Heltec HT-M2808 RSSI sensitivity

As other people have mentioned on discord and other forums, the Heltec hotspots seem to underperform compared to other brands.

I’ve made a test setup for a side-by-side comparison:

  • 2x sensecap
  • 2x heltec

2 meters spaced apart, all have 6.5dBi antenna and are on the same height.

I have run this setup for about 1 week:
The Sensecap earned ~1.5HNT per day each.
The Heltec earned ~0.2HNT per day each.

Checking the RSSI data I found that the Sensecap saw beacons with RSSI as low as -123.
The Heltec devices don’t have beacons below RSSI -95.

So I have setup a LoRa transmitter to check the raw log page of the dashboard.

To my surprise, the Heltec doesn’t sense anything below RSSI -95!
As soon as output drops below it, the hotspot simply does not see this data.

Are the developers aware of this issue?

[edit]
I just got the name of a hotspot of another Heltec that had low income but no issues before with another brand on the same location: it has exactly the same issue.

It seems that some hotspots don’t have an issue with receiving anything below -100 ~ -113 but some simply cannot see any LoRa packets below -95 to -100

4 Likes

Thank you for your feedback. We will do a detailed test on Monday.

Can you please tell me the two Heltec Indoor Hotspot’s names? That will help us analyze the problem.

For what its worth here is one of my results. Screenshot%20(202)

Antenna is 6dbi on the end of 10m of LMR-400

Yes, it seems that not all Heltec Hotspots are having this issue.

I have a small list of hotspots which cannot detect anything below -95dB to -100dB while others -like yours - can. This could be a hardware manufacturing issue as they all share the same firmware.

You can clearly see this back in the rewards. Your hotspot is doing very well, while the hotspots that have this issue are only generating 1/5th of the rewards they should earn as they miss a lot of beacons.

2 Likes

I have sent you more detailed information with hotspot names through email.

I’m experiencing this too, I have a seperate post about it although noone has replied to me. I am hoping someone will look into it this week and respond.

I am 99% sure this is a Heltec issue as I have other manufacturers in neighbouring hex’s with exactly the same setup variables performing 10x better.

I am trying to perform an analysis on the RSSI/LSNR of heltecs, if you wanna help pls send me the log of your heltec devices (both the working ones and not working ones)

Here you have my first analysis:


So here we have:

  • 2 working hotspots --> Dancing-Skarlet-Meerkat and Refined-Ultraviolet-Rook
  • all the other ones do not work properly and it looks like there is a problem with the LORA transceiver as their RSSI is not correct even though other hotspots in the same area show correct values. Is it maybe something that can be corrected with a calibration?
    Specifically, Gentle-Peach-Skunk just replaced a Synchrobit hotspot which was perfectly working, no changes in the setup (antenna, connection, etc.) were made.

I can share the analysis if you send me an email, its a simple spreadsheet which contains the lora logs from the Heltec devices.

2 Likes

Hi, I had looked into this issue in detail today. We found something strange:

  1. let’s look into this beacon and witness log (The two Heltec Hotspots mentioned in this topic are also in this witness log):
    https://explorer.helium.com/txns/s0bMu_0Op38FFqqew4x19L_jtKDRwwv8ZN24ELT0JsQ

Does anyone find something strange?

Whether it is a long-distance or a short distance, the RSSI is always between -115 and -120, which is more like taking a random number within a certain range… The distance of 900 meters is -115dBm, and the distance of 62Km only becomes -116dBm.

This beautiful place looks flat on the map, but the communication distance of 62Km still subverts my understanding of LoRa communication technology.

However, in a hex grid next to it, there are indeed two machines whose RSSI is too low (-121dBm) and the witness is invalid.

However, these invalid witnesses are closer to the data we actually tested (higher).
image

They are invalid because they do not satisfy Helium’s algorithm for PoC signal strength.


At the same distance, the higher the RSSI value, the better the radio frequency characteristics and impedance matching of the circuit.

Regarding the RSSI, we grabbed some Heltec Hotspot which approaching -120dBm.
image
image
image
image
image

Regarding the HNT earn… It has something to do with the network status, the deployment of other miners around, and luck… Yes, it has something to do with luck (such as your miner connected to the P2P network which is closer to your location and faster). Such as this one:

Or these from other manufactures:


The total number of HNT that can be mined every day is about 50,000, so ~0.18 is the current average level. If each miner can earn ~1 HNT per day, where does the extra HNT come from? These 3 FAQs may make sense to you:

1 Like

Personally feel that such a test is not comprehensive…

  1. At the same distance, the smaller the RSSI value, the better the impedance matching and signal strength of the circuit;
  2. It is possible that the lowest power that your device can transmit is limited (we assume it is a fixed value), and the lowest value may be submerged by clutter and noise. After this lowest value is received by the LoRa gateway, it will be converted to its respective corresponding RSSI value;
  3. You can try to set up the same antenna in the same location, and then use the same LoRa node as the transmitter to compare which gateway can cover a longer distance.

Yes, as I said: some heltec devices have no problem spotting lower than -100dBm. However, some cannot. The hotspot you are linking to is one of mine. It is a genuine hotspot with 6.5dB antenna. There is no data manipulation. The -115dBm at 900m versus -116dBm at 62km can be explained as well: one has a stock antenna while the other antenna is on the highest building in 200km area with good view to us. There is nothing strange.

Our country is very flat and we have direct line of sight to the other side of the country!
Sometimes we reach 236km with a 6.5dB antenna.

I fully understand PoCv10 rules. However, my complaint is not about this.

My complaint is about the Heltec LoRa transceiver is unable to pick up beacons with a RSSI lower than -100dBm. Not about valid or invalid beacons.

Yes, and you can find just as many hotspots which cannot sense lower than -95dBm.
We are showing you this in a graph!

This is not true.

Like I say in the start post: we have created a test setup because we are concerned.
We have identical test setup with 2 heltec hotspots and 2 sensecap hotspots spaced 2 meter from each other.

The sensecap are sensing more beacons, with lower RSSI and are generating 5x more rewards.
After multiple days of testing you should expect that the heltec can pick up the same beacons as the sensecap. They cannot! They can only sense >-95dBm.

It is no coincidence when there is a significant difference in behaviour between to devices.

Again, you are missing the point we are making here. Some Heltec receivers have a sensitivity issue.

1 Like

We own LoRa transceivers with adjustable attenuators which can attenuate the signal between 0dB and -65dB. Before sending a LoRa packet we lower the attenuator, then send the package and then see how the heltec and sensecap are receiving it.

This effectively mean that we can send out a signal to the hotspots that will be received as RSSI -60 to -125dBm. Below -95 to -100dBm the Heltec stops receiving LoRa packets while the Sensecap can still see them.

Yes I know you showed us a picture of a Heltec that can sense -112dBm packets. However, check the chart that @mrblue1995 has made. Multiple Heltec devices never see packets that go below -95dBm.

There is clearly a lot of variation in the receive sensitivity of your hotspot and this is in my eyes a defect.

This is a very comprehensive test as we made the setup for both Heltec and Sensecap exactly the same(same cable, same antenna) and the LoRa transceiver is at a fixed distance from both the Heltec and Sensecap hotspot.

Is there any way we can directly talk through chat? I feel the forum is not the best place to explain it.

@Supporter

Here is a schematic overview of how we setup the test.

  • All antenna identical
  • All LMR400 cable identical length
  • All vertical height identical
  • Horizontal distance as specified

We went from -60dBm to -125dBm in steps of - 5dBm by sending LoRa packets from transmitter on the right to the 4 hotspots on the left.

When RSSI dropped below -95dBm Heltec #1 and #2 stopped receiving LoRa packets.
Sensecap #1 and #2 continued to receive until about RSSI -110dBm to -120dBm.

As stated previously: not all Heltec hotspots have this issue. However, we have already identified 4 units with this problem(See chart by @mrblue1995 ). There is a clear correlation.

We see exactly the same results over the Helium network. Sensitivity of the 4 Heltec hotspot is worse and therefore number of beacons is significantly lower.

We know that some heltec devices work perfectly fine and you are doing a great job so far(enclosure is superb :slight_smile: ), we just think that there are some units which are defective on the lora transceiver side.

2 Likes

Thank you for your detailed feedback, I have submitted your question to our R&D team. They will have a comprehensive testing soon.

2 Likes

Some more data about a Pisces hotspot (grey/purple one) vs a Sensecap

1 Like

Some more data about a heltec which doesnt look to perform correctly, it replaced a synchrobit which was doing perfectly fine. here you can find the last lora transmissions after the resume of the blockchain:
18 00:18:15 JSON up: {“rxpk”:[{“tmst”:2506345740,“chan”:6,“rfch”:0,“freq”:867.700000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-15.8,“rssi”:-73,“size”:52,“data”:“QDDaAAEJ7hp8AG4jAK1Sv0oz+ozZJLhM/d76qkmJJBI7OYhtY6LC7ssB4cXuYaSvwFHIlA==”}]}
Nov 18 02:08:15 JSON up: {“rxpk”:[{“tmst”:568656348,“chan”:0,“rfch”:1,“freq”:868.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-12.8,“rssi”:-75,“size”:15,“data”:“4AAATE8VBQAAAAO/4pT/”}]}
Nov 18 04:42:11 JSON up: {“rxpk”:[{“tmst”:1196530244,“chan”:0,“rfch”:1,“freq”:868.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-17.2,“rssi”:-75,“size”:52,“data”:“QDDaAAHSG4C7ANTzANYCgm8chCSHl+mvLUylbwwL1fnUpCFNVvtoB2jyFXbDzs+Vsd7ffA==”}]}
Nov 18 05:43:45 JSON up: {“rxpk”:[{“tmst”:563401652,“chan”:0,“rfch”:1,“freq”:868.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:6.5,“rssi”:-68,“size”:52,“data”:“QDDaAAGFjf3jAPiWADA5CvJokQQD2CsWzOqguRS25ZPeAAcXcj3xq67i55nF0/QwUwpoGw==”}]}
Nov 18 09:30:45 JSON up: {“rxpk”:[{“tmst”:1350573540,“chan”:1,“rfch”:1,“freq”:868.300000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:9.0,“rssi”:-72,“size”:52,“data”:“QDDaAAFdV2PbAL7VAMv42YJpCojx8Fmxj0XADTP5McICA72p+Wns2ZjL1o15isyBqYRF+w==”}]}
Nov 18 11:59:15 JSON up: {“rxpk”:[{“tmst”:1632014076,“chan”:5,“rfch”:0,“freq”:867.500000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-14.2,“rssi”:-75,“size”:52,“data”:“QDDaAAE90E/8AEhpAHemg85cbTd4/Jie0NcUopWFY7V8/lgBGCmLE1hFe7HG4qSvKHEgaA==”}]}
Nov 18 12:13:09 JSON up: {“rxpk”:[{“tmst”:2540700220,“chan”:1,“rfch”:1,“freq”:868.300000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:8.2,“rssi”:-67,“size”:41,“data”:“QFwIAEgAAgAB0nNGoVavwbypeWPkBPvK0Z+rWRxh8ZqtW9HONPJz3Ho=”}]}
Nov 18 12:13:09 JSON up: {“rxpk”:[{“tmst”:2550709188,“chan”:0,“rfch”:1,“freq”:868.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:8.8,“rssi”:-67,“size”:41,“data”:“QFwIAEgABAABPe/b1+jQYsE5Km1R7nT+qAWuhny18Q8y74wpX+FlVuM=”}]}
Nov 18 12:13:45 JSON up: {“rxpk”:[{“tmst”:2565719604,“chan”:1,“rfch”:1,“freq”:868.300000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:8.2,“rssi”:-65,“size”:41,“data”:“QFwIAEgABwABUlz1Eal+1Fiu+r83B952pUypQnnIiyGAscSkR7kUAoQ=”}]}
Nov 18 12:13:45 JSON up: {“rxpk”:[{“tmst”:2570722484,“chan”:2,“rfch”:1,“freq”:868.500000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:6.5,“rssi”:-73,“size”:41,“data”:“QFwIAEgACAABCXnx/VUDkSeGgxSB9yw2RRLFNSIpKnczgJOWZ3n+5/w=”}]}
Nov 18 14:02:16 JSON up: {“rxpk”:[{“tmst”:496915788,“chan”:4,“rfch”:0,“freq”:867.300000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-13.2,“rssi”:-71,“size”:52,“data”:“QDDaAAESEPSDAKrQAMVWn4uabwp9C5gZiCfFZ6RS54IRtyGzlPReWluYaCaaQejUWsyBgQ==”}]}
Nov 18 14:54:13 JSON up: {“rxpk”:[{“tmst”:3533025228,“chan”:0,“rfch”:1,“freq”:868.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-16.8,“rssi”:-75,“size”:52,“data”:“QDDaAAEmn110ANnBAD+zaFU5bpKDXR/xBlkbfLEgTw69lCgR/FW465ZqpNQznK9QUbSS6Q==”}]}
Nov 18 15:07:16 JSON up: {“rxpk”:[{“tmst”:101884412,“chan”:2,“rfch”:1,“freq”:868.500000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-13.2,“rssi”:-75,“size”:52,“data”:“QDDaAAGrhemeAJ/RABClbwY+PELNoP5dWD53K2LLASsHJxXIomJBRYHNAZxfup1kkXPX/w==”}]}
Nov 18 16:22:16 JSON up: {“rxpk”:[{“tmst”:264493076,“chan”:0,“rfch”:1,“freq”:868.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:8.8,“rssi”:-67,“size”:52,“data”:“QDDaAAHAHHufACkdAP6a5MwLPlXahD3O8KMuc5IOUGw02tU43Uu49sifJwr958hZcctg9A==”}]}
Nov 18 19:21:16 JSON up: {“rxpk”:[{“tmst”:2464147356,“chan”:5,“rfch”:0,“freq”:867.500000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-18.0,“rssi”:-75,“size”:52,“data”:“QDDaAAEMO2K4APWUAJ8+M/EB/mW3ZIYX6Ba7l7HWBbOjhcSDNI77vkBpIDLjXJ1gwIg7Uw==”}]}
Nov 18 20:04:16 JSON up: {“rxpk”:[{“tmst”:729524116,“chan”:1,“rfch”:1,“freq”:868.300000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:7.5,“rssi”:-65,“size”:52,“data”:“QDDaAAG8ug3RAHhBACLwaOcQ5u6pGDlUqYsmp1yjwsc+wR9UN+OuTyjFmPV8mrGChOHXWw==”}]}
Nov 18 22:53:16 JSON up: {“rxpk”:[{“tmst”:2242615708,“chan”:4,“rfch”:0,“freq”:867.300000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-19.0,“rssi”:-71,“size”:52,“data”:“QDDaAAGGoGxiAPi8AGp0v0tPnJFoUbtlyi369JLyO1EsZdknfQsDoMVXPGzGn0yrcejXHA==”}]}
Nov 19 01:05:16 JSON up: {“rxpk”:[{“tmst”:1596085284,“chan”:0,“rfch”:1,“freq”:868.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-12.5,“rssi”:-75,“size”:15,“data”:“4AAFTU8VBgAAAAO/4pHA”}]}
Nov 19 03:49:03 JSON up: {“rxpk”:[{“tmst”:2809277092,“chan”:3,“rfch”:0,“freq”:867.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-15.8,“rssi”:-77,“size”:52,“data”:“QDDaAAESqfmzADUvADpjO43xt15jXzWP/O9B/jjrn/071U2WnXn1Syhj4muido96Ir5Eog==”}]}
Nov 19 05:37:46 JSON up: {“rxpk”:[{“tmst”:713826868,“chan”:0,“rfch”:1,“freq”:868.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-8.5,“rssi”:-75,“size”:52,“data”:“QDDaAAEx41tNAEDKAJL5/nr4qpolNOZeW0uhl7CDkQvdazeDG/R49aE2KkCs11pWbG9Prg==”}]}
Nov 19 07:14:36 JSON up: {“rxpk”:[{“tmst”:2260377260,“chan”:2,“rfch”:1,“freq”:868.500000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-19.5,“rssi”:-75,“size”:52,“data”:“QDDaAAHPxOaVAEDoALDxijhm3Y0j066FgOkmUm67ZGlaVfjrxIv12LGGcGbRHZ91IrRMQA==”}]}
Nov 19 07:34:17 JSON up: {“rxpk”:[{“tmst”:3496290340,“chan”:0,“rfch”:1,“freq”:868.100000,“stat”:1,“modu”:“LORA”,“datr”:“SF12BW125”,“codr”:“4/5”,“lsnr”:-8.2,“rssi”:-76,“size”:52,“data”:“QDDaAAFt5oBDAG1mAGWLR77Kk/6paoGIPo1Amy+AruUdLd4EuIHgB47CEYFH6vfsY9Y9tg==”}]}

ALL its rssi values are between -65 and -75 which suggest a low sensitivity of the device or a bad calibration. I have other heltecs which are doing perfectly fine, may we have a defective unit here?

1 Like

I’m sure to have the exact same issue.
I’m in the a goldspot on the map, everyone around me is earning a lot an i get 6$ a day.
My antenna is on the roof, swappes 6 8 and 10 dbi, 14m high without any obstacle and I’m getting nothing…