| View previous topic :: View next topic |
| Author |
Message |
PannetierJ
Joined: 18 Jun 2017
Posts: 31
|
|
 Posted: Sun Jul 16, 2017 2:05 pm |
 |
|
I join you the schematic :
 |
|
 |
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Sun Jul 16, 2017 3:52 pm |
|
|
The schematic looks OK.
While reviewing the schematic, I noticed that I made a mistake in the
program. I had the Rx pin for the hardware UART enabled, when pin C7
is used for the MOSI pin. I have modified the #use rs232() line below
to use only pin C6, as a soft UART for Tx only. Try this program on
the Transmit board:
| Code: |
// Register r/w test for the Transmitter board
#include <18F2550.h>
#device = PASS_STRINGS=IN_RAM //Admite ponteiros para constantes
#fuses HS,NOWDT,NOLVP,CPUDIV1
#use delay(clock=16000000)
#use rs232(baud=9600,xmit=PIN_C6) // *** Software i2c for Tx only
#define RF24_SPI_DISABLE_WARNING //for disabling nRF24 SPI warning message.
#define Led_Send PIN_A3
#define Led_Init PIN_A4
//********** DEFINE PORT NAMES
#define SPI_MISO PIN_B0 //SPI(Usar por hardware quando possivel)
#define SPI_MOSI PIN_C7 //SPI(Usar por hardware quando possivel)
#define SPI_CLK PIN_B1 //SPI(Usar por hardware quando possivel)
//Driver nRF24L01P.C
#define RF24_IRQ PIN_B2 //interrupcao nRF24L01+
#define RF24_CS PIN_C1 //chipselect nRF24L01+
#define RF24_CE PIN_C2 //chipEnable nRF24L01+
#define RF24_PERFORMANCE_MODE //performance mode ON
#define RF24_SPI STREAM_SPI2//Redirects SPI2 port to RS24_SPI stream
#use SPI(MASTER,DI=SPI_MISO,DO=SPI_MOSI,CLK=SPI_CLK,BAUD=1000000,MODE=0,BITS=8,MSB_FIRST,STREAM=STREAM_SPI2)
#include <nRF24L01P.C>
//==============================
void main()
{
int8 result;
printf("Start\n\r");
RF24_initPorts();
RF24_disable();
RF24_comm(W_REGISTER|SETUP_AUTORETRANSMISSION, 0x55); // Write 0x55 to Retry reg
result = RF24_comm_in(R_REGISTER|SETUP_AUTORETRANSMISSION); // Read it back
printf("Wrote: 0x55, Read: %x \n\r", result); // Display it
RF24_comm(W_REGISTER|SETUP_AUTORETRANSMISSION, 0xAA); // Write 0xAA to Retry reg
result = RF24_comm_in(R_REGISTER|SETUP_AUTORETRANSMISSION); // Read it back
printf("Wrote: 0xAA, Read: %x \n\r", result); // Display it
printf("\n\rTest done \n\r");
while(TRUE);
} |
|
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19244
|
|
| Posted: Mon Jul 17, 2017 1:17 am |
|
|
Just a quick update/suggestion to this.
I don't see anywhere what compiler version is being used?. Probably just missed it. However there was a problem some time ago, that if the fuse NOPBADEN wasn't selected, on a few compiler versions you got problems with using the portB pins for digital I/O unless you explicitly disabled the analog on these pins.
So I'd suggest adding this fuse to PCM_programmers code. |
|
|
temtronic
Joined: 01 Jul 2010
Posts: 9124
Location: Greensville,Ontario
|
|
| Posted: Mon Jul 17, 2017 4:51 am |
|
|
One item I didn't see on the schematic is a pullup resistor for the IRQ line of the nRF module.
Historically any I/O pin that can be 'multisourced' like I2C bus and interrupts were 'open collector' and so need a pullup. I do recall that the nRF module I used 3-4 years ago did not have a pullup(saved mfr 1 cent ?),so I'd check yours just to confirm/deny one exists.
Also the module does draw a lot of power during transmit so I'd have 1000mfd on the output of the regulator( 2 x 500mfd). Without a real good energy reserve( big caps ) ,VDD can 'dip', maybe not enough to trigger brownout but could cause weird,random problems.
just food for thought...
Jay |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19244
|
|
| Posted: Mon Jul 17, 2017 6:33 am |
|
|
I wondered about this, but a search found:
<https://electronics.stackexchange.com/questions/149110/is-the-nrf24l01-irq-line-open-collector>
However the supply is definitely worth looking at. Another poster having a similar problem online found the chip was getting reset during TX.
The consumption is not high (only a few mA), but the rate of change of demand is very high, so a good ceramic 22uF close to the pins feeding the module, is a very good idea.
Look at the final result of this thread:
<https://forum.sparkfun.com/viewtopic.php?f=13&t=8062> |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19244
|
|
| Posted: Mon Jul 17, 2017 9:24 am |
|
|
and another thing.
Add a 100mSec delay before calling the initialisation.
If you look at the chip state machine, it takes 100mSec to go from power being applied to being in the undefined 'ready to wake' state. Wonder if other users have all been using something like an LCD, where you typically have to wait for a good fraction of a second for it to be ready, before starting the operational code. |
|
|
PannetierJ
Joined: 18 Jun 2017
Posts: 31
|
|
| Posted: Mon Jul 17, 2017 10:47 am |
|
|
I'll test your suggestions in the week.
If I understand temtronic, I must add a pull-up on IRQ Line ?
And add 100ms before init. |
|
|
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Tue Jul 18, 2017 8:36 pm |
|
|
I ordered an nRF24L01P+ board, hooked it up and it works. It passed
the register read/write test. I ordered the Longruner NRF24L01+ boards
from Amazon, simply because they were cheap ($6.99 USD for three)
and the reviews said they worked. Their image shows a silkscreen, but
the delivered boards are missing the silkscreen. But it didn't matter.
Here are the results from TeraTerm:
| Quote: |
Start
Wrote: 0x55, Read: 55
Wrote: 0xAA, Read: aa
Test done
|
Here's a photo of my test setup. This board is running at +3.3v and
has a 16LF1847 on it. I compiled it with CCS vs. 5.071. I used the
16LF1847 because that was an LF PIC that I had in stock.
| Code: | // Register r/w test for the Transmitter board
#include <16F1847.h>
#device PASS_STRINGS=IN_RAM
#fuses INTRC_IO,NOWDT,NOLVP
#use delay(internal=16M)
#use rs232(baud=9600,xmit=PIN_A0) // Use soft UART for Tx to PC
#define RF24_SPI_DISABLE_WARNING
// These are the connections I used with a 16LF1847:
#define SPI_MISO PIN_B1
#define SPI_MOSI PIN_B2
#define SPI_CLK PIN_B4
#define RF24_IRQ PIN_B0
#define RF24_CS PIN_B5
#define RF24_CE PIN_B3
#define RF24_PERFORMANCE_MODE
#define RF24_SPI STREAM_SPI2
#use SPI(MASTER,DI=SPI_MISO,DO=SPI_MOSI,CLK=SPI_CLK,BAUD=1000000,MODE=0,BITS=8,MSB_FIRST,STREAM=STREAM_SPI2)
#include <nRF24L01P.C> // From CCS Code Library forum
//==============================
void main()
{
int8 result;
delay_ms(100);
printf("Start\n\r");
RF24_initPorts();
RF24_disable();
RF24_comm(W_REGISTER|SETUP_AUTORETRANSMISSION, 0x55); // Write 0x55 to Retry reg
result = RF24_comm_in(R_REGISTER|SETUP_AUTORETRANSMISSION); // Read it back
printf("Wrote: 0x55, Read: %x \n\r", result); // Display it
RF24_comm(W_REGISTER|SETUP_AUTORETRANSMISSION, 0xAA); // Write 0xAA to Retry reg
result = RF24_comm_in(R_REGISTER|SETUP_AUTORETRANSMISSION); // Read it back
printf("Wrote: 0xAA, Read: %x \n\r", result); // Display it
printf("\n\rTest done \n\r");
while(TRUE);
} |
|
|
|
PannetierJ
Joined: 18 Jun 2017
Posts: 31
|
|
| Posted: Wed Jul 19, 2017 1:46 am |
|
|
Hi PCM Programmer,
I like this sound.
I have a question about your schematic.
I noticed that you add a resistor, in the left on Pic.
It's a pull-up for IRQ Line ?
I'll do others tests this weekend.
Thank you for your time. |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19244
|
|
| Posted: Wed Jul 19, 2017 3:03 am |
|
|
As another comment, I still note you are showing Vusb left floating (PCM_programmer is using a PIC without this). Look at this thread:
<https://electronics.stackexchange.com/questions/27093/seemingly-unstable-basic-pic18f2550-circuit>
It is really silly to be using a USB PIC if you are not using it's USB abilities. Wastes pins. If you are running at 3.3v, then Vusb can be tied to the 3.3v rail. If not it really must have the capacitor. |
|
|
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Wed Jul 19, 2017 8:45 am |
|
|
There is no IRQ resistor. I'm doing an SPI r/w test. I don't even have
IRQ connected. The nRF24L01+ data sheet doesn't say that IRQ is
an open-drain pin. None of the schematics out there show a pull-up
on that pin. It's not required.
The resistor in my photo is the MCLR pull-up resistor. The capacitor is
the bypass cap for the PIC's Vdd pin.
I really suspect you have a connection problem. It will be that you have:
1. Swapped two signals, such as MISO and MOSI or CE and CSN.
-or-
2. You have misinterpreted the connector pin numbering, such as
swapping the two columns.
Either of those is equally likely. I will bet you did one of them.
You must get the SPI communication working before you can do anything
else. So find the bug. |
|
|
PannetierJ
Joined: 18 Jun 2017
Posts: 31
|
|
| Posted: Wed Jul 19, 2017 9:33 am |
|
|
Hy,
Good news.
I'll test this quickly (Without IRQ, if i understand, connected), same version of CCS. |
|
|
PannetierJ
Joined: 18 Jun 2017
Posts: 31
|
|
| Posted: Wed Jul 19, 2017 11:01 am |
|
|
PCM Programmer,
Did you take the driver of nRF24L01P.C on page #1 ? of this subject. |
|
|
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
|
PannetierJ
Joined: 18 Jun 2017
Posts: 31
|
|
| Posted: Sat Jul 22, 2017 2:48 am |
|
|
Hi PCM Programmer,
Now I have CCS 5.008.
I tried to import the driver nRF24L01P.h and nRF24L01P.c.
I modified this driver the last commands into the forum.
Nevertheless, it seems to have some errors into my modification.
Could you post your nRF24L01P.h and nRF24L01P.c which you use ?
To be ok on the version.
Best regard |
|
|
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
