Вставка
- Опубліковано 14 гру 2024
- LoRa and ESP32 are everybody’s darlings, and when a combination in the form of these small boards appeared on our platforms, it seemed they are a perfect combination. Is this true? Let’s test them.
A 433 MHz Heltec
An 868-915MHz Heltec
An 868-915 TTGO
And as a reference the Hope RFM95 868 MHz
These boards can be used as LoRa nodes and as TTN gateways.
In this video we have a closer look at these features:
The ESP32
The LoRa transceiver
The antennas
And the power management part
Links:
TTN Forum: bit.ly/2GFo0TN
ESP32 LorRa Boards: s.click.aliexpr...
LoRa32u4: s.click.aliexpr...
Whisper Node Lora: bit.ly/2GCB0td
dB vs VSWR: ww2.minicircui...
Antennas: s.click.aliexpr...
Supporting Material and Blog Page: www.sensorsiot.org
Github: www.github.com...
My Patreon Page: / andreasspiess
If you want to support the channel, please use the links below to start your shopping. No additional charges for you, but I get a commission (of your purchases the next 24 hours) to buy new stuff for the channel
For Banggood bit.ly/2jAQEf4
For AliExpress: bit.ly/2B0yTLL
For ebay.com: ebay.to/2DuYXBp
www.facebook.c...
/ spiessa
www.instructab...
Please do not try to Email me or invite me on LinkedIn. These communication channels are reserved for my primary job
Equipment in my lab: www.sensorsiot....
I do have to thank Andreas for this one; ain't many of us hobbyists that have the kind of hardware to test antennas like this properly, so it's very nice that he does it for us.
Nita Vesa agreed, I would have never realized this until I either test myself or see this video. Unfortunately I do not have anything exotic instrument nor detailed knowhow of RF field.
Thanks!
Who needs a lab or an external analysis company when you have Andreas as a friend ??!! Always torough, accurate and rigorous ! Danke !
You are welcome!
Perfect and in details analysis. Congratulations Mr. Spiess!
Thank you!
This channel has some of the best videos on UA-cam!!!
Thanks!
Sr.. im been looking your channel since 2 years ago. I really appreciate your efforts to present everything in your videos as crystal they looks and so simple in a matter that probably looks complex to many..... It is a shame that you have only 200K+ subscribers but would like to tell you in, a way to tank you, that all my customers are received better solutions thanks to you.
Thank you for your nice words. I am glad my work helps you!
This was extremely useful, Andreas! You saved me a lot of time. Again!
For those who need but can't afford a spectrum analyzer, it is possible to do inaccurate (but practical) testing of RF transmitters using $20 software-defined radios (SDRs) and Open Source software e.g., SDRsharp.
Thanks for your tip!
An SDR will, as part of its normal operation, display several useful characteristics of a given transmitted signal: signal strength, frequency, bandwidth, adjacent or overlapping signals that might squirrel up the S/N ratio of the receiver, etc. In fact, the only function of a spectrum analyzer that can't be directly reproduced with an SDR is precise quantification and measurement of the signal of interest. That functionality is seldom useful in experimentation, but it can be added to an SDR via external filters, amplifiers, and/or attenuators to allow very precise measurement of whatever you need measured. A spectrum analyzer just combines the SDR functionalities, the filters, amps, and attenuators in one box with an interface usable even by people who don't understand RF, but can follow a step-by-step recipe.
PS: SDRsharp is a software solution. It allows you to add plugin decoders, demultiplexers, transcoders, or any other instrumentation you can conceive. And, there are a lot of decoders already written. So, if you want to see the RF aspects of a signal AND the content of that signal, all you need do is pull up vim (the only editor worth its storage footprint) and start typing. :o)
I don't know of anyone else specifically using an SDR as a spectrum analyzer. It's kind of odd, as that was the whole reason I bought an SDR: to see the spectrum around a specific frequency of interest. They're used for all kinds of oddball things. I think the first one was sold as a TV receiver dongle, and someone else wrote different software for it, and thus the SDR fad was born. The desire for step-by-step procedures is a good indicator that you want to get a real spectrum analyzer. eBay, ham conventions, etc are your friend.
For things like casual LORA evaluation, an SDR is more than adequate. NOOELEC's dongle just works with Windows 10, Ubuntu/LinuxMint, and MacOS (I've used one on all four OSes). SDR gear, overall, is not a mainstream solution, and everyone involved in that sub-sub-hobby assumes that SDR users have a lot of radio electronics and software skills. If you're hitting brick walls hooking up a USB stick and running Windows software, then you probably don't want to waste your time in that hardware/software rabbit hole - you just want results. So, get a dedicated spectrum analyzer, get the analysis you wanted, make something insanely cool.
As a Patreon supporter it sure is nice getting the videos early! More of your subscribers should take advantage of the opportunity. The new boards have a very attractive of combinations of features. I reviewed them last week and came to the same conclusion that you did - not quite ready for prime time. Hopefully their shortcomings will be addressed in V2 hardware
Thank you for your support! The V2 is already on order. We will see...
Good Morning Andreas
Another well done and very informative vid. Very germain at the moment for me as I've just completed a Project Spec/Definition on a multi end-point surveillance system for my equipment trailer and SUV.
The SUV will act as a relay station for the trailer module which is positioned 200-300 feet away and the control end-point is at my side or in my pocket.
I plan on using an ESP32 based module for all points in the network and a second relay module so that all midules will be line-of-sight in a three hop system.
Your work on ESP8266/32, LAN/WAN/LoRa has transformed from being interesting to invaluable for me.
I always watch your vids with keen interest and intend to watch your entire series of comm-ctrlr vids.
I am also very appreciative of no music, no fluff, moocho info and a professional presentation.
You simply do awfully good work.
Regards/73
Russ in NH/KA1RUW
Thank you for your precise feedback. You have a very interesting project!
Excellent analysis! You truly live up to what we expect from engineers in the DACH region.
I’m so glad I saw this before purchasing a set of modules.
Glad I am able to help prevent bad purchases!
Loved the Maradona joke. Thanks so much Andreas for the technical analysis and for your Swiss sense of humour !
Thank you. Glad you liked the content!
Thanks Andreas for your videos. You also have "la mano de Dios" like Maradona on MCUs and sensors
Antenna measurements are always nice, you could do a full series on antennae and how to use them.
The measurement of the TTGO antenna shows how your setup will influence the outcome. You should always test an antenna in a situation that resembles the actual use and mounting the antenna directly on the reflection bridge is not "actual use".
I have developed a number of commercial products with 433, 868 and 2450 MHz radio solutions and have done antenna measurements (and actual range measurements) with these products. Our first 868 design was measured with the antenna mounted on a wooden pole measuring both the reflection and the antenna radiation pattern. Surprisingly, the measurements were worse than the actual product's radiation pattern. I then suggested to mount a piece of PCB material and a battery near the antenna in the same way as the actual product has - this gave a complete different view of the antenna.
It looks like the TTGO 868 antenna was designed to be used on top of a large steel cabinet - a typical use case for a short range 868 telemetry application.
P.s: reflection measurement is only part of the antenna measurement - you also have to measure the (3D) radiation pattern. From a reflection point of view, a 50 ohm resistor is the best antenna ;-)
You are absolutely right, also with measuring the pattern. But this is not easy. If the SWR is bad, then a good pattern does not help a lot. And this is what happens with many Chinese antennas because they are sold for completely wrong frequencies.The patterns for similar antennas should be much more similar than the SWR is.
2 years ago and still the best video about LoRa!
Thank you!
Hi! Andreas I see your videos frecuenly and sorry by my bad english but the goal of Maradona is the best moment of my life!! It was the best moment of my life, it is too much to say that I am Argentine, hehe, the electronics and soccer are my passion, I live in Tierra del Fuego Argentina! a big greeting from here for you!!
Glad you liked the video about Maradona!
Again an overwhelming video with deep technical backround. And I'm a little bit envious on your pool of measuring instruments.
This is the result of my age ;-)
Those antenna placements are ridiculous. Nylon mounting screws would help a little but you would probably get the best range by setting the boards on fire and using them to send smoke signals. Thanks for the testing.
:-)
XD
I love seeing these antenna tests! Please do more!
The problem is that Chinese merchants change their supliers without anouncement. So my tests might not be relevant for you :-(
thank you sir i was planning on getting one of these boards from ebay after watching your lora war driving video... now ill be getting a esp/hopefm setup and just doing the solder myself.
:-)
Hi Andreas, thanks for the great videos. I hope this isn't a redundant comment but your channel has so many comments I can't find if this was updated( hey whose fault its that ;))
I recently bought some of the Heltec WiFi LoRa 32 (V2). They have a new antenna, I attached one to my crude AAI RF vector analyzer(N1201SA). On the antenna I tried it I can measure ~1.2 for VSWR at 915MHz and at 868MHz I measure ~2.1 for VSWR. You can tell the antenna is better it has a molded plastic sleeve over the antenna bit and when you try to squeeze it it is solid. The resonant frequency looks to be 935MHz with a VSWR of ~1.05. Yes my measurements are not as accurate as a lab instrument but good enough for the girls I go out with. Thanks.
Thanks for sharing your measurements. This is why I do not give recommendations on antennas. Suppliers change them fast. Here obviously towargs the good side...
Once again, such GREAT information. I'm sad I hadn't found you earlier... would have saved me some headache.
Glad it was helpful!
This is a great video. It really shows some great investigation using tools that most of us mere mortals do not have access. The antenna results are truely staggering. Thanks for sharing!
You are welcome. The Spectrum Analyzer was a dream of young Andreas. So I bought me one when I became old ;-)
For me it was an oscilloscope. I'm not old enough for a tracking spectrum analyser just yet. :)
:-)
Stop sitting on the fence, get down and tell us what you really mean! :) This is exactly the resource and review users, who do not have the advantage of your knowledge and equipment need. Well said.
Thanks. An thank you again for the nice evening in Amsterdam. I think I will contact Stuart. I have an idea of building a automatic antenna tester based on his principles. Maybe you can help me getting the contact?
No problem - we can IM anytime. He only lives a kilometer away from me. Look forward to it.
:-)
Another excellent video, lightyears ahead of anyone else. I see so much potential for these low power ESP32 Lora boards . .
Thanks!
Thank you for sharing your vast knowledge and laboratory test equipment. Excellent !
:-)
Thanks very much for your summary. It shows that you need to be critical when choosing the right antenna.
True!
3 years later, still interesting and very usefull. Thank you internet, thank you sir with the Swizz accent for your energy in making this video and doing such extensive researches
Glad it was helpful!
Excellent analyse! Thank You, Andreas!
:-)
I really enjoyed your video. Thanks again for your methodical, analytical approach to evaluating these boards. I've been an amateur radio operator for years and years, and while I have no formal EE education in antenna theory, I have come to appreciate the empirical evidence I've noticed over the years. The analog and RF aspects of some of these amazing products these days tend to get inadequate attention and consideration. Thanks for shining a light on these aspects. Perhaps the vendors will do a bit better in the future if they know that people are paying attention to this aspect of the product's quality.
I ordered a V2 version. We will see what they learned...
@@AndreasSpiess So, did they?
I was just looking at these a few days ago, thank you for this great analysis. I dream of a spectrum analyser but my wife would certainly kill me if i bring yet another huge expensive piece of lap equipment to my desk in the living room :)
She might be right, I have to admit ;-) This is not exactly a device for the living room and I had to get 60 till I got one.
I found this video when I was trying to figure out why I could barely communicate on 915MHz between two heltec boards on the same bench. I have the nasty helical spring antennas and now I know that they probably hardly couple anything into the ether! Just watching this video in advance would have set me on a path to getting some better boards and saved a lot of pain and anguish.
Still, such a small range might have a different reason (e.g. you got a 433MHz version). But maybe you try with standard TTGO boards I use in my newer projects. They work well.
@@AndreasSpiess I have some of the Lilygo boards on order -- they have the better antenna and I hope are general better all round! Thanks
Andreas, you are a neat guy, a man after my own heart! I love your videos, I have learned so much about Arduino and module applications, please keep up the GREAT WORK.
Thank you very much for your nice words! Comments like that are the diesel for creators like me ;-)
Another exceptional analysis. Thank you for sharing!
:-)
Thumbs up again. To the point and informative video on right time. I just received rfm95 modules and was thinking should I really spend time in making pcb and soldering. But now I believe I better do it :).
I have been discussing same issue of antennae compatibility with another youtuber Informatica1. I got info same as you, Do not Trust Antennae from Aexp or banggood etc. Thanks again.
Happy soldering. It is a fast job.
You ROCK DUDE, most intelligent thing I've heard on UA-cam in a long time. Thank You
You are welcome!
Fantastic video yet again Andreas. Thank you so much for your continued great content!
You are welcome!
I junked the supplied HELTEC antenna in favour of a DIY 868MHz groundplane constructed on an SMA connector with brass rod radiator and radials (bent downwards by 45 degrees. I tuned using my on my SA starting with the radiating element slightly oversize and carefully clipping off tiny preces of the element until it resonated on 868. Far better.
Good idea!
awsome as usual Andi 👍🏻thx. i think these boards are great for preprototyping and development, no soldering, no worrying about power supply and so on, u can even ask the aliexpress supplier to solder the pins at no extra cost...
Thanks for the feedback
As for as measuring efficiency of LoRa antennas - I suppose the calibration should be made using calibrating blind BNC to IPX adapter. Than later you should test the whole assembly: BNC to IPX adapter IPX_CABLE_SMA SMA_LoRa antenna under test. I can guess the thin cable may change the whole fit of impedances between IC output and antenna. Of course I can not guess if it will be better or worse, but we should check it.
What you measured is just: spectrum analyzer output BNC to SMA adapter LoRa antenna. So you omitted quite significant influence of the thin connecting cable. You may tell you have a lot of experience and it is negligible :) I will answer - hats off, Master :)
In the end I am interested how much power tha antenna gets to transmit and how much is returned. I am not the designer of the PCB. He is interested in other things, of course.
@@AndreasSpiess I am discussing the IPX_CABLE_SMA cable, not PCB. We can assume if the PCB around LoRa chip is well designed there are no significant losses in path between LoRa chip and IPX socket.
But then there is a cable from IPX socket to the antenna - omitted in your test (5 cm @ almost 1 GHz).
And what if that cable moves resonance a bit? Making the whole efficiency better (I doubt) or worse - but it should be checked. Sorry Master, your method is not excellent, again, you made a great work but this point is not so good., no offence please. You have tools to test it.
For such measurements, you have to define a reference plane. And I defined my reference at the antenna. Of course, you can define your reference plane elsewhere.
Nicely done Mr Spiess. I hope Amsterdam was a lot of fun
Yes, it was. I was also in the „Flying Scotchman“ with a few Englishmen. If you know wat I mean ;-)
Andreas, thank you for all of your nice videos.
In the meantime I bought 3 pcs. 433MHz Antennas from the antenna shop you proposed and linked in your comments.
From today I have the N1201SA antenna analyzer you also used in later videos. The antennas show resonance frequencies of 411, 429 and 433 MHz instead somewhere in the 433-435MHz interval.
So I would not recommend the shop you linked to any more.
All three antennas are ok, two are excellent compared with what you usually get. A few Mhz does not matter too much. Just touch something or move something around the antennas and the resonance will move much more than only 3 MHz. Maybe you have a look at the antennas I got with other modules. They were 200 MHz away :-(
Hi Andreas, You are a true champion and a masterful maker of the most beautiful, Fascinating and intellectually stimulating videos, there is so much Rubbish coming from China, from sellers who show nothing but Contempt for the rest of us Idiots who should know Better than to buy their absolute Rubbish !! Unfortunately...
a lot of people (myself included) fall for their false advertising and end up wasting our hard earned money.
If it was not for people like you making these masterful, life Enriching Videos, society and the UA-cam community in general would be a very dull place.
Keep up the Great work!!!
Thank you for your feedback! It helps to keep me on track.
Well Played, well played! Maradona's mention makes my day!!! Great videos Sir.
Glad you enjoyed it!
Excellent especially explaining antenna reflection and how that needs to be considered. (The Maradona reference made it even better. :) )
Thank you for your nice words!
Exactly my experiences with the chinese antennas. The only antenna I have that is actually an 868MHz antenna is the one from IMST.
When you touch your hand to the base of the antenna it becomes sort of a ground plane. My guess is that is why it changed characteristics.
A piece of copper wire cut to 1/4wave mounted in a SMA connector usually comes out better than the antennas that come with the boards.
I did some experiments with ground planes but was not very successful :-(
Sometime you have to put the toys away and seek a more professional solution, thus you can order some of the 800 MHz antennas for Motorola MTS2000 handhelds (the long ones, not the stubby's). They are fairly cheap and reliable just add an adapter (a sma male to sma male coupler I think is what it needs) if you need to. They should say Motorola 800MHz or 700/800MHz but they are pretty broad and cover 868 MHz.
Super, thanks! Having the right tool for the job really does help. This confirms my experience with the Heltec helical antenna. Now time to find the post on TTN forum showing how to cut these and optimise a simple wire 868 antenna using a small piece of shield braid. Much obliged Andreas!
Or buy a better one...
I recently bought the Wemos TTGO Lora+ESP32 868MHz+Bluetooth+OLED from VNG Systems and tested it on the KPN LoRa network in the Netherlands, which worked fine for me. I think they have updated it, because the range was acceptable.
You see the difference of V2 if you look at the antenna position
Excellent video, helps us protect ourselves from being ripped off
Thank you!
really a quite excellent and informative video! thanks for posting it. I would love to see similar analysis of other boards such as M5Stack LoRa, and even possibly comparisons between libraries such as Sandeep Ministry and RadioHead.
M5 is here and will be covered. I might use the libraries mentioned in my videos (as I did in the past). But I do not plan a direct comparison.
On the SX1276 figure 6 schematic, it appears to be a balanced filter. The bottom components are the same as the respective parts on the top.
I did not look close to this filter. But some chips have balanced output which has to be matched to an unbalanced antenna
Yes. It is useful. Thank you for your great videos. Please continue to produce them. 😎
You are welcome
You really are a star Andreas, wonderfully detailed video, even allowing for the bad memories you brought back regarding the hand of God.
Thanks! I thought, the idea of a „unknown helping hand“ was similar to my experience;-)
It would be interesting to see what effect placing a four wire ground plane around the base of the final aerial on your test rig. I usually regard one as vital to remove any movement effect in the feeder.
I will include a GP in my antenna tests. I finished my test rig and did first outside tests today (we had a "springy" warm day here in the south ;-)
OMG!! What it's doing Maradona in your video!! From Argentina we also see all your excellent videos.
I thought, I did not understand why the antenna was much better with my hand than without. Similar to the 1986 situation.
I found that those annoying ham radio handheld whip antennas that TYT ships with their MD380 series HT's works great on the 433 MHz versions of those ESP devices.
Motorola MTS2000 HT 800MHz antennas with a male to male sma adapter also works well for the 868 MHz versions.
Thank you for sharing your experience!
Thankyou for all your hard work investigating these boards. Your work inspires me onto greater and more challenging projects. Thankyou.
You are welcome. Enjoy your projects!
Outstanding. This is your strength.
Thank you!
Nice video. You’ve just saved me about 100 hours of my time because now I don’t need to research it. Thanks
:-)
thanks sir, watching your channel is like attending classes ♥ with a good professor
You are welcome!
Thanks Andreas, you´ve saved me many hours of work and a box of aspirin pills. As soon as i receive my boards i´ll run similar analysis and share it.
I´ve got TTGO V 2.0 board, they said that some issues were corrected.
Thanks!
I will test the V2 board in one of the next videos.
My HELTEC's modules have the rubber antenna like TTGO for lora and the helical one for WIFI ( the pcb antenna remains but used ) .Providing a simple ground plane for the lora antenna improved a bit the lora reach.
Mauro Cordioli same here. HELTEC with the rubber antenna, but I still had trouble connecting to a LoRa network. It is the 868 version and I made some chnages to the default scetch. What scatch/plug-in are you using?
These are the scatch modifications I used www.thethingsnetwork.org/forum/t/big-esp32-sx127x-topic-part-1/10247/24
I used a test sketch becaus I needed a constant signal, not a connection to a gateway
I have used the same thread as a guide, I have also reduced the number il bands scanned... I remember also that changing the power supply affected the stability of the board,may by there is some EMI issue, I do not have yet the time to check in deep that question.
Great and interesting ! This will be of help when I put up met LoRa gateway. Thnxs !
:-)
Excellent analysis, thank you for this
You are welcome!
I wonder if you have have done any benchmarks to EBYTE Lora32 line?
I do not have one.
Always good stuff and I enjoy your channel. It is an abundance of information, well done.
Thank you!
this was a great tour
1) what is an FCC approval?
2) the current HELTEC board looks different then the one the one in the video
You can google FCC
Yes, Heltec has new boards.
Hi Andreas..love you're channel....you do a lot of Lora and TTN...Im looking for a private lora network. you did lora node to node but what are the possibilities for more nodes to one node? without TTN. something like dont send the data from te gate way to TTN but to a database that node red can acces and then do the decoding in js?.....maybe a vidoe idea?....
I mentioned the 1:1 LoRa in one of my videos. Then, LoRa chips are like any other transceiver chips which send and receive data. Maybe I will use this feature in one of my next videos.
Excellent video, especially the part about the hand of God!!! The only problem is that the user of the board will need you to be there all the time holding the antenna in just the right manner to get good performance...
You are right.
It would be so nice to see 433,868 and 915 external antenna tests. I'm looking for it and have no idea what to buy.
I left a link of a supplier which did not cheat me so far. So I assume, also his antennas on other frequencies are ok
Amazing resource! Thank you! I've subscribed. I was wondering though, when will you review the version 2 of these boards?
Welcome aboard the channel. I know, I should do it...
thanks a lot ,you have knowledge and equipements for testing .
You are welcome.
Hallo Andreas (I am the guy from the Arduino Hannover LoRa Group)
This time I don’t agree with most of the video content, you have done much better investigations so far
The board energy usage measuring must be differentiated between the LoRa standby mode, and the LoRa receiving mode and the LoRa sending mode, this would allow users to get an idea about their device battery lifetime.
Antenna measuring, the Heltec 868 antenna I received was a 868 SMA antenna (black plastic outside, spinning wire inside). The problem with this antenna is that there is no ground plane, therefore it must be mounted on a metal peace to work. Unfortunately this is not documented by Heltec and other vendors. Your antenna measuring was also wrong because the metal peace is missing. There are other antennas which have this metal tube inside the antenna. So a proper measuring setup is needed.
BTW. On our own Arduino LoRa board (www.radioshuttle.de) we have this ground plane on our PCB, therefore this simple 868 MHz SMA Antenna works pretty good.
Frequency testing setup failure. You cannot specify a range for a 400 to 900 MHz to the SX1276 chip in software and start measuring. The SX1276 has two antenna input and output paths, called LF for up to 500 MHz, and HF for higher frequency, usually 868-915 MHz range, Usually the boards use only one output path LF or HF to the antenna. A total wrong measuring results occurs when when you are using a 433 MHz designed board specifying a 868 output frequency (or the opposite) you receive a crosstalk within the chip and the wires on the PCB which will give you total wrong results.
Regards from Hannover
PS: I like to talk to you about this an other LoRa ESP32 stuff, maybe we can have a privat phone talk about it.
1. I measured the antennas I got with the boards. And I did not get a PCB antenna.
I was in Amsterdam and did not see one sensor with an antenna setup other than mine. Non had a ground plane, all had just an antenna plugged into the devices. I think, an antenna has to be tuned for this purpose. And if everybody just plugs it into a device, it should have its best performance in this setup. Your approach with an integrated ground plane is probably a good idea for some antennas. If you watch Andrew's channel ( ua-cam.com/channels/HqwzhcFOsoFFh33Uy8rAgQ.html ) he uses many antennas which do not need an additional ground plane because they use the shield or a tube inside the rubber to build something like a dipole. If you add a ground plane to this type, they do not like it.
2. Thank you for the info about the two "channels". I did not know (I did not look into the datasheet). I assume these two "channels" explain the frequency gap I measured but did not understand.
The main goal of my scan was to understand if they sell us 433 MHz boards as 868 boards. If you look at my tests from this angle, they are probably useful. Or how would you have done them?
You saw that the two Heltec modules behaved completely different. So I was able to show that the hypothesis of selling us 433 Modules was wrong, at least for the Heltec ones. (I am not so sure about the TTGO).
With only one output connected to the antenna, it is not easy to understand why the output power at the antenna on both bands was very similar. A not connected wire usually does not transport this amount of power.
We for sure can talk about this stuff. I am on facebook messenger and twitter. There I can give you my Skype name
The 433 MHz Heltec versions uses the SX1278 chip which is specified only for 433 MHz but if you set 868 MHz has output, it works, but with totally wrong/bad results. The Semtech LoRa chip is under the display to review it’s variant.
The 868 MHz Heltec versions uses the SX1276 chip, this one can do 430 and 868 MHz, but only the 868 MHz (HF path) is connected.
For Heltec I am talking about the boards within a 6 month timeframe of my purchase, there have been also earlier versions with different hardware.
PS: I don’t think the Hope RFM95 modules for 868 and 915 MHz are any different, for export limitations into Europe is not allowed to import 915 MHz devices. The frequency is revered to some US military applications in Europe, therefore vendors specify modules and chips often with 868 or 915 MHz versions. I will check if I get a 915 MHz RFM module to measure the RFM95 HF output part list versus the 868 version.
Some time ago I ordered via Mouser a 915 MH STM FM Module, they rejected my order due to export restrictions by the US.
Habe kein Facebook und kein Skype, nur Google+ und E-Mail, kannst mich aber auch gerne im Büro anrufen.
Hi @Andreas Spiess, acutally I´m playing around with two Heltec Wifi LoRa 32 V2 boards.
I have an sender and an receiver which sends back the RSSI from the last received packed. So far so good. Works as expected. BUT.... after an undefined time (sometimes minutes, sometimes hours) the receiver hangs up. I can indicate this as I have an LED which blinks as told by the sender. After hang up I can press the reset button and it keeps running as normal. I even have this problem with the provided examples from Heltec, even their basic tests hang up on the "receiver" board from time to time. Do you have an idea what can cause this issue and how I can prevent it?
Here´s my actual code, an modified version of one of the examples:
------------------------------------------------------------------------------
@ "heltec.h"
@ "string.h"
@ "stdio.h"
@ LED 25
@ BAND 868E6 //you can set band here directly,e.g. 868E6,915E6
char Readback[50];
String lastrssi;
void setup() {
Heltec.begin(true /*DisplayEnable Enable*/, true /*Heltec.LoRa Disable*/, false /*Serial Enable*/, true /*PABOOST Enable*/, BAND /*long BAND*/);
pinMode(LED,OUTPUT);
digitalWrite(LED,LOW);
LoRa.setTxPower(2,RF_PACONFIG_PASELECT_PABOOST);
LoRa.setSpreadingFactor(7);
LoRa.setSignalBandwidth(250E3);
}
void loop() {
int packetSize = LoRa.parsePacket();
if (packetSize) {
while (LoRa.available()) {
sprintf(Readback+strlen(Readback),"%c",(char)LoRa.read());
}
// Serial.print(Readback);
if(strncmp(Readback, "1", strlen(Readback)) == 0) {
digitalWrite(LED, HIGH);
}
else if(strncmp(Readback, "0", strlen(Readback)) == 0) {
digitalWrite(LED, LOW);
}
memset(Readback,0,50);
//Serial.print(" with RSSI ");
//Serial.println(LoRa.packetRssi());
lastrssi = LoRa.packetRssi();
//Serial.println("1_"+lastrssi);
LoRa.beginPacket();
LoRa.print("1_"+lastrssi);
LoRa.endPacket();
}
}
-great video, i was wondering is there a cheaper way to check antennas, if they are tuned for eg 868. Not that i dont want a spectrum analyser, but its a bit expensive :)
You are right! I just had a phone call to test a cheaper device (In the 100$ range). And I have some ideas for a very simple setup without any instruments. So stay tuned...
Another amazing video about LoRa!
Thanks!
Thank you very much for these great advices
You are welcome!
Andreas: great work, thanks a lot, this will be of help to my LoRa projects, for sure!
However, measuring these simple antennas without a proper ground plane and "free space" is skewing the results. As you demonstrated yourself by adding your hand as a makeshift ground plane: this alters the properties of the antenna dramatically. I think you should mount the antenna on a reasonably sized (bigger than wavelength) copper plate which would also shield it from your analyzer and make your measurements more reproducible.
Of course, such a ground plane should also be used later when the antenna is in production use.
You are right. But I think, we are not interested in how the antennas behave under good conditions, but on our devices. I never saw a sensor with a big ground plane ;-) I agree that my measurements are different from the real situation where the sensor is mounted somewhere.
BTW: I made a large ground plane to test the antennas. But this also did not provide reliable measurements.
But I think, if an antenna resonates at 500 MHz on my reflection bridge, the chance it resonates at 868 on the device is small.
Great video, i think the three different versions are because the non license frequencies of different countries can vary.
You are right.
I'm made my antenna using Hack RF one and GNU radio campanion, no necessary an expensive spectrum analizer for check!
Also a possibility. Thanks!
Perfect analysis! May I ask if there is a board that combines also a GPS module? Kindly asking, if you can present the creation of a complete (and small) network with 2-3 nodes and a couple of gateways? Thank you once more!
1. A Dragino shield for Arduino exists with a LoRa chip and a GPS
2. I made videos concerning LoRaWAN gateways and respective nodes.
Again another project of mine saved by Andreas thx
Glad the video helped!
Great Test! I wish to repeat this power measurement, to maybe update the output filter network, but also to test my own device. How did you measure the transmitted power without your SA missing the short pulse? Were you able to put the transmitter in CW mode? Or did you align your sweep with the pulse somehow? Do you still have code left for that? I try to keep myself from digging into the SX1276 registers just yet..
I did convince the LoRa module to send longer pulses (SF12)
I just received some V1 TTGO Lora32 boards 868/915 versions and after watching this I tried both a 433MHz and a 968MHz sender/receiver sketch to see the difference, the antenna I received is a plastic 40mm SMA version, (no brand, 2 notches on the top) at 968MHz from one room to another I get -58dB at 968MHz and -131db at 433MHz. So it looks like 433 is simply not an option with some boards.
You have to buy a 433 MHz board if you want to use it for 433 MHz. As I showed they are different.
Thanks a lot for your great videos. In this video you had a look mainly at the transmission performace. Do you also have any data for the receive-performance? I am trying to receive LoRa satellites with my heltec 433 board, but without sucess. The ttgo works perfectly in the same environment. I think the reason is to low sensitivity of my heltec, but I have now idea how to measure this exactly. 73, Stefan/OE6ISP
The transmit and the receive performance is related. If an antenna matches for transmission it also matches for reception. AFAIK the chips used on all transceiver boards are the same. So I do not know where the difference comes from. 73 de HB9BLA
Thanks for your interesting video..just check at 10:46..calibrating your analyser..It looks like you zeroed with an open connector..is that correct for this test?..When I worked in RF with the S-parameter test set we allways zeroed the instrument with a (very expensive) calibrated 50 ohm termination. I aslso woud be very intersted in a video on various antennas for both 433 and 868 MHz Lora..perhaps a yagi or corner reflector.
1. Yes, it is correct to zero a reflection bridge with an open connector. Open means 100% reflection.
2. This is for sure not the last video about antennas on this channel...
Thanks again Andreas for sharing your findings with all us. I’m making this question just in case. Are you planning some time in the future to analyze the new LoRa boards ? If so, please consider the the Heltec CubeCell. It is claiming very low battery consumption and I am very curious about how its antenna performs.
Currently I have no plans to do further tests. But you never know...
Excellent video yet again. Thankyou!
:-)
Extremely useful. Thanks for that. But it left a hole, no discussion about RA-02 chip. Can you do a video about it with esp chip.
I assume that the RA-02 also use standard Semtech chips as all other boards. So there is not a lot to write about it for the moment.
Thanks for sharing 😀👍
As always, you are very good to check out all the best way to get to the point
Thank you!
Where could I find some more information about how to set LoRa frequency with these modules? When I looked into library examples (e.g. arduino-LoRa) there are no comments to this. According to our regulation, a certain range is available and the only spec is that channel frequency should fit into 125 KHz if I am not wrong.
So, can I use in LoRa.begin(858.5E6) if I choose so? (arduino-LoRa library)
Maybe you go to the TTN page. There you should find the information ( www.thethingsindustries.com/docs/reference/frequency-plans/ )
I have been using the TTGO 433 Mhz board with the supplied helical antenna and get what is at least for me very acceptable range - with one unit on my desk and the other laying on the front seat of my car I get good reception at well over 500m in a heavily built up area at SF7. I'm sure if I repositioned the antennae that could be improved. The ESP8266 antenna on my unit is a PCB design, not sure if this is because this is a clone or they just designed the layout differently for this version - everything else appears identical.
Thanks for the feedback.
detailed and professional analysis, many thanks.
Could you please carry out the same test to a Pycom LoPy board? it is similar priced compared with heltec board, but has semtech sx1272 chip on board and is shielded, has FCC approval.
I have one but was never able to get it running. So, for the moment no tests :-(
Good to know. I was looking at these.
:-)
Hi such a great video! I assumed the antennas would have a little better than they did! 😅 Do you know if there are any Home Assistant Integrations or Add Ons that work with ESP32 and LoRa? I’m using ESPHome ‘s rc_switch library to control some rf relays but I would love upgrade for diy LoRa tech. Do you have any suggestions? Thanks!
I never checked if LoRa boards are supported by ESPHome.
@@AndreasSpiess Hello! I’m just researching it now but according to chatgpt (which has a way of making things up completely sometimes) there is a way. I still need to check this with esphome’s documentation but here’s the chatgpt yaml config ‘’’esphome:
name: lora_hub
platform: ESP8266 # or ESP32, depending on your board
wifi:
ssid: "your_wifi_ssid"
password: "your_wifi_password"
lora:
spi_id: SPI # Assuming you're using the default SPI bus
cs_pin: GPIO_NUM # GPIO pin connected to the LoRa module's chip select (CS) pin
reset_pin: GPIO_NUM # GPIO pin connected to the LoRa module's reset pin
frequency: 433MHz # Frequency of your LoRa module’’’
I’m really interested in building a poor-man’s Yolink system. 😅 The basic goal would be to build a Home Assistant compatible LoRa transceiver (that seems straightforward if the above yaml config works) that act similar to the Yoink Hub but then to also build super low powered (using probably something lower powered than an ESP) LoRa feedback or regular sensors (and maybe later a 5v or 12v relay) that that send state info (such as voltage) back to the ESPhome transceiver. I’m guessing that Yolink’s vibration sensor (that can last 1 + year off of two 1.5v AA batteries) is controlled by a very barebones microcontroller similar to the STM32 or MSP430? In theory an esphome lora transceiver could pick up data from an arduino lora transmitter device that use the latter microcontrollers but that’s the bigger mystery right now.
I love Yolink products (they are the most reliable consumer IOT I have bought so far) but I’m making something for a charity project, hence trying to diy something similar but cheaper. How has your battery performed on your mail box? I’m watching that video now and love it! Any thoughts would be greatly appreciated!
Very very useful! Thank you
Thank you!
This was so useful to me! Thanks!
:-)
Thank you for your thoughtful analysis, Andreas! I’m voluntarily building GPS collar trackers for wild animal conservation in the Patagonia. For my project I needed long LoRa p2p range, GPS, Persistent memory and very low power. Heltec was my first choice to try and was not very satisfied. Could I ask you for advice choosing the right components? 🙏
Maybe you search the internet. There are many projects similar to yours with cows, turtles, and others.
The new, improved Heltec 868-915MHz V2 is out on the market. The antenna is also new. Could you test them? Would be helpful to see if there is any improvement in the bad performance RF part, because otherwise it would be nice product.
So far I have no board available (and a lot of other LoRa boards laying around).
Thank you for the comparison. From the spectrum at 7:43 , does it mean I can use the TTGO at 433 MHz?
The chip can do both frequencies. The boards have to have the right parts for the chosen frequency.
Mr. Spiess thank you for checking out this modules. Although they might not be as hoped they diffidently serve the purpose bench testing and what now. I’ve been following along on the TTN forum I’m trying to flash the 1ch gateway code on my ttgo v1. I keep getting stupid errors like send Upd not declared and what not. You have any advice how to solve these silly errors that seems no one else is having issue with besides myself.
You are right with the bench testing. Concerning Upload: Sometimes it helps to reduce speed to 115000. And sometimes to press the boot and the reset buttons and release the reset before the boot button. This procedure should set the ESP into flash mode (as the old ESP8266)
Andreas Spiess uploading is no issue. Mostly compiling issues.
Very useful!, Thanks Andreas!
You are welcome
Congratulations!!!!, and thanks... so one interesting option (taking into account the total cost and the emission power) would be esp-32 lora 433Mhz board, with a real (not fake) 433 mhz antenna. I think that really there isnt 868 boards, i think all ttgo boards have the same design and only change the antenna (they have seen that with default configuration 868 and 915 "works". As you can see, signal of 433 mhz is near 7db, almost like hope rfm95
This is what I thought at the beginning. But the 433 and 868 Heltec boards are clearly different
I am thinking in ttgo, heltec seems a crap all versions...