Hi Karl,
Thanks for your reply. Sorry for I can't follow u right now. You did mention the increase the interval > 250ms, where can I change the value, or how? Can you show me where to change the interval >250ms so that I'm also can test it and let you know the result.
I've tried the demo program that I've download from the sourceforce website, where the name is "NE64_Vend_OpenTCP", and I just program into the eval board and when I connect from the PC using the TCP client software, I can see the program are keep sending data to the server. May I know how they do it? Thanks.
KahJoo
Hi KahJoo,
Hi KahJoo,
I think you may have misunderstood me. My suggestion was to change the code running on the PC, not the code running on the NE64.
The code you included in your first post is running in the NE64 and is similar to the code I have running on my NE64.
Freescale are still looking into this problem.
They are trying to investigate just using pings from the PC to the NE64 rather than using an application on the PC. I've tried this too and I can get the NE64 to time-out once the buffer length goes over 1600 bytes. For example...
800 bytes seems fine...
ping 192.168.2.3 -t -w 1000 -l 800
1600 bytes and you get time-outs
ping 192.168.2.3 -t -w 1000 -l 1600
This doesn't seem quite right to me. It may not necessarily be the same issue as the speed test but something's not right.
I've done this with 2 completely different sets of code in the NE64 and the result is the same. One set of code uses the OpenTCP stack but the other set of code is based on the example that comes with the demo box so it doesn't even use OpenTCP.
Perhaps you could try this ping test and see if you get time-outs too ?
Regards,
Karl
I too noticed my Pings to the NE64 would be dropped at certain times during the day.
Turned out it was checksum building alg that the NE64 using that has a bug in it, after I rewrote it, it worked like a charm now.
I don’t remember the details now, but if you need more info let me know.
Hi Karl,
I think I miss out your point. I'll tried to use the ping function that you show to me to test on the NE64 board, and also test on the module that I've develope up, and going to finish of the 8 bit MCU with Ethernet function. Right now my devices Ethernet module are under testing and up to now working quite good with no problem. But the limitation is the Ethernet module that I've develope are working on 10Bast-T, so is suitable just for data command sending.
Sorry for that I'm quite not understand what's is the meant of the red colour and bold term that you show me:- ping 192.168.2.3 -t -w 1000 -l 800 ?
By the way, right now I'm understand what's your concern about. However I'll test it on my side and will let you know the result. Besides, if you have any updated information, please let me know. Thanks.
KahJoo
For anyone still interested in this....
I've had a Freescale service request open for several weeks regarding my speed / buffer length issue. (Note that this may not be the same problem that Malaysia / KahJoo is suffering from)
Freescale managed to replicate the problem. The case was forwarded to the design centre so that the hardware could be checked. There was no fault found in the hardware. Freescale then tried a similar set of tests but the NE64 code was built using the CMX TCP/IP stack rather than the OpenTCP stack. The fault could no longer be replicated. The conclusion therefore is that this behaviour is due to a bug in the OpenTCP stack. Unfortunately, Freescale can no longer provide support for the OpenTCP stack since Viola Systems have started charging 20,000 Euros for the product. Since I don't have the time or expertise to debug the OpenTCP stack, we have decided to try the CMX stack which is not cheap but its a lot less expensive than the new (?) 'supported' OpenTCP stack.
Hi Karl,
Thanks for your update and sorry for late reply. On my side, I've test out the decive as below:-
1) I tried to ping the decives and it can be done and the reply time was ~10ms by using the cross cable.
2) I tried to send the packet from PC to the devices and the data was out from the COMM port, and it works even I tried to send every packet by delay of 50ms. However I don't know I right or not. Here are the code that I've wrote for the testing:-
for(i=0;i<par1;i++)
{
TCP_In_Msg[i]=RECEIVE_NETWORK_B();
while(!SCI0SR1_TDRE);
AS1_SendChar(TCP_In_Msg[i]);
}
if(TCP_In_Msg[8]=='9')
{
Bit1_NegVal();
}
For me I'm just received the packet from PC Server and then transfer out from the RS232. Am I follow you.
3) However I tried to resend back the packet, as long as reseived the data packet from PC Server, to PC server, but after a while the MCU was look hang on there and can't receiving any more packet. I've tried to increase the delay time between data packet sending from PC server to 1s even more to 2s, the situation was same that the MCU will stop receving packet and as well sending packet back to PC server.
4) Currently I want to test where I received the data from RS232 and send the data back to PC server and would like to see what will happens. By the way can you show me how to receive the RS232 in Interrupt Mode to received the data and send the data to PC server?
5) The final testing that I want to tried is the MCU will just sending the data packet to PC server, preset the data need to be send like the example code from NE64_Vend_Demo project file, and find out will the sending packet will occur any problem?
I've attached my project file to have a check if needed. Thanks.
Regards,
KahJoo