Retrofitting an '88 Legend FE display.

[GasSavers_bobski]

I'm not sure just how much it's helping, but it's definately a reminder. Once I'm satisfied with it installation-wise, I plan to do some experimentation - effects of vehicle speed on FE, aero mods, windows down vs. up vs. A/C on... The usual fare.
Currently, the info system is mounted in it's final position and about 90% wired in, though I haven't put together the brackets to hold down the display yet - it's just sitting on the dash. Now that the coolant temp gauge is connected, it's started telling me to check the oil level when I switch the ignition on. I guess that input is supposed to be grounded when the level is where it's supposed to be.
I'm going to pull my car's engine and transmission in the next couple days - the transmission has been acting kinda weird when cold... I think it's the torque converter lockup control components sticking. It also seems to have more drive line slack than I would like. I have a spare transmission sitting on a shelf (actually the original transmission for this car), so I'm going to swap that in and store my current one as backup.
My plan is to replace all the seals in the bell housing while I'm in there... Replacing the engine's rear main seal requires dropping the oil pan, so I thought I would go ahead and pull the oil level switch from my donor Legend and install that as well.
So, I spent most of today under a 18.5 year old Acura, trying to get the oil pan off. It's a transverse V6, so the downpipe for the front bank of cylinders runs under the oil pan and joins the one from the rear bank. That means both front and rear exhaust manifolds have to be unbolted from the downpipe in order to remove it. Lucky me, some mechanic replaced the 14mm downpipe nuts on the rear manifold with 1/2" fasteners. Not realizing what was going on (every fastener on the car is supposed to be metric), I tried a 13mm socket (1/2" is about 12.7mm) on the nut and proceeded to round off the points. I went home and got a 1/2" socket, but the deed was done. Downpipe nuts see alot of heat, which burns off any oil film that might protect them from rust or lubricate the threads, so they take alot of force to remove... That nut is on there until someone cuts it off.
I worked around the down pipe as best I could, then tried bending it once the pan was unbolted, but there isn't enough room under the car to pry effectively.
So, my plan for tomorrow is to take a jumpstart pack (12v source), a 400 watt power inverter, an angle grinder and a recip saw to the junkyard. I'm gonna show that downpipe hell.

Well... I'm going to try to remove it in one piece by grinding off the rounded nut, then I'll show it hell if it doesn't cooperate... Assuming the inverter can drive the recip saw. ^_^;

[GasSavers_TomO]

Any updates for us to drool over?

[Danronian]

I was thinking...I've seen these units on 4 cylinder fuel-injected chrysler cars (the optioned out "K" cars). I wonder how hard it would be to retrofit one of those to a Honda...

But yes, this legend one would be easy to find too, since I've seen a lot of these in the junkyard lately.

[GasSavers_bobski]

Originally Posted by TomO

Any updates for us to drool over?

Heh... Someone on the Legend forums asked the same thing not long ago.

Originally Posted by bobski

Yep.

It goes nicely with my car's frankenstein of the junkyard theme.

I still haven't hooked up the coolant reservoir, fuel door or fuel low switches. I gave up on the oil level switch when I realized there's almost no room in the oil pan for it and I would have to drill a hole in the engine block for it's bulkhead connector.
I started a thread about it on gassavers.org a while back... It's got a bunch more info.

That thread can be found here.

[Rower4VT]

What's the latest with the Legend computer? Get it finally installed and working consistantly? Any pictures of it installed?

[occupant]

I wonder if a later 80s Prelude would have something like this, maybe the 89-91 Si with 4-wheel steering, they got loaded down with gobs of gadgets.

[GasSavers_bobski]

Originally Posted by Rower4VT

Any pictures of it installed?

Scroll up a post.

Originally Posted by occupant

I wonder if a later 80s Prelude would have something like this, maybe the 89-91 Si with 4-wheel steering, they got loaded down with gobs of gadgets.

Possibly. I'll keep an eye out in the junkyards. Also, didn't the '92-96 generation have an optional digital dash? Maybe it has a MPG display built in.

Well, I think I have an ongoing accuracy problem figured out. My last tank of gas came out to 44.2 mpg, but the trip computer read 46.2 (once you add 50% as I described above someplace). The Legend and Civic injectors are supposed to have the same flow rates, so the meter should be accurate, right? Turns out the Legend and '92-95 Civic have different fuel pressure specs. The Legend manual says 36-41 psi is normal with the regulator vacuum line disconnected, while the Civic manual says 40-47 psi is normal. The Civic's higher pressure is forcing a little more fuel through the injector nozzles than would pass with the Legend spec. Since the computer judges MPG based on how long the fuel injectors are open, you get some inaccuracy. I guess I could try swapping in a Legend FPR, but then I would end up running a little lean in open-loop operation (hard acceleration).

[GasSavers_bobski]

Well, kinda bringing this thread back again. I've been PMing user nowhhs about his install over the past couple days... I thought it would be good to move the conversation here so others could benefit as well.

Originally Posted by nowhhs

I want to first thank you for the effort you made in figuring out that installation. Thanks to you, I've more or less also successfully done the same thing in my 90 hatch. I just have one problem I was wondering if you might be able to help me with. I didn't bother hooking up any of the diagnostic stuff, since I didn't want to bother with all of the missing sensors, but I'm having a problem getting the "distance to empty" display to work. When I followed the diagram you printed on that thread and hooked up the yellow/white wire to A8, I still get N/A for a "distance to empty" and my average mileage display stops working correctly. Do I have to maybe do something to A14? Since you've obviously dug into this installation more than I have, I was wondering if you had any suggestions? Does yours work correctly, or is this just something I have to live with? Thanks again, Norm

Originally Posted by bobski

What exactly do you mean by "stops working correctly"?
I have the same N/A reading for distance to empty. The Legend and Civic/CRX fuel level senders have the exact same resistance specifications, so I think you must need a sender assembly with a low fuel sensor on it. The info display probably uses the low fuel warning to calibrate or something. The fuel filler door switch could have something to do with it as well. The Legend owners manual says the display is supposed to estimate the distance to empty at all times based on average fuel consumption, then when the low fuel light comes on, show the fuel remaining.
On the bright side, the Legend fuel tank is about 50% larger than a Civic hatch/CRXs, so together with the 50%-lower-than-actual FE reading, it should come out fairly accurate (assuming we can get it working).
The low fuel sensor itself is kind of an odd device. The ones I've looked at in junkyard Accords are just a grounded can with one wire coming out. The low fuel warning light in my Integra gauge cluster comes on full-bright when I ground the signal wire, but when properly connected to a sensor, the light supposedly comes on slowly when the fuel gets low. As such, I don't know if you could easily wire something up that looks (from the info display's point of view) exactly like the sensor.
I think some european and japanese market Civics and CRXs may have come with the sensor and warning light. If you find a source for the sender assembly, let me know... The CRX and Civic hatch use the same gas tank so the senders should be the same as well.

Originally Posted by nowhhs

By not working correctly, I meant I was getting really low readings, like 2.8 or 3.1 for average mpg over 150 miles or so. I reconnected it this morning and it seems to be working ok now????? The reason I PMed you was I thought you had gotten everything but the oil level sensor working, I didn't realize you were still struggling with this as well. Maybe we can figure this out together. It seems to me that if you want to make your low fuel light work, all you really need to do is pull out a unit from an accord and take the sensor off and add it to your gas gauge assembly. It might take a little welding or soldering, but I don't think it should be too hard. I don't have a light, so I don't really care. What I would like to have is the distance to empty work correctly, excluding the obvious inaccuracy of 50% (actually worse for me, I'm using 190cc injectors, so its more like 90% off). But the display should get more and more accurate as it approaches 0 if its using the fuel level input. This is what I figured out so far. This morning I grounded A4 and got a miles remaining display. It seems that A4 must be connected to a switch attached to the fuel door. When you open the door to fill the tank, the switch closes and grounds it. That seems to be what triggers the display. It seems to need to be closed for a while though, not just a brief signal, more like a minute or so. Problem one: it doesn't seem to base the readout off the fuel gauge, but rather from calculating 18 gallons divided by the average mpg. It was showing 463 miles remaining, with an average mpg of about 25, even though the tank was only 2/3 full. I've also attached A14 to 12vdc through a 100ohm resistor (just guessed a value). When I grounded A14 in addition to having it connected to 12vdc, the display switched to 2.5 gallons remaining, in other words tripping the low fuel function. I went to a junk yard and scavenged one of the low fuel sensors from an accord. It seems to have about 5000 ohms in air and 3000 ohms in water, I didn't check with gas. So I'm thinking just put a 3000 ohm resister in place would work for me, not you though. Problem two: When the display switches to gallons remaining. I was hoping it would stay there until the tank was actually that low, but it also appears to be based on calculating 2.5 divided by average mpg. Even though my tank was 2/3 full, the gallons left kept dropping. (I wonder what happens when it gets below 0?) Under no circumstances does the processor seem to use the fuel gauge input. Question for you: any idea what the value of the resistor in parallel with the low fuel light is? Knowing that would help me fine tune my A14 input. I'm out of ideas for the time being. Your turn?
Norm

Originally Posted by bobski

Wow... Nice work.

Where are you tapping into power to drive the unit? I found that if you power it from an accessory circuit (stereo, cigarette lighter), it sees the injectors as running at 100% duty cycle until the fuel system power is switched on. This is simply a result of how the fuel injector circuit is set up and where the info display samples the state of the injectors.

My best guess regarding the resistor is that it allows some stray current to bypass the indicator light, delaying when it starts to light up.

The fuel door switch that I looked at was still installed on the car... I simply clipped my multimeter probe onto the ground and fuel door switch wires where I had cut the display and control units off the dash harness. The continuity function showed that the switch wire momentarily grounded when you pulled up on the fuel door release lever. It flipped back to an open circuit as soon as I let go of the release. It showed no change in response to re-closing the fuel door.

Do you have any objection to moving this discussion to my old install thread? It could be useful for anyone that decides to go the same route in the future.

Originally Posted by nowhhs

I think it would be a great idea to move it to the original thread once we get something definitive figured out. If we dead end on this, there isn't really any point. On the other hand, someone else may have something valuable to contribute (TomO seems very knowledgeable), so go ahead (typing while thinking).

So you're saying that the reset from the fuel door was momentary? I'll have to look closer at that. The reason I would like the actual value of the resistor parallel to the low fuel light bulb, is that it works as a voltage divider along with the low fuel sensor. That means that A14 doesn't see a full 12 volts, the actual voltage depends on the values of the resistors on either side of A14. Any chance you can access that resistor and check the resistance? I think in the thread you mentioned you had 2 of these clusters? I would like to duplicate that circuit with light bulb, resistor, and low fuel sensor to see exactly what A14 wants to see. Thanks again.

[GasSavers_bobski]

Originally Posted by nowhhs

So you're saying that the reset from the fuel door was momentary? I'll have to look closer at that.

Right, it's only triggered when the user is actually pulling on the lever.

Originally Posted by nowhhs

The reason I would like the actual value of the resistor parallel to the low fuel light bulb, is that it works as a voltage divider along with the low fuel sensor. That means that A14 doesn't see a full 12 volts, the actual voltage depends on the values of the resistors on either side of A14.

Yeah, ok... You have the parallel combination of the bulb and resistor on one side, and the sensor pickup on the other. The voltage would divide between the two, making the signal at A14 either battery voltage - the bulb/resistor voltage drop, or ground (zero) + the sensor voltage drop... They should be the same thing.

Originally Posted by nowhhs

Any chance you can access that resistor and check the resistance? I think in the thread you mentioned you had 2 of these clusters?

I have two '94-01 Integra clusters and an '88-91 CRX cluster. The CRX cluster doesn't have a low fuel lamp, and the Integra clusters don't have resistors as far as I can tell (I looked a few hours ago). I guess it's possible the resistor is built into the lamp or lamp socket... I'll go check that now.

[nowhhs]

bobski asked: "Where are you tapping into power to drive the unit? I found that if you power it from an accessory circuit (stereo, cigarette lighter), it sees the injectors as running at 100% duty cycle until the fuel system power is switched on. This is simply a result of how the fuel injector circuit is set up and where the info display samples the state of the injectors."

I tapped the VSS and injector signal directly from the ECU. The switched 12vdc, the constant 12vdc, the ground and the lighting dimmer I tapped from the factory clock plug and haven't had any problems. I had been using a DMM to monitor the duty cycle of the injectors before installing this also using the same tap point without a problem. I never saw 100% duty. It may be a problem if you are tapping the injector under the hood and get the wrong side. I'm just guessing though.

I also wanted to address TomO's comment as to where to find these units. They seem to be standard on the Legend LS models, irrespective of the Bose stereos. When I was searching for mine, I had to go through several yards and several times got excited when I found Bose stereos based on what he had said, but didn't find a mileage computer until I found the LS. Thats not positive, but I'm reasonably sure.