The urgent prolem of PLC-G3

Hi Long qin ,

In the above topology what is the hardware that is being used ? is it TI evm ?

what are the number of SN that are joining successfully ?

how many are you trying to work with ?

What Software are you using ?

Regards .

Hi MEGHANA MANAVAZHI

Thanks a lot for your reply.

MEGHANA MANAVAZHI said:

In the above topology what is the hardware that is being used ? is it TI evm ?

The hardware is TMS320F28PLC84 + AFE031,it is designed by ourself.

MEGHANA MANAVAZHI said:

what are the number of SN that are joining successfully ?

The number is uncertain according to the environment.

The number of  SN which are joining successfully is at most  50 in a DC.

MEGHANA MANAVAZHI said:

how many are you trying to work with ?

at most 300.

MEGHANA MANAVAZHI said:

What Software are you using ?

The software of SN is  g3_plc_aes_F2806x_AFE031_CRC16_MSB_LOADNG_image.hex,

and the version is 07.1.1.0.

The software of DC are  g3_plc_dc_F2806x_AFE031_CRC16_MSB_image.hex and g3_dc_AM335x_aes_msb_loadng.exe,

and the version is 04.1.0.0.

By the way,there is something wrong when higher version softeware are used,so we just used version 07.1.1.0 for SN

and version 04.1.0.0 for DC

Regards.

hi Long Qin,

The think is that the PLC83 is the device compatible with the older F28069 . the PLC83 and PLC84 have separate builds. So if you are using this hardware please use the PLC84 builds . Even the DC builds are different .

please use the TI-PLC-G3-CENELEC-SN-F28PLC84 code for PLC84 .

For the DC the build can be found here

We have tested the routing with about 8 hops and with about 64 nodes minimum .

Regards.

Hi MEGHANA MANAVAZHI

Thanks a lot for your reply.

I used TI-PLC-G3-CENELEC-SN-F28PLC84 code for PLC84,and i also used the DC the build for PLC84 in August.

At that time,sysbios_g3_plc_aes_F2806x_AFE031_CRC16_MSB_LOADNG_ROM_image.hex was used for SN,

 g3_dc_AM335X_aes_msb_loadng.exe was used for linux,and sysbios_g3_plc_dc_F2806x_AFE031_CRC16_MSB_ROM_image.hex was used for DC.

but a problem happened at that time.

the problem is follow.

the field of LBA Address in ATTACH.request has to be set  in new version,i did not know how to set it,

ATTACH.reply was always  wrong.I asked for help from local support team,they also did not know,

so i used back to version 07.1.1.0 that is used for TMS320F28069.

So how to set the LBA Address of  ATTACH.request to make sure ATTACH.reply be right, please tell me.

Regards.

hi Long ,

Are you referring to the LBA address or the LBS coordinator address ( in most cases the DC ) ?

ideally there is no need to specifically set the LBA address it needs to be taken care itself .

if it is the LBS address then there is a provision to change it in case required using PIB (adpCoordShortAddress) at SN and using “-l” option on DC but not mandatory , it would use short address 0x0000 by default .

Regards.

Hi MEGHANA MANAVAZHI

Thanks a lot for your reply.

In fact,even if it was set 0x0000 by default,ATTACH.reply also return error code,

so SN could not join successfully.

How can i set ATTACH.request correctly?

Regards

Hi,

A few things you should check.

1. What is the distance to the next G3-PLC modem closer to the DC on the same phase. Not many G3-PLC modems use all three phases, but DC's generally do if appropriate.

2. Is there any type of disturbance, attenuation problem.

Try using another DC or demo kit to test the connection at various points closer to the first modem that is having problems. The demo kits have pretty good link statistics that can help.

Fred

Hi,

Thanks a lot for your reply.

Fred Watrous said:

1. What is the distance to the next G3-PLC modem closer to the DC on the same phase. Not many G3-PLC modems use all three phases, but DC's generally do if appropriate.

In our street lighting application,the distance to the next G3-PLC modem closer to the DC on the same phase is about 100m at most.

please refer to the picture.

Fred Watrous said:

2. Is there any type of disturbance, attenuation problem.

The disturbance is very small. a private transformer for street lighting is using.

So I think it is right as MEGHANA MANAVAZHI said,but i don't know how to get correct ATTACH.reply.

Regards.

Hi MEGHANA MANAVAZHI

I tried once again.

ATTACH.reply returned for about 5 minutes after ATTACH.request was sent,

I think it did not work well,and the result was also not correct.

ATTACH.reply is follow.

ATTACH.reply.status was 0x50AE.

ATTACH.reply.Network Address was 0xFFFF.

ATTACH.reply.PAN ID was 0x7455.

Regards.

Hi MEGHANA MANAVAZHI

I have changed DSP from TMS320F28PLC84 to TMS320F28069 both SN and DC,
and also have changed software for new IC.
But the same problem happened.
Has the new version firmware officially been used by other custom?
I really hope the problem could be solved because our products are saled.
Please help me.
Regards.

hi Long ,

Yes the new version is being used by other customers .

the 50AE error is for ADP_STAT_JOIN_NOMORE_LBA_TO_TRY which means that the device is unable to find a suitable LBA to route to join to the DC .

As per the case where you are seeing the PAN ID getting set and then the further short address not being assigned can be seen when the device tries to join to the DC directly without trying to use HOPS . this could also happen if the max hops is reached hence it tries to do so .

There is limit to the number of hops that can be taken while routing , could you check from the routing table if the number of hops has reached the limit of 14 for those that have joined successfully ?

Regards .

Hi MEGHANA MANAVAZHI

Thanks a lot for reply.

The problem of new version software for PLC84 has been solved.

I will plan to test routing in real environment of street lighting next week.

Regards.

Hi MEGHANA MANAVAZHI

The new version softwares was tested in real environment,

V7.6.7.10 was used for SN,and V4.6.0.10 was used for DC,

it seemed the routing also don't work.

please help me to analyse it.

Regards.

Hi Long Qin ,

Are you still seeing the 0x50Ae error ?

the 50AE error is for ADP_STAT_JOIN_NOMORE_LBA_TO_TRY which means that the device is unable to find a suitable LBA to route to join to the DC .

As per the case where you are seeing the PAN ID getting set and then the further short address not being assigned can be seen when the device tries to join to the DC directly without trying to use HOPS . this could also happen if the max hops is reached hence it tries to do so .

There is limit to the number of hops that can be taken while routing , could you check from the routing table if the number of hops has reached the limit of 14 for those that have joined successfully ?

Could you look at PIBs adpLBAMaxAttempts and adpMaxJoinWaitTime / Duration for Network discover to give some more tries or time to complete join in case the network is not very stable in terms of attenuation.

Regards.

Hi MEGHANA MANAVAZHI

MEGHANA MANAVAZHI said:

Are you still seeing the 0x50Ae error ?

The problem has been sovled after psksvr_AM335X had to be recomipled and reinstalled.

MEGHANA MANAVAZHI said:

could you check from the routing table if the number of hops has reached the limit of 14 for those that have joined successfully ?

I think routing algorithm does not work and  SN can not be a repeater.so the number of hops has not reached the limit of 14.

Now,every SN joins DC just directly,SN can not join DC over about 400 meter distance .

MEGHANA MANAVAZHI said:

Could you look at PIBs adpLBAMaxAttempts and adpMaxJoinWaitTime / Duration for Network discover to give some more tries or time to complete join in case the network is not very stable in terms of attenuation.

Which software tool should i use to do this experiment.

Best Regards.

Hello,

I wanted to check if the question is still open or if it has been resolved. I apologize for the delay in responses over the December holidays.

Thank you
Lori

Hi Lori

Thanks a lot  for your care.

The question has not been resolved.

Best Regards.

hi Qin,

please use the " intermediate " gui of the Zero config GUI.to send raw packets to set / get the following PIBS.

PIBs adpLBAMaxAttempts

set to 2 with second last byte to 2 : as 0x0D 80 0E 00 D3 21 10 F9 40 00 0x00 00 00 00 02 00 02 00

get 0x03 80 0A 00 4D 4F 10 6A 40 00 0x00 00 00 00

and adpMaxJoinWaitTime / Duration

set to 3 with last byte to 3 0x0D 80 0E 00 D3 21 73 EE 21 00  0x00 00 00 00 02 00 03 00

get 0x03 80 0A 00 4D 4F A8 70 21 00     0x00 00 00 00

The algorithm has been checked and we do see a 64 node panel joining with minimum 1 hop .

Regards.

Hi MEGHANA MANAVAZHI

Thanks a lot for your reply.

I will do this next few days.

I want to know i can send above message by HCT message?

In our design,MSP430 as a MCU is used,so we have to fix the bug.

Which should HCT message be used?

By the way,can the new version firmware be download in the future?

Best Regards

Hi Qin,

yes this is the raw message for SET_MAC_PIB HCT command .

Yes the version available on web can be used .

Regards.

Hi MEGHANA MANAVAZHI

Thanks a lot for your reply.

I want to confirm below question.

Do the SET_MAC_PIB HCT command must be setted before SHUT DOWN Command or after SHUT DOWN Command?

If new version will be used,do the SET_MAC_PIB HCT command must be also setted?

Best Regards.

hi Qin,

Please set the SET_MAC_PIB HCT after the SHUT DOWN command .

Yes even for new version the command is to be set .

Regards.

Hi MEGHANA MANAVAZHI

Thanks a lot for your reply.

Best Regards