I'm sorry to report that FMA is no longer carrying Z sensors. They ran out of existing stock and because their basic Co-Pilot doesn't require them, they won't be making any more. They will continue to make XY sensors, which the Co-Pilot does use, and those should be back in stock in a few days.
We're currently exploring alternative sources for the Z sensors, but in the meantime, I suggest you buy them from Dean Goedde for $40 including cable (scroll to bottom for price list).[Update: Turns out that those may not be totally compatible (they're 3.3v and require special wiring to our 3.3v shield). Stay tuned....we're looking at other options.]
[Update 2: Dean writes and says that they are compatible: "My Z head is 100% fine at 5V. The analog output will be 1/2Vdd for a balanced horizon view, and skew away based on tilt. After all, myself, FMA, and Paparazzi are all using the same components and design. The AD8551 OpAmp is being used in all 3 of our Z heads and is fine to at least 5 volts.]
i think you should cancel your z axis order and I will call them up shortly after ;-)
so I assume this means we don't get to program our waypoints with the nice GIU interface that the 2.1 and above guys do? I have still to hook everything up but I am curious if the interface is different with each version.
On enquiry to FMA I received an email back today saying that they will ship my Z sensor and the XY sensor when the XY stock arrives. They predict later this week, lets live in faith :)
so what is the decision here? for those of us without the Z sensor. Jordi are you making some available for purchase? offline is fine with me. I need this up and running as soon as I can.... ;-) it's for a work demonstration.
William, as we've discussed before we will migrate ArduPilot to an IMU by the end of the year. The current board is powerful enough to handle the DCM algorithm. It will cost more than the thermopile, but only about 40% more.
The link I posted previously is for a "Atomic IMU 6 Degrees of Freedom". It has a Atmel ATMega168 so there's non need to process the gyro data. Sure in the long run re-designing the board will make sense, but what do you do with all of the current inventory? Hopefully Chris can make a PCB with thermopiles. Has the FMA design been reverse engineered?
Glad I ordered my Z sensor when I did. I've had it on my desk for a month now. The UAV project has fallen behind a little due to beta testing some products and lots to do at work. Will have to get back to it after the vacation.
I don't think it's an issue of cost William, I think it's an issue of computational power.
The cost difference is questionable I guess, the raw sensor cost of IR vs IMU (in the UAV DevBoard) is around the same, but then it depends on what gyros/accelerometers you use.
The ATMega's are only 8bit and the ATMega48/88/168/328 in particular can only do 20MIPS.
IR makes the attitude sensing pretty simple, just 3 ADC measurements and a (relatively) simple equation for pitch, roll, etc. With the IMU (gyros) you need to do quite a bit more to compensate for drift and all that comes with using an IMU.
The UAV Devboard is (from memory) 16Bit and can do 30MIPS, so has quite a bit more power than the ATMega's.
To do it properly, you would really need a separate processor (on a shield possibly like you say), which really just brings the whole project back to where it started (in a sense) as that's what the ArduPilot Pro was.
I think that the Ardupilot should just be converted to a 3.3v device (more common these days). Then you can use Deans sensors which are better than the FMA ones and "FMA Approved".
Gary, Yes! Balloons > UAV's.
As we both know, they also cost more, but are less frustrating (unless you build one) ;-)
Don't worry, we only used some of fat boy's advice, we are using a proper basket!!
Comments
so I assume this means we don't get to program our waypoints with the nice GIU interface that the 2.1 and above guys do? I have still to hook everything up but I am curious if the interface is different with each version.
thanks all!
-tychoc
The cost difference is questionable I guess, the raw sensor cost of IR vs IMU (in the UAV DevBoard) is around the same, but then it depends on what gyros/accelerometers you use.
The ATMega's are only 8bit and the ATMega48/88/168/328 in particular can only do 20MIPS.
IR makes the attitude sensing pretty simple, just 3 ADC measurements and a (relatively) simple equation for pitch, roll, etc. With the IMU (gyros) you need to do quite a bit more to compensate for drift and all that comes with using an IMU.
The UAV Devboard is (from memory) 16Bit and can do 30MIPS, so has quite a bit more power than the ATMega's.
To do it properly, you would really need a separate processor (on a shield possibly like you say), which really just brings the whole project back to where it started (in a sense) as that's what the ArduPilot Pro was.
I think that the Ardupilot should just be converted to a 3.3v device (more common these days). Then you can use Deans sensors which are better than the FMA ones and "FMA Approved".
Gary, Yes! Balloons > UAV's.
As we both know, they also cost more, but are less frustrating (unless you build one) ;-)
Don't worry, we only used some of fat boy's advice, we are using a proper basket!!
-
1
-
2
-
3
-
4
-
5
of 5 Next