A good night's work

[virgilmobile]

I have always got but sore on my 1100 and 1200 but the 1500 stock seat is much more comfortable.Almost like it has some gel padding in it.

As far as the exhaust...Well I tried a set of Harley pipes..Worse..so I went back to the original ones missing the resonator tip.
I made a set from 1-1/4 brass pipe and covered it with steel wool.
What a difference.It actually improved the performance of the bike.I guess removing the tip to put on the fancy chrome extensions changed the designed backpressure.

[WingAdmin]

Did you use the stainless steel wool? Hope so, otherwise plan to replace it soon.

I fixed my stupid heating gaffe - pulled the Gerbing plugs out, insulated them from the metal I mounted them in, put them back in, all works fine now - gloves and boots heated separately like they should.

Here's my cockpit now:

My GL1500 Cockpit

On the left is my GPS (removable, so I can stick it in the trunk when parked). On the left handlebar on top of the clutch reservoir are my lighting switches. On the right handlebar is my satellite radio (also removeable). On the right is my "Bike PC", which is permanently mounted. The controls for the bike PC are on the side of my lighting switch box.

Bike PC Main Screen

This is what it shows normally. Temperature on top left (showing 87.9 because it's parked and reading the temperature coming off the radiator), voltage on bottom left, current gear on top right. The bottom right shows four things:

- Current audio input (AM/FM radio, Satellite radio, Aux input - plug for my cell phone)
- Glove heat level
- Boot heat level
- Elapsed trip time - starts counting up from 0 when I first shift out of neutral

Bike PC Glove Heat Screen

Adjusting the glove heat, I hit my select button, then push up/down to increase/decrease the heat going to my gloves

Bike PC Boot Heat Screen

Same thing goes for my boots

Bike PC Audio Control Screen

Pressing up or down cycles through AM/FM, Satellite, and Aux In

Bike PC Brightness Screen

The PC has a light sensor that dims the screen at night, this screen lets you set how dim you want the screen to go

The screen is illuminated, but it's also sunlight readable, so you can see it in direct sunlight. I am thinking however of putting a little hood over it like I have for my GPS.

That's it! I'm happy with it, it's exactly what I wanted, and it works great!

[virgilmobile]

It's slick,and looks like it can be adapted for different functions. In the deep south,I wouldn't use some of the features you've implemented.Heated boots?I need a air conditioned helmet.
Will you share the basic technology components on the bike PC? I can't seem to locate the same PCB.

Oh,by the way.I just used plane old steel wool just to test the baffle idea.
It works so well,I will replace it with brass or SS this weekend.

[newday777]

Nice GS. How about some pics of the control switches for the PC? Did you do up a pictorial How To?

[WingAdmin]

It's slick,and looks like it can be adapted for different functions. In the deep south,I wouldn't use some of the features you've implemented.Heated boots?I need a air conditioned helmet.
Will you share the basic technology components on the bike PC? I can't seem to locate the same PCB.

You can't find the PC board anywhere because I designed it myself, it's a one-off custom job.

PC Board

The core processor is an Amtel 90USB1286. I've got some support circuitry such as a voltage divider to bring 0-20v down to 0-5v that the ADC can handle, to read the bike's voltage level. I've got it talking to a DS18B20 one-wire temperature sensor IC, I've got a dual optocoupler driving two relays to switch audio input, a 2N2222 to drive the LCD's backlight, and two IRLZ44N MOSFETs to control the heat. Some ancillary support components, some protective diodes, caps and so on to protect against the outside world. It's pretty simple, really.

[WingAdmin]

Nice GS. How about some pics of the control switches for the PC? Did you do up a pictorial How To?

I think this one would be a bit complex for a "How To" - and there's quite a bit of software I had to write in order to run it all. I'll take a picture of the control switches tomorrow - there's not much to them, really.

[littlebeaver]

Wow, that's some a real amazing stuff there Wingman..And most impressive..

[WingAdmin]

I had it lock up on me twice - and today I figured out what was doing it. When I turn OFF my driving lights, it consistently locks up. There's obviously a power spike that's making it through my power supply and into the processor. I'll pull it out tonight and add another, larger capacitor to the regulated side of the PSU.

[virgilmobile]

You could add a spike suppression diode on the offending relay,and a 0.01 uf across the contacts.
Or you may need to feed the pc with a LC network and sprinkle a few 0.01 uf to gnd. to knock out a short spike.The big cap helps to stabilize it,but may not react fast enough to absorb a micro spike.
Question....Did you have to provide heat to the LCD so it will operate in the Frozen North?

[WingAdmin]

You could add a spike suppression diode on the offending relay,and a 0.01 uf across the contacts.
Or you may need to feed the pc with a LC network and sprinkle a few 0.01 uf to gnd. to knock out a short spike.The big cap helps to stabilize it,but may not react fast enough to absorb a micro spike.
Question....Did you have to provide heat to the LCD so it will operate in the Frozen North?

Nope...but the LCD is rated down fairly low. I've had it out at 34 degrees, and it's obviously a bit slower, but still quite readable and functional.

I added a couple 1000uF caps, but no difference. I already have some small caps for small spikes. Started disconnecting things one by one to see which was feeding the spike in, and it was the heated clothing leads. Disconnected those, problem went away. Thing is, those leads weren't actually connected to anything (other than a disconnected coaxial socket) at the time! Head scratcher...

Then I realized those two leads were running physically next to the relay that controlled the driving lights. Aha! Induction!

Relocated the relay, put a couple ferrite chokes on the heater leads (as well as the power leads, for good measure), problem is fixed. Went out for a 45 minute night ride just to make sure.

I also installed a replacement fuel reserve sensor tonight, as mine was faulty. I now have a working reserve light!

[virgilmobile]

A length of wire with a battery arcing contact....Isn't that the first MORSE code???
RFI is fun to deal with, Glad you found it.
We had a Dodge charger cop car that would randomly start flashing the head/tail lights...RFI from the radar would trigger the BCM lighting.Fun,fun.
What did you have to give for the low fuel sensor? where,and is it the same as the 88?
Mine is removed as it failed,Weird piece of material inside.Resistance goes low when dry and hi when wet.I thought about putting in a mercury tilt switch....

[WingAdmin]

A length of wire with a battery arcing contact....Isn't that the first MORSE code???
RFI is fun to deal with, Glad you found it.
We had a Dodge charger cop car that would randomly start flashing the head/tail lights...RFI from the radar would trigger the BCM lighting.Fun,fun.
What did you have to give for the low fuel sensor? where,and is it the same as the 88?
Mine is removed as it failed,Weird piece of material inside.Resistance goes low when dry and hi when wet.I thought about putting in a mercury tilt switch....

I have a neighbor who is a HAM radio operator. I always know when he's keying up, because the paper shredder in my office (which has an "auto" setting) starts running as if I stuck a piece of paper in it.

The low fuel sensor is just a 1K thermistor. I got a Honda replacement for $40 (Cyclemax) before I found that out. When submerged in fuel it's cooled enough to have full resistance. When out of the fuel it warms up enough that the resistance drops, allowing the light to light up.

I have an LED on the dash, and it was lighting up even when it was submerged, so I put a 1K resistor in parallel with the LED to shunt some of the current around it. That works fine, but the thing doesn't glow very brightly when it should be on. Perhaps I'll cut out the 1K resistor and put a zener in series instead that will give me more of an all-or-nothing signal.

[virgilmobile]

Interesting...I think I've got a few around here somewhere to test with.Thanks for the details.
The last base radio I installed was in a unfinished security office at a Casino.
Antenna in the attic.
The last trip there,every time they keyed it,the paper towel dispenser would spit out one sheet.
I had to relocate the antenna.

[WingAdmin]

Hmm...now that I'm thinking about it...perhaps I have to have current flowing through the thermistor in order to get it to function, so a zener wouldn't work. I wonder if they count on the current flowing through the thermistor to heat it up, which lowers its resistivity, and allows enough current to flow to light the lamp? When it is submerged in fuel, that heat generated by the current would be dissipated in the fuel, keeping the resistivity high, preventing the lamp from lighting. That would make sense as to why the Honda manual says that the light should come on "within a few minutes" - the thing has to heat up. It would also explain why they seem to fail so often.

That said, I've got the 1K resistor in series with the thermistor now, which would only let 14mA through (plus whatever flows through the LED, which isn't much), so that's not going to generate much heat.

[virgilmobile]

I'll have a chance tonight to check.
I do think they use a current measuring circuit the same way as many cars do to indicate when there is a light out.
The circuit monitors current flow and volts.When there are both within specs,the circuit keeps the indicator off.
When either fail,the circuit triggers the light on.
I guess it may be that when the resistance goes up,it's like a lamp going open.All it has to do is reach a threshold voltage on a voltage comparator circuit.
So I don't think the actual sensor operates the light at all,and the initial lighting is a lamp test mode by the circuit module.
I know that if you remove the neutral lamp the others will not shut off.The lamps failed the test mode.
I'll experiment tonight.

[virgilmobile]

See what happens when you make me think without copious ammounts coffee.
It can't be a thermister.The resistance goes up when warm and thus the current should drop.The test is to short the terminal to ground to test the light.Meaning the resistance would have to go down and the current up.This leads me to one part.If I remember correctly,it's called a variactor.
I'll research tonight to find out.

[virgilmobile]

Ok,not a Varactor.I did find this article.It seems the most likely design.http://www.analog.com/library/analogDia ... nsing.html
The only thing now is to determine which one is used...the negative temperature coefficient (NTC) thermistors or the positive one.

[WingAdmin]

Ok,not a Varactor.I did find this article.It seems the most likely design.http://www.analog.com/library/analogDia ... nsing.html
The only thing now is to determine which one is used...the negative temperature coefficient (NTC) thermistors or the positive one.

Right, I was thinking that already, it has to be a NTC because it does lower its resistance when heated. A PTC would raise its resistance when it is no longer submerged, and it warms up, which would be backward from how it works now.

If I recall correctly, the circuit is pretty simple - +12 volts, through the lamp, then through the sensor, then to ground. Start-up lamp test grounds the sensor side of the lamp for a couple of seconds, to light the lamp. When the sensor is out of the fuel, by the book it's supposed to be from 900 to 1.3K ohms. Its resistance rises when it's cooled by the fuel. It's not the most ideal setup, as the light will come on more dimly when it's cold out, and will never come on full brightness by virtue that there will always be SOME resistance (being that the ambient air in the tank isn't going to be really hot - or will it?).

That's why I was thinking of a zener that would provide NO current flow until the resistance dropped enough that it overcame the zener, at which point it could perhaps trigger a small transistor that allowed FULL current flow, illuminating the lamp fully.

[virgilmobile]

I'm gonna test the theories this weekend.I'll hook up a pot to Simulate a thermister and record voltages and also see if the test does just manipulate the sensor lead.once I figure out the circuit,I can determine how to make it work.I don't like the "magic"explanation.

Well,I just went over the schematic and it does appear that the LCD module tests the light by grounding the lamp.It has 12 volts going to it from a relay and the other end goes to the little "magic" resistor.
Mine never worked,so I didn't realize that it would not switch between off and bright.
I'll continue with the test anyways,but in the meantime,I'll try to come up with a better switching method.I'll have to keep it small enough to fit inside the still well (original holder) or a add on switching circuit.

[WingAdmin]

Power comes in on the black/brown wire, to the fuel indicator light. The white/green wire goes off to the LCD unit at the top, where it is grounded during the lamp test. It also goes down through the 20P connector, where it is switched to a white/blue wire, which runs to the sensor in the tank - which just runs to ground through the thermistor.

FuelSchematic2.gif (2.78 KiB) Viewed 1814 times

[virgilmobile]

that's right.I can see where the lamp should never get real bright too either.
This is what I dreamed up in a pinch.It's a voltage controlled latch.When the volts fall below the threshold of the zener it should fire the trans that latches on the lamp.It won't cycle on and off or just glow,and has to have more fuel and the key cycled before it will reset.
The resistor values are up for grabs as is the zener value.I chose the 9 volt as 3.5 volts may not light the lamp too much.It can be changed after testing the voltage requirements.

This could be a neat little add on to the bike as 12 volts,ground and sensor leads are right under the seat.And it can be reset by operating the kill switch as it will kill volts to the pump(where you'd get the 12 volts from)

[WingAdmin]

A problem I see with your circuit is during engine start. The ground side of the lamp is grounded because of the lamp test. That means the zener is off, the lower transistor is not biased, which means your upper latching transistor is. From that point on the circuit is dead, because it's biased to ground, and your lamp stays on, regardless of fuel level.

Mine's going to be different because of the LED and the resistor I've got in the panel. I think I may just measure the voltage I've got at the thermistor terminal, then do it again when the tank is nearly empty. That will give me an idea of a zener value to use.

Hmm. I was originally thinking that I could pump +12V into the thermistor through a zener and use that to bias a transistor - but that could potentially put enough current into the thermistor to cause it to heat and fail in a bad way - creating a spark inside a nearly empty gas tank. That could be very bad. There obviously needs to be a fair sized resistor in series with the thermistor to prevent that.

Need to think about this some more.

[virgilmobile]

Sheesh.Just when you think...somebody comes along and whack......Hmmmm.unhook the resistor that operates the upper transistor and instead of 12 volts all the time,tie it to the acc line or the same circuit as the headlight.That would be a reset function and operate the circuit after the self lamp test mode.
Any ways....I don't even have one to play with....I'm looking into touch triggering a 555 to operate the lamp.

[WingAdmin]

Sheesh.Just when you think...somebody comes along and whack......Hmmmm.unhook the resistor that operates the upper transistor and instead of 12 volts all the time,tie it to the acc line or the same circuit as the headlight.That would be a reset function and operate the circuit after the self lamp test mode.
Any ways....I don't even have one to play with....I'm looking into touch triggering a 555 to operate the lamp.

Hey, there's an idea! A 555 to FLASH the light instead of just having it come on! That would get your attention!

[virgilmobile]

Yup,that's what I'm talking bout. blinking lights...who doesn't like them
Could be adapted to a working or non working unit.Just change the trigger method.
Mine has no continuity but still has the floating wire it the still well.Suitable for a touch switch.
More coffee needed now,brain hurts,close my eyes and see flashes of schematics.....got to focus.