFI tuners the other day about injector drivers, and how different ECU manufacturers handle PW current limiting. So me are really good... others Meh. 

[Chickenman]

Darn, so no solid answer.  I will try higher dead time numbers but my AFR problem happens at all RPMs, not just at idle.  This makes me wonder if anyone here in this forum actually has a solid tune.

 

From helping out some members here, I can tell you right now that you can't just Copy and Paste MS Tunes. Every engine is different and it seems every MS build is different. You really do have to start from scratch  with  the Default Tune for the Firmware version installed on your ECU.  

 

The ability to build MS with so many variations and build Options is one of it's greatest strengths. It's alo one of it's greatest weaknesses. Compared to retail boxes like HalTech, AEM, Link, Adaptonics etc...MS just has too many variables in the hardware builds and build quality to make any " Standard " tunes. Every Home built DIY box seems to be different . Used boxes are the worst, because you are relying on the building skills of someone who may not have a clue on how to build electronic components. And you may not be sure of what hardware is inside. 

 

It's  fun as a DIY science project. But trying to sort out a Tune can be very steep learning curve. It can be done, but you can't take shortcuts. . Build a Tune from scratch...  copying tunes from the Internet doesn't usually end well. 

Edited August 8, 2017 by Chickenman

[Chickenman]

PM sent...