SparkFun Forums

[NateW]

Solid-state gyro sensors boards with analog or serial output would complement your accelerometer boards very nicely.

There are a couple of vendors for the sensor ICs...
http://www.murata.com/ninfo/nr0283e.html
http://www.epson.co.jp/e/newsroom/news_2004_08_02.htm
http://www.siliconsensing.com/

FWIW, I have used gyros in radio controlled helicopters and products based on the Siliconsensing devices seem to be far superior to everything else (well, maybe the $1000 sensors work better, but who cares about those?). The "center" signal for other sensors drifts over time - after 15 minutes the gyro indicates that it's rotating when it really isn't. If you integrate the signal over time (as good heli gyros do) the error gets really obvious. The SS sensors don't seem to have that problem. I have no idea what the individual parts cost but if they cost a few bucks more it would be worth it.

[sparky]

Sounds good.

How about the new ADXRS gyro from Analog?

We're trying out a BGA sensor. If we can get our reflow oven to solder it, we'll get a board up immediately.

Btw, that's a sweet SMD gyro from epson - Dec. 04 we'll look for it.

-Nathan

[SOI Sentinel]

One of my concerns with the Analog gyros is the cost. $30 each for the gyro itself in small quantities, $50 for a single sensor demo board. And of course, hand soldering BGA was the other.

I've looked at the Murata and the Silicon sensing. I know the murata's an easier package to work with, with similar accuracies. Heck, my dual axis MG100's from Gyration are about as sensitive, they just don't work past 3G constant (and I need this too!). I've heard the Silicon Sensing ones are more accurate. However, despite the couple of projects I've seen using them online (like the intertial measurement unit and guidance system for R/C hellicopters on sourceforge), I've NEVER found a vendor or even a price for them (esp. the murata).

As for the Epson sensor, very nice, although the max voltage difference of ±0.067V seems a bit small.

[sparky]

Yep - the epson part would need an external amplifier which sort of defeats the small size and low cost, atleast for our puposes.

-Nathan

[NateW]

I'm told Futaba uses Silicon Sensing devices, so maybe the best way to get one would be to get one of these and crack it open:
http://www.aeromicro.com/Catalog/futaba ... 002404.htm
That's an expensive way to get a sensor, of course.

As SOI said, the ADXRS sensors ain't cheap. But if it beats hacking on something even more expensive.

Could you boost the Epson's output to a usable level with an op-amp? Or would +/- 0.067v just get lost in the noise?

[sparky]

Cool part - but $117, whew!

67mV is not hard to pickup at all. The ADC on PICs is 10-bit over normal operating range = 5V / 1024 = 4.88mv per bit change. Not too shabby.

I'm still trying to guesstimate if 5mv is small enough for the ADXRS401. It has 15mv per degree change per second. I'm thinking I'm going to need an opamp.

Of course we can always shrink the ADC window to 2V with Vref+/- settings but this requires atleast a 16F88 and I'm trying to use a smaller/cheaper PIC.

Tradeoffs, tradeoffs.
-Nathan

Oh yea, new IMU board is up : http://www.sparkfun.com/shop/index.php? ... itemid=292

[NateW]

Nice work!!!

Two questions:

1) does the .75" height include the pins?
2) have you any plans for a device with two gyro axes?

[sparky]

.75" is the height of just the PCB - pins add another 0.2"

No plans immediately. I would really like to create a full 6DOF board with two IMUs and an addtional Gyro - all sitting behind a SMiRF to get wireless readings. Expensive, but an interesting research device.

-Nathan

[NateW]

Thanks for the info! I have a biped on the drawing board that needs a little more room inside to fit this thing. I think I can make it work though.