I'm working on an inclinometer for my Xterra. I want to be able to incorporate it into my digital dashboard. I've looked on the internet for inclinometer manufacturers, but the prices are outrageous. I've decided to build my own. What I need is somebody like Electroken(hint, hint)to help me with the circuitry. I plan on building a capacitive resistance type. I'll build a circular tube with three magnet wire sensors and filled with glycol. Any suggestions on what current to run through it and how to change the output to digital using a microprocessor? I will be building one for each axis, so a circuit that can handle two inputs would be ideal. If I get this working, I'd be happy to help figure out how to make a stand alone unit for those interested. Thanks in advance.

Ok, I've been out in the garage experimenting with wire and glass and have determined that I can probably build one of these:


We use gallons of glycol at work(don't ask). I can get some large normal saline ampules that have expired from there as well. I should have a working model in 24 hours. Please let me know if any of you electrical engineers out there can help me. Let me know what info you need to make this thing tell me what angle my truck is leaning. I love Macgyvering crap like this.

My fingers are burned to a crisp, but I have one that is working intermittantly. I'll continue working on my technique and maybe I'll get something more dependable. Is anyone interested in this or should I delete this thread?

Always interested in people coming up with new ideas for the X. Just is a bit to high tech for me.

Don't delete this thread... I like Macgyvered stuff.

OK, now I know why they're so expensive. I think I'm going to give up on making my own for now. I have rigged an old RC servo to do the job. I just ripped the guts out and am using just the pot and gears to damp movement. I attached an old fishing weight to the servo arm to allow the weight to adjust the resistance. Now I just need that electronics help to change the output to digital. Any takers? I guess I can figure it out on my own if I have to. I'd like to bring the info into my computer via the LPT1 port. Since I don't have a printer the port is unused. On a side note I pulled my dashboard out of the truck this weekend and took measurements to replace it with two touchscreen LCDs. Can anyone say GPS in the dash? I think it will be a bit better than indiglo and will be completely customizable.

Xterrian,

I've got an old mechanical clinometer/altimeter pod from an older 4Runner that you might be able to hook up some electronical doo-dads to to take a reading from somehow. Hell, you can even see the internals of it here:
http://4crawler.cruiserpages.com/4x4/CheapTricks/index.shtml
Click on the Clinometer Tweaks link.

You want it, I'll send it to ya, no charge.
Brent

Thanks for the offer Brent, but being the high tech geek that I am, I insist on an electronic solution. I called Fredericks Sensors this morning and talked with one of their engineers. He pointed out the exact part number I'll need for my application and then connected me with sales. A very nice lady with a Bavarian accent talked to me and offered to send a free sample. I explained that I was an individual doing a one-off, but she insisted that I wouldn't pay for it. So now I need circuitry gurus to chime in. I actually got the very same sensor that I posted in the picture above, only with a metalic casing to withstand heat and not shatter in an automotive application. Two axis of measure + - 60 degrees for the price of none! I can't wait to get it! Thanks again for your kind offer Brent.

Xterrian,
Take a look at this it is a four channel A-D converter. Should do everything that you want. Now with this it connects to Dallas Semiconductors propritary 1-Wire bus but from there you can go to serial or printer port.

-Mike

That looks like a begining. I'll need to do more research before I commit to any one idea. It took me most of the day yesterday and all night just to pick a sensor type. Thanks for the link. It is now bookmarked.

OK, it looks like, from my research, that I have to build an oscillating circuit with a schmitt trigger. If anybody here has ever heard of these things let me know. Otherwise, I will have to go spend time on an electronics message board. I really don't want to go somewhere else and be the newbie that gets told to do a search. I will if I have to, but I'd rather not. Any takers?

If I understand correctly the principle of the sensor you want to use, it's a glass ampule filled with a conductive solution. There are 4 wires around and one in the center, the more a wire is immerged in the solution, the more current (or voltage) it conduce.

I this correct ? If yes, I can help you with the circuit. Again, if I understand well, you want to do two separate circuit (one for the left-right, the other for front-rear).

Xterrian,
I'm sorry that I haven't read this thread until today. Here at work we began a tilt sensing project that used the exact same glass sensors that you're trying to use. It got shelved for cost reasons.

The math needed to make sense of the signal from the center pin of those sensors gets hairy. It turned out to be too much for any microcontroller in our price range. Another obstacle is that we were trying to resolve angles to 1/10 of a degree.

Only one A/D converter is needed. A square wave is applied alternately to the pins at opposite corners of the sensor. The resulting signal at the center pin carries all of the information you need to derive the tilt in two axes.

Some advice about that sensor:
The output from the center pin is high impedance and must be followed with a good op amp before you go anywhere with it. The distance between the op amp and the sensor must be short. Don't apply DC to the sensor or you'll kill it.

Another and possibly more suitable sensor is the ADXL202E from Analog Devices. Its output is two PWM signals that carry the tilt data for two axes. This device is much easier to tame.

Thanks to 20-100 and Electroken for responding. I left out the DC killing the sensor part in my posts just to make sure that the person responding knew what the hell they were talking about. I looked into the ADXL202E, but it seemed to me that it was actually an accelerometer and could have problems centering properly if started while on an incline. I may be wrong, but I couldn't get anybody where they were sold to state emphaticaly that this wouldn't be a problem. I'll be happy with a +-2 degree solution, so it doesn't seem like that hard of a sensor to get usable results for me. If it's too complicated to do it myself, I can always purchase their solution for $100. I was trying to keep the project under $100 total. I'll still have to build a serial adapter and buy a serial to usb adapter as well. I'm easily looking at $150 if I go with their solution. There is a downloadable data sheet at the Fredericks Sensors site I posted above, with test circuits and some information that may be helpful in building a reasonable circuit. Electroken, you are the shiznit when it comes to electrical questions here, so if you say scrap it and go with the ADXL202E, I probably will. Let me know what you think. I am probably getting too high tech anyway. What do you think about the gutless servo with just the pot and gears to damp movement?

Yeah, the ADXL202 is an accelerometer but in this case it's measuring the accereration due to gravity. In my evaluations of the part I found it to be quite repeatable. Once the microcontroller was taught where straight and level was it always found it again.

Grab application note AN715 from Microchip:

http://www.microchip.com/1010/suppdoc/appnote/all/an715/958/index.htm

It's interesting stuff.

I bow before you ElectroKen.

And maybe you can answer me: The circuit I was thinking about was derived from a Wheatstone bridge to detect very small changes on resistances between the side pins and the center pins. Did running a very small DC current (in the range of 10 micro amps) can ruin such a sensor ?

I don't really know the threshold for damage to the electrolytic sensors due to a DC bias. The folks that make them are very secretive about the magic goo and aren't too specific about the mode of destruction.

Very interesting paper, but once again it never specifically states that the zero will be there if the device is powered down and up again. I think they are fuzzy on this information on purpose. Yes, once it's calibrated at zero, I have no doubt that it will be able to tell when it gets to zero again. My problem occurs when I park on an incline and get out of my truck to take a picture or spot terrain. I get back in and start the truck while sitting on a 10 degree slope. Now the inclinometer thinks I'm at zero and I'm ten degrees off until I stop again.

Xterrian,

The ADXL202E that Ken suggests is an absolute, not a relative, device. The zero is always "at the same spot". And since it's two-axis, you can also determine your longitudinal pitch, and therefore, with a bit of math, your power...

Regards,
Richard.

My power?

I've tried the ADXL202 and I'll swear on the hood of my X that it will still know what's level even after you cycle the power to it. It has some other handy features too, like being able to low-pass its response right on the chip to reduce output changes from bumps and vibration.

I'm sure that many people have made the electrolytic sensors work, but I found them to be a major pain in the ass and not worth the trouble.

Added for clarification:
The microcontroller that decodes the PWM signals is what actually remembers where "flat" is upon power up.

Can you get me one, or suggest a supplier?

The whole part number is ADXL202AE. It comes in a leadless SMT package so I hope your soldering skills are good!

I have two samples coming, which is the maximum that Analog Devices will send because they're $12 each. I'll gladly send you one of them. You can order 2 more of your own as samples if you can sound reasonably convincing at their sample request page.

Go here:

http://products.analog.com/products/info.asp?product=ADXL202

Okay,okay! Your vehicle's rear-wheel power, then!

Regards,
Richard.

Does the sample include the micro controller? What else will I have to buy to make an inclinometer that can input to the serial port? Do I have to get a hexidecimal converter? $12 isn't bad at all. I won't lie to them to get free samples, my integrity is worth more than $24. I'll try calling and see what they say when I explain what I'm doing. Is anybody interested in a stand alone unit with a character LCD output?

Here is where the hard stuff starts. The ADXL202E comes with nothing - it's going to come as a lonely IC in a anti-static box. No microcontroller, no PC board, nothing.

Now you'll have to round up all of the stuff you'll need to build the rest. If you want to use the Microchip application note as a starting point you'll need most of the parts shown in the schematic (PIC16F84A, crystal, voltage regulator, etc.) less the LCD. In place of the LCD you'll need some circuitry for the RS232 interface to your computer. The PIC they used doesn't have a hardware UART so I'm digging through my stuff looking for some serial bit banging code for the PIC. Between the app note code and the software UART you'll have all the code that's needed.

Will I need some way to feed the code into the controller? Do I have to buy a programer? Thanks for all of your help Ken.

Yes, you'll need a programmer. Fortunately the PIC16F84A is a very popular part in the hobbyist commununity and there are several low-cost programmer kits out there. You will also need the code assembler from Microchip called MPASM. They give it away on their web site.

dang you guys are good.... !

Xterrian, if you are going to go "glass cockpit" then perhaps this will work for you. Might blow your budget though!
Crossbow Gyro AI

Yeah, that's basicaly what I'm building. No heading info though, except for my GPS. I guess I could add that in to my program. Wow, one program accessing three serial port devices, that's just asking for trouble.

What do you think of this Ken? Do you think I can put together something similar for less than $100. That's all they want for it, and it seems to meet all my requirements.

Kayak, I just noticed that if you go to the second page of the pdf file linked to above, there is a picture of the Crossbow software running. Coincidence? I think not.

I prowled around the xbow site some more and it looks like they have a model for auto use. It also looks like you can download the software for free but I didn't try.
Anyway, I'm interested in what you are doing. At one point I was going to mount an old vacuum gyro driven AI that had been decertified for flying. Not much panel space for a 3 inch gauge though. They are pretty chunky.

Another link for you of somebody doing inertial guidance using the chip you are looking at using. You could browse his code etc. IGS

That's the evaluation board that Analog Devices makes to get engineers familiar with the part before they start their own designs. If you're only building one it's nice because you don't need a PIC programmer and you don't need to learn PIC assembly language. It's probably a faster path to your goal and hard to beat for $99. It could be all you need but you need to know some of its limitations.

Unlike the PIC16F84A in the app note circuit, the PIC16C63 has no non-volatile data memory. This means that the calibration values and offsets for "level" will have to be stored in the host PC rather than on the eval board. Probably no big deal.

The other limitation is that it uses the old version of the ADXL202 which is less accurate that the "E" part. The old part may be just fine for an inclinometer.

Damn Electroken, I wish I could give you another 5 rating. You are one hell of a resource for the Xterra community. Your encyclopedic knowledge of elctronics is priceless.

Well, I've gotten my tilt sensor samples from Fredericks. They look so cool, and the circuits to use them look so simple. An oscillator to give it some AC then an anolog to digital converter to give me PWM and then into the computer as hexidecimal value. No worries about it knowing which way is up. Fredericks even sells a board for it that puts out PWM for $100. I don't need perfect accuracy, so I believe I might be able to do something cheaper. Electroken, can you recommend some basic electronics books for building oscillator circuits? I really want, "The Art of Electronics" by Horowitz and Hill, but it's way too expensive. Know where I can get a used copy cheap? Sorry Ian, I want to support authors, but for a hobby, I can't pay $70 for a book. I already spend over a thousand dollars a year on medical textbooks.

Give me some idea of the frequency, duty cycle and amplitude you're after. There are many options.

I have the book you mentioned. You can't have mine.

I ordered a used copy for $25. I'm looking at running about 1KHz. The test circuit says 25 Hz to 20 KHz. Not sure about the duty cycle or amplitude. I picked 1 KHz out of my butt so I'm open to suggestion. My total lack of electronics knowledge makes it easy for me to believe it will be easy to do this. Maybe after I get my copy of, "The Art of Electronics" I'll realize how hard it really is.

Just so those that are interestd still know I'm working on it. I've included this link to a guy that built one for his personal radio telescope.
I'm still working on digesting, "The Art of Electronics", but I can only spend so much time in the bathroom.

Interesting link. $100 for only one axis.
How sensitive is your device going to be to acceleration? Dip a wheel in a gully and your reading could swing off the chart unless you damp/avg it somehow.

ok i just wanted to throw a post in here and say.

Ken, you are the man... i only understood about half of what you typed, and i find that damn impressive

Just to let you know Kayak, I still plan on building my own. It's just nice to see what others have done. Since I've never even seen an inclinometer in person, other than the Gyro actuated ones in small aircraft, it's good to get the nitty gritty on all of the options. I could probably build an anolog one for $25. I just want to integrate the output into my virtual dashboard. I'm still looking for a vga capable LCD to replace my real instrument cluster. I need something 15 inches wide by 5.5 inches tall. Multiple LCDs are ok. Please don't send me links to video input LCDs, VGA only please. I will nickel and dime it together until it works like I want it and looks stock. I just painted and mounted a touchpad next to my steering column and the cordless mouse is in the back seat for the kids. I have both of them working together at the same time. I just don't want to post too many pictures until I'm happy with things.

Very cool. When I talked to you at GOX I never thought you'd try to go through with the virtual dashboard. A lot of work, but I'm sure it will be worth it in the end. I'm looking forward to seeing pics. I just bought a new computer and am thinking of trying the indash computer thing with the old one if I can fiberglass some sort of mini case. Good luck!
John

Here's a crappy photoshop version that gives an idea of what I'm trying to do. The actual screen would be able to be customized for GPS, etc... No DVD to the dash screen for those that were worried.

xterrian, you are a sick puppy.

But i like it

Xterrian, why not rename your Xterra "KITT" and program in some AI so it will interact with you and make wisecracks as you drive.

When I was working on voice control of the computer I did have it saying some funny things back. "Oh great master of Visual Basic, I have started Mapsource as you ordered". The cabin is too noisy for current voice control technology though.

i can see it now...

"hello Randal"

"hello HAL"

"what are you doing"

"driving Odessa canyon"

"i can't let you do that"

this could get ugly

Hey Ken,what do you think about this cicuit for my electrolytic sensors?

Edited to add:
I ordered samples of the LCD and ADC from Maxim. Can you help me decide which PIC I need. There are too many choices when I put in PIC16C54C with Microchip.com

At first I couldn't understand why someone would use a separate A/D converter when most of the PICs have them built-in. The answer was in the code: it was written by a Maxim guy to demonstrate the Maxim A/D converter. Definitely the hard way. The code seems to support only one axis.

The PIC16C54C shown in the schematic is an OTP meaning "one time programmable". You have one chance to get it right or it goes to the landfill. They offer UV-erasable windowed parts for code development but you need a UV eraser to use them. The '16C54 is the oldest surviving member of the PIC family. Newer PICs are flash-based and can be reprogrammed many times.

If your heart is set on the '16C54:

Erasable = PIC16C54C/JW
OTP = PIC16C54C-04/P

My heart isn't set on anything except getting an inclinometer. Can you recommend a flash based PIC with on board ADC that can fit in this schematic or one like it?

Here is where engineers get their reputation for being "difficult to deal with", "not a team player", etc. Pointing out all of the alternatives and their consequences is often interpreted as making things hard.

Option 1: Use design as-is.
This is possibly the quickest path to a goo-sensor based inclinometer because the hardware and software already exist. This is assuming that the posted code is correct. One of the cons is that not all hobbyist device programmers support the PIC16C54. Another is the OTP aspect mentioned earlier.

Option 2: Replace PIC with a flash-based A/D PIC.
A more elegant but more difficult solution. The PIC16F818 includes the A/D converter, but you will need to write the code to run the A/D converter. Since it is flash based it can be reprogrammed without an eraser. Most of the '16C54 code will drop in. This option is only valid for someone who wants to master the PIC instruction set.

Option 3: Lose the goo-sensor.
They work great on the bench. Interpreting a signal from a sloshing vial of goo in a moving vehicle is another story. I'd hate to see all of your effort result in something that doesn't do the job.

"Goo sensor"
Sorry but that just made my day.