TOD discussion

[bobmumgaard]

Tone-
Your axle breaking could shed alot of light on this whole subject. I did a search but couldn’t come up with anything. What exactly happened? I assume it was on a dyno pull. So it was probably in third gear and strapped down with no power to the TOD ecu. Where did the axle break? What rpm? (and thus speed of rear wheels, and speed differential between front and rear) What did the break point look like? Was the break plane smooth, jagged, notched in a pattern, did it look like it snapped or did it deform ductilly? I have a few theories about how it could have happened but they depend on the exact situation, and I could be absolutely wrong on them.

[mbeach]

I know where you're going with that thought. I considered it as well. It would be extremely difficult to do a failure analysis at this point, with no photographic or video evidence of the break. I also would like to see the fracture, measure the diameter and wall thickness of the driveshaft, and see if I could come up with a torque value that could have produced the failure.

I can guarantee that it was more than 15% of 230 lb/ft for sure.

[VehiGAZ]

So I've been wondering (as I've been learning)...

The normal operation of the ToD system on a slippery road will transfer torque up front to keep the front wheels spinning about the same as the back wheels, right?

Then it backs off of that torque bias when you come to a stop - partially because you don't need it when stopped, but also (and I am speculating here) to protect itself if you make a sharp turn from that complete stop - in that case, the full 50/50 bias setting may not allow enough rotational speed differential between the sets of wheels and you'd get wheel spin from an axle unwinding, you'd slip the transfer case differential clutch (if you were on dry, grippy pavement that didn't allow for any wheel spin). It makes sense that you wouldn't want any torque transfer to the front while turning at very low speeds (i.e., when starting from a stop).

Well, if we come up with a switch & circuit that will tell the transfer case differential clutch to keep applying a full 50/50 torque split all the time, won't we be asking for trouble in those types of situations? On a snowy road, an axle can unwind pretty easily by letting a tire slip, but if you were turning onto a clear road, wouldn't that stress on the transfer case differential still be there? And wouldn't that be not-so-good?

I freely admit to speaking in total ignorance on this subject (as I said, I've just started learning about how differentials and transfer cases work) - I'm just trying to follow along with the other discussions.

I have come to one conclusion, though - the Trooper ToD conversion in the first post is not as easy to implement for the VX.

Any thoughts?

[mbeach]

The Trooper TOD conversion (as documented and implemented by Tad (Nazrat)) is a fine piece of work in its own right, however, a few of us have not used it because it has a critical flaw.
The e-clutch uses a very low resistance coil, 1-5 ohms maximum. This coil was designed for a PWM input from the TOD computer. By locking it up with straight DC voltage, you risk burning out the coil because it no longer has the ability to modulate its holding power (at the millisecond rate, which is not detectable). You are simply forcing it to stay locked. I had to look this up in order to better explain it. Here's a quote that I found that seems to put it in better terms:
"...This is why PWM drive is the best for solenoids. A solenoid designed for PWM drive will have quite a low coil resistance. If DC was applied to it, the coil would heat up and burn out. It is the inductance of the coil on a PWM signal that restricts the current to safe values. Of course this means that if for any reason the plunger gets stuck in the "out" position, where the inductance is low, the coil may burn-out..."

Now, this has not happened yet, but I fear that prolonged use in 4hi could cause this to occur when using the 'power mod'.

What I am proposing, and I suspect Bob is doing as well, is to modify Tad's work by replacing his 12V signal with a PWM signal. We are NOT reinventing the wheel here, we are simply making it safer and more flexible.

By using the PWM generator, we can vary this pulse/duty cycle to provide a variable torque split as opposed to the all-or-nothing 4hi. We can also tie this controller into the dash display with Bob's transistor arrangement.

As for your 4hi question, Yes, you are correct. 4hi should not be used on dry pavement. You should leave the TOD in the AUTO mode for such applications.

I have started streamlining the circuit, and should have a paper copy done by Monday. I have drifted towards a single (dash mounted?) pushbutton switch, labeled MANUAL that will turn off the TOD Auto mode. When MANUAL is engaged, the PW can be varied from 0-88% by means of a knob, mounted next to the winter/power buttons. All with appropriate dash displays.

Just keep in mind, we cannot request LESS torque than what is required for the driving situation.

[VehiGAZ]

Okay, so I'm learning something.

So the 555 you speak of is some sort of signal generator capable of creating a PWM signal? Can it modulate it too?

[mbeach]

We can vary the duty cycle of the square wave produced by the timer circuit by changing the control voltage that is applied to the 555.

In theory, we can have an infinite amount of variation to the solenoid control signal this way. In reality, I would like a (now) 3 position (knob) switch, that will correspond to the dash indicators as well as the 15/30/50% torque values that we want to select.
Push the MANUAL button (maybe use the spare foglight switch that I have laying around), and turn the knob all the way to the right for 50% (or 4hi).

If I should forget and leave it in MANUAL mode (signified by the light on the switch, and the lack of an AUTO displayed on the TOD display), it should be alright as long as the speed differential (aforementioned donuts and dynos) is not so great as to activate the ball/ramp mechanism. After all, the truck is in 15% mode 99% of the time anyways.

EDIT: My first sentence was in response to VehiGAZ' question.

Bob,
I'm still fairly certain about the ball/ramp's ability to transmit large amounts of torque to the front driveshaft. I thought the same thing about the redundancy of the e-clutch, but I believe that its primary purpose is to proportion/modulate the amount of torque to the front, rather than to simply link the two shafts together.
I don't believe that mere residual friction among the clutch packs could be responsible for the broken driveshaft. A clutch pack would've slipped long before steel failed.
Also, why would the Trooper's designers implement a Shift-on-the-fly mechanism if they could simply release the e-clutch and achieve the same result?

[bobmumgaard]

I've spent a few hours studying the patent and look at the maintenance manual and I think I have a pretty good idea about how this thing is assembled and how it operates. Though I would love to get one on the bench to take apart and examine.

About torque:
mbeach is correct that you can only transmit as much torque as something (say a wheel) can hold. This means that if the wheel was on ice, you could send torque to the wheel until traction faltered and then the wheel would just spin, So if you are sending torque to the rears and they are spinning, trying to send more torque causes the wheel to spin faster, nothing more. Now if the wheel is locked, say the VX is strapped down on a dyno and you send alittle torque to the front, nothing will probably happen. If you send more and more torque eventually the wheel will either start slipping against the ground, the straps will snap or you will break something in the drivetrain, say an axle.

Construction of the transfer case:
It seems to me that the clutch works very much like a Cusco or Kaaz limited slip differential. This means that they do allow a difference in rotational speed front to rear. And this makes sense because otherwise it would be the equivalent to a true mechanical lock. If the front and rear HAD to spin at the same speed it would be like driving in 4Low all the time, lots of scrub.

The transfer case appears to take the output from the engine and transmits it to the rear wheels through a shaft, much like there is just a driveshaft and no transfer case. This means the rear is powered all the time, no matter what and always spins at the same rate as the transmission output. (unless we activate the planetary gearset for 4low) On this shaft is a sprocket, this sprocket is connected by a chain to another sprocket on the front driveshaft. Inside the first sprocket (the one on the mainshaft) is the clutch. The clutch appears to consist of 13 friction disks. 6 of these are attached to the mainshaft, (and thus spin at the speed of the rear wheels) and 7 are attached to the sprocket (and thus spin at the speed of the front wheels) These disks alternate, one for the rear next to one for the front next to one for the rear and so on..

There is then an inductive coil that creates a magnetic field when we send current through it. This coil squeezes all the friction disks together. So when we power this coil up the squeezing action increases the friction between the front and rear shafts, this means we take some of the torque we were sending to the rear and send it to the front. Now the front might not move, if the wheels were in a tough spot then the clutch would just slip. Squeeze harder and you get more friction=more torque. Keep squeezing and eventually the front wheels will either move, something will break or you will be squeezing your hardest and the friction disks in the clutch will still just slip. However, in normal situations, if you squeeze the front will start to spin at same rate as the rear, and if it doesn’t, the TOD ecu just squeezes harder until they spin at the same speed (or the ecu is happy with the speed difference)

The system is the same in a clutch type limited slip differential like Cusco or Kaaz, we just modulate the squeezing electronically; they do it with elaborate shims. And the same principle as the multidisk clutches you see on high torque cars.

In fact, this is almost the same system we see on any car’s clutch. In that case, your foot controls the squeezing, and the clutch slips, it is still transmitting some torque even though it’s slipping. Hopefully the clutch eventually stops slipping and is then transmitting 100% of the engine’s torque to the wheels.

The ball ramp system appears to be there to make sure that the clutch doesn’t self activate. (Imagine if it did at 75 mph, dangerous liability there) and it doesn’t appear to be able to transmit very much torque, atleast not enough to break a driveshaft. Though I am still looking the whole ramp thing over, but this is my initial assessment, though I could be wrong. In addition to its small size, I don’t think that the ball ramp system is involved with any automatic torque transfer because if it was the clutch would be a VERY expensive redundancy.

So… the moment we are waiting for….can we get true 100% 2wd? In my opinion not 100%, but maybe 95%

Because the friction disks are always in some contact, and there are so many of them they will always create some friction, even when there is no squeezing from the solenoid. How much friction (and thus torque) probably depends on how fast everything is spinning.
This means that if you take off the front driveshaft, the front output of the transfer case will still turn. Also, if you break a CV joint and have the TOD disconnected everything will still spin due to this residual friction. An interesting experiment would be to disconnect the front driveshaft and connected the transfer case output to a brake (like an engine brake used my manufacturers to dyno engines) and measure this torque.

A more practical experiment would be to put the TOD control we are talking about in a VX (I think we are on the right track with electronic systems discussed above). And take the VX to a brake based dyno. In fact, dynopack is the most common brake based system. This dyno attaches a large brake to each hub of a 4wd car and applies a variable resistance to the wheel. This resistance measures the torque each wheel is putting out. (most dynos are inertia based (use rollers) and measure power and then infer torque, the opposite strategy as a load (brake) based system)

If someone really wants to find out the exact characteristics of the TOD system, find a dynopack system. They are used for rally cars most often, so the best bet is to find a place that works on STi’s and Evos. Unfortunately for me I’ve never located a dynopack system within a few states of Nebraska.

Load based dyno info:
http://www.mustangworld.com/ourpics/News/nowheel.htm
http://www.dynopack.co.nz/

Limited slip info:
http://auto.howstuffworks.com/differential8.htm

Normal car clutch info:
http://auto.howstuffworks.com/clutch.htm

About the mechanical lock
When we put the transfer case in 4Low it slides a mechanical lock between the front and rear axles. (actually the mainshaft and sprocket) In this situation the clutch is completely out of the loop. This means both axles rotate at the same speed no matter what. So if both front wheels where on the ground and the rears where not (imagine the VX hanging off a cliff, scary aint it?) then the car should be able to pull itself up (assuming enough traction and engine power) in this situation the fronts receive 100% of engine torque because of the mechanical lock (there can’t be any slipping) and because the rears cannot transmit any torque because they would be spinning against air at the same rate as the fronts pull against the ground.

Well alright, sorry this was so long, just trying to hash out the exact nature of the things we are working with here.

And again, I could be completely wrong on all this. What are your guy’s thoughts?

[bobmumgaard]

I agree with VehiGAZ and mbeach that 4hi would be bad in dry situations, and the TOD handles theses with ease. It is the times when TOD is overactive, trying too hard on snow and ice that we are looking to correct. Not to mention the need of the mechanically inclined to have COMPLETE control over the machines they use.

So the 555 you speak of is some sort of signal generator capable of creating a PWM signal? Can it modulate it too?

The 555 is a timer chip. You can use it along with capacitance, resistance and a transistor to create a PWM signal. This signal is varied by varying a control voltage into the 555 chip. So yes it is exatly what you speak of.

I agree with mbeach's assesment of the trooper mod. The coil is not meant for the prolonged current a constant 12V source provides. And a PWM signal would be more inline with an OEM solution.

I was working toward a knob with 5 poles (and thus 5 positions): Auto, 0% torque, 15% torque, 30% torque and 50% torque. This makes wiring the lights up easier (since youre working with poles instead of an infinte knob). And it can be wired to elimate the manual button.

I really need to clean up my circuit diagram so I can post it.

[mbeach]

Poop.
Posts are out of sequence.

[mbeach]

Here's what I can do:
-A MANUAL/AUTO switch. Tie it in to the same place as the Trooper Power Mod has it located.
-Replace the relay in the power mod with an appropriate timer circuit/PWM generator.
-Implement a knob, that by varying the values (vdc) applied to the timer circuit, will vary the PW of the output signal to the e-clutch.

Things that I need.
-Values, voltage and duty cycle, from the TOD to the e-clutch. This weekend I am splicing in taps to some (o.k., a bunch of) wires. I will be driving around with a Fluke 97 on my passenger seat for a few hours.
-I also want to gather all of this data for each condition (15/30/50/4lo/N) from every conceivable input source (TPS, ABS, Front and rear VSS, etc.). I also want to peek at the 4hi and 4lo switches that I see on the schematic.
-A plan to control the dash lights. Transistors is the way to go for sure. I need to think this through later though. Must collect data first.

I'm not going to promise anything, but I should have a handle on those duty cycle readings by monday.

[VehiGAZ]

I think you two are on the right track here! Thanks for the layman's explanation of how the transfer case works, and what happens when you shift into 4WD-low, Bob. VERY helpful to me!

So I think we should be fine if we come up with the circuit that allows us to generate the right signal to the ToD transfer clutch. It is used to having some current on it all the time - but it's best to mimmick the "natural" signal it gets - VDC is probably a bad idea in the long run.

I think my concerns about having the full-4WD-Hi activated constantly are not a big deal as long as we pay attention to what we are doing. Whatever stresses we put into the drivetrain will undoubtedly release themselves in that clutch pack.

Alrighty then - you EE types figure out that circuit. It would be nice if we could find extra ToD ECU wiring harness connectors - that way, it could be set up to plug right in in-line.

Thanks again!

[bobmumgaard]

I'm still fairly certain about the ball/ramp's ability to transmit large amounts of torque to the front driveshaft. I thought the same thing about the redundancy of the e-clutch, but I believe that its primary purpose is to proportion/modulate the amount of torque to the front, rather than to simply link the two shafts together. mbeach

After further review, there is no indisputable conclusive evidence that my analysis of the ball ramp thing is right and the play stands as called on the field by mbeach

Afterall, the techs at Isuzu went through alot of trouble to put the ball-ramp mechanism in there. And SOMETHING had to cause Tone's broken axle. And I certainly don't think it was residual friction from a few ~5 inch disks.

I think someone should call Merlin and see about finding the supplier for those connectors, they would make this a fool-proof modification. Right now I am looking at tapping or modifying 9 different wires (though i am trying to reduce this #). I talked to a contact at Bosch but they weren't very helpful.

mbeach, I drew up my circuit diagram on SmartDraw and am looking it over. I think the next step is to find those PWM values for the solenoid, and Voltage values for the other important wires And while youre in there, some data on the real currents the solenoid is drawing might be helpful to us when it comes time to select timer components. If I had a portable scope I'd be in the VX confirming your numbers but right now I am using aTektronix TDS 210 that is chained to the lab.

VehiGAZ, glad to help, and I hope I got everything right on my analysis.

I'm an MEtech working in an EE's position. I think Bob's an ME as well. That's why we're overanalyzing everything

If only ME's built everything: We'd run out of titanium, parts would start at 1million dollars, everything would fix itself and the world would be bombproof.
Mechanical Engineering: the definition of 8 gallons in a 5 gallon bucket.

[Tone]

Front driveshaft snapped at about 90MPH on the dyno during my second pass. It broke about 2 inches out the the transfer case (required replacing it also) and was a bit jagged. Made a loud pop when it happened.

[mbeach]

I'm working on uploading photos and data (spreadsheets) now. All that I can say is AWESOME.
Everything worked like a charm (a tribute to my cable-building and soldering skills ).

There were some VERY interesting occurences during testing.

Things of note (to get us started before the pictures are up):
-The frequency of the PWM signal to the e-clutch is 50 Hz. Duh, 20 milliseconds. I remember reading that in the literature for the VX.
-In "2wd" mode, below 15% torque to front, the TOD is sending current to the e-clutch. The PWM signal = 3.2V. This is a 22% duty cycle based on my input voltage.
-My voltage into the TOD was 14.0V even. This will skew results a bit, but it still works. Plus the math is easier.
-The pulse width varies with speed, more speed = greater duty cycle, even if it's still in 15% mode. At 30mph, the PWM signal = 5.5V. This is a 39% duty cycle.
-Anything above a 68% duty cycle will put the truck into 4hi (spinning tires, low speed). This signal should be increased proportionately with respect to vehicle speed. At 68%, we are seeing only 9.5V at the e-clutch.
-Contrary to my reading in the patent description, ABS does NOT turn off the 4wd.

I have plenty of photos, plus some screenshots of the scope and TOD in action. I'm trying to put in into my gallery along with an excel file with my data collection.

[mbeach]

O.K., this is a bit unrehearsed.

This was the cable that I built to interface to the TOD:
[img]
http://www.vehicross.info/gallery/data/500/thumbs/P2130001.JPG[/img]

This was the cable map, it shows what parameters were monitored:


Here is the TOD connector:
[img]
http://www.vehicross.info/gallery/data/500/thumbs/TODview.JPG[/img]
I spliced into this with my cable. Connections were made inside of the plugs (which are impossible to dismantle by the way. If anyone knows how to take these things apart, please let me know. I was forced to solder INSIDE of the holes underneath the white clips.)

TOD interface cable, ready to go.

Here's the road we used for testing. Old Murphy Dome Rd., about 18 miles north of Fairbanks, Alaska. This road starts out as packed gravel, frozen with patches of hard snow and ice. It quickly peters out into a narrow trail, all ruts and deep snow, about 5 feet wide. Yes, I scratched the truck -a lot.


This is my lovely co-driver, whose patience was sorely tested this day. "...I said SPIN the tires woman!"




This is a scope shot taken at idle, in Neutral (transmission). Notice the pulse width.

At 60mph (on ice) in 15% mode:

Random scope shot:

I need to attach an .xls file. how do I do this?