Hello,
I recently bought a Facet and have really been loving it. I was thinking about getting a second unit and I really like the idea of the tilt compensation on the Torch. I was curious if there was any additional hardware changes expected from the beta version to final version other than the ones that are listed on the product page (screws and leds). Also is there any expected cost difference to final version? Thank you for your help!
Sincerely,
Matt
In my opinion, it will be wise to expect the price of the final version a bit higher.
Hi Matt - Almost nothing will change between beta hardware and full hardware. The expected changes are in the product description:
The PPS, Bluetooth, and battery LEDs are a different color from the final version.
The pocket in the carrying case is too tight, making removal of the device less than ideal.
These units have four T9 security screws securing the antenna lid. These will be replaced with Phillips head screws in the final version.
IMO, not a big deal or a big change, we just wanted to not have to document two versions (this one has a green LED, this other version has a blue LED), while also giving a useful life to the early units.
If we have left over beta units, weāll discount them. I suspect the full version will be priced at the same price as the beta version.
Thank you for your reply! I ordered one 
Looking forward to using it.
Sincerely,
Matt
Hey, I just received my Torch today and I got a chance to try it out and Iāve been running into a couple issues.
The accuracy is showing that it is somewhere around 28mm (converting from feet) when Iāve consistently been getting 14mm using the Lefebure Ntrip client and Qfield. Is there something that I need to change to increase accuracy?
When enabling the tilt compensation it seems as though Iām loosing the datastream because both swmaps and Qfield no longer show my position.
The GNSS indicator always seems to be flashing between blue and orange even when Lefebure is reporting RTK Fix. I read that this indicator should be solid when RTK fix is achieved. Any thoughts on this?
Thank you for your help!
Sincerely,
Matt
Hey,
I was just re-reading this and needed to clarify:
The horizontal accuracy on the RTK Torch is showing that it is 26mm when Iāve consistently been getting a horizontal accuracy of 14mm with the Sparkfun Facet using the Lefebure Ntrip client and Qfield. What do you think would be causing almost double the accuracy error?
Even after a few more attempts and following the Tilt compensation video IMU initialization steps Iām still losing positioning once tilt compensation is activated.
After further monitoring it appears that the led is solid orange flashing blue, so this may be the intended behavior.
I would really appreciate some help with points 1 and 2. I purchased this unit with the belief that accuracy would be better than the Facet and that tilt compensation would be functioning, but this hasnāt been my experience so far. Iām really hoping to get these resolved because I really like the build quality and goals this unit is trying to achieve. Thank you for your help!
Sincerely,
Matt
Hi Guys
First post. Not sure if I should start a new topic for the RTK Torch, battery replacement. I would think after a few years the LiPo cells will need replacement. How is this done. Tahks. Ian
Hi Matt, Iām also interested in purchasing the RTK TORCH. Have you been able to use it as an NTRIP base or with radio? Could you provide me with your contact information so I can ask you some questions about the equipmentās operation?
1. The horizontal accuracy on the RTK Torch is showing that it is 26mm when Iāve consistently been getting a horizontal accuracy of 14mm with the Sparkfun Facet using the Lefebure Ntrip client and Qfield. What do you think would be causing almost double the accuracy error?
2. Even after a few more attempts and following the Tilt compensation video IMU initialization steps Iām still losing positioning once tilt compensation is activated.
Interesting post, thanks Matt.
Your point 1. Does the accuracy reduce when you engage ātiltā mode or does it never get better then 26mm when using it in ānormalā mode too?
Your point 2. Does it always loose lock when you enter "tiltā mode. So form open sky and in āfixedā mode, then looses lock when ātiltā mode engaged still under open sky?
Iāve also noticed a decrease in reported accuracy with the Torch in comparison with the Facet.
Reported accuracy with the Facet has typically been ~14mm with RTK Fix for both horizontal and vertical.
The best Iāve been able to achieve with the Torch is 17mm horizontal accuracy and 29mm vertical. Although most of the time vertical is more like 60-70mm.
RTK corrections are provided by a CORS network. Rovers are approx. 9 km from nearest base.
Shooting a known benchmark, both the Facet and the Torch were within ~7mm ft of the horizontal position.
Vertically, the Facet was off 25mm and the Torch was off a whopping 75mm.
Torch ARP is set to 115 for L1 and 118 for L2.
Current UM980 config:
$command,config,response: OK*54
$CONFIG,ANTENNA,CONFIG ANTENNA POWERON*7A
$CONFIG,NMEAVERSION,CONFIG NMEAVERSION V410*47
$CONFIG,RTK,CONFIG RTK TIMEOUT 120*6C
$CONFIG,RTK,CONFIG RTK RELIABILITY 4 4*74
$CONFIG,BESTPARA,CONFIG BESTPARA 16.00 10.00*02
$CONFIG,PPP,CONFIG PPP TIMEOUT 300*6C
$CONFIG,DGPS,CONFIG DGPS TIMEOUT 0*6F
$CONFIG,RTCMB1CB2A,CONFIG RTCMB1CB2A DISABLE*70
$CONFIG,ANTENNADELTAHEN,CONFIG ANTENNADELTAHEN 0.0000 0.0000 0.0000*3A
$CONFIG,PPS,CONFIG PPS ENABLE GPS POSITIVE 200000 1000 0 0*69
$CONFIG,SIGNALGROUP,CONFIG SIGNALGROUP 2*16
$CONFIG,JAMMING,CONFIG JAMMING AUTO*2B
$CONFIG,AGNSS,CONFIG AGNSS DISABLE*70
$CONFIG,COM1,CONFIG COM1 115200*23
$CONFIG,COM2,CONFIG COM2 115200*23
$CONFIG,COM3,CONFIG COM3 115200*23
Changing the RTK RELIABILITY and SIGNALGROUP parameters didnāt appear to affect results significantly.
Next step is to set up a local base (within ~1 km).
Is the reported accuracy supposed to take into account the distance from the base?
With RTK Fix, the Facet almost always reports ~14mm regardless of distance to base. Doesnāt matter if its 10 km or 60 km away.
I wonder if Ubox and Unicore are calculating these accuracy numbers differently.
Could the helical antenna be negatively affecting performance in comparison with the Facet?
Hey,
@Victorjim05 Please feel free to send me a private message on the forum. I havenāt tried using it as a base yet. Iām lucky that I have base stations close by to my locations so I just receive corrections data over bluetooth.
@IanThomson I havenāt been able to achieve better than 22mm with the Torch in āNormalā mode even though I have been able to consistently achieve 14mm with the Facet. Iām hoping Iāll hear something from Sparky soon on the subject. When I enter tilt mode I lose position data, the only data that shows is horizontal and vertical accuracy.
@ssmcd I really appreciate you chiming in. My facet is consistently 14mm horizontal and 20mm vertical. Interesting the accuracy you have been able to achieve with the Torch. Iām usually between 22 and 26mm.
Iām using the a CORS network as well with the bases about the same distance that you are reporting. When I was checking accuracy with a NGS point I found that the Facet was 15mm off horizontal and the Torch was 30mm off. Iām very curious as to what environment Sparkfun was able to achieve the advertised 8mm accuracy. In one of the release videos it seems like they are near their base station and the Torch was only reporting a 12mm accuracy. Iāve also wondered about the helical antenna, but I trust that the Sparkfun team had done some thorough testing before advertising it as a better accuracy than the Facet. Iām hoping to hear something from them soon!
Have you had any issues with losing position data when you enter Tilt mode? Iām running firmware 1.0
Thank you!
Sincerely,
Matt
Hello,
Iām interested in buying an RTK Torch, but Iām not sure if the RTK torch is compatible with the South H6 or similar controller. Can anyone help clarify this? Since I am from Latvia, I also wonder if there will be any problem to fully use the RTK Torch here in EU? Thanks for your help in advance!
I look forward to some useful information>
Thank you,
Andris
Hi Shonky2 - Sorry I missed this thread notification!
The accuracy is showing that it is somewhere around 28mm (converting from feet)
8mm is reported by Unicore in their datasheet (https://raw.githubusercontent.com/spark ā¦ asheet.pdf).
download/file.php?mode=view&id=8984
This is ideal conditions. How far are you from a base station? I expect to see ~10mm if you are 10km or closer. I generally add 10mm to the (in)accuracy for every additional 10km, so if youāre 10-20km from base station expect 20mm+ accuracy. Iām going to run some additional RTK Fixed tests on the Torch here at SFE and report what SW Maps is reporting.
- When enabling the tilt compensation it seems as though Iām loosing the datastream because both swmaps and Qfield no longer show my position.
Help me replicate this. Are you connected over Bluetooth? Are you obtaining corrections over Bluetooth or Wifi or other?
- The GNSS indicator always seems to be flashing between blue and orange even when Lefebure is reporting RTK Fix. I read that this indicator should be solid when RTK fix is achieved. Any thoughts on this?
This is normal. The docs are not great (I just updated them to reflect this: https://docs.sparkfun.com/SparkFun_RTK_ ā¦ #power-usb) but you should see the GNSS LED blink once per second pretty much all the time (indicates GNSS fix). Once you have RTK Fix, it will illuminate green as well. So you should see a green light with a yellow blink on top at 1Hz.
Just as a reminder folks, letās not muddy threads by mixing questions. Iāll answer some of the questions inline but please start new threads if further discussion is needed.
@IanThomson - Battery replacement is very straightforward on the RTK Torch. Itās 7 Phillips head screws. We will be updating the repair docs (https://docs.sparkfun.com/SparkFun_RTK_ ā¦ air/#torch) over time as well as offering replacement parts (including batteries) for sale down the road.
@ssmcd - How are you configuring the UM980 in the Torch? It is behind the ESP32 (https://raw.githubusercontent.com/spark ā¦ c_v2.1.pdf) so sending config commands is not trivial. Also, the ESP32 will re-config the UM980 at every boot so Iām not sure your configuration commands are sticking.
@andkiv - The Torch will work in Latvia just fine however, the onboard LoRa radio is 902-928MHz so you wonāt be able to use the internal radio once we have it up and running (source: https://www.thethingsnetwork.org/docs/l ā¦ y-country/). Iām unfamiliar with the H6 data collector (https://www.southinstrument.com/product ā¦ d/199.html). The H6 has WiFi and Bluetooth 4.1 (which supports SPP) so yes, you should be able to connect to the RTK Torch just fine, but I canāt absolutely confirm without testing one. If you want to send me an H6, Iāll happily test and ship the H6 back to you.
Facet reported HPA vs Torch HPA (horizontal positional accuracy):
On the Facet we display what the ZED tells us its HPA is - this is nearly always 14mm for an OSR based RTK Fix. However, if you use SW Maps to inspect the HPA, it will report 10mm. This is because SW Maps calculates its own HPA based on the NMEA messages.
On the Torch, we display what the UM980 tells us its HPA is - this is nearly always 10mm when weāre testing in and around SparkFun HQ. Iāve seen this vary a bit more than the Facet based on the distance to the base. Iāll try to go for a drive later but Iām seeing 10mm solid today.
download/file.php?mode=view&id=8986
Notes to other readers: The best accuracy comes from a local base station (aka an OSR service https://www.geopp.de/ssr-vs-osr/), not an SSR service like PointPerfect. SSR is great but has less accuracy.
Iām setting up my new RTK Torch.
I cannot get it to access my WIFI network to get firmware updates or PointPerfect keys.
My SSID has spaces in it. I have tried the standard Linux escapes ( single quotes, double quotes, double single quotes and backslashes)
I still get Connecting WiFiā¦ No friendly WiFi networks detected.
WiFi not available.
Any thoughts?
Hey,
No worries! Thank you for getting back with me.
Ok, thank you for that clarification. It looks like the station Iām connecting to is about 13.8km away. So should I expect an HPA of approximately ~14mm? If so Iām still a bit off of that as the best Iāve seen so far is 22mm. Could the difference be related to the antenna design? Iām still trying to reconcile how the Facet is performing better when checking against NGS Control Points. Facet was within 0.05 feet of the point while Torch was 0.10 feet.
Yes, Iām connected to the Torch over bluetooth and using the Lefebure NTRIP Client for corrections data. Everything works fine up until the IMU is initialized after entering tilt mode. It is beeping every 10 seconds but QField and SWMaps arenāt showing any XYZ information. Also, when exiting tilt mode (single button press) I noticed that the serial data (was connected via Putty) said something along the lines of ESP Panic and no there was beep to indicate tilt mode exit. Also, running firmware 1.0
Thank you! After watching the indicator I noticed that this was most likely intended behavior. (LED Color different with my Beta unit)
I look forward to hear what you find in your testing in regards to the accuracy and if you can replicate the tilt mode issue.
@NerdyOldMan - Just as a reminder, letās not muddy threads by mixing questions. Please start a new thread if the question doesnāt pertain to the original forum topic.
I cannot replicate your issue: I can connect to WiFi networks with a space in the name just fine. I entered them using the serial terminal and the WiFi config page.
https://forum.sparkfun.com/download/fil ā¦ ew&id=8988
If you continue to have problems, please start a new thread or consider reporting firmware/bug issues on the repo (https://github.com/sparkfun/SparkFun_RT ā¦ are/issues).
It looks like the station Iām connecting to is about 13.8km away. So should I expect an HPA of approximately ~14mm?
The effect a long baseline has on an RTK Fix is far from linear; more distance = bad is all we can say.
Iām still trying to reconcile how the Facet is performing better when checking against NGS Control Points. Facet was within 0.05 feet of the point while Torch was 0.10 feet.
They are entirely different GNSS receivers with different antennas with a baseline greater than 10km. Your millage will vary greatly in this zone. But I love the idea of pitting the Facet against the Torch with a ~5km baseline. I shall work on this.
Everything works fine up until the IMU is initialized after entering tilt mode. It is beeping every 10 seconds but QField and SWMaps arenāt showing any XYZ information.
Odd. Iāll retest.
Also, when exiting tilt mode (single button press) I noticed that the serial data (was connected via Putty) said something along the lines of ESP Panic and no there was beep to indicate tilt mode exit. Also, running firmware 1.0
That sounds like a separate issue. Iāll also monitor the output upon exit.
Tilt is working fine for me along side Bluetooth. However, I am seeing weirdness when I exit (single tap) tilt compensation mode. Not a full reset, but Bluetooth definitely drops. Iāll dig in.
sparky:
@ssmcd - How are you configuring the UM980 in the Torch? It is behind the ESP32 (https://raw.githubusercontent.com/spark ā¦ c_v2.1.pdf) so sending config commands is not trivial. Also, the ESP32 will re-config the UM980 at every boot so Iām not sure your configuration commands are sticking.
I modified the [um980ConfigureOnce() function and recompiled the firmware, so the changes are persistent. Been testing out the various config parameters in the UM980 [documentation to try and increase accuracy. It looks like the UM980 is running an outdated firmware version, so at least one of the commands (enabling multi-path mitigation) doesnāt work. It seems FW updates for the UM980 arenāt possible yet on the Torch for the reason you described above. There is a CLI menu option that allows for ādirect connectionā with the UM980 via the ESP32, but itās somewhat crippled at this point and only allows sending a couple pre-configured commands. Iāll open up an issue on Github. It would be nice to have full bi-directional serial access to the UM980 for connection to UPrecise (and firmware updates). Not sure if this is on the radar for development - I might take a stab at it myself.
sparky:
On the Facet we display what the ZED tells us its HPA is - this is nearly always 14mm for an OSR based RTK Fix. However, if you use SW Maps to inspect the HPA, it will report 10mm. This is because SW Maps calculates its own HPA based on the NMEA messages.On the Torch, we display what the UM980 tells us its HPA is - this is nearly always 10mm when weāre testing in and around SparkFun HQ. Iāve seen this vary a bit more than the Facet based on the distance to the base. Iāll try to go for a drive later but Iām seeing 10mm solid today.
This is consistent with my observations so far. Iām curious how your testing goes at 5km.
shonky2:
Have you had any issues with losing position data when you enter Tilt mode? Iām running firmware 1.0
I havenāt experimented with tilt compensation mode yet but will take a look when I can the chance.](https://docs.sparkfun.com/SparkFun_UM980_Triband_GNSS_RTK_Breakout/assets/component_documentation/Unicore_Reference_Commands_Manual_For_N4_High_Precision_Products_V2_EN_R1_1.pdf)](SparkFun_RTK_Everywhere_Firmware/Firmware/RTK_Everywhere/UM980.ino at main Ā· sparkfun/SparkFun_RTK_Everywhere_Firmware Ā· GitHub)
I modified the um980ConfigureOnce() function and recompiled the firmware
You know what youāre doing and have joined a small group of people in the world (less than ten) that have compiled from source. Congrats! SparkFun should have achievement stickers or something! Nice work.
It looks like the UM980 is running an outdated firmware version,
Itās quite possible. Let me know where you are seeing recent Unicore firmware for the UM980. I also plan to add Galileo E6 PPP support shortly. Avinab (SW Maps author) reported getting RTK Float with no other corrections (!).
There is a CLI menu option that allows for ādirect connectionā with the UM980
Yes, that is indeed an outdate menu/feature that could use improvement. I welcome a PR.
@shonky2 and @ssmcd - Iāve pinned down and fixed the āexiting tilt causes system panicā bug (https://github.com/sparkfun/SparkFun_RT ā¦ issues/278). Fix is in todayās RC. You can load from the firmware update menu (be sure to enable beta).
5km and 11km reports to follow momentarily.