22 May, 2019 at 14:26 #62014
@kevin_erik yeah that’s what I actually did. I suspect Holybro changed their configuration at some point and GetFPV didn’t realize it.
@novulon Yeah, the LIS3MDL is what comes in my old PixHawk GPSs if I’m not mistaken. Unfortunately, the custom dRonin build that I’m using that was made by @jihlein only supports the HMC5983 and HMC5883 at the moment.
23 May, 2019 at 03:22 #62024
- This reply was modified 3 months ago by swissfreek.
Hey I bought the gps/mag combo from the amazon link given in this thread. Can anyone confirm the wiring to the kakute for me thanks.23 May, 2019 at 04:15 #62027
Mine gets here Friday but the product listing looks like it has the pins labeled at the connector. I use UART6 for the GPS (so T6 -> RX on the GPS and R6 -> TX on the GPS), VCC to any 5V, GND to any GND, and SCL and SDA are on the bottom of the board (not through hole, just pads).23 May, 2019 at 19:31 #62039
Thanks Swissfreak. The module comes with a connector or on the bottom it has matching solder pads. Since I managed to roll my chair over the included harness and wrecked it I will be using the solder pads.
I wasn’t sure we needed to connect SCL and SDA and what the voltage was. Thanks for the info. Is it SCL-SCL and SDA-SDA I assume?23 May, 2019 at 20:30 #62040
The existing driver for the LIS3MDL is an SPI driver, which will not work with external mags, as they are typically connected via I2C. A driver would need to be written that supports the I2C interface.
It looks like modifying that existing HMC5883 driver would not be too difficult, but I don’t have a QMC5883 to test with. I’m betting the GPS/Mag unit I returned to the manufacturer for non-functioning mag was in actuality a QMC5883 and I didn’t realize it. This certainly explains a lot of the issues I was seeing early on, as I’ve used that HMC5883 driver before and know it works…..
On these GPS/Mag boards, the GPS is serial, and the mag is typically I2C, so you need all 6 connections (including power and ground). It is SCL-SCL and SDA-SDA connection wise.23 May, 2019 at 21:08 #62042
Thanks jihlein and swissfreak. Much appreciate all your hard work in getting this adopted into dronin!23 May, 2019 at 21:32 #6204324 May, 2019 at 01:37 #62044
Ok I guess you can stack the GPS antenna on the clean side of the PCB. Do you solder it at the corners where they match up with the solder pads? Or just double sided tape and clear heat shrink?24 May, 2019 at 01:39 #62045
Personally I’d go with Option 2. Not sure why I’d ever want to take it back apart, but I like the thought that I’d have the option. Multiple solder points can make things mighty hard to remove.24 May, 2019 at 05:33 #62046
@swissfreek – shoot me and email to jihlein at hotmail.com26 May, 2019 at 03:53 #62069
Hi guys I am finished with the build can you point me to the latest dronin inav compatible tricopter firmware thanks.
26 May, 2019 at 04:21 #62071
- This reply was modified 3 months ago by billyd.
If you’re using Windows, I believe this post has your link (Page 2 of this thread, post #61725):
If you’re on MacOS, for sure the link you want is two post below that one. That’s what I’m currently using (and as of yesterday I’ve got a functional GPS and magnetometer using the one @jihlein linked).26 May, 2019 at 05:19 #62074
Yeah I need to get this new build pulled together when I get home. I’ve got a possible fix for the altitude solution drift. I’m going to pull out the SmartAudio for now because I can’t make it work (I’m unsure if it’s the code base I’m using, my VTX, or some combination of the two). I’ll leave the rest of the OSD menus in place.
Regarding the SmartAudio, I can see the proper msg formats going to the VTX, and I get a response, but the response is garbled and is not accepted. I don’t have a true TBS VTX, I’m using a HolyBro Atlatl HV V2 FWIW (supposed to work with BetaFlight SmartAudio). When I get my Tri LR fully wired, I’ll try it with both BetaFlight and iNav and see if that tells me anything.
I also think I’ve got a driver that will work with the QMC5883. Thanks to @swissfreek, I’ll be able to test it.26 May, 2019 at 05:53 #62075
I would bet a not-insignificant sum of money that your problem is the VTX. There was an issue in the past with SmartAudio on non-TBS VTXs. Don’t quote me directly on this, but as I recall the situation went somewhat thusly: before TBS made the protocol “openly” available to third parties (or it was third parties who didn’t want to go through the process, or pay the money, to formally get access to the protocol, but anyway), those third parties (I believe it was AKK that did it first) reverse-engineered the SmartAudio protocol. And it worked. Only TBS had a bug in their protocol and so they reverse-engineered the protocol with the bug. Then TBS found that bug and corrected it, and so those VTXs no longer worked with SmartAudio because anyone who implemented SmartAudio correctly made it incompatible with those VTXs. I think in some cases, the companies that found themselves in this situation (like RaceDayQuads, who licensed the AKK VTX and were at the center of the whole “controversy”) replaced the early runs of their VTXs with updated ones that properly implemented SmartAudio (the whole thing is what led TBS to make the protocol open, or “open-ish” at least), but I sort of stopped listening after that because it was just drama and I only use TBS stuff anyway, so it didn’t affect me.
So there you go. I don’t know if the Atlatl was one of those affected VTXs, but it, along with the AKK VTXs, were some of the first non-TBS VTXs to have SmartAudio, so it’s very possible that they fall into that category.
Would you like me to send you a Unify to test with? :-p
26 May, 2019 at 16:58 #62082
- This reply was modified 3 months ago by swissfreek.
“Would you like me to send you a Unify to test with?”
If you don’t mind, it would be great to see where this issue lies and if it could be made to/or does work.
You must be logged in to reply to this topic.