Where to get legit nRF24L01+ modules?

[d00616]

@gohan said in Where to get legit nRF24L01+ modules?:

I need to get some nrf24 but I'm unsure to start moving to the nrf5x or not.

IMHO the only missing feature to start is OTA updates. Implementing OTA updates is a question of (my) time. On the other hand the ATMEGA limitations like PROGMEM dealing, 2k RAM or slow speed cannot be fixed. With nRF5 you haven't care about used RAM or Flash. The radio protocol can easily changed to support encryption at the RFM level.

[NeverDie]

@d00616
Well said. I think we're all going to be so much happier with the nRF5 series that later we're all going to be wondering why we didn't switch sooner.

[NeverDie]

@Nca78 said in Where to get legit nRF24L01+ modules?:

They also have through hole version that use a much smaller crystal

It turns out I have this product already. I purchased it two years ago from a different seller on ebay. A couple small observations about it:

1. I really do doubt that it's a 4-layer board as they are claiming. The board is very thin, and I don't see any evidence of there being more layers than just top and bottom.

2. The NRF24L01 chip on the module I have is rotated 180 degrees relative to the one in the Ebyte rendering.
   
   However, all the lettering on the chip is exactly the same.

Here are some photos of it:

[NeverDie]

Here is what I wrote about it on August 3,2015 in the "Which are the best nrf24l01+ modules?" thread:

I received the NRF24L01+'s that are on the red PCB's (above), and when I saw they were using the now notorious 1242AF chips, I had little hope. However, I tested them at 1mbps over the same challenge distance as the others, and so far they're doing very well: I transmitted over 200,000 packets, and there were only 0.03% lost packets. Average round trip time was 2.2ms.

As before, I'm using the RFToys to do the testing. The modules seem more finicky about their orientation than others that I've tested, and moving things just a little can make for much, much worse results.

I bought them from MDFly on ebay.

[NeverDie]

It seems that the red modules were fakes. Here's what I wrote on August 15, 2015 in the other thread:

Based on the mA and uA measurements above, it's likely that the Addicore modules and the red modules are based on the Si24R01, because they are a reasonably good match for the electrical specifications on page 22 of the Si24R01 datasheet: https://www.dropbox.com/sh/kdenpdg60v5hzbd/AACG1jxQR71fkzX-U4a7CIh0a/SI24R1 (cn).pdf?dl=0

I haven't tested the red modules from Ebyte, so I can't say whether the same will apply to them. Maybe Ebyte is using genuine Nordic nrf24L01 chips, as they claim. However, given the above, I wouldn't take it for granted that they necessarily are.

[NeverDie]

For reference, here's a better photo of the red nRF24L01+ board that I had previously tested in 2015:

[NeverDie]

I see that there's another Aliexpress store that's selling what looks like the same module for almost half the price:
https://www.aliexpress.com/item/E01-ML01D-Imported-components-nRF24L01-2-4G-wireless-transceiver-RF-data-transmission-module-industrial-grade-imported/32817210922.html?spm=2114.search0104.3.31.yf0kqS&ws_ab_test=searchweb0_0,searchweb201602_5_10152_10065_10151_10068_10130_10084_10083_10119_10080_10082_10081_10110_10178_10137_10111_10060_10112_10113_10155_10114_10154_438_10056_10055_10054_10182_10059_100031_10099_10078_10079_10103_10073_10102_10120_5360020_10189_10052_10053_10142_10107_10050_10051,searchweb201603_2,ppcSwitch_4&btsid=7cb3695c-3e0d-40fe-8806-e962a3d54db1&algo_expid=2d68395c-5638-4e8a-8b27-4dacf8ca0318-4&algo_pvid=2d68395c-5638-4e8a-8b27-4dacf8ca0318

Anyhow, it's conceivable that eByte is simply reselling these red modules. Its other modules are blue, so I'm hoping that's the case, and that its blue modules might be better.

[d00616]

@NeverDie said in Where to get legit nRF24L01+ modules?:

Anyhow, it's conceivable that eByte is simply reselling these red modules. Its other modules are blue, so I'm hoping that's the case, and that its blue modules might be better.

It's also the same module name. This could be also a fake of the eByte modules.

I't looks like, there is no trusted source for nRF24 modules.

[NeverDie]

I think consistency probably matters more than genuine vs. fake. In other words, if you end up with fakes, try to make sure that all your radios are all the same fake. It gets confusing when you start to mix different kinds together, because different variants can have different idiosyncrasies, and then you start to wonder whether a problem you might be having is due to a chip difference, or mix of differences, or something else. Very quickly the combinatoric complexity rises and troubleshooting becomes more time consuming.

[gohan]

@NeverDie the problem is to be able to find good and fake one so that you could mark them with different colors. Not everyone have the tools

[NeverDie]

@gohan
So what? Some people live in rural Africa without electricity or running water, and they don't have soldering irons or even solder or even any money. Are we going to worry about whether they can do this as well? I mean, where do you draw the line? The way I see it: if a hobbyist really wants to know if a particular chip is fake or genuine, then they can either buy or borrow or rent an o-scope for long enough to figure it out. The other thread tells how to figure it out using an o-scope. Prior to that thread, I had never even used an o-scope. So, the required skill level isn't very high at all.

[Nca78]

@NeverDie but even low end oscilloscope is over 200$/€, it's expensive for a hobby.
I don't have one, because it's too much money for the use I would have of it at the moment, and I wouldn't even have a place to store it

And I don't know who I could borrow an oscilloscope from, so the quest to find a reliable supplier is worth it for me. Even if I think it will only end with buying NRF5.

[NeverDie]

@Nca78
In that case you could get the job done using one of the ARM chips that comes with a fast sampling AD converter and, preferably, lots of RAM. You could get it on one of the cheap demo boards to make the project easier. Fortunately, the fake nrf24's are different enough from the genuine in their Tx and Rx current draws that you won't need to split hairs on your measurements to tell them apart.

I did something similar once for capturing and decoding RF signals using an Arduino Due, so I'm sure it can be done. In fact, you could probably use a Due for this if you wanted to, but no doubt there are better options available now.

[Yveaux]

@NeverDie said in Where to get legit nRF24L01+ modules?:

Fortunately, the fake nrf24's are different enough from the genuine in their Tx and Rx current draws that you won't need to split hairs on your measurements to tell them apart.

Nope. Some of them might be, but certainly not all of them...

[NeverDie]

@Yveaux said in Where to get legit nRF24L01+ modules?:

@NeverDie said in Where to get legit nRF24L01+ modules?:

Fortunately, the fake nrf24's are different enough from the genuine in their Tx and Rx current draws that you won't need to split hairs on your measurements to tell them apart.

Nope. Some of them might be, but certainly not all of them...

Which ones? Even the fakes have names and datasheets. The datasheet tells the tale. The ones I tested (see other thread for the details) were pretty noticeably different.

[Yveaux]

@NeverDie said in Where to get legit nRF24L01+ modules?:

Which ones? Even the fakes have names and datasheets. The datasheet tells the tale.

I've ordered many nameless modules (regular black ones) through Ali express the past years, which didn't come with a datasheet or any specs. A number of them are proven fakes (some even x-rayed by Nordic), but I've not seen any of them having a current consumption significantly different from original Nordic modules.
I just want to stress again that although some fake modules have a higher current consumption, this certainly isn't true for all fake modules.
A module with regular consumption still could be fake. X-ray is the only way to be 100% certain.

[NeverDie]

@Yveaux

What was it about the chips (the ones that had identical current draws to the real thing) that you sent to Nordic for x-ray that motivated you to send them?

If your requirement is 100% certainty, then you may just have to make your own modules with chips procured either direct from Nordic or from a trusted supplier, like Digikey.

[Yveaux]

@NeverDie said in Where to get legit nRF24L01+ modules?:

If your requirement is 100% certainty

That's not my requirement; I didn't start this thread
I only have (presumably) fakes deployed around the house and they've never failed me. I won't bother getting 100% genuine ones.

[d00616]

I have received two of the red modules. This is an X-Ray image of the chip (montage)

[NeverDie]

@d00616 said in Where to get legit nRF24L01+ modules?:

I have received two of the red modules. This is an X-Ray image of the chip (montage)

Meaning?

[d00616]

@NeverDie @Yveaux said in Where to get legit nRF24L01+ modules?:

Could be that different clones of the nRF exist of which some have an increased power consumption, but currently the only way to know for sure (also for Nordic ) is to x-ray them.

I hope somebody can say if this looks like an original.

[Yveaux]

@d00616 the xray is rather vague.
Here are die shots of the original and a fake: https://zeptobars.com/en/read/Nordic-NRF24L01P-SI24R1-real-fake-copy

Maybe you can match them.

[d00616]

@Yveaux Thank you for the X-Ray images. I can't get images in that resolution.

[Yveaux]

@d00616 they aren't xray images. The die shots were obtained by eating away the chip package with acid.

[Bogus Exception]

@hek Thanks for writing. I got 4 from itead the other day (2 each of 2 types), and they are in the boxes in front of me. FOr one thing-they have a box! I am making a video for my blog comparing them to the clones (as best I can), and I'm going to test out that current statement I found online...

Thanks for the tip!

[Bogus Exception]

@tbowmo Thanks for the tip. I have some in front of me, still in box. Will post results.

[Nca78]

I received my Ebyte modules, not a warranty of authenticity but the nrf chip has a square instead of a dot like all the fakes I have been using until now. I have a few standard SMD modules and a PA LNA version, I will try to make some basic range tests soon compared to the SMD clones.

[Bogus Exception]

OK! I followed your suggestions and tried itead. These are the pics of what I got 4 of (sorry no pro rig-just a loupe & module in one hand & mouse in the other!

(all are hosted on one of my servers, not here locally)

[gohan]

@Bogus-Exception let us know how they perform

[Bogus Exception]

@gohan I'll do my best to measure meaningful metrics between these & clones. Also, I have 4 eBytes coming (2x blue, 2x red)...

[Bogus Exception]

So far I've noticed on all versions of the legit nRF24 chips that the 'square' is actually 4 vertical lines, covering a square area, at the 'top left' of the chip.

Just tried over the past 2 hours to replace the clone in my MyController gateway with either the itead or ebyte legit modules, and neither one of them would communicate with the other nodes. I put the clone back in, and it worked. Weird.

Not like it matters much, but here is the sketch (all comments removed) using a BME280 and a light sensor module, on a Nano, USB powered:

#define MY_DEBUG
#define MY_RADIO_NRF24
#define MY_RF24_PA_LEVEL RF24_PA_LOW
#define MY_GATEWAY_SERIAL
#if F_CPU == 8000000L
#define MY_BAUD_RATE 38400
#endif
#include <Adafruit_Sensor.h>
#include <Adafruit_BME280.h>
#define BME_SCK 13
#define BME_MISO 12
#define BME_MOSI 11
#define BME_CS 6
#define SEALEVELPRESSURE_HPA (1013.25)
const float ALTITUDE = 18;
Adafruit_BME280 bme(BME_CS); // hardware SPI
#define LIGHT_CHILD 0
#define BARO_CHILD 1
#define TEMP_CHILD 2
#define HUM_CHILD 3
boolean metric;
uint8_t decimals;
#include <MySensors.h>
MyMessage lightMsg(LIGHT_CHILD, V_LIGHT_LEVEL); // 0
MyMessage pressureMsg(BARO_CHILD, V_PRESSURE); // 1
MyMessage tempMsg(TEMP_CHILD, V_TEMP); // 2
MyMessage humMsg(HUM_CHILD, V_HUM); // 3
#define LIGHT_SENSOR_ANALOG_PIN 0
int lastLightLevel;
unsigned long SLEEP_TIME = 60000; // Sleep time between reads (in milliseconds)
void setup()
{
  #ifdef MY_DEBUG
    Serial.begin(115200);
  #endif
  decimals = 2;
  metric = getControllerConfig().isMetric; 
  bool status;
  status = bme.begin();
  if (!status) {
    #ifdef MY_DEBUG
      Serial.println("Could not find a valid BME280 sensor, check wiring!");
    #endif
    while (1);
  }
}
void presentation()
{
  sendSketchInfo("Master Nano", "1.0");
  present(LIGHT_CHILD, S_LIGHT_LEVEL);  // 0
  present(BARO_CHILD, S_BARO);          // 1
  present(TEMP_CHILD, S_TEMP);          // 2
  present(HUM_CHILD, S_HUM);            // 3
}
void loop()
{
  int16_t lightLevel = (1023-analogRead(LIGHT_SENSOR_ANALOG_PIN))/10.23;
  if (lightLevel != lastLightLevel) {
    #ifdef MY_DEBUG
      Serial.print(F("Sending lightMsg: "));
      Serial.println(String(lightLevel));
    #endif
    send(lightMsg.set(lightLevel));
    lastLightLevel = lightLevel;
  }
  float p_Mb = bme.readPressure() / 100.0F;
  float p_Hg = p_Mb / 33.86;
  #ifdef MY_DEBUG
    Serial.print(F("Sending pressureMsg: "));
    Serial.println(String(p_Hg));
  #endif
  send(pressureMsg.set(p_Hg, decimals));
  float t_C = bme.readTemperature();
  float t_F = (t_C * (9/5)) + 32;
  #ifdef MY_DEBUG
    Serial.print(F("Sending tempMsg: "));
    Serial.println(String(t_F));
  #endif
  send(tempMsg.set(t_F, decimals));
  float h_Pct = bme.readHumidity();
  #ifdef MY_DEBUG
    Serial.print(F("Sending humMsg: "));
    Serial.println(String(h_Pct));
  #endif
  send(humMsg.set(h_Pct, decimals));  
  #ifdef MY_DEBUG
    Serial.print(F("Sleeping for: "));
    Serial.println(String(SLEEP_TIME));
  #endif
  sleep(SLEEP_TIME);
}

It works with the clone... I am using a regulator board for the nRF24-the one it plugs into & breaks out the 8 pins...

[Yveaux]

@Bogus-Exception said in Where to get legit nRF24L01+ modules?:

So far I've noticed on all versions of the legit nRF24 chips that the 'square' is actually 4 vertical lines, covering a square area, at the 'top left' of the chip.

You cannot rely on the print of the nRF to distinguish fake from genuine. Nordic is fabless, so they use different factories to produce their chips, each with its own specific marking.

[gohan]

@Bogus-Exception may I ask why are you using spi for bme280 sensor instead of i2c?

[Bogus Exception]

@gohan I already have the 3 pins for SPI in use for the nRF, so adding the BME was just one more pin. Simply conserving pins.

[Bogus Exception]

@Yveaux Fair enough. I won't rel on that, but did want to mention the observation.
On a relatively side note, I got 4 SMD nRF24 cards from eByte, but can't locate any boards to mount them on! The holes are 1.27mm pitch, so I just ordered some 1.27mm headers.
Has anyone grabbed a board to surface mount solder them at this pitch?

[NeverDie]

@Bogus-Exception said in Where to get legit nRF24L01+ modules?:

@Yveaux Fair enough. I won't rel on that, but did want to mention the observation.
On a relatively side note, I got 4 SMD nRF24 cards from eByte, but can't locate any boards to mount them on! The holes are 1.27mm pitch, so I just ordered some 1.27mm headers.
Has anyone grabbed a board to surface mount solder them at this pitch?

What's the pitch on most other SMD nRF24L01+ modules? Is it also 1.27mm, or something else? I hadn't paid attention.

[Bogus Exception]

@NeverDie Great question. Mine look like these from a post earlier in this thread, and they are the first I've seen that didn't have their (breadboard-unfriendly) pinout.

[NeverDie]

Not sure if it's what you mean, but for prototyping purposes you could solder them to generic 1.27mm proto-board, such as this: https://www.aliexpress.com/item/6-X-8CM-spacing-1-27-universal-board-thickness-1-6mm-sided-HASL-PCB-test-board/32774106087.html?spm=a2g0s.9042311.0.0.eD40xE

[gohan]

I think there are adapters to solder the smd version and make them 8 pin compatible or there are generic adapters that can change the pitch

[Nca78]

@Bogus-Exception said in Where to get legit nRF24L01+ modules?:

On a relatively side note, I got 4 SMD nRF24 cards from eByte, but can't locate any boards to mount them on! The holes are 1.27mm pitch, so I just ordered some 1.27mm headers.
Has anyone grabbed a board to surface mount solder them at this pitch?

You can use NModule, the footprint highlighted in red color in the image below is compatible

https://www.openhardware.io/view/364/NModule

[Bogus Exception]

@kalina made these DIP-to-SMD-NRF24L01-adapter, but my need isn't enough quantity to buy the chemicals, pay someone to make a rum of them, or convince my wife that, "We'll sell them and make our money back!" :-[

[Bogus Exception]

@Nca78 Not just the module (10 for $15, right?), but the page concept is exactly what I am doing these days. Built in flexibility where it matters!
You've clearly spent a lot of time at my stage of tinkering with this!

Little touches like PA pinout, etc. makes this cake have more icing then cake!

[NeverDie]

@Bogus-Exception
I'm sure the Kalina adapter would be inexpensive if you purchased 3 through OSH PARK because the surface area is small.

[Bogus Exception]

@d00616 I have to ask...

How did you create this image?

[d00616]

@Bogus-Exception said in Where to get legit nRF24L01+ modules?:

How did you create this image?

The image is created with a medical X-Ray machine.

[Bogus Exception]

@NeverDie What I meant was, I don't have anything in 1.27mm, so now I've ordered enough stuff to keep my 2.54 and 2.0 company!

[Bogus Exception]

@d00616 Yeah, but I mean, that image doesn't show detail, does it? I thought it was thermal/IR when I first saw it-like it was running.
First time I've seen this tried. Has it worked for other electronics? Is that image you posted what is usually expected?
And most importantly, how did you talk your dentist/radiologist into doing it?!

[Bogus Exception]

@gohan Was this wrong? Is there a readon to change?
Ya kind of left me hanging there, buddy!

[gohan]

@Bogus-Exception I don't know, I just could not figure out what you were trying to achieve

[Bogus Exception]

With the inability to get signal strength from the nRF24, I'm open to any suggestions to fairly evaluate the boards I've collected. Don't think the wife will let me spring for a spectrum analyzer...

[gohan]

Well usually you could test how good they are by running a sketch that keeps sending data and, while increasing distance, look when you start getting NACKs on the serial monitor.

[NeverDie]

@Bogus-Exception said in Where to get legit nRF24L01+ modules?:

With the inability to get signal strength from the nRF24, I'm open to any suggestions to fairly evaluate the boards I've collected. Don't think the wife will let me spring for a spectrum analyzer...

Somewhere in the other thread I posted a sketch for automatically counting packet loss percentages. That's the best way that I've been able to think of. Signal strength, bit error rate, and packet loss rate should all be very strongly correlated to one another.

[gohan]

@Nca78 where did you get the ebyte modules? From aliexpress they are charging quite a bit for shipping.

[Nca78]

@gohan they don't charge me for shipping, but I'm in Vietnam so it's not far for them

[d00616]

Some news. After playing with NRF5 and the NRF24 protocol. The red EBYTE modules are fakes.

Over the N of NRF is a dot. I have ordered these modules: https://aliexpress.com/item/Wireless-Module-NRF24L01-original-2-4-Ghz-RF-transceiver-SPI/32712174920.html they have an rectangle over the N and responding line an original NRF24L01. Maybe this helps to identify originals.

The difference between both modules is in NoACK attribute handling. Fakes always responding with ACK and sending inverse NoACK attributes. So fakes have problems communicate with original modules.

EDIT: This was an issue at the NRF5 side.

[Nca78]

@d00616 said in Where to get legit nRF24L01+ modules?:

Some news. After playing with NRF5 and the NRF24 protocol. The red EBYTE modules are fakes.

You mean you bought those red modules from Ebyte shop and they sent you some with a dot on the chip ?
The ones I received from them (blue SMD ones) have a rectangle, that's in fact made with very close lines.

The difference between both modules is in NoACK attribute handling. Fakes always responding with ACK and sending inverse NoACK attributes. So fakes have problems communicate with original modules.

Is there an "easy"/"not too hard" way to test that ?

[gohan]

@d00616 Are you sure about that?

[d00616]

@Nca78 said in Where to get legit nRF24L01+ modules?:

@d00616 said in Where to get legit nRF24L01+ modules?:

Some news. After playing with NRF5 and the NRF24 protocol. The red EBYTE modules are fakes.

You mean you bought those red modules from Ebyte shop and they sent you some with a dot on the chip ?

Yes. I had ordered the module in the form factor like this posted above.

@gohan said in Where to get legit nRF24L01+ modules?:

@d00616 Are you sure about that?

Yes, about the ACK/NoACK handling. With the NRF5, I see raw packages. I'm not sure about the dot vs. rectangle, but my modules with a dot violating the protocol.

[gohan]

how did you setup NRF5 "sniffer" ?
BTW, my red modules are working really great; the link you provided from aliexpress are really genuine?

[Nca78]

@gohan do you have rectangles/squares or dots on yours ?

[gohan]

I don't remember, I'll have to look tonight

[d00616]

@gohan said in Where to get legit nRF24L01+ modules?:

how did you setup NRF5 "sniffer" ?

At the moment, I reimplement the protocol for NRF5 chips to fix an NRF5 hardware issue. I don't have a "sniffer".

BTW, my red modules are working really great; the link you provided from aliexpress are really genuine?

My red modules are good working clones, but they are clones which violating the protocol. Using clones resulting in sending more packages over the air and never receiving an ACK package from original NRF24. So it looks like a duck, but not quacks like a duck over the air.

EDIT: Sorry, Cannot identify clones at the moment.

[NeverDie]

There's no reason to continue using nRF24L01's anyway, except for legacy situations. Their only virtue is cheapness, and the nRF51's offer a good substitute for that.

[d00616]

@NeverDie said in Where to get legit nRF24L01+ modules?:

There's no reason to continue using nRF24L01's anyway, except for legacy situations.

...when https://github.com/mysensors/MySensors/issues/949 is fixed. If not, a node requires an hard reset from time to time.

[Yveaux]

@d00616 said in Where to get legit nRF24L01+ modules?:

they have an rectangle over the N and responding line an original NRF24L01. Maybe this helps to identify originals.

Here we go again... Nordic is a fabless company. Therefore the laser marking on their genuine models can be different with each batch. There is no known way to tell if a module is genuine or not from the outside of the module.

[NeverDie]

@d00616 said in Where to get legit nRF24L01+ modules?:

a node requires an hard reset from time to time.

First I've heard of this. How often is "from time to time"?

[d00616]

@NeverDie said in Where to get legit nRF24L01+ modules?:

@d00616 said in Where to get legit nRF24L01+ modules?:

a node requires an hard reset from time to time.

First I've heard of this. How often is "from time to time"?

<0,1% of packages. I have three versions of the ESB protocol code. With the published version, the module hangs when PAN#102 is triggered. My second version was a try to fix this issue by adding some timer code. With this fix, the NRF5 looses more packages than before. The version I currently working on looks more stable, is easier to debug, supports the fast ramp up mode partially, but it's not complete yet.

[d00616]

Sorry for my NRF24 clone post yesterday. The ACK bit was an issue with the NRF5 platform. I had switched the memory region to handle TX/RX/ACK packages separately. Sometimes the received packages are written in the wrong region. I had seen my testing ACK bit. Now, I have changed my NRF5 code and I can't see no difference between the modules, I have. ACK and timing are exactly at the same level.