Pm3 Error Code 380

[old] ChrisW · Post by **[old] ChrisW** » May 16th, 2005, 5:04 pm

The PM3 screen (App Ver 976) displays this error when I attempt to issue a PM3 specific command from my application. The strange thing is the expected values are returned to my app - but the PM3 displays this error screen each time a command is issued. Does anyone have any ideas what is happening? Thanks

[old] ChrisW · Post by **[old] ChrisW** » May 16th, 2005, 5:08 pm

To be more specific, the actual error code is 380 -35. Chris

[old] mlyons · Post by **[old] mlyons** » May 16th, 2005, 7:58 pm

<table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><div class='genmed'>QUOTE(ChrisW @ May 16 2005, 04:08 PM)</div></td></tr><tr><td class='quote'>To be more specific, the actual error code is 380 -35. Chris </td></tr></table> Chris, Please list the specific command frame you are issuing to cause this error. The SDK document has a list of error codes in an appendix. Error 380 indicates an "invalid CSAFE command". If you are using a PM3-specific command, you may not be "wrapping" it properly with the appropriate public CSAFE command. Mark

[old] ChrisW · Post by **[old] ChrisW** » May 17th, 2005, 3:33 pm

Mark, I am attempting to use CSAFE_PM_GET_STROKESTATE using the CSAFE_SETUSERCFG1_CMD wrapper: //get the stroke state int[] cmdDataX = new int[4]; cmdDataX[0] = 0x1A; cmdDataX[1] = 1; cmdDataX[2] = 0xBF; cmdLen = 1; i = PM3DDI.tkcmdsetCSAFE_command(0, cmdLen, cmdDataX, replyLen, replyData); val = replyData[4]; This approach works for commands like CSAFE_GETHORIZONTAL_CMD, perhaps I can't use this approach for PM3 specific commands? The thing is, the call to get the stroke state works (ie it returns the correct data), the problem is that it immediately causes the PM3 error 380. Thanks for any ideas or comments you may have. Chris

[old] mlyons · Post by **[old] mlyons** » May 17th, 2005, 4:56 pm

<table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><div class='genmed'>QUOTE(ChrisW @ May 17 2005, 02:33 PM)</div></td></tr><tr><td class='quote'>Mark, I am attempting to use CSAFE_PM_GET_STROKESTATE using the CSAFE_SETUSERCFG1_CMD wrapper: //get the stroke state int[] cmdDataX = new int[4]; cmdDataX[0] = 0x1A; cmdDataX[1] = 1; cmdDataX[2] = 0xBF; cmdLen = 1; i = PM3DDI.tkcmdsetCSAFE_command(0, cmdLen, cmdDataX, replyLen, replyData); val = replyData[4]; This approach works for commands like CSAFE_GETHORIZONTAL_CMD, perhaps I can't use this approach for PM3 specific commands? The thing is, the call to get the stroke state works (ie it returns the correct data), the problem is that it immediately causes the PM3 error 380. Thanks for any ideas or comments you may have. Chris </td></tr></table> Chris, I'll have to check V976 behavior to confirm what you are seeing. There have been numerous bug fixes in the area of "'wrapped" PM3-specific commands. I have verified proper behavior of your command frame using the current Beta firmware release. I would suggest you update your PM3 with the Beta firmware release (see <a href='http://www.concept2.com/05/rower/servic ... update.asp)' target='_blank'>http://www.concept2.com/05/rower/servic ... update.asp)</a> and continue your development. Let me know if you have further difficulties and continue to visit this forum for additional Beta firmware updates. Mark

[old] haboustak · Post by **[old] haboustak** » May 18th, 2005, 10:17 am

<table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><div class='genmed'>QUOTE(ChrisW @ May 17 2005, 02:33 PM)</div></td></tr><tr><td class='quote'>//get the stroke state int[] cmdDataX = new int[4]; cmdDataX[0] = 0x1A; cmdDataX[1] = 1; cmdDataX[2] = 0xBF; cmdLen = 1; i = PM3DDI.tkcmdsetCSAFE_command(0, cmdLen, cmdDataX, replyLen, replyData); val = replyData[4]; </td></tr></table> Shouldn't cmdLen be 3 in this case, not 1? I don't know how that will affect the packet that's assembled by CSAFE_command, but it very well could affect the CRC and stuff. As a sidenote - if you're writing C++ or C#, I like to assemble my frames this way: UINT32_T cmd_data[CSAFE_FRAME_LENGTH]; UINT32_T cmd_len=0; cmd_data[cmd_len++] = 0x1A; cmd_data[cmd_len++] = 0x01; cmd_data[cmd_len++] = 0xBF; It eliminates some counting on my part and some chance that cmd_len is not equal to the actual data in the buffer. Mike

[old] ChrisW · Post by **[old] ChrisW** » May 18th, 2005, 7:19 pm

<table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><div class='genmed'>QUOTE(haboustak @ May 18 2005, 09:17 AM)</div></td></tr><tr><td class='quote'><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td><div class='genmed'>QUOTE(ChrisW @ May 17 2005, 02:33 PM)</div></td></tr><tr><td class='quote'>//get the stroke state int[] cmdDataX = new int[4]; cmdDataX[0] = 0x1A; cmdDataX[1] = 1; cmdDataX[2] = 0xBF; cmdLen = 1; i = PM3DDI.tkcmdsetCSAFE_command(0, cmdLen, cmdDataX, replyLen, replyData); val = replyData[4]; </td></tr></table> Shouldn't cmdLen be 3 in this case, not 1? I don't know how that will affect the packet that's assembled by CSAFE_command, but it very well could affect the CRC and stuff. As a sidenote - if you're writing C++ or C#, I like to assemble my frames this way: UINT32_T cmd_data[CSAFE_FRAME_LENGTH]; UINT32_T cmd_len=0; cmd_data[cmd_len++] = 0x1A; cmd_data[cmd_len++] = 0x01; cmd_data[cmd_len++] = 0xBF; It eliminates some counting on my part and some chance that cmd_len is not equal to the actual data in the buffer. Mike </td></tr></table> Mike Thankyou - you were absolutely correct - fixing cmdLen eliminated my error! Thanks again, Chris