Looks like no one posted this yet. Good to see some innovation and improvements from HK
Improvements:
1) 100uf cap in Vcc and 10uf tantalum capacitor added to stop 3.3v bad gyro health issues
2) Replacement of the (not used/redundant) 1.0mm 6p GPS port with df13-4P I2C external compass port, putting the GPS and external compass ports next to each other, making for easier connections.
3) Multifunction MUX port added in the old I2C port location, this lets users select an output from the bonding pad (Uart0, Uart2, I2C, OSD “default is OSD”. This new layout makes data transmission more stable with an isolation resistance.
4) 3.3V and 5V Dataflash/Flashrom swapped trace problem corrected to the AT45DB161.
5) External Compass enable/disable jumper added, verses having to cut a trace wire.
Comments
Warning about goodluckbuy: I got some stuff from goodluckbuy before.
0) They sold me an APM 2.6, but it was really a 2.5 (I found out much later that there was an internal compass and I had to cut the trace). I told them, but they didn't seem to care.
1) 2nd set of 3DR radio clones didn't work
2) They sent me a GPS that was missing the GPS module (one side of the board was empty, no kidding)
3) It took over a month giving them a chance to fix stuff and they didn't fix it all. I got my money back from Ebay, and they banned my account from buying from there again.
So, with goodluckbuy, you do get what you paid for. Sometimes ok, sometimes not at all.
Sadly I also gave them feedback on things that could be changed in their kit (like not the right OSD cable), and they didn't seem to care. Again, furthering the not always deserved stereotype against chinese sellers.
Mine works. I'm using it with an external compass and the jumper removed.
Hobby King Live video with detailed explanation of updates to 2.7 board has been added.
The APM 2.x boards won't suddenly stop working. They will continue doing what they have been doing for as long as you like then to. So if you already have the features and performance you need, there is nothing to worry about.
The only difference is that in the future there won't be updates for the APM 2.x with new features every 2 weeks like we have been spoiled with so far. And even this has some nice benefits, since it will lead to a more mature and stable system.
All of that could be done off-board from the APM with a seperate unit, Ravi. Pixhawk and the MIPS race isn't the only way to go about it.
time is to move on to pixhawk. after all lot of hard work has gone into it. with 252MIPS and 168 Mhz we have a lot to look forward to. LIDAR, image processing. object tracking. only pixhawk can provide this.
I believe that if you take a distributed controller model of an UAS, the APM still has a lot to offer. Sure, it can't handle the EKF and image processing and and and.., but it is a known quantity with an increasing number of flight hours experience and with known issues. I think all the fancy stuff should be shovelled off-board from the APM to another high-power processor.
Once the clever algorithms have done their thing, they could feed the aggregated/regressed data back to APM as more Arduino-digestible messages that APM can act upon. When it gets into a tizz, the APM can continue to "fly the plane" (as student manned-aircraft pilots are exhorted and reminded to do.)
So, tell me, why is a Linux super computer guiding an APM a bad architectural model for future development...?
Should we better to wait next PX4 ?
"commit 8710922a02ae9c3dc9ae54121d2c4a1699f262f1
Author: Andrew Tridgell <tridge@samba.org>
Date: Fri Jul 4 13:58:38 2014 +1000
PX4: fixed orientation of internal hmc5883 on FMUv3"
Best not let mine know that either! I have ordered one of these HK 2.7s to see what all the fuss is about though!
Looks cool!
I hope the APM's stay around for a long time to come.
I don't use my Pixhawk much since there isn't an easy way to use the AUX pins as relays for CHDK.
-
1
-
2
-
3
-
4
of 4 Next