Dear 3DR I love pixhawk FC and used it for years,, But Solo *********

[Rich]

Nonsense or not, it's what the 3DR tech told me in response to my question of why, when tested at the same time and in the same physical location (sitting next to each other), the Solo GPS takes longer to find satellites (often minutes), achieve lock, and give an OK to fly, where my 3DR uBlox used with my Pixhawk will get satellite lock, 3D fix and OK to fly in less than a minute.

If you think this is nonsense, then free free to take it up with 3DR tech support.

Here is a direct quote of his reply, cut from the support email

Regarding the GPS sensitivity, we're using the same, unaltered, "un-updated" GPS module since the first batch.
What is true is that we fixed some issues with the GPS signal being lost and the Solo handling that event correctly, but that's all software related.

I agree that it is a longer wait than with the uBlox module on our former products and, unfortunately, power cycling the Solo, as you do already, is the only troubleshooting I have for that.

I was as surprised by this answer as you seem to be, as I thought the GPS in the Solo was the same unit as the uBlox GPS that I have from 3DR, but I haven't disassembled them to compare part number or revisions numbers.

 

[Ian [P13]]

The tech isn't saying what you said in your summary at all, he's saying the hardware hasn't changed (true) the software has been improved (true), I did give you an example of one of the bugs fixed which he could have been thinking of.

An external GPS has less background noise so will at times lock that bit faster, but there's not usually all that much in it.

The pre-arm check in the Solo Arducopter is stricter than that in normal Arducopter as it checks the ublox sAcc (speed accuracy) value, so again the Solo would taken longer as it's being (rightly) more cautious as it relies more heavily on GPS for the smart shots. Normal Arducopter allows you to take off simply from HDOP being below 2.

Maybe I should apply for a tech support management role, however, that would mean me moving to Mexico and I'm used to the clouds and rain here in the UK.

 

[rrmccabe]

Maybe I should apply for a tech support management role,

That would be awesome Ian !

I don't put a lot of weight in what the 3DRs technicians say anyway. I have been given bad info on a number of occasions. If it comes from Vu then I would consider it good info. Otherwise it needs checked out

 

[RichWest]

Maybe I should apply for a tech support management role, however, that would mean me moving to Mexico and I'm used to the clouds and rain here in the UK.

Sometimes I wonder how you know all this stuff, very informative either way.

Tijuana for work, live in San Diego...best of both worlds....

 

[Ian [P13]]

Sometimes I wonder how you know all this stuff, very informative either way.

I had the Iris since that was launched so I learned a lot from that, I also pester Philip who designed the Pixhawk 2 if I hear contradictory information. Philip is straight up, so he's a reliable source for technical information.

I do get things wrong now and then, but I will own up if that's the case and correct any info I post.

 

[Ian [P13]]

I have been given bad info on a number of occasions. If it comes from Vu then I would consider it good info. Otherwise it needs checked out

I believe most of the tech support guys are required to have an engineering degree, so they'll be a smart bunch.

There's a lot to pick up and know with Arducopter and also there's some variation between the copters 3DR sell in how they behave, so it's inevitable advice now and then might not be 100% on the money.

 

[rrmccabe]

Well I am sure you will agree having an engineering degree does not always mean much. Not to mention the difference between customer service and tech support is pretty blurry for the average user as CSR and tech support gets grouped into one bunch by most people

All growing pains. I have no doubt.

 

[SloppyDingo]

The tech isn't saying what you said in your summary at all, he's saying the hardware hasn't changed (true) the software has been improved (true), I did give you an example of one of the bugs fixed which he could have been thinking of.

An external GPS has less background noise so will at times lock that bit faster, but there's not usually all that much in it.

The pre-arm check in the Solo Arducopter is stricter than that in normal Arducopter as it checks the ublox sAcc (speed accuracy) value, so again the Solo would taken longer as it's being (rightly) more cautious as it relies more heavily on GPS for the smart shots. Normal Arducopter allows you to take off simply from HDOP being below 2.

Maybe I should apply for a tech support management role, however, that would mean me moving to Mexico and I'm used to the clouds and rain here in the UK.

I'm curious what the difference between the previously used GPS modules and the new module used on the solo. I might have to compare some data sheets when I get home from work. Could be an indication as to why it takes so long to get a fix.

I was thinking last night that it could have something to do with the way the GPS shuts itself down. With the units we use at my work (Athena and h.764 LRUs) if you power up the plane in a location with no GPS signal, such as a hanger, the GPS will default to somewhere off the coast of Africa. I'm pretty sure this is true of all GPS. If you shut down the aircraft with the GPS in this state, it will record that location (almanac and ephemeris data) to NV memory. The next time you power up the aircraft in a location where there is good LOS for GPS it will take forever for the GPS to acquire because it has to download all new almanac and ephemeris data... Sometimes it can take hours! For this reason, we have a user enabled setting that stops the GPS from writing its current position to memory so that the previously recorded (good) position remains in memory. If this is used properly, once the aircraft is rolled outside and powered up it takes seconds for the GPS to acquire a good sat lock with a GDOP <1 as long as the plane is relatively close to it's last recorded location. If it gets shipped somewhere new, far away from the recorded position it doesn't matter what position data is saved, you will be waiting a while to get a fix for the first time.

I'm pretty sure this applies to all GPS. I also remember an engineer once telling me that the better the accuracy of the GPS unit, the longer it can take to get a good fix after a cold start. With the DGPS units we use it can take over 24 hours to get a fix on the first, initial power up of a new system. So it makes me wonder if the delays people are experiencing are due to powering up Solo in the house or in a location with poor signal strength or maybe flying one day in one location then flaying another day in a location that's a significant distance from the previous location.

If anyone wants to know more, here is a pretty good explanation of what I'm talking about:
jeepx: GPS Discoveries: Cold Start and Aided GPS

 

[Rich]

The tech isn't saying what you said in your summary at all, he's saying the hardware hasn't changed (true)

Ok, I thought that's what I said, I was referring to the GPS module (hardware), if someone interpreted that to include firmware/software, then I'm sorry about that. Of course we're all using the same software, or I assume we are, if everyone stays updated. With the 3DR system, It's kind of hard not to stay updated, unless an update comes out while you're in the field and not connected to the internet.

Tech support also told me (because I asked) that the 'cardboard mod' doesn't help in their opinion and testing. Now to be sure I don't misinterpret their words, and attract flames, here is a quote:

We're not doing anything different with the shielding.
We've seen the mods some customers have done, like the cardboard one, we have tested the Solo many hours in the RF chamber and we don't really see a difference in GPS performance with those mods.

I know some will say the cardboard mod helps them, but others have said it doesn't help them. One or two have even said that it made things worse (this was probably something strange occurring).

For me, I'm trying to get 3DR to tell me what to do to improve things. It seems to me they are telling me that what I'm seeing is normal for the Solo, but to me it seems that the Solo is sub-par to my other 3DR systems.

Could it only be the Solo software is too fussy and rejects the slightest glitch in the GPS signals? Yes, I agree it could, but that's for 3DR to say. For me, as a customer, it just looks like I'm spending a lot of time sitting around waiting for a GPS lock with the Solo, while my other 3DR GPS based multi-rotor would have been already flying for several minutes under the same conditions.

 

[Rodney Johnson]

If anyone has a bashed Solo, By now I'm sure there it's 1or 2.
I would like to know what happens if you move the stock GPS reciever about 4 to 6 inch vertical from its position just to get it away from other electronics and motor mag field..

If the signal gets better and things smooth out that would be a good indication that there is an internal RF or other noise issues coming from withing Solo causing signal issues... maybe changing out the reciever isn't the answer just relocation the stock unit in a vertical sence.

 

[Ian [P13]]

The cardboard mod is often being done wrong (the video showing the mod for one doesn't cover the error), people are still touching the feedline, I have posted warnings and corrections on a few threads here.

The information about the GPS shielding came from a 3DR hardware lead on the Mod Club, the best performance comes about making sure the feedline isn't touched and ideally the shielding is 4mm back from the ublox. Ian Soler's mod is about the cleanest I've seen, so is the one I'd recommend to people if they decide to do this, I wouldn't do it unless you have a problem first.

The photo above is a mod I did for a friend's Solo where I've made sure the feedline isn't touched and the 4mm gap is present.

Ian Soler's 3D printed mod:
Solo GPS Isolation Plate Mod Kit

 

[ChrisM101]

Well I can tell you other 3dr developers have stated on diydrones that the ublox7 was problematic, and it was a known issue for other projects. They chose to use the U8 in those projects because of the ordeal of trying to make the u7 work. They know it's bad hardware when used with ardupilot and arducopter etc.

 

[Ian [P13]]

Well I can tell you other 3dr developers have stated on diydrones that the ublox7 was problematic, and it was a known issue for other projects. They chose to use the U8 in those projects because of the ordeal of trying to make the u7 work. They know it's bad hardware when used with ardupilot and arducopter etc.

Post a link to a single hardware developer saying that.

Philip Rowse who's the lead on the Pixhawk 2 hardware has been recommending the 7 over the 8 since last year, well before the Solo was even known about.

I could show you numerous comments from people using the M8N having problems. The M8N failed 3DR's testing the, NEO-7N proving just as good in practice. The M8N has a wideband antenna which is susceptible to noise, so it's not a good option for placing internally.

 

[Rodney Johnson]

Post a link to a single hardware developer saying that.

Philip Rowse who's the lead on the Pixhawk 2 hardware has been recommending the 7 over the 8 since last year, well before the Solo was even known about.

I could show you numerous comments from people using the M8N having problems. The M8N failed 3DR's testing the, NEO-7N proving just as good in practice. The M8N has a wideband antenna which is susceptible to noise, so it's not a good option for placing internally.

View attachment 1265
View attachment 1266
View attachment 1267

I have flown with the M8n all summer now with pixhawk without a single issue.
But again it's in the air away from noise so mounting a 8 in the Solo is not the answer being so close to RF and other noise . Evidently the 7 isn't doing so redhot in the situation either..

 

[Ian [P13]]

I have flown with the M8n all summer now with pixhawk without a single issue.
But again it's in the air away from noise so mounting a 8 in the Solo is not the answer being so close to RF and other noise . Evidently the 7 isn't doing so redhot in the situation either..

I did suspect people on the Pixhawk that have been using the M8N were doing so on a mast.

 

[ChrisM101]

Post a link to a single hardware developer saying that.

Philip Rowse who's the lead on the Pixhawk 2 hardware has been recommending the 7 over the 8 since last year, well before the Solo was even known about.

I could show you numerous comments from people using the M8N having problems. The M8N failed 3DR's testing the, NEO-7N proving just as good in practice. The M8N has a wideband antenna which is susceptible to noise, so it's not a good option for placing internally.

View attachment 1265
View attachment 1266
View attachment 1267

Im sure Phillip would find it hard to admit that the location of the GPS internally of SOLO is a poor choice. You think he's going to admit the GPS failings of SOLO? If he's the lead on Pixhawk 2 it its his ass to blame for most of the issues, especially if Neo-7N was the outright best choice and its not a GPS hardware issue. Which is about as obvious as it gets since the thing seems to think it can move its GPS points 40+ ft at a time.

So if M8N is out and Neo-7N isnt getting the job done, someone should have sourced another or went with the M8N externally on a mast. seriously what is the cost to do that in manufacturing? $0.25 versus 3 and 4 exchanges .

People on other projects have had issues with 8 and 7. Yet someone thought putting it in and skipping GLONASS was a bright idea. It wasnt.

 

[Ian [P13]]

Im sure Phillip would find it hard to admit that the location of the GPS internally of SOLO is a poor choice. You think he's going to admit the GPS failings of SOLO?

Did you not read my post? Philip was stating that before the Solo was even a product, this goes back to his advice for Pixhawk 1 users making their own builds on the APM group, so his advice has been consistent.

Read Philip's post on GLONASS scenarios, it's not a magic bullet and can present its own problems.

 

[Rodney Johnson]

I did suspect people on the Pixhawk that have been using the M8N were doing so on a mast.

Yes it works very well on a mast. Again being out and away from anything that might cause interference is a plus. SPACE is not a luxury in the Solo so I can't wait to see how 3DR will get around it.

 

[ChrisM101]

Every one knows s/n matters but guess what more properly filtered signals equals less noise.

 

[Retired2Fly]

Lan and Philip are correct . The u7 with a narrow band antenna is often better suited for UAVs than the U8 using a generic broadband GPS/GLONASS antenna for the reasons already stated. Many here often confuse the number of satellites acquired (more satellites acquired is always better) as the single determining factor when determining the performance of a GPS or GLONASS module when in practice it is not necessarily always the number of actual satellites but the consistency of the positional data being received by the FC and the intelligence within the FC to process that satellite positional data combined with additional data form the IMU(s) and Baro(s) that is the real determination of positional accuracy when flying in GPS modes.