Recommend a good accelerometer

[daveyfire]

What consensus? the LCX wasn't even mentioned until your post.

I've also heard rumors that GWiz is no longer in production... Can anyone more informed than I confirm or deny them? I'm definitely eyeing MARSA as a replacement, Gus just loves them!

 

[cerving]

That price is from Analog Devices' web site, none of the ECIA suppliers are quoting prices or availability yet. That's what I meant by "readily available"... this is going to be a hot part when it finally comes out, though.

ADXL375 3-axis 16-bit 200 G digital accelerometer, $5.74 in quantity over 100 available ub September.

You don't need an accelerometer for a dual deployment altimeter. Current baro units work fine to 100 kft and use Kalman filters to eliminate Mach delay timers, and cost less than $100. Stick a $5 accelerometer into a $100 altimeter and you get a Raven, or similar altimeter for ~$150.

Bob

 

[troj]

I've also heard rumors that GWiz is no longer in production... Can anyone more informed than I confirm or deny them? I'm definitely eyeing MARSA as a replacement, Gus just loves them!

I don't know if they're still in production, or not. I have heard from several folks who've had problems getting in touch with them over the years.

As for the MARSA, I've never had a problem getting in touch with John. Very much on top of his game, when it comes to responding to customer inquiries.

I have one that I have very limited experience with. A guy in our club has a reasonable number of flights on MARSAs and really likes 'em.

-Kevin

 

[Larry Curcio]

Larry, no reason to get depressed.

Here is a plot of data from the MARSA54 on Gerald Meux min diameter N5800 flight at LDRS. He posted this data in his video file and was kind enough to share the data with me.

Blue is inertial altitude, red is barometric. Flight to 30K pretty decent integration wouldn't you say? He also used this data to characterize the motor and it estimated the thrust curve and total impulse of the N5800 within 10% of the certification values.

View attachment 143345

You're right. I have to look around more. Some instruments that are apparently very good don't allow export of the data, however. Nevertheless, thanks for the cheerful news.

 

[4kids49]

I don't know if they're still in production, or not. I have heard from several folks who've had problems getting in touch with them over the years.


-Kevin

Kevin,

I have contacted these guys within the last year. I traded emails with Robert Briodi. He moved to Michigan and took a full time job. G-Wiz is in business, but production is highly scaled back. I was hearing last Oct. that they were going out of business also, but they are not. They don't communicate too well. Robert sent me his personal email ([email protected]) you might try that if interested. The G-Wiz website lists some vendors, but most of these were out of stock. I helped someone locate one at Sierra Fox recently, but ships from overseas. Apogee just got some back in stock, so they are still around. Apogee only has a few units. Thanks.

Chris

 

[emckee]

With respect to using accelerometers for altitude determination, isn't there an implicit assumption that the rocket's long axis remains parallel to the vertical? As soon as the heading deviates, you will start overestimating the altitude, assuming your accel doesn't re-orient as a gyro would. I would think that comparing accel and baro data would give the best estimation of altitude, with the knowledge that neither would be perfectly accurate.

I did a bit of playing-around with an adxl345 and an arduino last year (low to mid-power stuff, since it's limited to 16g's) and got some interesting thrust curve and speed data, but without consideration of the other axes via vector addition, you won't have solid alt data, and even then, you have to make assumptions for a known thrust curve in order to determine orientation from the deviation. I didn't have much filtering set up (aside from the visual elimination of obviously erroneous data), but the accel data I had matched the published thrust curve quite well, so I was generally happy. My subsequent project is trying to integrate a 3-axis gyro (St micro L3G4200, I think) into the system, so you could get the heading explicitly. Or at least less implicitly.

ADXL375 looks like it uses the same interface as the 345, so I might have to upgrade!

 

[cerving]

Theoretically, you could build something to tell you where your rocket is going with a good single-axis accelerometer, a 3-axis gyro, and a compass. The compass is only needed to tell you which way the rocket is pointing pre-launch, the accelerometer and gyro can tell you (noise notwithstanding) how far it went downrange and in what compass direction from the launch site. You'd need some kind of radio, 3DR/XBee of course, to track it in flight and/or report back the landing spot. Compared to a GPS, it has the theoretical advantages of much higher update rate, and it won't lose lock during high velocity or high-G flights, and it should be cheaper to build too.

This is all theory, of course... I don't know if anyone has tried it yet.