Looking for info on 1800 Valkyrie Digital Display - for custom GL1200E

[Miklayn]

So I have a GL1200E- ye' old Valkyrie Granddaddy, and I'm planning some minor mods to make it a little faster, better handling, lighter and generally cooler looking.



One of the things I'd really like to do is replace the original, rather large analog tach and speedometer with something lighter, less electrically intensive, and sleeker. Like the svelte digital display from the '14+ Valkyrie, which is the best looking multifunction display I have found anywhere.



I'd like to think that it wouldn't be too hard to get the functions to work, but I can't seem to find much info on the gauge itself, and nowhere can it be found for sale used.

Whatcha got?

[Wayne Basso]

You can find the digital display on Bike Bandit called Meter, combination.  Gives 2 part numbers - 37100-MJR-771 and 37100-MJR-671.  Retail price $758.45.  Probably will need covers as well.  It won't be cheap.  Could contact Honda and see if it will fit.

[Colin]

Have a look at the Koso instruments. They are much cheaper than $750, they are available in the US from JP Cycles and others. I have fitted the DB-01-R+ to my Valk.

http://www.koso.com.tw/products.php?id=185&lang=en

http://www.koso.com.tw/products-detail.php?id=889&lang=en

Photos of it installed on my Valk http://rebuild.turnspain.com/

[Miklayn]

Yeah I've seen all the KOSO gauges and a number of others, like Dakota Digital. The DB-02r and the DB-03 are pretty cool but they're a bit smaller than I'd like (yes I know I'm being picky) and I really love the inverted LCD look of the 1800 multi. It's just that there aren't any for sale anywhere!

PS your build looks pretty awesome if I say so myself. Very good job on the pipes and the intake horns

[Bill Havins]

Miklayn,

I know little about the GL1200, so take this post with a grain of salt.

I have a shop manual for the GL1800C here at my elbow.  I have also been active in designing human interface devices for vehicles (e.g., gauge clusters) so I am reasonably familiar with the issues noted below.

To me your biggest challenge is going to be the interface between your GL1200's sensors and the inputs into a GL1800C Display.  The speedometer section of the GL1800C gets its input directly from the Vehicle Speed Sensor - I don't know what the speed sensor is like on the GL1200 (i.e., variable reluctance, Hall Effect, cable drive, etc.).  There may be a need to create/develop/buy an interface of some sort so the GL1800C Display will be able to accurately display speed.

The tachometer display on the GL1800C Display gets its input directly from the ECM (engine control module) on the bike.  As above, I don't know where the tach signal for the GL1200 comes from (i.e.,  coil lead, variable reluctance sensor on the crankshaft, etc.).  Since the GL1200 is not fuel-injected its ECM (if it has one) may be pretty simple.  So, again, you'll probably need to create/develop/buy an interface to change the GL1200 tach signal into something the GL1800C Display can "read."

As above, I know next-to-nothing about the 1200 so consider the source of these comments.

Good luck!

Bill

[Miklayn]

That is what I was asking about. On the Standard and Interstate GL1200 the speedo and tach are analog, and the Speedo is cable driven off the front hub. The tach I believe gets its signal from one of the coils.

On the 1200 Aspencades the digital speedometer is driven by a hall effect sensor that I believe is interchangeable with minor modification from the cable-driven unit, not certain about this. Otherwise I'd like to believe it wouldn't be too hard to integrate an 3rd-Party rotor mounted sensor, but the problem as you state is whether and how the display would take these signals.

If the tach on the GL1800C gets the signal from the ECU, how does the ECU receive it? Likewise,  what is the configuration of the vehicle speed sensor? I'll have to try to pull a parts fiche or get ahold of a shop manual for the 1800.

I wonder if it would work to use an ECU just for the functions I want, if I could somehow disable whatever error codes it would throw. Not that that would make it any less expensive 

[Bill Havins]

We'll pretty quickly get into the area of "black magic" talking about this stuff.  All the iterations of sensors and interfaces can begin to make your head spin!  And it can turn into a tremendous "time hole" that absorbs your every waking hour.  It's a curse being interested in such things.

When I installed my Rostra Cruise on my GL1800C I discovered that the speed sensor produces an unusual number of pulses per mile (don't have my notes in front of me, but the number was really high, maybe 50K+).  The folks at Rostra make a 4-to-1 pulse "divider" for such situations.  So, by installing one, I was sending pulses to the Rostra at an acceptable rate (<18,000 per mile).  Now remember, the Rostra Cruise "knows" nothing about "how fast" you are going (i.e., your speed in MPH).  All it knows is the frequency of the pulses at the spot where you push the "set" button; the Rostra simply adjusts the throttle to keep the frequency of the pulses approximately "constant" with the "set speed."

Some years ago there was a "standard" that speed sending units would output 8000 pulses per mile (IIRC).  But not every motor vehicle manufacturer adopted that standard.  Most were pretty close; technology and cost really set the limits on what they did.  The GL1200 falls generally into that time period when manufacturers changed from cable driven to speed sensor "driven" speedometers.  If the Interstate sensor is a "direct replacement" for the cable drive unit, by all means use one of those.  It will make life simpler.  

But 8000 isn't close to the 50K+ pulses per mile that are output by the GL1800C Speed Sensor.  So, to raise the number of pulses per mile  you'll need a "multiplier."  IMHO, that ain't good - multiplying as much as you'll need to introduces "variability" which can cause speedometer readings to fluctuate too much (depending on the multiplication factor you have to use).

Okay.  So you get your Interstate sending unit.  Install it and then determine how many pulses per mile it produces.  Use this formula:  ((5280/circumference of your tire) X pulses per revolution).  In the formula the circumference of the tire should be in feet, not inches (or you can change 5280 feet to inches and work in inches - it really doesn't matter).  Once you have the "pulses per mile" figure you can begin to consider the multiplication factor you'll have to use to move the project along.  It may approach a multiplication factor of 8 - that's pretty big!

The tach output on the GL1800 ECM is a simple 5 12 Volt square wave (minimum of 10.5 Volts required).  If I remember correctly the wave varies from about 0 Volts up to near 5 12 (it's been awhile since I looked at it on my oscilloscope).

Coil-driven tachometers "may" get hit by voltages as high as +60 Volts!  That will blow a pre-microcontroller input designed to receive up to 5 12 Volt inputs.  So, depending on the voltage of your tach signal, you will have to find a "converter" to reduce the voltage to no more than +12 Volts.

Okay.  You really like the design of the GL1800C display.  I understand that.  But, to tell you the truth, my sense is that interfacing it will be a challenge.  It can be done, but the issues above are the ones that you'll have to resolve.

Oh, crap!  I just remembered about the odometer/trip meter issue.  On a modern car those data are stored in the ECM.  But, now that I've checked the schematic, it appears they are stored in the GL1800C Display itself.  So, by resolving the interface issue for the speed sensor you'll take care of the odometer/trip meter, too.

My head hurts.

Bill

EDIT:  I checked my numbers and made some changes to the above - it's been awhile since I dealt with the multipliers involved in this.  I have also gone back and checked the manual.  The above adjustments to voltage measurements are accurate now. -BH

[Bill Havins]

Found my spreadsheet with the data I collected on the output of the sending unit on the GL1800C.

In the rather crude measurements I did it appears the sending unit outputs one pulse per 6.67 degrees of rear wheel rotation.  That equals about 54 pulses per revolution of the rear wheel.  With a stock and near-new rear tire that's 43,920 pulses per mile (+/-).

If you switch to a sending unit that outputs, say, 8000 pulses per mile, you will need an interface for the GL1800C Display that multiplies the sending unit output by about 5.5.  Ugh.  But you won't know what you need until you have the Interstate Aspencade sending unit in hand and can measure it's output.

The "problem" with the GL1800C Display is the software in its microcontroller (its CPU so-to-speak).  The software is written for a sending unit that sends out 43,920 pulses per mile (according to my crude measurements).  The software counts the number of pulses it receives in about a second or so, calculates what speed that is, and displays the result on the LCD.

If you interface the Display to a vehicle speed sending unit that outputs a different number of pulses per mile the software will still count the pulses, calculate a speed, and display a number on the LCD.  But the number displayed will not be an accurate measure of speed because the "frequency" of the input (something other than 43,920 pulses per mile) is not the "standard" used by the software.

Aftermarket speedometers have a system that allows you to "capture" the "frequency" of the onboard sending unit.  To do this you stop at the start of a measured mile course, press a button, drive the length of the course (a mile), and press the button again.  After the first button press the software begins to count the pulses it receives.  When the button is pressed the second time the software stores the total number of pulses it has counted and then uses that as its "standard" for transforming pulses received into a measure of speed.  These aftermarket speedometers allow you to change tire sizes and "recalculate" the speedometer so it displays speed correctly with the new tire size.  They also allow users to mount the same speedometer in cars that use sending units with different pulses-per-mile outputs.

The GL1800C Display does not have the above "reset feature."  It expects to see 43,920 pulses per mile.  So, if you use a sending unit with a different output, you have to transform that output such that it is approximately 43,920 pulses per mile.  Not fun.

The software could be re-written to use a different "standard" (say, 8000 pulses per mile).  But that is a real "hack" and, if done incorrectly, could completely ruin the Display.  That's an expensive "Oops!"

Bill

[Bill Havins]

Regarding the Tach input on the GL1800C, as noted above, the schematic indicates the Display gets the Tach signal from the ECU.  On a fuel-injected vehicle there are one or more sensors on the crankshaft/camshaft(s) that "measure" or "time" where the crankshaft/camshaft is in its rotation.  The ECU uses data from these sensors to "fire" the fuel injectors at the right point on each cylinder's intake stroke.

The ECU on the GL1800C gets its Tach signal from one of these sensors (I haven't chased down which one).  I don't know if there is a crankshaft sensor on the GL1200E - I doubt there is since it isn't fuel-injected.  But, if it uses electronic ignition (as opposed to points) there may be.

There are many ways to capture a "timing mark" (e.g., a specific tooth, a magnet, a hole, etc.) on a spinning shaft.  There are riders out there who have added "modern" tachometers to vintage bikes by adapting one of these techniques.  Since the GL1800C Display is expecting a 5 12 Volt square wave on its Tach input, a 5 12 Volt Hall Effect sensor seems like the easiest adaptation to use.  There are a number of waterproof and easy-to-mount options out there (I have one sitting on my desk, a 5 volter, that I got from www.mouser.com for about $30).

These issues are "Black Magic," and "time holes."  

Bill

[Robert]

Bill awesome explanation and information  I am going to put that information in my Valk file. I was surprised to see the amount of pulses per mile and wondered if that could be wrong. But the more I thought about it the pickup is in the engine and I believe measure's off the output shaft or drive shaft so that could very well be true. I have also heard that the lack of a good pulse in the speedo will cause a check engine light.

There was one other thing I remember Honda said that if you see a particular pattern on the display send it in for warranty. I had never heard that kind of thing before and wondered if it was an engineering problem with the cluster. But the more I hear information like yours I am not surprised since the level of complexity and durability that have to be designed into this cluster is pretty interesting.

I dont think this project is very feasible or cost effective since there are so many other displays that would look good provide the information needed and be easier to adapt.

[Firefight100]

I would look into the Trail Tech line of dash pods.  They are far easier to make work than the factory one.
http://www.trailtech.net/digital-gauges/vapor

[Bill Havins]

Thanks, Robert.  I've been "tinkering" in this area for too many years.  As an aside, in 2011 I presented a paper to the Society of Automotive Engineers at a Chicago conference - the paper was entitled, "Application of Human Factors and Cognitive Neuroscience Principles in the Design of Vehicle Information Displays."  Now if that doesn't bore the spots off your dalmatian I don't know what will.

Yes, a weak speed sensor pulse will cause the GL1800C Display to throw a "check engine" light (a "trouble code" in the vernacular).  I watched this happen as I was installing my Rostra Cruise.

Brian (bscrive on this forum) did a write-up on this forum about installing his Rostra Cruise some years(?) ago.  He mentioned that a 47K ohm resistor had to be installed between the output of the speed sensor on the bike and the input of the Rostra Cruise.  At first I didn't understand why (and wanted to find out).  So, after installing my Rostra Cruise (without the resistor) I headed out for a trial run.  As soon as I shifted into third (fourth ?) gear the "check engine" light came on.  It turns out the Rostra Cruise was "loading" the speed sensor output/ECU input and, as the "frequency" of the speed sensor output increased, the voltage the ECU was seeing was too small to allow the ECU to differentiate one pulse from the next.  So, the ECU "threw" a trouble code indicating the speed sensor was having problems.  It was funny, though, because the bike continued to run okay.

Adding the 47K ohm resistor between the speed sensor and the Rostra Cruise prevented the cruise from "loading" the speed sensor (it changed the path from "low" impedance, or a "direct route," to "higher" impedance).  So, for all intents and purposes, the ECU doesn't even "see" the Rostra Cruise after the resistor is installed.

Eventually I "may" get around to building an interface that allows me to replace my GL1800C Display with an Android-based tablet.  I've been tinkering with a Samsung Tab-S for this purpose.  It has a Super AMOLED display that is easier to see in many lighting conditions than the negative FSTN LCD used in the bike's current display.  And the Android-based tablet would allow me to apply my patented "cognitive mapping" paradigm, while also allowing for such things as "split screen" view of vehicle information and GPS at the same time.

We'll see if I ever do it - but doing this is one reason I bought this particular scooter.

Cheers!

Bill

[Bill Havins]

Okay, please let me clear up some erroneous data (I've tried to fix it in the posts above.

The speed sensor sends 5 Volt pulses.

My crude measurements indicate the speed sensor sends 43,920 pulses per mile; there is no documentation anywhere I have found that specifies exactly what that number is.

The Tach input to the Display expects 12 Volt pulses (something >= 10.5 Volts).

The above numbers are correct.

I did most of the posts above relying on my memory....MISTAKE!  That's why I take notes and either digitize the notes or transfer the data to a spreadsheet - there is too much running through my head for me to hang on to data I don't use everyday.  So, DON"T TRUST MY NUMBERS until I indicate I have reviewed notes or references to confirm their accuracy. 

Bill

[Robert]

Okay, please let me clear up some erroneous data (I've tried to fix it in the posts above.

The speed sensor sends 5 Volt pulses.

My crude measurements indicate the speed sensor sends 43,920 pulses per mile; there is no documentation anywhere I have found that specifies exactly what that number is.

The Tach input to the Display expects 12 Volt pulses (something >= 10.5 Volts).

The above numbers are correct.

I did most of the posts above relying on my memory....MISTAKE!  That's why I take notes and either digitize the notes or transfer the data to a spreadsheet - there is too much running through my head for me to hang on to data I don't use everyday.  So, DON"T TRUST MY NUMBERS until I indicate I have reviewed notes or references to confirm their accuracy. 

Bill

Application of Human Factors and Cognitive Neuroscience Principles in the Design of Vehicle Information Displays   ??? ??? remind me never to head to head with you on a description 


The new cars have the tablet displays you are talking about with no analog gauges on them and I'm not a big fan. I agree that your description of one for the Valk could be good, but question the durability of them especially being on a motorcycle.

Now I understand the why the data was so meticulous and thorough, just awesome. I am so unaccustomed to ever seeing that kind of data unless I do it myself or someone else that is specific to a project. Its so hard to get some information these days to repair or modify something. Plus to see it on something so obscure as the dash for a 1800c is great.

I have a chart on my wall of o rings with numbers, diameters and thickness. I love that chart, but not to many do but its like honey to bees. Had an engineer walk in he noticed it right away and stood there looking at it. I was laughing, I guess its a engineer, motor head type of thing.

[Bill Havins]

Robert,

You and I have to get together some day and have a beer....wait....I don't drink (60 years of diabetes won't let me drink, if I want to live  ).  Anyway, we could have some laughs about our apparent shared interest of data and detail.  I'll admit to being a geek.

I won't belabor the point but will say that tablets can easily display analog data.  My Apple Watch is...a tablet.  Nothing more.  And it has an analog face.  It can even be a Mickey Mouse watch if I want it to be (had one of the originals when I was 5).

We are at a point in our technology where vehicle information displays can display analog data (including icons) about our vehicles even before we are aware we need to pay attention to it.  And by presenting that information in the place where we instinctively look first (when we "address" a display), those important data will be the first thing we see.  As an example, when you enter a School Zone, what is the first thing you look for when you address your gauge cluster?  Your speed.

But when you're zooming down a lonely country road and you want to know "everything's okay," what would you want to see in the middle of the display if something was "amiss"?  You'd want to see the errant datum, like the oil pressure indicator if the pressure had crept low.

Tablets can display almost any information in an easily recognized form "almost instantly" (with the right CPU installed).  As far as speed of recognition is concerned icons and analog data are where it's at - text is "too slow."  And knowing where to display it in the gauge array (the operative word there is "array") dramatically affects recognition speed.

Crap!  Talk about hijacking a thread!  Sorry!

Bill

[Robert]

Bill

Iced tea would be great as I dont drink either.

 AS for the display I agree but as a thought the new Land Rovers have a tablet display, its colorful, accurate, and looks like gauges but there is no depth to the gauge so the needle is the same depth as the numbers and as the rest of the dash. Not a big issue just something that I'm not used to. Its kind of like looking at a picture that changes. So is this where the Human Factors and Cognitive Neuroscience Principles in the Design of Vehicle Information Displays comes in? 

[hungryeye]

I love those old necked goldwings, they are the original Valk. I'd leave it stock but thats just me. I have a showroom condition 83 GL650I with 10,000 miles on it and its wine berry color almost matches my 14 necked goldwing Valk.

Almost forgot, I love your Dart, 66?

[Miklayn]

Thanks for your responses everyone! I understand this would/will be a frustratingly difficult, not to mention exorbitantly expensive project, but eh, that's what custom stuff is sometimes right?

I have seen the Trail Tech and basically every other aftermarket gauge system out there. I've been scouring the internet for the past week toying with the idea- and Koso and Dakota come the closest but none of their gauges match this one for style, simplicity and pure functionality.

Bill- it's really difficult to know what questions to ask, as on the one hand I don't have access to that which I'm asking about, nor to the correct documentation, and on the other hand, I'm not an electrical engineer nor have I any programming experience.

That being said, how would one go about attempting to re-program the display microcontroller? I assumer there is some sort of specialized interface?

None of this seems entirely intractable, I guess is what I'm saying. The voltage convertor for ignition, or reprogramming the microcontroller of the display itself, even programming a processor to extrapolate x-number of pulses into a signal the Display can interpret into an accurate speed measurement, while I don't presume I could do any of them in a night, in principle they seem... doable...


And so to that end, I strongly encourage you to expedite your AMOLED tablet-display development! For if you are able to produce a working system, perhaps you would sell me your stock display!  AND even if I can't ever get it to work, I can ask to use your Tablet software instead! 

[Miklayn]

@Hungryeye-

I too love the GL1200 Std, and it is a rare beauty. At least as rare as the Valk 1800, which I don't see on the road. And likely faster, owing to the 150lb weight advantage. I'm trying to make that advantage even greater, and add a little subtle flair while preserving the integrity of the original bike..

That Dart is not mine, I have a unit in a shared garage space. I thought it was a '65 but I don't know for sure. It's in superb condition, can't tell if the paint is new without brushing off the dust and looking closer but I'm just gonna stick with not touching it at all 

[Robert]

Honda is not going to release the software for the display, and I doubt it can be programed or even diagnosed by any scanner. So you will be forced to incorporate and use the current inputs and outputs to make the gauge work. You will have to adapt the bike you install it on to mimic the same signals and values. If you get a wiring diagram you should be able to see what goes in and what comes out, and Bill has done an awesome job of giving you some very hard to come by values to mate with. Put the two together and you should get the gauge to work. It sounds easy, its not at all in practice.

[Bill Havins]

Miklayn,

Don't even go down the rabbit hole of wanting to reprogram the software in the GL1800C Display.  If the display uses a flash-type microcontroller it might be reprogrammed.  But most people and companies who write software for commercial uses encrypt what is flashed into the microcontroller (I do).  If the software is encrypted, it would be impossible to know where to make changes and then....oh, forget it!  It's an absolute rabbit hole/time hole that leads to disappointment!

Come up with a tach signal, and find a speed sensor that has an "adjustable" output frequency (does Dakota Digital make such?).  Or, go the electronic speedometer route - take a look at Summit Racing or Jegs to get started on the search.  There is some really neat and adaptable stuff out there.  That's your easiest route.

Good luck!

Bill

[Miklayn]

So to be clear, are we saying in principle can I use the display without the 1800's ECU?

What about fuel and temp display? Are the warning lights simple circuits?

To this end, Bill could you upload a photo of the wiring diagram from your manual?

[Bill Havins]

So to be clear, are we saying in principle can I use the display without the 1800's ECU? [snip]

In principle, yes.

I briefly scanned a parts breakdown for the '84(?) GL1200 and it looks like it has a "distributor" that uses sensors instead of points (an electronic ignition).  There is the "potential" to pick up a signal there that, after "treatment" can be used as a tach signal.  It depends on how many times the "distributor" spins per revolution of the crankshaft.  That's about as far as I want to go with that concept.

[snip]  What about fuel and temp display? Are the warning lights simple circuits?  [snip]

Fuel level sensor...ugh!  I'll keep this short; the fuel pump and level sensor are a combination unit inside the GL1800C tank.  The combo unit uses a 3 pin connector; the pump and level sensor share a ground (one of the three pins).  A green/yellow wire comes off the GL1800C Display and goes to the other pin on the fuel level sensor inside the tank.  So, the level sensor has a "hot wire" of sorts coming from the Display, and a ground.

The fuel level sensor is a "variable resistor."  As the level of fuel in the tank changes so does the "resistance" measured across the green/yellow wire and ground.  Pretty simple.

Here's the tough part - I don't know exactly how the GL1800C Display is "measuring" change in resistance of the level sensor and turning it into a fuel level "reading" on the Display.  If I was doing the design I'd have a "current sense resistor" on the Display circuit board.  I'd attach +12 Volts to one end of that resistor and I'd attach the green/yellow wire to the other end.  The connection to the fuel level sensor would complete this circuit from +12 Volts to Ground.  We'd have voltage, and more importantly, current would flow at a level consistent with the total resistance of the circuit.  (For the sake of clarity I've described an overly-simplified version of what needs to be on the circuit board.)

I'd then attach the "sense" connection on the current sense resistor to one of the analog-to-digital converter (ADC) pins on the microcontroller on the Display circuit board.  My software would then sample this ADC periodically (i.e., "read" the level of current flowing through the resistor) and, then, transform that reading into the reading you see on the GL1800C Display.  Simple!

Here's the rub - the fuel level sensor and fuel pump are a "unit" [16700-MJR-671 PUMP ASSY., FUEL (MSDS) $340.63].  They are not separate pieces.  And, the float is designed to "fit" the GL1800C fuel tank - it will give inaccurate readings if used in a tank with a different size and shape.

There is really no cost-effective way to develop a substitute for the fuel level sensor.  And, somehow, you have to get a suitable sensor into your tank - uh oh!  So that kinda' kills using the fuel gauge.

Oops!  You already have a fuel level sensor in your tank [37800-MG9-771 METER UNIT, FUEL (NOT AVAILABLE)].  So, if the resistance is the same and the "direction-of-change" (i.e., high-to-low) is the same it ought to work.

If you have a temperature sensor in your cooling system on the GL1200 you're probably good.  The GL1800C temp sensor is a simple switch.

It's a rabbit hole - Black Magic - go there at your own risk!  

Bill