Atmega328p @ 8mhz @ 250kbps = fail

[cimba007]

int pipe = wait(1500);
if(pipe != -1)
{
  uint8_t tc = radio.testRPD();

Using Serial << "setDataRate: " << radio.setDataRate(RF24_250KBPS) << endl;

tc is always 0 and I get a very very poor reception. Using the same code with RF24_2MBPS my results are fine. 1MBPS is working great too. Have I missed an important point?

PS: I got 2x 1µF ceramid on my 3,3volt ldo, voltage seems pretty solid. Tested both normal modules and pa+lna modules .. 250kbps stays "broken". In Addition to the caps the nrf24 got a didicated 1500µF

My whole init:
`
void rf_setup()
{
radio.begin();
radio.enableDynamicPayloads();
radio.setAutoAck(1); // Ensure autoACK is enabled
//radio.setChannel(100);
//radio.setPALevel(RF24_PA_MAX);
Serial << "setDataRate: " << radio.setDataRate(RF24_2MBPS) << endl;
radio.setCRCLength(RF24_CRC_16);
radio.setRetries(15,15); // 5 retries with 15 * 250µS = 3.75ms delay between each

if(settings[ce::myid] == 0){
Serial.println(F("I am station"));
// Station
radio.openWritingPipe(station_to_node[0]);
radio.openReadingPipe(1,node_to_station[1]);
radio.openReadingPipe(2,node_to_station[2]);
radio.openReadingPipe(3,node_to_station[3]);
radio.openReadingPipe(4,node_to_station[4]);
radio.startListening();
}else{
Serial.print(F("I am node "));
Serial.println(settings[ce::myid]);
// Node
radio.openWritingPipe(node_to_station[settings[ce::myid]]);
radio.openReadingPipe(1,station_to_node[settings[ce::myid]]);
}
}
`

[Greenstreem]

I am having the same exact problem with the atmega 328p.

[TMRh20]

How are you powering the radios? 250kbps on-air time is 8 times longer than
2mbps, so uses more power.

Sent from my iPod

On 2016-06-05, at 12:34 PM, DasHammer [email protected] wrote:

I am having the same exact problem with the atmega 328p.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#264 (comment), or mute
the thread
https:/notifications/unsubscribe/ACfVHBxq8fLIGFF-Cvsp49hJWy7qLedVks5qIxajgaJpZM4IuYQ-
.

[cimba007]

I use an 250mA LDO with 3300µF cap on my receiving node and 1500µF on my sending node. I will hook up my osci to check the voltage.

thx TMRh20, will reply soon

[cimba007]

okay .. I measured Peak-to-Peak ripple of 60mV during send. Adding another 1500µF on my sending node (3000µF parallel) decreases the ripple to 30mV (only when sending).

Still the same Problem. I replaced the PA+LNA-Board with a "normal" one and got still the same figures.

EDIT: Receiving node has even smaller ripple of 21mV peak2peak. (on average cause there are no packets send)

PS: I can even see 15spikes on my tx node trying (sometimes not always) the retry .. so 250kbps has validly no good connection at all.

[Avamander]

Okay, now, if you take the example sketch and only add the line to lower tha datarate, does it work?

[cimba007]

I will try this out tomorrow. thanks for your help

[cimba007]

Nothing changed .. contrary to what might be sane I get the worst results on 250kbps.

I changed the example sketch to include RPD:

http://pastebin.com/2V1hvDQY

I looked many hours into the source and found nothing .. My last resort would be to completly redesign my pcb although oscilloscope reading show nothing wrong with the power supply.

I even powered both nodes from an 3,7V -> LDO -> nrf24 .. giving even more stable supply but simply .. no ..

[TMRh20]

A couple of questions:

1. Since the title is 'Atmega328p @ 8mhz' have you tested with any 16mhz 328s? Difference or no difference?
2. How about if you comment out lines 111, 112 & 113 in the gettingstarted example and just do one-way communication? Please post the serial output from both devices.

[cimba007]

I narrowd down the problem to the auto-ack feature although still under investigation. I use these settings:

radio.setAutoAck(true);
radio.setRetries(15,15);

radio.setAutoAck(false); ==> seems to work better

remember .. 1mbps and 2mbps works fine ..somehow autoack doesn't work as expected on 250kbps.

I will test your two suggestions now.

[cimba007]

mycode: http://pastebin.com/V9t1MYVR
changing baud-rate for different tests

Test 1: NO OK

• Both modules @ 8mhz internal oscillator
• Both modules set to 250kbps
• Reply code removed so the communication is just "node1 (ping) -> node2 (pong)"
• Node1 and Node2 are PA+LNA with a distance of 15-20cm, stable supply and RF24_PA_MAX

Output on node2: http://pastebin.com/ATssyu2Y

Test 2: OK

• all the same as Test1, both modules set to 1mbps

Output on node2: http://pastebin.com/W1p0mqZ8

Test 3: OK

• same as Test1, "radio.setAutoAck(false);"

Output on node2: http://pastebin.com/JD8DLw4Z

Test 4: NOT OK

• same as Test1 but "radio.setRetries(15,15);"

Output on node2: http://pastebin.com/zPkGbJMv

Test 5+6 prove difficult cause I can only get one of my two nodes connected to my IPS so I tried this:

Test 5: NOT OK

• same as Test4, but receiving node2 @ 16mhz

Output on node2: http://pastebin.com/zPkGbJMv

Test 6: NOT OK

• same as Test4, but sending node1 @ 16mhz

Output on node2: http://pastebin.com/Hk2s9JBK

From my understanding the waiting period for auto-retries must be higher on 250kbps but the default settings already comply with this requirement. setRetries(15,15) let no doubds that the delay is high enough.

[trojanc]

Hi

I'm not that technical about the inner workings...and do not have the solution for your problem, but I've had endless issues with RF24_PA_MAX when the modules are too close to each other (like in the same room). Somehow they interfere with each other. When the modules are further away I use RF24_PA_MAX, otherwise RF24_PA_MIN while developing with the modules close by....don't know if that help 😃

[cimba007]

hi trojanc, thanks for this tip. I don't want to retry all tests but I did for Test1 and moved the sending node 1 room away, all settings staying the same .. same result

tried out RF24_PA_MIN again on the same room, same result

Edit: To be extra sure I just opened a pair for PA+LNA modues from my latest shipping. The ones I used for my tests are like 2-3 years old and they show the same phenomena: 250kbps@autoack = working so lala, 250kbps@noautoack = working fine.

As I already said before my last resort will be to ditch my pro minis and do an extra testbed with 5V @ 16mhz with the two new rf24-modules but this will have to wait until the weekend.

I am pretty sure this is unrelated and I did not fully read the thread but I will just keep it here for reference (http://forum.arduino.cc/index.php?topic=198471.15), but the OP mentions something from changing from 250kbps to 1mbps

[Avamander]

This again sounds like you have stumbled upon counterfeit modules that perform not as they are supposed to. Clones tend to have issues with autoack and n+1 other things.