TOD discussion

[Nazrat]

http://forum.australia4wd.com/lofive...php/t1153.html

-Tad

[bobmumgaard]

The guys on the link above and it's links have got the TOD system pinned down and solved.
If our system is the same as theirs (and it must be pretty close, same manufacturer, overal design, and basic operation) then the TOD brain sends a 50Hz PWM (pulse width modulation) signal to the transfer case solenoid with a duty cycle between 0% and 88%.

(0% being no torque to front and 88% being 50/50 split)

Aside for those not electrical engineering prone:
PWM sends a square wave to the load, this square wave has a peak voltage of about ~13.6V(in our case at least). By altering the width of the square wave you can change the effective voltage the load sees. So with an 88% duty cycle, the square wave occurs over 88% of the period and the average voltage to the load is about 12V. So if we wanted the load to see 7.2 V (this would be about 30% torque to the front) we would send a duty cycle of around 53%.
The advantages of PWM make it very common in automotive apllications, these advantages include a resistance to electrical noise, ablity to integrate digital and analoge components and a lower strain on electrical components (the current in PWM only flows for a fraction of an analoge only system). Infact many sensors, including most speed sensors use PWM.


Onward with the TOD side
This means we can alter the amount of transfer case lockup by altering the PWM signal the solenoid sees.

So heres a proposed idea:
If we want to have 4hi, 2hi, or anything in between (say 30% torque to the front) we could disconnect the TOD ecu from the solenoid (and connect the TOD ecu to a false load, say a 5? ohm resistor) and connect the solenoid to a PWM controler that we can vary. This PWM would be variable from 0% to 88%. (0V to 12V rms) this way we could dial in the amount of transfer case lock-up from 0 (2hi) to 50/50 (4hi).

The implications for those that want to full control over their TOD are huge. You could have a dial that selects the amount of torque to the front much like a rally car would. And the signal would be the same configuration the TOD auto mode sends, so no adverse affects. And if you wanted TOD auto mode you would reconnect the TOD ecu to the solenoid (with a relay) this would set the system up to operate as usual.

Modification could be made from there:
You could directly alter the TOD display lights with a series of relays in conjuction with the PWM controler so the display would display the front torgue correctly. You could aslo use a clever set of relays to set the sysem up so the TOD ecu automatically regains control over the transfer case in certain situations like 4 low or when the abs activates. Though i have not thought these all the way through yet they remain tantalyzing possiblities.

And, theorectically, the system could be done even simpler. The soleniod might not NEED to see a PWM signal to operate, you could simply alter the voltage it sees with a poteniometer instead of a PWM controller. Though this idea is alittle hairy because the potential for excessive wear and adverse affects because the signal the solenoid receives is not of the same form as it is design to use. I would have to do some extensive tests with the transfer case outside the vehicle to be sure how this could work.

So it looks like this winter I will be driving around in the snow with no passenger seat and a multimeter and oscillosope hooked to the TOD ecu to soleniod wire. Hopefully this research will confirm what the aussie and spanish Terracan drivers have discovered. From there it will be a matter of constructing the right circuit and we could have a viable torque split selector.


Anyway, if you read this and think I'm wrong, right or a moron, shoot back your ideas. With alittle brainstorming I think we could come up with a pretty nifty, inexpensive and reversible solution for those that want complete control over their transfer case.

[Joe_Black]

The Isuzu rally teams had an electronically selectable TOD ratio in their 5-speed VehiCROSSes, so it's basically knowing what/how to connect. From what I was able to read about the competition set-up indicated simplicity and plug-n-play reversibility to stock.

[Ldub]

Bob M,
YOU DA MAN...wish I had the tech. abilities reqd. to figure something this complex out in the first place. Seriuosly, your aside for the non elec. eng. prone was too deep for me.
However, if you get a handle on this thing & produce a few "plug & play" units (harness, controller, DETAILED instructions, etc.) You can count me in as your first customer.

Hope ya get it to work, Ldub

[mbeach]

I couldn't open the Terracan links, but I have been suspecting the same thing (PWM signal) for a couple of weeks.

I have plenty of snow for practice, what wire(s) should I monitor?
I think that a few inputs would be in order (front axle, rear axle, ABS, 4Lo) to the TOD ecu.
I'd definately monitor the TOD to clutch solenoid as well.

I figure that this weekend I could get started checking voltages and frequencies under various conditions. It will be difficult to catch a 70/30 or an 85/15 condition however. It seems that TOD is all-or-nothing on ice and snow.

I'll think about a nice spreadsheet for gathering all of this data.

If we get the right information together, I have an extensive inventory of tools and parts to make this happen.
Heck, if it works I might not sell the VX after all.

[mbeach]

I was discussing this with my lead tech (super genius).

We can make this happen, BUT:
I need a TOD computer. They pop up on eBay all the time, except for today. I will pay up to $100 for one, I cannot use of of mine right now (too much ice on the road still).

We were thinking about catching the signal INSIDE of the TOD computer, before it hits the amplifier section. This should eliminate the need for a dummy load. We can make a PWM generator with a timer (555?). By altering the control voltage into the timer, we should be able to modify the square wave output. This means a KNOB, with infinite variability or presets (whichever you want).

HELP ME FIND A COMPUTER!!!

[bobmumgaard]

mbeach-
I imagine you have the snow up there to try this, I fortunatly haven't got the white stuff down here NE yet.

If you want to monitor what the TOD ecu is doing I think monitoring these wires would be the most important:

1.

TOD to Solenoid - On a 2000 it is the light blue wire that exits the TOD ecu from pin 4
If you hook a multimeter to this wire and monitor voltage it should hover around 0-1V most of the time and jump when the TOD dislpay shows power to the front wheels. An ocsiloscope would confirm if this is indeed a PWM signal.

2.

Front and Rear vehicle speed sensors- On a 2000 TOD ECU the front is pin 24 (green/white or grey/white), the rear is pin 23 (green/yellow or grey/yellow) and the reference for these sensors is pin 15 (green or grey)
The TOD ecu's logic probably looks at a differential between these two sensors and then decides how much torque split to enact. We could probably gain alot of insight on the function of the ecu by doing the same thing. I assume these two are probably PWM signals as well. If you use an oscilloscope with two channels of imput you can see this PWM differential occur when the rears slip.

If you want to look at the circuit diagrams yourself you can use Tone's maintaintence cd (if you don't have one just e-mail him or post and I'm sure he'll hook you up) The most useful pages are:

Driveline/axle > Driveline control (TOD) > Basic diagnostics flow chart > Checking failed pin > Connector pin assignment
and
Body and acceserories > Wiring diagram > circuit diagram 19 (the second 19)

I agree, at this stage a spreadsheet would be very helpful here, the way I'd set it up is with columns for this data:

Front speed sensor duty cycle
Rear speed sensor duty cycle
TOD-solenoid duty cycle
TOD-solenoid rms Voltage
What the solenoid is doing (0 front, 50/50 etc.)

Alright, so after we get this information together (hopefully it will confirm what the Terracan guys have discovered) it is a matter of making the modifications

I talked to a few of my EE collegues and they all agreed that substituting our own controlled PWM signal to the clutch instead of the ECU's signal would be the best route. This could certainly be done with a 555 like your tech said and it would be fairly inexpensive. Though I personally don't have the indepth know-how to build the actual circuit, but I know people that do, and you sound like you might have the knowlege and tools.

After thinking about the actual mechanics of modifying this system I think we have three basic options:

1. Modify the signals coming into the TOD ecu
This basically tricks the TOD ecu into thinking the car is doing one thing, and we get the desired result from the TOD ecu outputs. So if we modify the front or rear speed sensor signals before the ecu, then the ecu thinks there is slipping and we get lock-up. The trooper guys have kinda done this by cutting the front sensor wires. Then the ecu sends power to the front axle, but also throws a code. This method would produce the correct dash display. Though it scares me because if something goes wrong the TOD ecu would get fried....$$$

2. Modify the action inside the TOD ecu
I think this is the approach you and your tech are talking about. It would be the most elegant because it would all be contained inside the existing black box. If you can score an extra TOD ecu it could be taken apart and investigated, reprogramed almost. However, this approach would be hard to duplicate for the electronically timid among the VX community. Though it seems to be the most technically interesting.

3. Modify the TOD ecu's outputs
This would consist of modifying the signal from the TOD ecu to the clutch like I mentioned before. It is not the most elegant because it wont display correctly on the dash display but it would be the least risky and easiest to duplicate. The risk is minimized because you couldn't hurt the ecu (since the modification is after the ecu, you would need a dummy load though), and the soleniod is probably robust enough to withstand any minor malfunction we casue. The electronic circuits could be contained in a project box that would require cutting only the wire going from the TOD ecu to the solenoid. This makes it easy to install and produce for members on this board (might as well spread the wealth ehhh), you would have to cut one wire, install the knob, hook up power and ground and you'd be set. I think this approach is the one I'll persue.

Well thats my take on the whole deal right now, throw back your opinions on it and we'll keep brainstorming, this shouldn't be that hard. Plus, once the research is done and the modification tested there are plenty of guys on this board that could benefit from the knowlege obtained.

[mbeach]

My initial idea was option 1. I liked the fact that the display would be correct. However, once the vehicle were stopped, the TOD would be thrown into confusion. So I abandoned it.

Option 2 is the most elegant, yet it is not reversable (easily), removable, or easily installed. I'm going to try to veer from this, but we'll see. My goal is to have a correct dash display, and 5 modes of operation: Auto, 2wd, 15, 30, 4hi. All selectable with a knob. I think that by keeping it simple (5 modes, or outputs rather than infinite) I can make the display work as well.

I would need to:
a) turn off the AUTO display when the knob is switched to any other position.
b) turn on the appropriate lights with each mode.
c) default to AUTO mode whenever 4Lo is activated -or maybe not. 2Lo is an interesting option, if it's even possible.

I guess that I'm looking at option 2.5. I want to be "inside" the TOD computer for the display benefits, but I want it to be accessable and easy for a novice to hook up.

I pick up a better (portable) scope this week. This weekend I'm going to pull the harness(s) off of the TOD computer and make extensions (to minimize my crawling under the seat) that I can easily tap in to. I have the workshop CD, I'll be looking it over tonight. I think that I have a few spare connectors around the shop.

[VehiGAZ]

I really appreciate this conversation, guys!

I want only one thing - to have a switch to flip to go into 50%/50% (4WD Hi) mode, which would be best on snowy roads when going into 4WD Low is overkill, while leaving it in Auto ToD mode means it is CONSTNATLY (and annoyingly) engaging and disengaging the front wheels.

I figure I can use a light-up switch to indicate that it's in the forced 4WD Hi mode. That means I should go with option 3, but without the multi-position selector.

So circuit-wise, my switch would be put a dummy load on the output to the ToD, correct? Any advice on how exactly to wire it, and what to put on the "alternate" circuit?

Lastly, I may have that extra ToD ECU you are looking for, mbeach. I will PM you.

[VehiGAZ]

mbeach, your PM box is full. I sent you an e-mail to your yahoo address instead.

[mbeach]

I cleaned out my PM box, please re-send your message -I have been without home internet for a few months thanks to a bum antenna for my wireless provider. I can't check webmail at work -I could use my phone but it takes forever.

I sat down with schematics and diagrams from the workshop CD last night. After about an hour, I think that I may have worked out a fair solution. One that will settle my desires as well as be easily installed and removed by anyone with fingers. I'm also looking at a 4hi only option as well (VehiGaz).

Just for kicks I looked up the TOD patent, just to see if there was some additional information that would be helpful. Wow. This is one impressive piece of engineering. It's patent #4,989,686 dated 5 February 1991 for those who care. There's tons of info in there, also some diagrams which I would really like to see (logic flowcharts, schematics, graphs of various clutch command signals). The patent also confirms the suspicion of a PWM signal from TOD to clutch, as well as a feedback signal from clutch to TOD.
I could not view the drawings however. If anyone can figure out a way to view them (free), please snatch them and post them here -if it's legal.

[VehiGAZ]

One of our patent attys is on the case right now. I'll post the diagrams if he can get at them and its legal.

[VehiGAZ]

Sweet! I got it. And it's all nice-and-legal. Apparently, all anyone needs is to be set up for viewing .tff files to see the diagrams.

I have it in a 1.9 MB .pdf file, but I can't upload it for some reason. Is it too big? Is the site just having some difficulties? Anyone know?

[bobmumgaard]

Thanks for digging up that patent mbeach. It sure has alot of information about the logic and control systems of TOD. The techs at Borg-Warner put together quite a system here. And the price of the TOD ecu now seems somewhat reasonable. The amazing part is that they didn't put in any user control, it certainly would have been pretty simple compared to the amount of work they put into the system. (maybe it was a liability or marketing issue)

After looking through the patent it appears that the signal to the clutch is PWM and if we cut that wire we need a dummy load (to satisfy the feedback requirements of the TOD computer) Other than that there might be still be more than one way to skin this cat.

I think I am going to apply my modifications after the TOD ecu only, as I dont want to fry such an expensive component. After studying the wiring diagrams for awhile I think it might be relatively simple to alter the dash lights along with the TOD ecu-solenoid signal.

Heres my plan:
I want to control the sysem with a knob, probably 5 positions like mbeach stated earlier: Auto, 2wd, ~15%, ~30%, and 50/50

I want to create a pin-to-pin harness between the wiring harness of the vehicle and the TOD ecu (to minimize cutting stock wires), this would make the modification very reversable and easy to install for most anyone. (literally "plug and play")

I plan to cut the ecu-solenoid wire with a relay, so when the knob is in auto mode everything functions normally. Then when the knob is anywhere other than auto mode the ecu is connected to a dummy load and the solenoid receives the signal from a PWM controller.

The PWM controller will probably be a 555, or 556 timer set up to send a 50Hz signal. The controller will be voltage controlled, with the control voltage coming from the 5 position switch. (with the descreet voltages set up by finite resistances)

To accompilsh the correct display lighting I was thinking about interputing the stock wires for the auto, and 3 bar lights with transistors, wired so when the knob is in auto mode the lights are controlled by the TOD ecu normally, but when anywhere other than auto, the lights are disconnected from the ecu. (this darkens the auto light, and allows the 3 bar lights to be controlled by the knob position; see next paragraph)

Then another set of transistors would connect each individual bar light to the illumination wire depending on what position the knob is in. (this lights the 3 bar lights up correctly)

And the whole thing could be bipassed by a relay controlled by the 4low switch, so if you took the transfer case shifter out of Auto the TOD ecu would have all electrical control

I have a rough circuit diagram drawn up right now, and it seems to look like it should work. I plan on spending some time in the University's electrical engineering lab tonight or tommorrow setting up the circuit and performing tests. I also still need to look at a few things in the VX before I go too far.

[mbeach]

...Took the words right out of my mouth.

That's exactly how I plan on doing it. The dash display is not going to be too difficult either. Thanks all for getting the ball rolling on this -it seems like we might have a plan on the way.

Also, I'm sure that you noticed the 4hi switch on the schematic. I mentioned to vehiGAZ (whose TOD ECU I am buying -just in case) that turning this on with the 4lo switch off, should activate a 4hi input to the TOD. There are only 3 states listed in the diagram, with the missing state being a 4hi.

I'm splicing in a few bullet connectors this weekend. They should make diagnostics/testing easier. I'm also charging up the batteries on the Fluke 'scope as we speak.

I would also like to have some OEM type connectors for a true plug and play option -I had hoped that Newark or Tessco would have these as I deal with them often. No such luck however. Do we have an Isuzu hookup that can I.D. and locate the C43 and C44 connectors?