2007 FG cruise control project.

[pugslyyy]

Have you considered using SSRs?

For this application I need a clean mechanical throw. I'm trying to adulterate the analog signals as little as possible.

I'm incorporating a feedback loop to verify the output of the arduino processor. It is PWM output (not a true analog) so I will have to smooth it.

 

[boostin]

Ok, I really wanted to help but I dint think I have any input, I don't really know remotely what you guys are talking about

 

[SkiFreak]

Ok, I really wanted to help but I dint think I have any input, I don't really know remotely what you guys are talking about

Solid state relays, opposed to mechanical relays.

 

[pugslyyy]

Solid state relays, opposed to mechanical relays.

the other thing is that I need the system to operate normally when my cruise hardware is not energized - the spdt mechanical relays are the best way I know to do this. I like solid state when I can use it, just not convinced this is the right application.

Phase 1 is to splice the new wiring in and still have everything work as normal. I've ordered some molex connectors and a a couple of compact Arduinos that I can use as a daughter board to the relay board I am speccing.

Once I get the system functioning, my next step will be to put the controller in line during normal driving. I need to get a feel for the compensation that will required for smoothing the PWM output to a truer analog - and I'm also not sure how touchy the Fuso ECU will be about all of this.

Finally, after I can get the cruise to reliably mimic accelerator inputs to the ECU I'll worry more about starting to manipulate the signal (cruise, cruise up, cruise down, etc). Gotta walk before you can run and craw before you can walk. I'm at the stage where I am beginning to contemplate eventually crawling at some point.

 

[pugslyyy]

Finished laying out the prototype PCB today. Because I'm not sure exactly what I am going to encounter I provisioned for pullup and pulldown resistors on all of the inputs. I won't need them all, but it will be simple matter of soldering in what I need - or tie something high/low with a jumper. In theory.

 

[pugslyyy]

Was able to spend part of today working on the software. There were a couple of changes to the hardware because certain pins support certain functions and others do not, but I'm still able to route everything so it's all good (so far).

I've got the basics programmed, but am now starting to dive deep into the proportional-integral-derivative controller which is the algorithm you use to decide how much throttle to increase/decrease to achieve the desired speed. Basically you have to give more throttle the further away you are from target, but ease up as you approach target so as not to overshoot. In terms of lag, the Fuso is much more like an oil tanker than a speed boat, so once I get everything set up I imagine I'll be playing with the constants for a while. It's new territory for me as I've never spent a lot of time on industrial control systems, but it's been fun to dredge up my knowledge of integral / differential maths.

 

[boostin]

So question, is this something you would sell if it works good?

 

[pugslyyy]

So question, is this something you would sell if it works good?

Yes and no. I'm not planning on going into business with these, but there's a minimum order of 10 PCBs, so assuming everything goes reasonably well I should have some left over that I would sell to folks that want them either as bare boards, kits, etc.

Of course, it's still too early to know if this project will actually even work.

 

[boostin]

Is it going to be fairly plug n play, or cut into harnesses?

 

[pugslyyy]

Is it going to be fairly plug n play, or cut into harnesses?

I'm going to cut/splice into the harness. There's no native support for a cruise control in mine so you have to intercept and inject, and I'm happier splicing wires with heatshrink butt connectors than most any other approach.

I've been working on the programming today and already revised my PCB layout 4 or 5 times. I also spent quite a lot of time reading papers on PID control theory / lambda compensation - it's pretty interesting stuff and actually helps me understand why my cruise control on my other vehicles works the way that they do.

I should get a couple of headers, etc delivered midweek and then I'm going to splice into my wiring harness and start looking at signals. Some of the signals are already logic level (5VDC) but others are higher so probably will need to use some zener diodes to make sure I don't blow up anything.

The Fuso ECU is pretty rugged so I'm not too worried about damaging it (famous last words), the arduino nano I'm planning on using is the likely weak link.

Tomorrow I'm hoping to have time to look at the VSS (vehicle speed sensor) and figure out how to capture that data. That's the last input that is a big question mark.

 

[pugslyyy]

I got the connectors I needed in today, and spliced the ECU wiring harness.

The ECU is now back in place and hopefully I won't need to mess with it again.



This is my manual bypass connector for the new wiring. I have to tap three signals (brake, clutch, VSS) and intercept inject three others - you can see the loopback on the connector.

 

[SkiFreak]

Yellow tape where there should be black tape would send my OCD into overdrive.

 

[pugslyyy]

Yellow tape where there should be black tape would send my OCD into overdrive.

Lol, the yellow tape happened to be the closest at hand. I was just delighted when it started and threw no codes. I need to get out and drive it next and see what the signals are like.

The prices on the hardware are crazy cheap these days.

arduino nano - $7
8 relay PCB - $8
20x4 serial LCD module - $11

I'm using .093 molex connectors for most of it, with Dupont connectors to the LCD display. The Molex connectors are kinda overkill, but they are a tried and true standard and I like how they work even if it takes three separate crimps per pin to make it happen.

 

[boostin]

I am so impressed and excited at the same time, I hippie this works and works awesome cause I would love to buy a kit. Mine being auto though might make things a little different

 

[pugslyyy]

I've gathered about as much information as possible without actually rolling down the road. Hopefully tomorrow I'll have a chance to drive with the oscope on the VSS and gain a fuller understanding of what that signal looks like.

I captured the clutch switch and brake switch today - they are both at truck level (12VDC) so I am going to use an optoisolator on each to bring them down into the 5V logic range. I'm using both the brake and clutch as interrupts, so if they get touched the cruise drops out - which seems like the right thing to have happen.

My current big questionmark is to capture what the VSS signal looks like so I can figure out how best to use it.