Highest Altitude GPS?

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I sent an email to Ublox a while back asking for SAM and CAM M9 modules...still waiting for a response.

SAM and CAM are going straight to M10. Will be available later next year. Max is available now for samples in M10.
 
I had one such flight. We were getting packets and then they suddenly stopped. We weren't watching the altitudes in real time though (tracking 3 gps transmitters), and for a few minutes, I essentially lost situational awareness on the flight. It was confusing when the packets stopped and then confusing again when they started a minute or so later.

Jim
Did you have real-time IMU telemetry so you could at least tell what it might be doing?
 
Did you have real-time IMU telemetry so you could at least tell what it might be doing?
No, I didn't have anything other than the gps. There were a lot of things going on with four pieces coming down. The radio I was using was tracking 3 gps transmitters and 2 RF beepers simultaneously, and we had another Marshall tracker running two transmitters as well. I was trying to move volunteer humans around in a circle permimeter to mark the directions to the various pieces as they fell. It was quite exciting. By the time I got to looking at the sustainer, which I thought has crashed or shreaded, it was still pretty high up. I sent Kip out to pick up what I thought would be pieces and I was surprised he came back with a complete rocket. That 50km limit was a big part of the confusion, and it would have been nice to know about it.

Jim
 
M10 has been optimized for ultra low power designs, not so much for performance. It is a nice upgrade over M8 though. M9 will continue to be the “power” receiver. Seems like they are really introducing two families that will alternate upgrades with each other. Both will continue to have pros & cons for our applications.
 
M10 has been optimized for ultra low power designs, not so much for performance. It is a nice upgrade over M8 though. M9 will continue to be the “power” receiver. Seems like they are really introducing two families that will alternate upgrades with each other. Both will continue to have pros & cons for our applications.
Good to know, thanks guys. Well, so much for that hope of awesomeness. I guess our hobby will take what we get & be happy with it. :)
 
A quick update on the Ublox M9N. I haven't been able to test its upper limits (that will be a long ways off), but I did swap out my Ublox M8N for the M9N on a flight this weekend. To keep it "apples to apples" I used the same patch antenna I have been using with the M8N and I am logging locally to SD card every two seconds. Overall, the M9N performed slightly better, but it didn't blow my socks off. Here is a summary of the comparison between two flights with the same profile:

Flight: A 4" rocket launched to about 15K feet with a 4 second motor burn time (7700 N-sec), max speed of 1,250 Mph (1833 fps), max gee force of 23 G's, and 30 seconds to apogee.

Both GPS units consistently track a solid 12 satellites, while waiting on the pad, but both lose GPS shortly after take-off.

M8N: The M8N loses GPS immediately after launch and regains it about 28 seconds later with 4 satellites, two seconds before apogee. It takes a full 42 seconds of descent (avg 75 fps) to get back to a full 12 satellites.

M9N: The M9N held on to reception longer following launch (about 6 seconds) and reported 2D coordinates with only 3 satellites, but then also lost reception. It came back online 10 seconds before apogee (with 4 satellites), a full 8 seconds earlier than the M8N. Following apogee, it only took 10 seconds for the M9N to regain the full 12 satellites.

So, with the sample of one, the M9N seemed to perform better than its predecessor, but it still drops out during the high speed portion of the flight. I'll keep flying with the M9N and see if the trend continues.
 
So, short of being tracked by radar, is the only current way to know what altitude you've reached above 250,000' is by 'calculations' based on speed/time at last known altitude and time to return?

Also, what would be the most reliable way to deploy drogue at or near apogee...accelerometer? timer?
 
Last edited:
So, short of being tracked by radar, is the only current way to know what altitude you've reached above 250,000' is by 'calculations' based on speed/time at last known altitude and time to return?

Also, what would be the most reliable way to deploy drogue at or near apogee...accelerometer? timer?

The Mx150 “Kate 2“ product from Multitronix will report GPS altitudes above 250,000’. With the optional Mx152 pyro board it can also deploy the drogue right at apogee using those GPS readings.

www.Multitronix.com
 
The Mx150 “Kate 2“ product from Multitronix will report GPS altitudes above 250,000’. With the optional Mx152 pyro board it can also deploy the drogue right at apogee using those GPS readings.

www.Multitronix.com
The Mx150 “Kate 2“ product from Multitronix will report GPS altitudes above 250,000’. With the optional Mx152 pyro board it can also deploy the drogue right at apogee using those GPS readings.

www.Multitronix.com
So with the optional pyro board it will record/transmit apogee altitude and fire drogue at say 340,000' or more?
 
So with the optional pyro board it will record/transmit apogee altitude and fire drogue at say 340,000' or more?

YES. The GPS has unlimited altitude capability. All data is transmitted to the ground in real time. That data is also recorded on both the transmitter and receiver. The pyro board will use GPS altitude as the primary source for triggering apogee deployment. You also have the option of enabling the accelerometer, timer or even manual deployment as backup contingencies.
 
A quick update on the Ublox M9N. I haven't been able to test its upper limits (that will be a long ways off), but I did swap out my Ublox M8N for the M9N on a flight this weekend. To keep it "apples to apples" I used the same patch antenna I have been using with the M8N and I am logging locally to SD card every two seconds. Overall, the M9N performed slightly better, but it didn't blow my socks off. Here is a summary of the comparison between two flights with the same profile:

Flight: A 4" rocket launched to about 15K feet with a 4 second motor burn time (7700 N-sec), max speed of 1,250 Mph (1833 fps), max gee force of 23 G's, and 30 seconds to apogee.

Both GPS units consistently track a solid 12 satellites, while waiting on the pad, but both lose GPS shortly after take-off.

M8N: The M8N loses GPS immediately after launch and regains it about 28 seconds later with 4 satellites, two seconds before apogee. It takes a full 42 seconds of descent (avg 75 fps) to get back to a full 12 satellites.

M9N: The M9N held on to reception longer following launch (about 6 seconds) and reported 2D coordinates with only 3 satellites, but then also lost reception. It came back online 10 seconds before apogee (with 4 satellites), a full 8 seconds earlier than the M8N. Following apogee, it only took 10 seconds for the M9N to regain the full 12 satellites.

So, with the sample of one, the M9N seemed to perform better than its predecessor, but it still drops out during the high speed portion of the flight. I'll keep flying with the M9N and see if the trend continues.

I’ve been off the grid for awhile on this but holee cow Mike, this is incredibly helpful to me and others. I was aware of the loss of GPS reception due to speed limitations and thought Rf reception issues were a big deal.

I think your data might suggest if the rocket is changing geometry quickly (spinning or tumbling) it makes it a bit more difficult for the GPS to get a lock in some of the latter phases of flight. Once it settles down under drogue or main, a better lock can be had.

This is exactly what I’ve seen with even much lower altitude flights. Loss on ascent and then a variable re-acquisition of signal on descent. Sometimes seems like a long wait to me. Nonetheless, enough positions are received to find the rocket. Never lost one with GPS tracking except with a core sample. In that case the metallic paint inhibited the Rf from getting out. Was my first flight. Got the remains back a year later. A later core sample got one position out before impact and that’s where the fincan was found sticking out of the ground. New nosecone, new tracker and the rocket still flies. (Was fiberglass rocket)

Will have to keep the M9N in mind for possible modifications for trackers in the future.

I have some AP510’s https://www.ebay.com/itm/AVRT5-APRS...507521?hash=item4220e1c141:g:LgsAAOSwzkld1XJV
I’ve been itching to fly and yes it‘s on the Ham band but can be switched from 500mW to 1 watt power.
Only issue is since it’s APRS the best refresh/transmit rate is once every 5 seconds. Not a problem with the intention of recovering the rocket but not ideal if one would like to have a live data stream with the recovery of as much data as possible. With a bit more Rf output, it helps with reception at a distance. Also Yagi antennas on the 2 meter band have a wider bandwidth so it is easier to point the Yagi in the general direction of the rocket. Even on the 70cm band (400Mhz) a Yagi is easy to point and can give a more reliable downlink at a distance.
The 900Mhz, 100mW trackers do have their limitations and I wouldn’t recommend that one primarily depend on them for extreme high altitude flights. For sport fliers (like me most of the time) they’re fine. The Multitronix Kate can put out 1 watt on 900 Mhz hence the exceptional performance that is seen.

Your research shows some of the limits and will allow others to stop chasing zebras with data recovery.

Did some live map tracking with APRS and the NMEA one position per second trackers. Especially if one has a photo map program it’s incredible to see the tracked rocket positioned on a map while in flight. Using APRS and the Linux program Xastir, there will be a rocket icon shown with a bubble that shows bearing to, GPS altitude and horizontal speed right next to the icon in live real time flight! Can record the flight for later replay in real time. Is that cool or what?
Once one’s really high flying rocket disappears on ascent, might as well watch the laptop screen for the first incoming position packet to be plotted. We’re talking soo high that the apogee/drogue event/smoke can’t be seen from the ground.

It got to the point where dual boot Windows/Android tablets were coming out of China. I took a chance and snapped up two different ones that had an onboard GPS chipset in them. They were hard to find. They work. Can do Windows tracking programs or boot to Android and use the GPS rocket tracker program. I have the devices connected to an outboard battery as the computing burden has a significant drain on the onboard battery. With onboard GPS, I don’t have to dink with connecting up an outboard GPS to the tablet to get the local position when proceeding to the downed rocket. That helps a lot although some of the outboard B/T GPS units I’ve tested on plain windows laptops have their own battery supply. No big deal as like mentioned, I just plug in a larger battery and carry the shebang in a one handed box.

If one gets beyond a bearing arrow and distance, it’s pretty neat to track on a live map with the ability to store the flight for later plotting.

Kurt
 
Don't blame you as that is proprietary information. Doubt any hobbyist is going to copy your efforts! (Can't be too careful though;))

Ummmmm, The 60,000 feet limitation on GPS should be workable for 99.99% of fliers. I believe some chipsets maintain a lockout if one is going above 1000 knots or are above ~59-60k feet and the unit will transmit nothing. The deal here is that some GPS chipsets will transmit properly above 60k as long as the speed is less than 1000 knots.

Some years ago I used to follow the flights of high altitude APRS
(Automatic Packet Reporting System) balloons that were GPS tracked and flown by Ham radio aficinados. Picked up positions sometimes from my home station directly from the digipeater that was 1200 feet from my house that had a 100 foot antenna in the air. Some of the balloons were 450 miles from me but hit the digipeater directly from altitude. The digi re-transmitted the data I would pickup on my home receiver/laptop.

The balloon folks know which chipsets would work above 60k because their balloons were not certainly traveling at rocket speeds. Many times they'd get above 100,000 feet and get accurate positions. The key there was the chipset used was an either/or proposition. Fly slow and up high and positions would be transmitted just fine.

Some chipsets would lockout at greater than 1000knots OR > 60k feet no matter what. These are the ones the high altitude balloon hobbyists avoided like the plague. They wanted a chipset that would work at >60k as the balloons flew very slowly. Actually, a balloon at 100,000 feet sort of hangs at one place in the sky as there aren't really any high winds at that altitude. Little air to move in a near space environment. Put a balloon at about 40k or less and sometimes the jetstream hits it and away it goes at 175mph! Makes for a long chase until the balloon bursts!

After launch, once the rocket slows down re-acquisition of the data link will occur as long as the rocket is not tumbling so badly that makes it hard for the GPS to re-acquire the satellite signals and/or the Rf power of the transmitter is enough to make it back to the receiving ground station. If the rocket gets too far away for the the Rf to make it to the receiver, the signal will be lost.

Kurt
 
A quick update on the Ublox M9N. I haven't been able to test its upper limits (that will be a long ways off), but I did swap out my Ublox M8N for the M9N on a flight this weekend. To keep it "apples to apples" I used the same patch antenna I have been using with the M8N and I am logging locally to SD card every two seconds. Overall, the M9N performed slightly better, but it didn't blow my socks off. Here is a summary of the comparison between two flights with the same profile:

Flight: A 4" rocket launched to about 15K feet with a 4 second motor burn time (7700 N-sec), max speed of 1,250 Mph (1833 fps), max gee force of 23 G's, and 30 seconds to apogee.

Both GPS units consistently track a solid 12 satellites, while waiting on the pad, but both lose GPS shortly after take-off.

M8N: The M8N loses GPS immediately after launch and regains it about 28 seconds later with 4 satellites, two seconds before apogee. It takes a full 42 seconds of descent (avg 75 fps) to get back to a full 12 satellites.

M9N: The M9N held on to reception longer following launch (about 6 seconds) and reported 2D coordinates with only 3 satellites, but then also lost reception. It came back online 10 seconds before apogee (with 4 satellites), a full 8 seconds earlier than the M8N. Following apogee, it only took 10 seconds for the M9N to regain the full 12 satellites.

So, with the sample of one, the M9N seemed to perform better than its predecessor, but it still drops out during the high speed portion of the flight. I'll keep flying with the M9N and see if the trend continues.

I missed this post earlier. I have one of these units but have yet to integrate it onto a board. I'm curious about your test conditions:

1) Was it configured for a 4G Airborne environment?
2) What constellations did you have enabled? All 4 GNSS or just the defaults?
3) What update rate did you set? It can do up to 25 updates per second, and if you had it at 25, I'm curious if it actually achieved it.
4) What is your assessment of the ground track accuracy? Did it look about right, or were there unusual deviations?
 
1) Was it configured for a 4G Airborne environment?
2) What constellations did you have enabled? All 4 GNSS or just the defaults?
3) What update rate did you set? It can do up to 25 updates per second, and if you had it at 25, I'm curious if it actually achieved it.
4) What is your assessment of the ground track accuracy? Did it look about right, or were there unusual deviations?
Hi Sparky,

Yes, 4G Airborne. All 4 are enabled by default with the M9N. I am updating at 19hz, but I only log and transmit GPS data every 5 seconds in my solution, so I don't have a high resolution sample set. I do get data at 19hz, as I can see the time to get new sentences is about 48ms. In my limited datasets the ground track accuracy is pretty good. There is a setting to enforce 3D quality (vs. 2D with variable altitude), so sometimes the altitude is wonky coming out of the launch phase. With 3D quality set it takes longer to lock.
 
Hi Sparky,

Yes, 4G Airborne. All 4 are enabled by default with the M9N. I am updating at 19hz, but I only log and transmit GPS data every 5 seconds in my solution, so I don't have a high resolution sample set. I do get data at 19hz, as I can see the time to get new sentences is about 48ms. In my limited datasets the ground track accuracy is pretty good. There is a setting to enforce 3D quality (vs. 2D with variable altitude), so sometimes the altitude is wonky coming out of the launch phase. With 3D quality set it takes longer to lock.

Would it be possible to have enough onboard memory to log at the higher rates and still transmit once every 5 seconds? Me thinks that would be accuracy to the point of absurdity but perhaps interesting to download the data after a flight. Memory size and latency might be a problem but I'm not an electronics geek.

The Ham radio APRS regime trackers are indeed once every 5 seconds. It takes time to transmit the data. I did take a stationary laptop, radio, TNC and the Xastir APRS tracking program to transmit once every 3 seconds and it worked on the decoding end of the APRS packet with a Kenwood D72A. If I cut it to 2 seconds or 1, the packet transmission couldn't be completed before the next set of data was composed hence nothing was transmitted. (I call it packet "collision" and the radio couldn't transmit it fast enough.)

I had to go into the code by rote and change "5" to "3". It worked but pretty hard to fit a laptop, full sized TNC and full sized radio into a rocket just to transmit an APRS packet once every 3 seconds. ;) It was a fun exercise during my downtime to show me the limitations of APRS.

That said, the NMEA trackers like the Eggfinders transmit the raw NMEA sentences once a second. They are limited a bit by the 100mW Rf output but work really well with finding sport rockets. Reception can be improved on 900Mhz by using what is called a patch antenna on the receiving end. I'll post a picture later of the one I use. (Too dark to take a pic now.)

The units are economical and easy to experiment with. During flights, the positions are locked out during the high speed portion on the upside and then it takes awhile for positions to start coming back in after apogee. When I say "awhile" it means several seconds but during a flight it seems like forever.

I've been told that the doppler effect can have a bearing on the reception end of GPS signals. Plus positioning of the tracker inside the rocket might make a difference. Most of us have the GPS patch facing out to the side of the rocket as that is the way to mount the board primarily. A straight up mounting of the GPS patch receiver would give good view of the sky on the upside but after apogee deployment would be facing the ground and not so good for reception. Facing out the side is the best compromise.

With all of this said, I didn't see much difference between the once every 5 seconds transmission of position data via APRS trackers and the once a second transmission of the NMEA trackers. The positions disappeared on the upside of the flight as expected and took varying times for the datalink to be re-established on the "downside" of the rocket flight. Either way, once the rocket settled down under drogue or main chutes, positions were transmitted to effect the successful recovery of the rocket. The 2 meter and 70cm Ham band APRS trackers are likely better in range over the 33cm/900Mhz unlicensed trackers but sport fliers won't notice the difference. Extreme flights expected to land many miles away would benefit from a 2 meter/144Mhz or 70cm/400Mhz tracker. Except if they invest in the "Kate" system they would be good as it can output on 900Mhz with 1 watt of Rf. No other 900Mhz system I know of does 1 watt output.

Anyone who flies on the salt flats knows that the salt absorbs Rf like a sponge. The ground range of a tracker is ridiculously small as is the range of handheld walkie-talkies. Up in the air a tracker range is nominal as it is flying anywhere. Out on the salt it's the last known position as the rocket is still up in the air that's the key and allows for successful recovery. Once on the ground, the footprint of any Rf tracker is poor as opposed to landing in the dirt of a cornfield.

Kurt
 
Last edited:
Would it be possible to have enough onboard memory to log at the higher rates and still transmit once every 5 seconds? Me thinks that would be accuracy to the point of absurdity but perhaps interesting to download the data after a flight. Memory size and latency might be a problem but I'm not an electronics geek.

Yes, some systems can do this. I code my own system so I do exactly as you suggest: capturing all the data onboard at the highest possible rate, but transmitting the GPS coordinates over the air only 5 times per second. I only transmit the coordinates, not the full NMEA sentences.

There is a setting to enforce 3D quality (vs. 2D with variable altitude), so sometimes the altitude is wonky coming out of the launch phase.

This is good to know. I'll code that up and see if it makes a difference. I definitely see "wonkiness" in the altitude data right after launch, mostly in flights that exceed 8Gs or so. Interestingly, I notice a difference between the MAX and NEO modules. The NEO seems to hold up a lot better. I wonder if the integral SAW filter and LNA are accounting for this difference.
 
Interestingly, I notice a difference between the MAX and NEO modules. The NEO seems to hold up a lot better. I wonder if the integral SAW filter and LNA are accounting for this difference.

Is it an M8 Max vs the M9 Neo? The receiver works by recording what looks like random noise and then testing that noise to see if it is correlated with the PRN code for the satellite it is searching for. It shifts that recording in time and Doppler space until it gets a valid peak correlation. Then it attempts to track the signal by searching a smaller space around that peak for each consecutive time period. If there is a big jump in position or speed, the signal will be outside of that search space and it will fall back to a broader search to try and re-acquire. Since it's now a moving target, it will take a lot longer than a normal stationary hot start. I would guess you're seeing improvements in the correlator speed not so much the RF path. You'd still get dropouts with perfect signals unless the receiver is designed to accept that environment.

One of the really interesting upcoming applications is GPS nav for small cube sats. Obviously most of us don't need orbital grade hardware but hopefully we see something greater than "airborne, 4g" in the future...
 
Is it an M8 Max vs the M9 Neo?

I see the difference between the MAX-M8Q and NEO-M8N/Q. Only a handful of samples to compare on different days, so it could be random. Thats very interesting about the cube sats. It would be nice.

I'm going to program them to force 3D mode and see if there is a difference. It shouldn't be that hard and I'll test it at this weekend's launch.
 
Hi Sparky,

Yes, 4G Airborne. All 4 are enabled by default with the M9N. I am updating at 19hz, but I only log and transmit GPS data every 5 seconds in my solution, so I don't have a high resolution sample set. I do get data at 19hz, as I can see the time to get new sentences is about 48ms. In my limited datasets the ground track accuracy is pretty good. There is a setting to enforce 3D quality (vs. 2D with variable altitude), so sometimes the altitude is wonky coming out of the launch phase. With 3D quality set it takes longer to lock.
A bit late, but how is the neo m9n going? Did you try some of the the sam-m10q modules? They are a fair bit cheaper and have the same specs as far as altitude and speed limits go.
 
A bit late, but how is the neo m9n going? Did you try some of the the sam-m10q modules? They are a fair bit cheaper and have the same specs as far as altitude and speed limits go.
I have been using the Ublox M10S exclusively the last year with a Maxtena "Helicore" antenna and I am very happy with it. I do not know its true highest limits, but for all my normal flying it performs exceptionally well. Below is data from a launch last month that was "low and slow" -- about 500 mph and 8700 feet. You can see I start on the pad with 22 satellites locked and I only drop to 19 sats for two seconds and then come back up to 22 sats for the duration of the flight. I didn't hit a high enough velocity to get a user limit lock-out on this flight, so the drop was all doppler drift. Starting with 22 sats and only losing a few during motor burn is exceptional.

Screenshot 2023-06-21 at 5.54.48 PM.png
 
I have been using the Ublox M10S exclusively the last year with a Maxtena "Helicore" antenna and I am very happy with it. I do not know its true highest limits, but for all my normal flying it performs exceptionally well. Below is data from a launch last month that was "low and slow" -- about 500 mph and 8700 feet. You can see I start on the pad with 22 satellites locked and I only drop to 19 sats for two seconds and then come back up to 22 sats for the duration of the flight. I didn't hit a high enough velocity to get a user limit lock-out on this flight, so the drop was all doppler drift. Starting with 22 sats and only losing a few during motor burn is exceptional.

View attachment 587788
Interesting, thx for the data/info. Why didn't you use the sam-m10q?
 
The uBlox "SAM" series includes a module-mounted patch antenna, the "MAX" series does not. If you're going to use an external antenna anyway (and it's hard to beat anything made by Maxtena) then there's no point is spending the extra money for the SAM module.
 
Back
Top