I noticed that Digikey is now selling Honeywell's newest aerospace-grade IMUs, which cost $1,328 each (note that's just for the IMU; it's not a full autopilot). How do the specs of these aerospace IMUs compare to those we use here? Are they worth the extra money?
In terms of overall approach, the Honeywell IMU seem very similar to modern autopilots such as Pixhawk 2.x and 3.3: they both have MEMS sensors with internal environmental isolation and temperature compensation.
As for the sensors themselves, I'm no expert on specs, so I'll just post the basics here, comparing the Honeywell sensor to the Pixhawk 3.
On the face of it, the Invensense and ST sensors in the Pixhawk 3 appear at least as good, if not better. But I imagine that there are some other factors that may be more important, such as gyro drift and vibration filtering. The Honeywell specs in drift are shown here:
Meanwhile the Invensense ICM-20602 sensor in the Pixhawk 3 gives its drift in different units: ±4mdps/√Hz. I really don't know how to compare those.
Finally, I'm sure that a lot of the performance depends on the software running on the Pixhawk boards, be it PX4 or APM, both of which use GPS to augment the raw IMU data to compensate for drift, along with a lot of other smart filtering.
So for those IMU experts out there: how do you think these two approaches compare? Are aerospace-grade IMUs worth the extra money?
Comments
Your fridge might have the best compressor in the world without doors (north pole = great performance, desert = bad preformace)
Put your money where your mouth is and pay to verify that the system specs match the component specs.
I think I this is going of topic, will make a new post with facts.
Let's be serious. The specs are provided by the manufacturer using a valid power source.
Have you tested this assumption with an oscilloscope?
Assumption is free. (might be the difference in price tag!)
It assumed that the sensor has a sufficient power supply. Typically what you are saying Anthonie is a non-issue. Food in my fridge won't stay cold if I have no power!
I read through all of the apm2 sensor data sheets the other day trying to determine the accuracy. The interesting thing I remember is the the accuracy is tied to the quality of the power supplied to the sensor (the MEMS sensors have analog amplifiers sensitive to power fluctuations, even if the output is digital). So the specs mean nothing if you have noise on the power rail!
Note that there are various grades of these devices; commercial, aircraft, automotive, aerospace, etc.; see Vector Nav. Simply asserting the sensor is better because of specs in not the whole story. In addition, the high cost of the Honeywell IMU might be just marketing play (price is higher; must be better) and partly base don low volume costs.
Specs wise this Honeywell device rates in the commercial grade but manufacturing methods play a big part too. It is also true that the big IMU company's see their marketing share begin eroded by companies making $8 IMuy sensor chips (Invensense) that really are JUST AS GOOD!!
I just noticed that the Invensense ICM-20602 with its 0.004 deg/sec/root(Hz) is a single sided PSD measurement. This means that the ICM-20602 gyro angle random walk(ARW) is actually better than I showed in my early posts:
ICM-20602 ARW 0.17 deg/root(hr)
This is 3.5 times less gyro rate noise(ARW) than the best HG1120!
For single sided rate noise PSDs the conversion is to multiply by 60/root(2).
So the MPU6000 with it's 0.005 deg/sec/root(Hz) has gyro ARW of 0.212 deg/root(hr), only 2.8 times less than the best HG1120. :-)
The device itself must contain aerospace graded IC components inside. The components graded for aerospace are -55C to 125C (with a standard being -40C and military grade being -65C). Price difference is 20-40%.
Another thing is tripple a redundancy requirement, which must be realised via the separate ICs. So, unless they have IO pins eady for fault tolerance, the black box is completely overpriced.
Can I add in "supply chain" to the equation?
One of the slight frustrations of the open source AP's is that in the rush to add the next great feature hardware is quickly made redundant and, eventually, unavailable.
Compared with an IMU from a known, stable supplier developing products around APM/Pixhawk is taking a gamble on future availability
This all reminds me of a time before rigid-body dynamics was readily available for CGI and video game developers. It also reminds me of Arthur C. Clarke who said, "Any sufficiently advanced technology is indistinguishable from magic." If you didn't know how it all worked, it seemed really difficult. But once you look at the code, you might think, "Is that all?"
That said, what I'm seeing is every chip company on the planet has an IMU product. Oh, look, we have 9 degrees of freedom and the sensors are of the highest quality. So what? None of these products include source code for the Extended Kalman Filter. None of them show you how to properly calibrate magnetometers on all three axes. None of them deal with transient magnetic fields being generated by high current wiring or big motors. That, folks, is where the money goes.
I spent months with an MPU-9150 based device trying to figure out why I couldn't get the magnetometers to calibrate correctly. Turns out the the manufacturer's algorithm was totally wrong. They never fixed it. I ended up writing my own based on that Russian paper that's been discussed on this site.
-
2
-
3
-
4
-
5
-
6
of 6 Next