RESISTORS (no, it's NOT a soccer team....)

[Zorlac]

"Since they only use a a 20k resistor as the filter, I would try some inductance as close to the controller as possible, an easy way, if you have extra lead, would be with Radio Shack 273-104 or if you don't mind soldering (or just wire nuts to try it) 273-102 in series with the tach lead."

I had suggested this awhile back, but after reading that the magnetic pickup was flaky also, I was not so sure the tach signal was the culprit.

Andre's suggestion of a tach signal prescaler (divde by 4) was interesting and two flip flops in series would be a cheap experiment.

The DC signals to the vacuum actuator need to be measured during the failure mode to see if indeed the control unit is actually telling the vacuum actuator to disengage.

 

[bmwhd]

"Since they only use a a 20k resistor as the filter, I would try some inductance as close to the controller as possible, an easy way, if you have extra lead, would be with Radio Shack 273-104 or if you don't mind soldering (or just wire nuts to try it) 273-102 in series with the tach lead."
I had suggested this awhile back, but after reading that the magnetic pickup was flaky also, I was not so sure the tach signal was the culprit.

Andre's suggestion of a tach signal prescaler (divde by 4) was interesting and two flip flops in series would be a cheap experiment.

The DC signals to the vacuum actuator need to be measured during the failure mode to see if indeed the control unit is actually telling the vacuum actuator to disengage.

Bingo. I'm doubtful we'll ever solve this until scope signals of the control unit are taken during a failure and we have access to diagnostic information from Audiovox that decodes the failure mode.

I'm keeping my chips on the marker that reads: "heat is causing something to expand that in turn causes binding on the actuator cable." I bet that if the failure mode could be decoded from the scope trace, it would show a shut down caused by excessive resistance to movement of the cable.

 

[2WheelTiger]

Based on the info about the variability of the inputs, I would have to say that my prescaler (divide by 4) circuit idea has no validity. Coil and mag pickup pulse widths aren't very narrow and their frequency isn't very high.

 

[wfooshee]

"Since they only use a a 20k resistor as the filter, I would try some inductance as close to the controller as possible, an easy way, if you have extra lead, would be with Radio Shack 273-104 or if you don't mind soldering (or just wire nuts to try it) 273-102 in series with the tach lead."
I had suggested this awhile back, but after reading that the magnetic pickup was flaky also, I was not so sure the tach signal was the culprit.

Andre's suggestion of a tach signal prescaler (divde by 4) was interesting and two flip flops in series would be a cheap experiment.

The DC signals to the vacuum actuator need to be measured during the failure mode to see if indeed the control unit is actually telling the vacuum actuator to disengage.

Bingo. I'm doubtful we'll ever solve this until scope signals of the control unit are taken during a failure and we have access to diagnostic information from Audiovox that decodes the failure mode.

I'm keeping my chips on the marker that reads: "heat is causing something to expand that in turn causes binding on the actuator cable." I bet that if the failure mode could be decoded from the scope trace, it would show a shut down caused by excessive resistance to movement of the cable.

BZZZZZZZZZZZZZ!! Thanks for playing!!!!!!!

How would it be Gen-II's only?

 

[ionbeam]

The /4 circuit would have some use to drop a Gen I VSS signal to a usable frequency. But, that still wouldn't address the root cause of the Gen II problems. In suspicion of the coil signal, others have installed the magnetic pick-up to replace the coil signal. No help. This pretty much absolves the 20k resistor in the coil wire too.

Important information (once I talk to Smitty I hope to be able to declare them facts), Smitty has taken 4 failing AVCC units off of Gen IIs and installed them in Gen Is where they seem to work just fine. This means that there is no overt defect of the AVCC units. The AVCC does work in some Gen IIs so it isn't a 'hard' problem caused by changes to the Gen II.

The cable assembly attaches to a vacuum diaphragm that is cup shaped and it is at least 16 sq/in 10.28 sq/in so there is the capability for a lot of force. If the cable were dragging, then cable wouldn't release and the throttle would stay on. The vacuum is controlled by three solenoids; VACUUM, DUMP, LEAK. One of these solenoids may be responsible for Ross's soft cable failure but probably aren't related to the majority of the failures. The vacuum cup is positioned so that with zero vacuum the end of the cup is pressed against the cable end of the servo. As vacuum is applied, the bottom of the cup is drawn toward the connector end of the servo, thereby pulling the cable. When the AVCC is either shut off or sees an error the DUMP solenoid opens and dumps all vacuum. When the servo is activated the VACUUM servo pulls on the vacuum cup and LEAK regulates the position of the cable by managing the vacuum level.

Surprisingly, the AVCC control circuit has a very small number of active components. There is significant +12 volt filtering to stabilize main 12 volt power for the AVCC. All the smarts for the cruise are contained in just one custom made IC fabricated by FreeScale someplace in the Far East. There are three transistors that switch the three vacuum solenoids, and that is essentially it. Every signal that goes into the AVCC's 10 wire connector first go through diodes, resistors and capacitors to clamp and massively filter the signals, then they go straight to the CPU. Given the amount of filtering for the signals something would really have to be out of whack to cause problems.

I hope to be able to get back to tracing out the circuit and analyzing the signal filtering circuits next week. There is also a chance that we will find the failing units to have common date codes of manufacturing or common date codes on the CPU causing some units to be more susceptible to noise than others.

If it begins to look like a noise filtering problem, I will look into building an in-line signal driver circuit that will kill *any* noise for all signals including the ones from the key pad. The key pad signals have not yet been given a clean bill of health. There is a chance that if the key pad were replaced with switches like Brundog did, the units may miraculously work.

Edited 3/11 to correct vacuum cup size.

 

[2WheelTiger]

Maybe I should mention that I'm not using the stock keypad. Mine works fine. My control pad has two waterproof switches with gold contacts. My AVCC is always ON.

 

[BrunDog]

As I am reading this thread (and its predecessors), I keep thinking that why are we speculating, when it is 2008 and modern electrical test equipment like OSCILLOSCOPES and SPECTRUM ANALYZERS readily exist. Slappy stepped up and mentioned it first.

Let's put a scope on a Gen II (from an AVCC known not to work) and a Gen I (with the same ACC known to work) and compare. I'll bet something will make itself obvious then.

I also just had a brain fart: does the different pully have anything to do with this? The AVCC with its PDI control circuit may be freaking out with non-linear throttle pull. Unless, of course, the AVCC cable is being connected to a moment arm off the butterfly rod, upon which it would not matter...

Oh, also, thanks to Alan for pointing out the resistor's 20k rating, not 2k. Here's how I remember the color code. Disclaimer: This does not reflect any personal opinion and was something my electronics shop teacher taught us back in the day. Today, he would be fired for this, I am certain. Starting with zero:

Black Boys **** Our Young Girls But Violet Gives Willingly

-BD

 

[2WheelTiger]

You forgot the word "young" for yellow.

And the modern version is "Bad Boys........."

 

[v65]

Maybe I should mention that I'm not using the stock keypad. Mine works fine.

Gunny, I used Brundog's switches.

 

[ionbeam]

I keep thinking that why are we speculating, when it is 2008 and modern electrical test equipment like OSCILLOSCOPES and SPECTRUM ANALYZERS readily exist. Slappy stepped up and mentioned it first.
Let's put a scope on a Gen II (from an AVCC known not to work) and a Gen I (with the same ACC known to work) and compare. I'll bet something will make itself obvious then.

I also just had a brain fart: does the different pully have anything to do with this? The AVCC with its PDI control circuit may be freaking out with non-linear throttle pull. Unless, of course, the AVCC cable is being connected to a moment arm off the butterfly rod, upon which it would not matter...

I currently have a 4 channel digital O'scope with which I can take screen shots. What I don't have is a FJR in my garage and don't know of a failing AVCC install in New England. Once my FJR does come home (next week???) it has been defarkled and I have to reinstall 90% of the electrical and mechanical doodads. I do have Ross's failing AVCC and I'm working with that. My business work load has been relentless recently, I hope to get some lab time later next week where I can get back to evaluations. It will be interesting to see if I can make a bench setup that mimics a dynamic road situation.

The non linear throttle pulley may indeed cause a negative effect resulting is shut-off. The AVCC is setup to recognize when there is excessive error between command (pull the cable) and response (rate of change in the timing pulses from either the coil or magnetic pickup) and when this happens it is programmed to shut off. What I can't explain is why this would be OK when cold but cause a problem after 15 minutes or more later. It is easy to make a test for this though. The non linear aspect of the pulley is only for a specific range of throttle openings. Once the AVCC starts to fail, run 'er up to something like 50 mph in 1st gear and set the cruise while the pulley is out of the non linear area. If it works then, we know that the pulley profile is at least partly responsible. Um, why do some Gen IIs work fine?

 

[wfooshee]

The non linear throttle pulley may indeed cause a negative effect resulting is shut-off. The AVCC is setup to recognize when there is excessive error between command (pull the cable) and response (rate of change in the timing pulses from either the coil or magnetic pickup) and when this happens it is programmed to shut off. What I can't explain is why this would be OK when cold but cause a problem after 15 minutes or more later. It is easy to make a test for this though. The non linear aspect of the pulley is only for a specific range of throttle openings. Once the AVCC starts to fail, run 'er up to something like 50 mph in 1st gear and set the cruise while the pulley is out of the non linear area. If it works then, we know that the pulley profile is at least partly responsible. Um, why do some Gen IIs work fine?

The "non-linear" is in the grip, not the throttle pulley, isn't it? I'm thinking the CC would never notice, with it's cable and chain attached to the throttle pulley.

 

[2WheelTiger]

C'mon guys! How does the AVCC know the throttle response is non-linear? It simply senses that accel or decel is necessary with no attention paid to the power characteristics or the throttle response of the motor. Horsepower and torque are not linear by rpm.

 

[ionbeam]

In the '06 and '07 FJRs the pulley between FI throttle bodies 2 & 3 is not round like the Gen I FJRs. If you search out SHIM and filter by member name Fred H. you will find some threads where he 'fixes' the problem by putting a copper shim between the throttle body pulley and the throttle cable.

The G2 hand throttle tube has a cam which is profiled to be the inverse shape of the non-linear part of the FI throttle body pulley, thereby canceling out the throttle body's cam curve, and creating the equivalent of a round pulley. :wackosmiley:

Fred 'fixed' the problem by making the non round throttle body pulley round with a shim, G2 fixes the problem at the other end with their hand throttle cam.

 

[ionbeam]

C'mon guys! How does the AVCC know the throttle response is non-linear? It simply senses that accel or decel is necessary with no attention paid to the power characteristics or the throttle response of the motor. Horsepower and torque are not linear by rpm.

The AVCC doesn't directly know that the throttle response is non linear, it only knows that the controlling action of the cable is too divergent from the coil signal response. This is how the AVCC automatically releases the throttle when the clutch is pulled in. When the clutch is pulled in there is a sudden, very rapid increase in rpm even though the throttle cable was not commanded to move. This error condition causes the AVCC to activate the DUMP solenoid and vent all vacuum. You have to either re-set or resume again after the shift.

The AVCC CPU is based on an electric motor PID loop controller (Proportional, Integral, Derivative, a mathematical control algorithm), which is an error feed back system. When you press the SET button the CPU takes a frequency count of the coil signal. It then starts to pull on the throttle cable until it returns the coil frequency back up to the set value = no error. As you go up a hill the coil signal starts to slow down; the AVCC pulls on the throttle cable until the frequency comes back up to match the set value. As you go down hill the coil signal starts to be faster than the set frequency; the AVCC starts to release the throttle cable until the frequency matches. Part of the magic of the PID loop is the way it controls: GAIN - how fast the system responds; STIFFNESS - how aggressive it is in trying to maintain matching values; DAMPENING - where it applies a bit of 'softness' to keep the loop from oscillating, you would feel this as surging or hunting to maintain a regular speed. When SET is first pressed you may detect a small drop in speed, then it comes back up to set speed. This is the AVCC easing the throttle pull up to match the coil frequency.

When you set the silly little DIP switches for light vehicle, high power you are telling the AVCC what terms to use in the PID loop control equations. When you set the number of cylinders, it is supplying info on what the PID controller can expect for a frequency range. There are a number of command/response rules that define when the process is getting out of hand and will result in an immediate shut-down. If you want a wild ride, tell the AVCC that you have a heavy, low powered vehicle. This will cause the AVCC CPU to aggressively pull on the throttle cable, have abrupt throttle inputs and may even oscillate instead of producing steady speed. There is a very good chance that the FJR's speed response will be so rapid and strong that it will exceed program limits causing the AVCC to shut off.

 

[BrunDog]

You forgot the word "young" for yellow.
And the modern version is "Bad Boys........."

Oops. Edited. Thanks.

Got so hung up on whether I should post the old school version, I forgot the damn Y!

-BD

 

[bmwhd]

"Since they only use a a 20k resistor as the filter, I would try some inductance as close to the controller as possible, an easy way, if you have extra lead, would be with Radio Shack 273-104 or if you don't mind soldering (or just wire nuts to try it) 273-102 in series with the tach lead."
I had suggested this awhile back, but after reading that the magnetic pickup was flaky also, I was not so sure the tach signal was the culprit.

Andre's suggestion of a tach signal prescaler (divde by 4) was interesting and two flip flops in series would be a cheap experiment.

The DC signals to the vacuum actuator need to be measured during the failure mode to see if indeed the control unit is actually telling the vacuum actuator to disengage.

Bingo. I'm doubtful we'll ever solve this until scope signals of the control unit are taken during a failure and we have access to diagnostic information from Audiovox that decodes the failure mode.

I'm keeping my chips on the marker that reads: "heat is causing something to expand that in turn causes binding on the actuator cable." I bet that if the failure mode could be decoded from the scope trace, it would show a shut down caused by excessive resistance to movement of the cable.

BZZZZZZZZZZZZZ!! Thanks for playing!!!!!!!

How would it be Gen-II's only?

Because the Gen-II's have a different arrangement under the tank including the heat shield. Space is ultra tight under there and the actuator cable has to make a rather tight turn. If the support bracket isn't just so, the cable can bind with or without heat. I'm suggesting that perhaps something is heating up that impacts the actuator cable's physical resistance. We've eliminated damn near everything else.

Still, all speculation until a failure mode can be observed on a scope.

 

[wfooshee]

"Since they only use a a 20k resistor as the filter, I would try some inductance as close to the controller as possible, an easy way, if you have extra lead, would be with Radio Shack 273-104 or if you don't mind soldering (or just wire nuts to try it) 273-102 in series with the tach lead."
I had suggested this awhile back, but after reading that the magnetic pickup was flaky also, I was not so sure the tach signal was the culprit.

Andre's suggestion of a tach signal prescaler (divde by 4) was interesting and two flip flops in series would be a cheap experiment.

The DC signals to the vacuum actuator need to be measured during the failure mode to see if indeed the control unit is actually telling the vacuum actuator to disengage.

Bingo. I'm doubtful we'll ever solve this until scope signals of the control unit are taken during a failure and we have access to diagnostic information from Audiovox that decodes the failure mode.

I'm keeping my chips on the marker that reads: "heat is causing something to expand that in turn causes binding on the actuator cable." I bet that if the failure mode could be decoded from the scope trace, it would show a shut down caused by excessive resistance to movement of the cable.

BZZZZZZZZZZZZZ!! Thanks for playing!!!!!!!

How would it be Gen-II's only?

Because the Gen-II's have a different arrangement under the tank including the heat shield. Space is ultra tight under there and the actuator cable has to make a rather tight turn. If the support bracket isn't just so, the cable can bind with or without heat. I'm suggesting that perhaps something is heating up that impacts the actuator cable's physical resistance. We've eliminated damn near everything else.

Still, all speculation until a failure mode can be observed on a scope.

I can see that, and I had come back after to add that statistically we'd have just as many stuck throttles as non-engaged CC if it were a cable friction problem, but another post after mine had already pointed that out. There's not been a single report of CC-derived (or other, for that matter) stuck throttle that I've seen.

 

[jayke]

You guys are thinking too hard. I work in the electronics industry. When my techs get stuck I'm the one that has to solve the problem. It's human nature to think electrical problems are really complex, usually it's something simple.

Remember Smitty said that the Audiovox cruises' that didn't work on the Gen II's worked fine on the Gen I's. To me that rules out the Audiovox Cruise as the root cause. Since Smitty, a proven installer, has had problems we can rule out installation error.

The Audiovox Cruise has only two connections to the FJR not counting the power connections. The blue tach wire and the purple brake wire. Several forum members with non-working cruise on the Gen II tried using VSS off the rear wheel. The cruise still didn't work. To me that rules out the blue tach wire as a cause.

All that's left is the purple brake wire. I know from personal experience if the purple wire is not grounded or with the wire just sitting there the cruise won't turn on. If there is any voltage present on the purple wire the cruise won't turn on either.

My conclusion is a intermittent ground or voltage on the purple brake wire.

If I am correct , connecting the purple wire to chassis ground should make the non-working cruise operate.

A big caution here... the cruise may work with purple wire connected to chassis ground but it won't shut off unless you pull in the clutch or turn the cruise off at the keypad. Please be very careful if you try this!!! Do this only as a test ... operate the bike with the purple wire connected to ground only long enough to see if the cruise works

If the cruise now works add a relay to switch the purple wire from ground to +12VDC to shut the cruise off normally before you ride in traffic.

Edit 3/10/08: Don't do this !! Apparently grounding the Purple wire to the chassis has been tried and the cruise still doesn't work. Read the posts below.

 

[jayke]

Check out Zippers post from the Audiovox poll post. (He posted before I did)

I was saved today from voting No by the wisdom of the collective. I installed a CCS100 almost 2 years ago and couldn't get it to engage. I spent so many hours on the install and troubleshooting that I didn't have the heart to rip it off and toss it. That's a good thing, because I just read in this thread (or someplace that led from here) that a Back-Off unit will cause my issue exactly.
You got it, I disconnected the purple wire from my the wire coming from my rear brake switch, grounded it (the purple wire) and presto I now have a functioning CCS100!

I'll add a relay tomorrow to automatically take care of switching the ground 'on' when I'm not braking and I'll be all set.

Many thanks to everyone who participates in the forum, you're an invaluable resource.

 

[kaitsdad]

All that's left is the purple brake wire. I know from personal experience if the purple wire is not grounded or with the wire just sitting there the cruise won't turn on. If there is any voltage present on the purple wire the cruise won't turn on either.

Can you give us more specifics? What does 'Wire just sitting there" mean?

The purple wire is connected to the brake light wire - not grounded. The idea is that it disengages the unit when the brake is applied.

So - if you ground it, you are negating the main safety feature?

EDIT - OK - I figured out what you're talking about - having a LED bulb or some other device that's injecting power onto the brake light circuit, and it creating a problem. In other words, a non-standard situation in where a user added accessory is creating the problem.

This is not the problem we are chasing with the Audiovox poll nor in this post.

While I applaud the use of a relay to overcome the problem or the use of a diode to do the same thing, this post is about resistors.