Engine/Transmission (1998.5 - 2002) Anatomy of an APPS

[Gary - K7GLD]

Originally Posted by rweis

What do you think about using 2 OMRON D2VW-5L1B-1MS Long Lever, with wire leads, mounted side by side (total package width . 82") over in the area where the TV cable for us ETC folks connects?

As Karl pointed out, one switch should do the job just fine - and thanks for the pointer - I just ordered one from Mouser to use instead of the far cheaper/poorer ones mentioned earlier - those were of poor quality, with far too much slop in the actuating lever, and after digging another VERY good quality one outta my junkbox yesterday, found it was too large to install reasonably in the mounting area I plan to use.



By the way, the one I ordered from Mouser is #653-D2VW-5L2A-1MS - which is exactly like the one you supplied a number for, and was pictured, while the one your number matches is a different type than displayed, with a longer lever and NO roller tip. I figure the shorter the lever, the less potential mechanical slop and erratic actuation.



This all points up earlier comments from me about the cost of rounding up quality parts - the rotary potentiometer I figure to end up using was a $19 item - but FAR better quality than cheaper types that can be had for around $2 - or the pair if inline types that were also cheap in quality as well as price.



It can be difficult at times to obtain parts at a decent mix of price and quality, in the size and shape needed - and in single piece quantities - I located some very nice, but expensive, pots rated for 10 million operations - but only available in large quantity wholesaler lots. The one I ended up with, is a 2 million operation precision type - not quite as good, but 1/5th the price, and easily located and replaced if required.



With all that in mind, I'm focussing on installing virtually ALL the needed parts - the potentiometer and switch, as well as the filter elements, on the bellcrank side of the bracket. That allows far easier and convenient initial testing, adjustment, and eventual maintenance or replacement as needed.

 

[rweis]

The long lever so I can trim the length back to what I need, rather than trying to figuer out how to make it 1/8" longer.



Only need 1?



Then put the normally ???? (open? or closed?) on the idle validation, the normally ???? (open? or closed?) on the throttle validation, the 3rd wire is the ECM voltage feed.



D2VW-5L1B-1MS, D2VW-5L2A-1MS now is that scary or what? The only difference is the type of lever.



I could use some lessons in building a noise filter for it.



Bob Weis

 

[Gary - K7GLD]

The long lever so I can trim the length back to what I need, rather than trying to figuer out how to make it 1/8" longer.



Only need 1?



Then put the normally ???? (open? or closed?) on the idle validation, the normally ???? (open? or closed?) on the throttle validation, the 3rd wire is the ECM voltage feed.



D2VW-5L1B-1MS, D2VW-5L2A-1MS now is that scary or what? The only difference is the type of lever.



I could use some lessons in building a noise filter for it.



Bob Weis

Bob, here's a pic showing the mechanical/electrical schematic of what the IVS amounts to:



#ad




That IVS portion at the lower half of the diagram is essentially what you would have with a SINGLE MS like you are considering, and I just sent for.



You're forgetting that the "voltage feed" as used in the DC APPS is there ONLY to operate the IVS solid state circuit - but a MECHANICAL MS eliminates that need entirely - and ALL the MS does (or NEEDS to do!), is MECHANICALLY open/close the switch contacts to signal the ECM as to on/off idle condition!



My noise filter will be four . 01 uf ceramic bypass caps, and a pair of 47 ohm resistors - very simple, and purely to control any IVS switch contact "spikes" on open/close that might confuse the ECM - and they're also helpful in keeping RF from my mobile Ham gear out of the ECM as well...

 

[Batphreak]

Here's the SAE spec, hope it helps a little.

SAE J1843

 

[KarlC]

Batphreak,



Thanks a lot for making this available. This is a big help and confirms a lot of what we were guessing about the APPS operation and the mechanical and electrical interfaces.

 

[KarlC]

The long lever so I can trim the length back to what I need, rather than trying to figuer out how to make it 1/8" longer.

Only need 1?

Then put the normally ???? (open? or closed?) on the idle validation, the normally ???? (open? or closed?) on the throttle validation, the 3rd wire is the ECM voltage feed.

Looking at the bottom of page 3 of the datasheet for the microswitch it has three wires: red (normally closed), blue (normally open), and black (common). The normally-closed wire connects to common wire when the lever is released, and the normally-open wire connects to common wire when the lever is depressed. The DC APPS circuit electronically connects pin 6 to pin 1 when at idle, and pin 2 to pin 1 when not at idle (I checked this on mine). Since the switch lever is going to be depressed when the bellcrank is at idle, I would connect the NO (blue) wire to pin 6 of the connector, the NC (red) wire to pin 2 of the connector, and the common (black) wire to pin 1.

It looks like the DC APPS was made to meet the the SAE spec that Batphreak kindly posted (many thanks, ). According to the spec, the IVS should switch at a point somewhere between 3% and 10% of the bellcrank range. This gives quite a bit of leeway, ideally the microswitch should actuate just a bit before the bellcrank hits the idle stop.

 

[Gary - K7GLD]

Here's the SAE spec, hope it helps a little.



SAE J1843

That info is VERY helpful - thanks! Oo.



This blurb, related to design voltages at accelerator extremes was especially helpful in regards to design and setup of alternate APPS methods, as we are doing:

6. 3 Output Range—See Figure 7.



a. Span = 67. 5% ± 7. 5% of supply voltage

b. Minimum APS Position = 15% ± 5% of supply voltage

c. Maximum APS Position = 77. 5% ± 7. 5% of supply voltage

That, plus the rotational specs for the start of the off-idle and WOT position put the frosting on the cake!



It also explains the apparent typical remaining, unused, resistance at each end of the APPS rotational travel...

 

[rweis]

Then the idle validation switching to the throttle validation (and vice versa) can happen simultaneously? As long as there is a switching of the pins of the ECM then it should be good to go.



Bob Weis

 

[Gary - K7GLD]

Then the idle validation switching to the throttle validation (and vice versa) can happen simultaneously? As long as there is a switching of the pins of the ECM then it should be good to go.



Bob Weis

Not sure I perfectly understand your question, but look at the APPS function as being 2 different, separate operations - the resistance portion that communicates accelerator position to the ECM, and the IVS switch function that is keyed to, but NOT electrically connected to the APS function, and is ONLY there to signal to the ECM at the appropriate rotation point, that the throttle is off idle and in the accelerate mode - then is again keyed when the pedal returns to the idle position.



The only critical issue in PHYSICALLY coordinating between the IVS and the ACP function, is the TIMING between the 2, and that has now been established to occur at about the first 5% or so of accelerator pedal/APS rotation. There is NO electrical connection between the APS and the IVS, and the confusion on that issue apparently is caused by the fact that in the DC APPS module, BOTH functions are in a single module, and get their operating power from the same source - but they are 2 totally INDEPENDENT functions!



And for the record, the ARM of the IVS switch is at GROUND potential, the same common ground as used by other engine sensors attached to the ECM- so actually, the function of the IVS is to ground either the IDLE or OPEN THROTTLE side of the IVS leads, depending upon throttle position.

 

[KarlC]

Then the idle validation switching to the throttle validation (and vice versa) can happen simultaneously? As long as there is a switching of the pins of the ECM then it should be good to go. Bob Weis

The two contacts of the switch can both change at the same time, one opening while the other is closing. The DC electronic IVS circuit has both change within milliseconds of each other. They really wouldn't need two contacts as they do, but it's not hard to do and provides a bit more redundancy in case of a bad connection.

 

[rweis]

OK, then.



As soon as my MS gets here I should be able to test the bus APPS and the MS approach. The "adjustment" seems to be to get the idle / throttle IVS to switch in the . 7 - . 8 range (I'll shoot for . 75). The switches are in stock.



Bob Weis

 

[Gary - K7GLD]

The Wilson APPS is 2. 09K ohms to . 262 k ohms. The . 262 k is at zero throttle, the 2. 09 k ohms is at full throttle.



Bob Weis



As mentioned before the Wilson APPS is an exact fittment to the DC APPS mounting plate. As mentioned above, MAYBE there are several APPS that have the same mounting?

Bob, it looks to me as though the resistance range of that unit is right on, and if you can easily pick off the potentiometer function by itself as can be done with the DC APPS, and then use an outboard MS for the IVS function, you'd have it made! The cost is more than a separate rotary pot like I'm using, but FAR less Rube Goldberg adapting - might be worth the effort, since you already have it anyway...



Didn't see it if you posted it, but what is the full resistance, end-to-end of that pot? As mentioned earlier, one of my OEM DC APPS units measures a tad over 2. 5 K. and the other is a tad over 2. 3 K full end-to-end resistance...

 

[timbo]

Got a thought here. A DC APPS is slotted to allow for some adjustment to get the voltage to the proper voltage. Did you try turning in the idle stop screw to achieve a higher voltage before you added the resistor? If the resistence range is right on, then could it be possible the DC bell crank is just a few deg off from the bus petal assembly and the sensor is down in the fault range. Maybe it's just as easy as turning in the idle stop screw to get in the . 55v range and if your really lucky the IVS might just work also. Just a thought.

 

[rweis]

No, I had not thought of that. I only had to add 100 ohms to the pot to get it to . 60 (my dc APPS sticker is . 595) . I'll try changing the stop a tad to see what happens (counting the number of turns to see what it does to the idle rpm). I was excited that I could duplicate the dc APPS voltage of . 595 so closely so easily.



Gary - to get the really close . 6 volts at idle I added 100 ohms so it went from . 362 - 2. 19 functional range. But anyone could buy the bus APPS and add whatever ohms it took to get to . 6 over at C3 pin 23 and the pot part would be standardized for everyone.



I like the idea of the seperate IVS because you have some control over when it kicks in. If I can standardize it to . 75 volts (C3 /23) (if that works), then everyone would have a taget to aim (. 75 volts) at using only a pair of needle nose pliers and a volt meter.



At the most I'll use the pot from the bus for the APPS and an MS to run the IVS. My dc APPS is still good and eventually I'll run the dc APPS and the MS IVS.



I think we are going to have several choices here. Especially if the pot part of the dc APPS is fairly reliable, all most members with "failed APPS" would have to do is change the IVS over to a MS. Not too bad.



Just have to be as universal and repeatable in the IVS design as possible. I'm thinking a 1" piece of Al angle about 1 1/2" long with a slotted mount for the MS so there is some adjustment as to when it switches. It has to be easy to make / duplicate and easy to mount reliably and accurately hopefully using nothing more than hand tools.



If we can get this so you could fix it on the side of the road in a snow storm at night, that would be the "cat's meow".



Bob Weis

 

[mhenon]

And for the record, the ARM of the IVS switch is at GROUND potential, the same common ground as used by other engine sensors attached to the ECM- so actually, the function of the IVS is to ground either the IDLE or OPEN THROTTLE side of the IVS leads, depending upon throttle position.

Gary, That's the main difference between the DC APPS and the Williams bus APPS that Bob is working with... the bus APPS wants 5volts at the arm. Scrool to page 4 to look at the bus APPs schematic:

http://www.abc-companies.com/servic...elerator pedal - Throttle Position Sensor.pdf

This might be a problem with using the bus APPS. I suggested to Bob in a PM to try feeding the IVS in the bus APPS the 5v it wanted. Then let the "idle active" and "throttle active" actuate SPST relays which throw to "ground" when they're activated. With this wired properly the ECM will get "ground" from idle and accelerate when it's supposed to and might just work.

Don't know if Bob's already tried this. We've been on the road for the past few days and I just jumped in to see what progress has been made.

Mike

 

[rweis]

I have been thinking which way is the quickest way to get results or failure (which is equally as important).



1. Do the +5 to the bus IVS and relays as mentioned above

2. Figure out a bracket etc to use a MS for the IVS



I think I can do the +5 and relays quickest.



FYI on the MS for the IVS I have shifted my thinking from mounting the MS in the pocket from the TV cable to the APPS frame to a location on the other side of the bell crank over where the Idle set screw is because of more room and easier to get to for adjustment.



As I work both puzzles, I need some opinions on material choice if we go the bracket route.



Over on the APPS bracket by the idle set screw there is a 2" length of the bell crank that would make a good place for the MS to act against. However, the structure of the rotation mechanism for the bell crank has a 1/4" rise in the rotation casting (sort of like a step up). There are 2 mounting bolts that could convently be both ends of the MS atachment bracket so we do not have to drill any holes.



My idea is to lay flat a 1" thick wide (this wide because it is a standard size, and we are going to drill 1/2" holes for the dc APPS attachment bolts (which leaves 1/4" strap material on both sides of the attachment drill hole)) 1/4" piece of xxxxx (steel or aluminum or ??) and make it go to both ends where the attachment bolts are and use the existing top and bottom atachment bolts to attach the MS holding bracket. Drill the aligning holes in the 1" wide x 1/4" thick for the top and bottom existing bolt holes and that forms the basis for the MS attachment bracket. The 1/4" thick has some value to get the MS attach bracket level with the step up in the APPS rotating hub where the MS will be mounted.



The actual MS bracket will be welded 90* to the atachment strap (above) and is basically a rectangular piece to hold the MS perpendicular to the bell crank edge.



I have changed my idea of how the MS actually actuates on the bell crank arm. It can not be by direct 90* touch because if you need to adjust the MS to actuate sooner or later than the physical bell crank would touch the MS there is no adjustment.



Gary's idea of a MS with a roller on the switch would allow the roller to roll under the bell crank sooner than the actual physical edge would get to the MS. Later I do not think is a problem, but the sooner consideration than the physical bell crank arrives at the physical MS I think is needed because I think the IVS needs to be actuated a couple of hundred rpm (@ 1100 rpm?) before the mechanical set screw stop idle. This means you have to touch the bell crank before it physically gets to the MS. A roller lever MS could touch the bell crank arm and roll under the bell crank arm and still actuate the MS before the physical bell crank arm physically go to the MS. If you used a lever switch for the MS like I was going to do the physical bell crank arm is not there yet. If you "adjusted" a lever arm to touch the bell crank early, then as the bell crank arm continues toward the idle stop is bends the MS lever back on itself and the lever arm is not in proper position for the next actuation.



Now the materials for the bracket. The MS bracket would be 1/8" x 1 1/4" (tall) x 2" (long)



Aluminum Pro

Aluminum would be lighter, easier to drill the 2 mounting holes for the MS, Aluminum Con

Cost more (but probably not a whole lot because it is small), harder to find someone to weld the MS bracket to the 1/4" attachment strap, not as firm a material, may not hold as well as steel, not as firm a material if you needed to "oval" the mounting holes to get a final fine adjustment.



Steel Pro

Little heavier material, firmer material could be "worked" more to get an exact adjustment, easier to find someone to weld the 2 pieces together

Steel Con

Harder to drill the 2 mounting holes for the MS, little heavier material, rust?



I am inclined to go with the 1/4" steel mounting strap and a 1/8" steel MS mounting bracket.



I want to be able to have a fab shop make 100 of each piece and then mail both brackets to you when you need to fix your APPS. You have them welded together, you drill the MS atachment holes (you may use a different MS than I use and therefore need the MS attachment holes in a different place), mount your MS, get your Wilson bus APPS, swap it out, wire it up (wire diagram will be included for bus APPS and MS) and you are done.



OK, so which material?, and if the relays work it may be a mute point anyway



Bob Weis

 

[KarlC]

Bob,



I didn't follow everything in your description, so you may have already mentioned this. What about attaching a small block (aluminum?) to the bellcrank with a couple screws to press the MS lever (or roller). Might give more options for mounting the MS. The holes could be oversize or slotted or maybe the block could swivel to get the needed adjustment. Disadvantage: requires drilling the bellcrank.



For sheet metal brackets I don't like aluminum because of strength and steel because it rusts. Auto makers use plated steel but for us plating small quantities is expensive.



For blocks where strength isn't an issue I like aluminum, for all the reasons you mentioned plus it's easy to tap.



I like stainless for brackets, 1/16 or 1/8" thick. It's more expensive (but like you said there's not a lot needed here), harder to drill or tap (have to go slow & use a sharp bit with oil), it can be bent (with care and/or heat) and welded (easier than aluminum). Biggest advantages: strength, thin pieces can be drilled & tapped, and no rust or corrosion.



A lot of this is personal preference, what someone is used to, and what tools are in his shop .

 

[mhenon]

Bob, We'll be home in another week or so. My engine is already rewired with weather pack connectors so I can flop back and forth from OEM APPS to Williams APPS with little effort. I'm looking foward to digging into this thing especially now that you guy's have done most of the work.

Still believe we're right on the edge of a workable solution... maybe more than one! Hang in there guy's... we're light years ahead of where we were just six months ago.

Mike

 

[Gary - K7GLD]

FYI on the MS for the IVS I have shifted my thinking from mounting the MS in the pocket from the TV cable to the APPS frame to a location on the other side of the bell crank over where the Idle set screw is because of more room and easier to get to for adjustment.

BINGO! :-laf:-laf



Exactly my plan - and I also plan to mount the MS in a relatively fixed position, and use the existing idle-side belcrank adjustment stop screw for the final adjustment - the APS pot will have it's own, independent adjustment, if all goes as is planned. The pot(s) arrive tomorrow (Friday), and the MS on Monday or Tuesday - if you hold off 'til then, I'll have pics to post, and all will be clear...

 

[Gary - K7GLD]

Why APPS modules go bad Part II - my version!

Here's what has evolved, as the result of this thread, as my personal opinion as to why many/most APPS modules fail prematurely;



While much of the internal design of the DC APPS seems cheap and flawed, the FINAL insult to poor design seems to be the method used to connect the APPS internals to the external pin connectors - the best internal design in the world is WORTHLESS if permanent and reliable contact to the connecting point is lacking. Here's a pic of the neck of the APPS cut off, showing the components and method of attachment between the inner APPS foil strips, and the connecting pins:



#ad




A pure pressure contact, using the intermediate "buffer strip" of conducting material to allow passage of contact between the elements on either side - pin bases on one side, foil strips on the other - exactly like that used in a kids toy or a $2 digital calculator display. AND, if you've ever seen erratic or missing lines on those displays, it's usually because of THIS type of design, and resulting failure, as the intermediate contact strip fails to maintain pressure and contact, or otherwise looses conductivity.



It's a fast, cheap, and easy assembly line construction method, but sure is a POOR one in this type APPS service, where components are exposed to constant vibration and heat cycling!



Here's another pic, with the intermediate strip removed, showing all 3 major components, the pin assembly, the thin buffer/conducting strip assembly, and the exposed APPS inner membrane trace strips:



#ad




My purpose in showing all this, other than to display the weakness of the design, was to see if it's possible to make a more permanent and reliable contact with soldered connections, and total elimination of the intermediate contact strip. My basic thought, is that the APS resistance potentiometer section may well be fully functional still, and usable for that function - and for that matter, the whole module - a supposedly "failed" one replaced with a new one by another owner - may well also be returned to service, if a poor or erratic contact at this point WAS the reason for failure and replacement.



Besides, I like to tear things apart to see what makes them work - or not... :-laf:-laf:-laf