Engine/Transmission (1994 - 1998) High idle from cruise control

[mwilson]

Scientific Wild Arse Guess.

That's a good one, I'm keeping it for later use... .....



Mike.

 

[Mhannink]

Prior to Jan 1998 any feasible speed control was a manual implementation. There is no computer to control rpms that has fine enough control to work very well. An electronic one that is really a servo and cable is about the only option.

i think its possible. if the computer can control the engine electrically then there has to be a way.



as for the presision, we will find out. i dont think it will be much of an issue. if it is then ill tackle it then. first i need to get it to work at all though. i tried grounding the wire but it didnt work. im thinking i need to power it, i havent tried it yet, i am waiting on a powertrain diag manual that goes into a lot more detail about the cruise control. what i might have to do is put a diode into the setup to keep voltage from backfeeding the computer.



my other thought is get the plug and socket that are used and build a splice harness. that way all the factory wires would be diode protected, then i can bring my own controls in with wiring and switches.

 

[fest3er]

You need to apply +12V to pin 3, ground to pin 4. Then momentarily apply +12VDC to open the vacuum solenoid to increase the engine speed or momentarily apply +12VDC to pin 2 to open the vent solenoid to decrease engine speed.

If you leave the servo connected and tap into the wires on pins 1 and 2, you can 'return to idle' by turning CC off, by tapping the brake pedal, or applying +12VDC to pin 1 until engine is at idle. As you said, you probably want to diode-protect the PCM's pins.

IIRC, ~0. 2 sec is good for 300-500 RPM;; it's pretty quick.

 

[Mhannink]

thanks fest3er, that will help a lot i think. let me run through it and see if i have it right. your saying i need to supply my own power to pin three? i cannot use the cruise control? based on schematics and descriptions in the service manual it sounded like pin 1 or 2 was grounded through the ecm depending on input. is it actually power that is needed?



since i have a switch the is off in the center and momentary up and down im going to hook up both sides so as to adjust the idle as i want. it sounds like i will need another on off switch though to activate the system. that is power pin 3. it should automatically ground through pin 4. then the momentary will let me adjust idle as i see fit.



as for controling rpm, i used to play with stop watches, my reaction time from start to stop was between . 13 and . 26 seconds. if . 2 gives me 3 to 500 rpm then i dont think it will be a problem at all. thats exactly the rpm adjustment i want to make for high idle.



i might have some time tomorrow, ill try and source another plug and wire it up with the diodes and then plug it in, mount my switch, try it out, and let ya'll know what i come up with. i might not have enough time to get it all done but we will see. anyone know where i might find the duetch(i think)(sp?) plug and socket to build my own harness.

 

[cerberusiam]

i think its possible. if the computer can control the engine electrically then there has to be a way.

That is the point I am making, the computer on 12V doesn't control anything about the engine operation. Idle, acceleration, decel, is all mechanical. The CC simply runs a cable that pulls the throttle open.



In an electronically controlled engine speed scenario it is very fine control with a feedback loop working in the millisecond range to raise and lower and maintain rpm's. Quite likely it is a PWM signal that is used as that is the standard.



CC uses wheel speed to maintain its setting, it doesn't care about engine speed so the settings are much courser. You would have to override the computer setting watching wheel speed to get it to work then provide a signal that the CC algorithm would recognize as a hold point or it just keeps going and going and going.....



By the time you do all that you have another electro-mechanical control seperate from the CC, like I mentioned. I just don't think the CC has the logic or fine enough control to set engine speed on an unloaded engine with any accuracy.

 

[ewcmr2]

My EE and AE kids would like this thread. I'd use a stick between the seat and throttle pedal

 

[mwilson]

Alright, I have been sitting back reading this and now I must throw youse guys another theory. .



Never mind trying to make the somewhat dense 50 pin ECM do your bidding, it is quite happy as it is. . (Kinda' like me)...



I have sold several "Bimba" valves over the years, that was the preferred high idle control of Cummins Engines for PTO use waaaaaay before electronics came along.

Searched "Bimba high idle" and two threads from the Cummins Forum popped up.



See post #94 on this one...

Diving in head first with my new 12v - Page 8 - Dodge Cummins Diesel Forum



And this one...

My high idle setup works like a champ! - Dodge Cummins Diesel Forum



He is hooking vacuum to the back side as to pull it in. Clever.



Bimba site here.....

Bimba Manufacturing - Products & CAD



Have a look, I gotta' go back to work, the baby needs new shoes... .



Mike.

 

[fest3er]

... CC uses wheel speed to maintain its setting, it doesn't care about engine speed so the settings are much courser. You would have to override the computer setting watching wheel speed to get it to work then provide a signal that the CC algorithm would recognize as a hold point or it just keeps going and going and going.....



By the time you do all that you have another electro-mechanical control seperate from the CC, like I mentioned. I just don't think the CC has the logic or fine enough control to set engine speed on an unloaded engine with any accuracy.

Ah, but the primary point is not to create a cruise control. Rather, it is to create a high idle controller. And the using the CC servo in an open-loop (or, actually, non-looped) circuit will fit the bill nicely.



The intent is to increase the unloaded engine speed above idle to, say, around 1200 RPM. Clearly a feedback loop would be required to set the engine speed to some RPM and hold it there.



From my simple experiment, I do know that a very narrow 12VDC pulse on the vacuum and vent pins is needed to finely control RPM changes. I never measured it, so I am only guessing that the pulse width probably needs to be less than 50ms based on my bare-hand test.



MH, I think you can use the OEM power system for the servo; I tested it that way once, IIRC. That is, let the PCM supply power to the servo through the brake switch; then tapping the brake, turning off CC or pressing the cancel button with disengage and let the engine return to idle. You only need to access the two control pins (1 and 2). I vaguely recall that applying power to the vacuum solenoid for a half second raised engine speed *far* too much; slapping the two bare wires past each other as fast as I could got about 100 RPM or so increase/decrease. This is the simplest method, but is the least predictable control method.



If an electronics wizard here could design a simple circuit that would output a 12VDC, 5 Hz clock with a 1-10ms pulse width, we might find that that would yield adequate high idle speed control: push and hold + button to increase speed or push and hold the - button to decrease speed, or tap the brake or turn off CC to return to idle. This is a more involved design, but will yield much more predictable control but still lacks accuracy, precision and repeatability.



This of course has no feedback. The P-pump's rack position is fixed (meaning that as the engine warms up, it will speed up). A 6502 microcontroller could be programmed to read engine RPM from the ODB port, allow the user to dial in a speed and, thus, close the loop to provide true engine speed control. This is a much more complex undertaking, but will yield accurate, precise and repeatable RPM setting and could double as a CC (for a manual trans or an auto trans with TCC engaged).

 

[cerberusiam]

Ah, but the primary point is not to create a cruise control. Rather, it is to create a high idle controller.

An good analogy for this would be banging nails with a screw driver. It can be made to work but the efficacy is always the issue. Likewise, you cannot use a CC to effectively control rpms because of their inherent design.



Using the standard CC to set engine speed is not doable because the CC has no input from a CPS. It is concerned with wheel speed and has large window that signal can fluctuate in to operate. CC is dependent on load to somewhat regulate speed for best results. In order to use the CC you would need to fake a signal it can understand based on what rpm you want or you get 2 reactions, it never sets or it never quits revving.



Any type of rpm control must be a feedback system operating on a relatively fast cycle. That is completly opposite of how CC is built to function. Any type of rpm control on these trucks is always a function of electronic fuel control to be effective. The CC is simply not up to the task.



In some 20 years nobody has come up with a feasible electronic system to this, that should be a obvious factor. No matter what you use to actuate the throttle you will still have to build a controller to make it work and be safe. As Mike pointed out, the solution is simple. A Bambino. :-laf

 

[Mhannink]

That is the point I am making, the computer on 12V doesn't control anything about the engine operation. Idle, acceleration, decel, is all mechanical. The CC simply runs a cable that pulls the throttle open.



In an electronically controlled engine speed scenario it is very fine control with a feedback loop working in the millisecond range to raise and lower and maintain rpm's. Quite likely it is a PWM signal that is used as that is the standard.



CC uses wheel speed to maintain its setting, it doesn't care about engine speed so the settings are much courser. You would have to override the computer setting watching wheel speed to get it to work then provide a signal that the CC algorithm would recognize as a hold point or it just keeps going and going and going.....



By the time you do all that you have another electro-mechanical control seperate from the CC, like I mentioned. I just don't think the CC has the logic or fine enough control to set engine speed on an unloaded engine with any accuracy.

i get the impression we are both dancing around the same subject but are not hearing or understanding each other.



yes the computer does NOT control the engine in anyway. it is sending voltage to a servo to operate either a vacuum or vent solenoid therefore adjusting speed.



in an electronic engine the computer simply adjusts the fuel rate untill it sees the rpm it wants. probably works similar to the auto fueling on the common rails.



here is where i think we are having the disconnect. the cruise control is not its own standalone computer. it is just a servo that is operated by the pcm via voltage. not even voltage signals just straight up 12 volts. the whole thing works based on the switches on the steering wheel. with the switch on voltage is sent to the cc servo. it also tells the computer to look for inputs from the switches. when you press a switch a voltage is sent to the computer at different levels depending on the resistance built into the switch. once vehical speed is set the computer adjusts the servo based on the vss input.



here is what i am planning/attempting to do. the cc servo is just that a servo. it does not have any logic algorithims built into it. the electro-mechanical control you mentioned is already installed and hooked up, its the cc servo. all i am doing is sending it signals from another source other than the computer. since the servo operates on 12 vdc there is no need to step the voltage down. probably i dont even need to install diodes but i am working on doing that in order to safe guard the computer. once i have tapped into the wires and i can power pin 3 off a toggle switch connected to the battery, then pin 1 and 2 can be controlled on the momentary switch. the cc servo will act just like any other solenoid. the only question is sensitivity. how fast/much will the cc servo react to input. that is where a controled "shot" might come into play. once i have time to really get into this and play with it i will know.

Alright, I have been sitting back reading this and now I must throw youse guys another theory. .



Never mind trying to make the somewhat dense 50 pin ECM do your bidding, it is quite happy as it is. . (Kinda' like me)...



I have sold several "Bimba" valves over the years, that was the preferred high idle control of Cummins Engines for PTO use waaaaaay before electronics came along.

Searched "Bimba high idle" and two threads from the Cummins Forum popped up.



See post #94 on this one...

Diving in head first with my new 12v - Page 8 - Dodge Cummins Diesel Forum



And this one...

My high idle setup works like a champ! - Dodge Cummins Diesel Forum



He is hooking vacuum to the back side as to pull it in. Clever.



Bimba site here.....

Bimba Manufacturing - Products & CAD



Have a look, I gotta' go back to work, the baby needs new shoes... .



Mike.

mike that is an iteresting device and i think it would work well. my only problem is im trying to do this with as little mod work as possible, and spending as little as possible. wire and butt connectors are cheaper than that solenoid is.

 

[mwilson]

mike that is an iteresting device and i think it would work well. my only problem is im trying to do this with as little mod work as possible, and spending as little as possible. wire and butt connectors are cheaper than that solenoid is.

Just checked flea-bay, some similar Bimbo actuators on there in the $30. 00 range.



The electric solenoid that Freightliner uses on the air supply for the clutch fans would work as the solenoid valve for the vacuum flow.



That will cost $75. 00 or so unless you can get one out of the scrap in the warranty room.



I have much more checking to do but I am leaning towards the Bimbo thing.



Cold weather is coming and I certainly miss not having any high idle on this '97 after running the '06 in the cold weather with it's high idle and engine brake feature.

As I get older, I get colder... ... ...



I'm still cheering for you to find a direct electrical solution, I just don't think it will be constant enough.



You could always prove me wrong and I hope that you do... ... .



A lot of people would like a simple high idle for these things.



Mike.

 

[cerberusiam]

here is where i think we are having the disconnect. the cruise control is not its own standalone computer. it is just a servo that is operated by the pcm via voltage. not even voltage signals just straight up 12 volts. the whole thing works based on the switches on the steering wheel. with the switch on voltage is sent to the cc servo. it also tells the computer to look for inputs from the switches. when you press a switch a voltage is sent to the computer at different levels depending on the resistance built into the switch. once vehical speed is set the computer adjusts the servo based on the vss input.

Yes, we are together up to operation.



From my exprience, admittedly it was a 1st gen but the systems are very similar, the 12 volt signal is allowing vacuum to the servo but not setting any position. Its an energize and supply signal for the solenoid. The position set is done via a solenoid with a PWM signal generated by the PCM. This sets the solenoid position that controls how much vacuum is applied to the servo to pull the cable. All this is done on the ground side of the circuit not the power side so the PCM is protected.



Set, accel, and decel are controlled this way via the VSS with a very broad operation range, at least 10 mph. Resume is a memory position stored in the PCM to go back to.



Essentially you would have to emulate the PCM's CC control algorithm to make a functional rpm control. That is another controller that is not just a simple switch, there is logic and references need to make it funtion correctly. Then, you have the problem of maintaining a set rpm with no engine load and I think at that point you will run into a problem.



The set point needs to be a damped movement. It cannot adjust for minor fluctuations or it will just sit there and cycle make the engine lope back forth several hundred rpm's. More logic to detemrine how much and when to move the solenoid.



In order to achieve what oyu suggest you then need to cut the CC completely out of loop to use it or you wil create problems. Now, if you are suggesting adding another servo and using it then the problems you face will only be control, not integration.



By all means continue your pursuit. Just offering what I have tried and deduced for appraisal.

 

[Nyoest]

Would it be easier to produce a fake vehicle speed? As long as it reads above 35 mph the steering wheel controls would work and shouldnt affect normal operation

Of course this might be completely wrong. In that case I'll be in the truck with my helmet on.

 

[mwilson]

Would it be easier to produce a fake vehicle speed? As long as it reads above 35 mph the steering wheel controls would work and shouldnt affect normal operation



Of course this might be completely wrong. In that case I'll be in the truck with my helmet on.

You could use the guinea pigs from the Geico commercial to generate the fake speed... ...



Row..... Row... ... Row... ..... Row... .....



Plus, it will take 8 months to teach the little chubby one... ... .



For those that haven't seen it yet...

Guinea Pigs Row Tiny Boat - Easier Way to Save - GEICO commercial - YouTube



Sorry, couldn't resist...



Mike.

 

[Nyoest]

You could use the guinea pigs from the Geico commercial to generate the fake speed... ...



Row..... Row... ... Row... ..... Row... .....



Plus, it will take 6 months to teach the chubby one... ... .





Sorry, couldn't resist...



Mike.

Now I gotta clean my monitor! Good one mike. I love that commercial. And those mayhem commercials are pretty good as well.

 

[mwilson]

Now I gotta clean my monitor! Good one mike. I love that commercial. And those mayhem commercials are pretty good as well.

"I'm your blind spot" is my new mayhem favorite. .



Mike.

 

[cerberusiam]

Would it be easier to produce a fake vehicle speed? As long as it reads above 35 mph the steering wheel controls would work and shouldnt affect normal operation.

Sure, that the only way to use what is already there effectively, but... ...



What mph equates to what rom when the trans is in neutral?



What does the VSS signal the PCM is getting look like? Is it analog or digital? What is the shape of the signal?



First problem is locate a machine capable of analyzing and presenting the signal in a form that makes sense.



Next, how to generate the same signal? Now we are back to building a small computer to generate the needed signal and be adjustable enough to tune it on one of those expensive machines so it can match what the PCM expects.



On to the next problem, how to tie this into the existing system and override any possible cutouts programmed into the PCM. Does the PCM need to have a signal telling it the truck is in gear or ??? What other safeguards are built in to keep the operation dummy proof enough somebody can't run over themselves or somebody else?



ROI on doing something that has never worked correctly or buying a solution, that ends being the question.

 

[Nyoest]

Cerb you'll get no argument from me. Designing and engineering is no easy feat I just use my exhaust brake.

It is interesting to ponder and discuss though. It beats the hell out of some first gen threads I've been a part of.

As far as what rpm at what mph I don't think it would matter. The cruise just wants to maintain a set mph regardless of rpm.

I've got my helmet ready.....

 

[Mhannink]

i did have some responses to your comments, i went back to verify what i was going to say and found some more info that i need to read and figure out. ill be back later with a reponse.

 

[JKosten]

I go away a for a day or so, and this thread takes off. Sweet. I have not had any time myself to poke at this system, but I might tomorrow after reading all of these posts.



I will use my PowerProbe III to give a quick pulse to either pin 1 or 2 and see what happens. I just really want to have a warmer truck in the morning.



And a side thought: if a short pulse is needed, can't i just hijack an unused circuit from my remote start (Python 1401) that can do momentary pulses at low amperage?



Thanks everybody for getting interested in this topic Oo.