AVR Freaks

Alex, does LW mesh relies on any patents?

Thank you for the quick response, I had the same impression but I just want to ask. For sure it would be stated somewhere in the document or in the source code.

Is there any way to detect address conflicts? Will the stack deliver a message to the application layer if the source address matches the destination address?

Hi Alex, I just downloaded the March version of the Lightweight Mesh Software Stack v1.0.1.
I have been playing around with the STK-600-Atmega128RFA1 and Bitcloud but manipulating the GPIO is very cumbersome and slow (ADC, IRQ's etc...)
My first impression is that the mentioned ready to use binaries to start testing are not in the zip file.
Am I missing some other download that I should be aware of?
Thanks,

I'll do that. I was just afraid I could be missing something else.
Thanks for your incredible quick response :)

Hello Alex,
I have a few questions about the LwMesh. I was under the impression that I would be able to just program the boards with the Peer2Peer solution provided, and just change a few parameters to achieve a wireless UART configuration. After testing this, it is proving to be more difficult. I have tested two atmega128rfa1 dev. boards with attached incoming RX and TX lines. It seemed communication was not correctly made, and I tried programming a few different ways:
within config.h:
-one board is coordinator, other is router
-one board is coordinator, other is end point
-both coordinators
-differing PanIDs for the devices
-same PanIDs for the devices
-set HAL_UART_RX_FIFO_SIZE and TX to 1, to hopefully just send the data as it is received
within haluart.c:
-set 2 stop bits
within fuse settings:
-0xFE,0x9D,0xC2 (Extended,High,Low)
-0xFE,0x91,0xE2 (Extended,High,Low)

Sorry to overload you with information, but I figure I'm better off giving as much detail possible. Sorry I have no experience with networking, and am still familiarizing myself with micro controllers. Do you think LwMesh is a good choice for our UART to Zigbee needs?

Thank you for any help you may provide!

Edit: Also, I just noticed there is a newer version of the LwMesh (1.0.1). I will have to test that in the next few days, because I have been using 1.0.0.

Thank you very much for the speedy reply Alex! I will have to debug our boards a little more to make sure all components are working as expected, and test it out some more. I will report back if there are still issues or questions.

Thank you for the speedy replies again!
I have tried a few different possibilities, but to no avail.
Am I correct in thinking that HAL_UART_RX_FIFO_SIZE in the config applies to the buffer to be filled before wireless transmission? For example, if I wanted a byte by byte transmission, sending each byte as it is received; I should just set both of these values to 1?

Hm, OK. Thank you very much for your patience and help! I may have to see if I can figure out what is causing this. Again, thanks for the help!

Hi, Alex!
Last two weeks I tried to make OTA service working, but with no success...
I fall back to Your WSNDemo, but... nothing...
I tried many things. Made modifications in PHY level, reverse engeneer PHY part of bitcoud stack (by the way, phySetState there doesn't uses FORCE_TRX_OFF state. It issues FORCE_PLL_ON, and according to the documentation this will improve performance of the channel), and so on... But nothing helps...
But the last day, I found in NWK level that data request made from Ind callback is places in NWK frames queue before the ACK frame, and though is sent BEFORE ACK frame...
I.e.: Server sent START_UPGRADE packet and is waiting for ACK, but receive START_UPGRADE_RESP packet, which is filtered put, as state of OTAUs sever is not yet OTA_SERVER_STATE_WAIT_START_RESP...
I think, that code of TX state machine is to be modified. ACK frames must be processed first...

Oh! I tried so many things, that I don't remember all of them...
But problem was very simple... NWK_MAX_ENDPOINTS_AMOUNT wasn't defined in config.h of OTAUServerDemo... :-)
But with modifications mentioned in previous post...

Of course, in that case we need to define timer with expiration time ~5min to initiate transition to TRX_OFF state
for recalibration to take place (according to documentation)
But I thing that it’s worth the cost as transition time from TRX_OFF state to TX_ARET_ON or RX_AACK_ON takes 110 us, but from PLL_ON to them only 1 us
and

CSMA_SEED_1_REG_s.aackSetPd = 1;
there is from Bitcloud stack
TRX_CTRL_2_REG_s.rxSafeMode = 0;
TRX_CTRL_2_REG_s.rxSafeMode = 1;
From documentation... to release buffer and to set protection again...
In addition in PHY_Init I add
HAL_RfRxTxIndicatorOn()
which is defined in halPhy()

and PHY_RxTxSwitcherOn()
in phy.c

and modification in nwkTxTaskHandler to force SERVICE_ENDPOINT frames to send first.
I can send it to You if needed...

Hi Alex,

I got the WSNDemo ported to work on a new project using ASF in AS 6.1 with STK-600. I am trying to use the ASF driver/service for the UART so I took out all of LwMesh UART code and place the ASF.
My problem is that there must be some collision with the clock handling of ASF because if I run the initialization nothing works, and if I just skip the "sysclk_init()" LwMesh works perfectly but of course the UART output is garbled (I am using usart_serial_putchar() instead of HAL_UartWriteByte())
The init function is like this:
void sysclk_init(void)
{
#if !MEGA_XX_UN0 && !MEGA_XX_UN1
uint8_t *reg = (uint8_t *)&(POWER_REG_ADD);
uint8_t i;
/* Turn off all peripheral clocks that can be turned off. */
for (i = 0; i < NUMBER_OF_POWER_REG; i++) {
*(reg++) = 0xFF;
}
#endif
#if !MEGA_UNSPECIFIED && !MEGA_XX
/* Set up system clock prescalers if different from defaults */
if ((CONFIG_SYSCLK_PSDIV != SYSCLK_PSDIV_8) ||
(SYSCLK_PSDIV_8 != CLKPR)) {
sysclk_set_prescalers(CONFIG_SYSCLK_PSDIV);
}
#endif
}
Could you please point me into what could be colliding?

Thanks,

However, there are still cases when the last received packet was e.g. Seq= 136 and the receiver does answer the following broadcast with Seq=1. But according to your reply 1 < 136. Capture here http://goo.gl/adBg2

BTW: Dont you know, Alex, what value of Duplicate Rejection TTL uses BitCloud? I'd like to compare these two systems in practise.

Hello Alex,
Now I am trying to setup Lightweight Mesh with RS-485 (Half Duplex). I am using an RS-485 breakout board to achieve this, but it must be triggered to let it know when to transmit or receive. I believe I must use a TXc (transmit complete) interrupt, to trigger from transmit back to receive. I do not know where I would put the other trigger, or if this will even work with how the Lightweight Mesh is setup. Any help would be greatly appreciated.
Thanks!

Hey, I am also using RS-485 in my project with LightWeight Mesh. I would recommend to copy the UART library source used by LightWeight Mesh and modify that to make the flow control work with your RS485 driver.

Going through the stack, it's not clear where the variable TRX_STATUS_REG_s is being assigned? It's used in conditional statements in a few places, but I can't find where it's assigned any values, is it memory mapped?

Hi Alex, TRX_STATUS_REG_s.trxStatus is set to 9 (TRX_STATUS_PLL_ON?) when it's stuck in the while loop.

Thanks Alex. The frequency is 8 MHz, I'm running off the internal RC clock for the CPU. One thing to note is that I have several ISRs running in addition to the stack, Timer1, Timer3 and UART0/1 TX and RX. There are a lot of instances in code where I block interrupts (for short durations). Also, the ISR for Timer 3 takes about 500 uS to complete and is called every 2 ms, so significant CPU time is dedicated to it.

I will look into the FORCE_TRX_OFF issue and get back to you.

That makes sense. I'm wondering if my application code is causing this, but as you say, you've never seen this problem before. What's interesting is I'm running a similar application code for the RF on my "master" LightWeight Mesh node, and it has no issues. I'm also debugging a very simple network, just one master and the slave that's hanging. Could noise issues on the transceiver clock be responsible for something like this? I'm not too familiar with how the PLL works.

Hi Alex, I switched the CPU clock to run off the transceiver clock, and I tried a different PCB for the device getting stuck.

The problem still persists, it transmits several hundred packets successfully, then gets stuck while waiting for state 22.

Alex, in this ISR in the stack:

ISR(TRX24_RX_END_vect)
{
TRX_STATE_REG = TRX_CMD_PLL_ON; // Don't wait for this to complete
phyRxRssi = (int8_t)PHY_ED_LEVEL_REG;
phyRxSize = TST_RX_LENGTH_REG;
phyState = PHY_STATE_RX_IND;
}

The state is set to TRX_CMD_PLL_ON. Isn't there a possibility that this interrupt changes the state while in the phyTrxSetState(uint8_t state) function is running (waiting in while loop)? The state is being modified in both the regular stack code and in the ISR without any protection between the two from what I can see.

This "solution" is just a temporary hack, but it seems to do the trick.

TRX_CMD_FORCE_TRX_OFF is the brute force way. Keep in mind that any ongoing transmit or receive actions are canceled immediately. Normal state changes are queued, that means
ongoing transactions are finished and after that the new
state is entered.

Depending on the transceiver (e.g. RF230) a transition over PLL_ON is necessarry before entering RX_AACK_ON or TX_ARET_ON.

Thank you Alexru for answers !

Again I have few questions to ask:

1. Can I use following function:

on routing nodes to disable their routing function for some time for testing purposes?

My goal is to find out how to place nodes to achieve redundant paths. By using this function I would like to find out how good is link(LQI and RSSI) to some end device when I turn OFF TX on R1 and after that turn OFF TX on R2.

2. Does it make sence to let's say build network topology on routing records?
Like comparing BitCloud with LW Mesh, in BitCloud there is

There is nothing similar in LW Mesh.

3. In network I have moving nodes and e.g. one moving node is present in one part of network(in reach of "COORDINATOR") and other nodes in that part of network have this node in their routing table. Now
moving node moves to other part of network(NOT in reach of "COORDINATOR" ) and other nodes in that part don't have it in their routing table.

What happens if I send data to this moving node from COORDINATOR ?

Message is stuck in part of network which has moving node in it's routing table records?

It's more question about routing algorithm, hope you understand what I'm trying to ask.

4. Is in LW Mesh implemented CSMA/CA algorithm, because I would like to send a BROADCAST messages and I await responces form nodes?

5. Do I have to manually disable NWK_OPT_ACK_REQUEST when I send BROADCAT message or it is disabled dy default because of broadcast address 0xFFFF?

Thank you once again!

Thanks Alex !

Few questions came to my mind:

1. In LW Mesh datasheet is stated that:

So there is no POLLING mechanism like in BitCloud.

Is there going to be something slimiar to POLLING technique in future releases of LW Mesh since in new release there is API for neighbor/routing table?

When sending BROADCAST message there are NO retries(deliveries to sleeping NODEs), which nodes are awake they receive BROADCAST and send it further(in case or routers) and that's it ?

2. Suppose I have situation like this :

C sends message to ED1 and on ED1 I look at LQI and RSSI of the received frame.

How can I know if message came through R1 or R2 ?

Do I have to disable routing on e.g. R1 with

and that way I know that message to ED1 was delivered through R2 or there is some other way?

3. Can I use stack functions(in comment) like this (in int main()):

Functions(commented) are used before while(1) loop which is main thing in app?

Thank you !

Hi Alex,
I am facing another weird situation. I have tried many scenarios including two simultaneous sending nodes and one receiver - gateway. Packets are sent with no delay - after the previous was successfully sent out.
Each test was successfully done except the case when

. After receiving say 200 or 400 packets the gateway stuck on while loop in the function HAL_PhySpiWriteByteInline in halPhy.h waiting for the SPIF flag. No matter how big the message was (I have tried 1 to 105 B data payload). I am using IRIS motes (ATmega1281+RF230).

Do you know how to avoid this state please?

PS: Link to download software stack v1.1.0 doesn’t work.