Date   

Re: New list user

Clint Bradford
 

Save that memory card, and snag a quality 16GB card from your favorite electronics store - then make up your own 
system boot disk -

https://www.raspberrypi.org/downloads/

You are SO close to having a bunch of fun!


Re: New list user

Joe K2UF
 

PS is good and R-pi is new from Amazon.  The write up on the sd card said it could be used with the 3b.  Maybe that was not 100% true.  I'll have to find a local group and have them point me on the right path.
 
joe 


From: RaspberryPi-4-HamRadio@groups.io [mailto:RaspberryPi-4-HamRadio@groups.io] On Behalf Of Clint Bradford
Sent: Tuesday, April 10, 2018 10:51 PM
To: RaspberryPi-4-HamRadio@groups.io
Subject: Re: [RaspberryPi-4-HamRadio] New list user

Is your power supply at least 2.5A?

Was this a purchase of a new unit? I have read that if you see the old square symbol,
you're trying to boot an outdated Raspbian with boot code that doesn't support the 3B.
Flash a recent image.

And in the future - take heed as you format your little cards for use in the Pi. I believe 
32GB and larger is the breakpoint where the card HAS to be formatted one way - DIFFERENT 
from cards with less capacity. I did mine wrong, and the little card heated up so much I could 
not hold it! But read the instructions (hams, of course, are notorious for NOT reading manuals (grin)), 
and all will be well.

Clint

Virus-free. www.avg.com


Re: New list user

Clint Bradford
 

Is your power supply at least 2.5A?

Was this a purchase of a new unit? I have read that if you see the old square symbol,
you're trying to boot an outdated Raspbian with boot code that doesn't support the 3B.
Flash a recent image.

And in the future - take heed as you format your little cards for use in the Pi. I believe 
32GB and larger is the breakpoint where the card HAS to be formatted one way - DIFFERENT 
from cards with less capacity. I did mine wrong, and the little card heated up so much I could 
not hold it! But read the instructions (hams, of course, are notorious for NOT reading manuals (grin)), 
and all will be well.

Clint


Re: New list user

Joe K2UF
 

Hi Clint,
I just purchased a PI-3 and a 5in monitor that mounts on the PI.  I also purchased a "noobs" sd card.  when I hook up everything all I get is the 'rainbow' square.
 
I did build an android AZ/EL rotor controller that couples with SatPac32 on a windows notebook.  Would like to do something similar with the R-pi but I am at the very beginning with R-pi.
 
Been chasing DX for the past several decades and am now one DX qso from the honor roll and need new challenges ;o)
 
73  Joe K2UF


From: RaspberryPi-4-HamRadio@groups.io [mailto:RaspberryPi-4-HamRadio@groups.io] On Behalf Of Clint Bradford
Sent: Tuesday, April 10, 2018 10:16 PM
To: RaspberryPi-4-HamRadio@groups.io
Subject: Re: [RaspberryPi-4-HamRadio] New list user

Hey, Joe - What are you doing with your Pi? I have GPredict satellite trackker on mine (it looks GREAT) ...
and outfitted one of my units with the SENSE HAT and camera - replicating what is aboard the International
Space Station. I show it off at my satellite presentations ... audiences love the SENSE HAT's LEDs (grin).

Any ham-related projects on your Pi on the schedule?

Clint K6LCS
http://www.work-sat.com

 

Virus-free. www.avg.com


Re: New list user

Clint Bradford
 

Hey, Joe - What are you doing with your Pi? I have GPredict satellite trackker on mine (it looks GREAT) ...
and outfitted one of my units with the SENSE HAT and camera - replicating what is aboard the International
Space Station. I show it off at my satellite presentations ... audiences love the SENSE HAT's LEDs (grin).

Any ham-related projects on your Pi on the schedule?

Clint K6LCS
http://www.work-sat.com

 


New list user

Joe K2UF
 

Hi All,

New to list and R_pi. Just making sure I have set up group correctly.

73 Joe K2UF


---
This email has been checked for viruses by AVG.
http://www.avg.com


Re: TNC-PI Checksum error

John G8BPQ
 

Adam,

I’m not sure if you are asking me or John Hansen who supplies the TNC-PI and
maintains the docs, but I don’t maintain a GitHub repository,

73,
John

________________________________________
From: RaspberryPi-4-HamRadio@groups.io
[mailto:RaspberryPi-4-HamRadio@groups.io] On Behalf Of J. Adam Sowers
Sent: 09 April 2018 22:01
To: RaspberryPi-4-HamRadio@groups.io
Subject: Re: [RaspberryPi-4-HamRadio] TNC-PI Checksum error

Hi John,

Is there a repository (GitHub or something like it) where we can keep track
of the latest version of the pitnc_[g|s]etparams programs? I had a tough
time finding a valid link a month or so ago (the one in the docs didn't
work) and I'd love to have one place to go to to get the latest version, add
feature requests, etc. 

Thanks,
Adam

On Sun, Apr 8, 2018 at 11:31 AM, John Wiseman <john.wiseman@cantab.net>
wrote:
Ok, Steve.

Thanks for letting me know.

73,
John

________________________________________
From: RaspberryPi-4-HamRadio@groups.io
[mailto:RaspberryPi-4-HamRadio@groups.io] On Behalf Of Steve Jones via
Groups.Io
Sent: 08 April 2018 14:22
To: RaspberryPi-4-HamRadio@groups.io
Subject: Re: [RaspberryPi-4-HamRadio] TNC-PI Checksum error

I have tried the new version (PITNCParamsApr18) and it looks good.
It now displays version of the pitnc_(g/s)etparams as 0.0.0.5

Below is the output from a couple of my tests:

************************************************************ Output start

pi@raspberrypi:~ $ ./pitnc_getparams 0 0
pitnc_getparams Version 0.0.0.5

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     14
8 1 0 40 a 0 0 0 0 e 4d c0 sum 0
pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2
pitnc_setparams Version 0.0.0.5
pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20
pitnc_setparams Version 0.0.0.5

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     14
8 1 14 40 a 0 0 0 0 e 59 c0 sum 0
pi@raspberrypi:

************************************************************ Output end


I will be doing more testing and will let you know if I find anything funny.

Thanks for your work and looking forward to moving forward with this!

Let me know if I can help in any way.

Steve KA6PRW






On Saturday, April 7, 2018, 2:12:43 PM PDT, John Wiseman
<john.wiseman@cantab.net> wrote:


Steve,

I think this will fix it.

http://www.cantab.net/users/john.wiseman/Downloads/PITNCParamsApr18.zip

Could you test and let me know?

Thanks,
John G8BPQ
On 07/04/2018 18:34, John Wiseman wrote:
I think it is a bug in pitnc_get/set when the checksum (or in fact any other
character) is 0E. That character is used in i2c polling to indicate nothing
available, but is being being incorrectly ignored in get/set. I'll looking
into a fix.

73,
John
On 07/04/2018 17:45, ka2dew Tadd wrote:
I thought I saw a 1 in his PIC firmware version.  Could that be the deal?
Tadd --- Sent from Planet X
http://tarpn.net   

On Apr 7, 2018, at 12:43 PM, John Wiseman <john.wiseman@cantab.net> wrote:
Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For
some reason your TNC isn't sending the checksum, so the final sum is
non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but
it would be nice to know what is happening. Do you get the same problem if
you change some of the registers?

73,
John


On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone
into the ether.

Thanks John for your update.  The new version does print PIC version (1),
but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was
pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I
am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I
also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 =
0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d =
0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the
sources that were in the zip, but I am getting an 'undefined error:
FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW


Re: TNC-PI Checksum error

J. Adam Sowers <adam.sowers@...>
 

Hi John,

Is there a repository (GitHub or something like it) where we can keep track of the latest version of the pitnc_[g|s]etparams programs? I had a tough time finding a valid link a month or so ago (the one in the docs didn't work) and I'd love to have one place to go to to get the latest version, add feature requests, etc. 

Thanks,
Adam


On Sun, Apr 8, 2018 at 11:31 AM, John Wiseman <john.wiseman@...> wrote:
Ok, Steve.

Thanks for letting me know.

73,
John

________________________________________
From: RaspberryPi-4-HamRadio@groups.io
[mailto:RaspberryPi-4-HamRadio@groups.io] On Behalf Of Steve Jones via
Groups.Io
Sent: 08 April 2018 14:22
To: RaspberryPi-4-HamRadio@groups.io
Subject: Re: [RaspberryPi-4-HamRadio] TNC-PI Checksum error

I have tried the new version (PITNCParamsApr18) and it looks good.
It now displays version of the pitnc_(g/s)etparams as 0.0.0.5

Below is the output from a couple of my tests:

************************************************************ Output start

pi@raspberrypi:~ $ ./pitnc_getparams 0 0
pitnc_getparams Version 0.0.0.5

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     14
8 1 0 40 a 0 0 0 0 e 4d c0 sum 0
pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2
pitnc_setparams Version 0.0.0.5
pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20
pitnc_setparams Version 0.0.0.5

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     14
8 1 14 40 a 0 0 0 0 e 59 c0 sum 0
pi@raspberrypi:

************************************************************ Output end


I will be doing more testing and will let you know if I find anything funny.

Thanks for your work and looking forward to moving forward with this!

Let me know if I can help in any way.

Steve KA6PRW






On Saturday, April 7, 2018, 2:12:43 PM PDT, John Wiseman
<john.wiseman@...> wrote:


Steve,

I think this will fix it.

http://www.cantab.net/users/john.wiseman/Downloads/PITNCParamsApr18.zip

Could you test and let me know?

Thanks,
John G8BPQ
On 07/04/2018 18:34, John Wiseman wrote:
I think it is a bug in pitnc_get/set when the checksum (or in fact any other
character) is 0E. That character is used in i2c polling to indicate nothing
available, but is being being incorrectly ignored in get/set. I'll looking
into a fix.

73,
John
On 07/04/2018 17:45, ka2dew Tadd wrote:
I thought I saw a 1 in his PIC firmware version.  Could that be the deal?
Tadd --- Sent from Planet X
http://tarpn.net   

On Apr 7, 2018, at 12:43 PM, John Wiseman <john.wiseman@...> wrote:
Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For
some reason your TNC isn't sending the checksum, so the final sum is
non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but
it would be nice to know what is happening. Do you get the same problem if
you change some of the registers?

73,
John


On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone
into the ether.

Thanks John for your update.  The new version does print PIC version (1),
but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was
pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I
am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I
also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 =
0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d =
0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the
sources that were in the zip, but I am getting an 'undefined error:
FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW








Re: TNC-PI Checksum error

John G8BPQ
 

Ok, Steve.

Thanks for letting me know.

73,
John

________________________________________
From: RaspberryPi-4-HamRadio@groups.io
[mailto:RaspberryPi-4-HamRadio@groups.io] On Behalf Of Steve Jones via
Groups.Io
Sent: 08 April 2018 14:22
To: RaspberryPi-4-HamRadio@groups.io
Subject: Re: [RaspberryPi-4-HamRadio] TNC-PI Checksum error

I have tried the new version (PITNCParamsApr18) and it looks good.
It now displays version of the pitnc_(g/s)etparams as 0.0.0.5

Below is the output from a couple of my tests:

************************************************************ Output start

pi@raspberrypi:~ $ ./pitnc_getparams 0 0
pitnc_getparams Version 0.0.0.5

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     14
8 1 0 40 a 0 0 0 0 e 4d c0 sum 0
pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2
pitnc_setparams Version 0.0.0.5
pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20
pitnc_setparams Version 0.0.0.5

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     14
8 1 14 40 a 0 0 0 0 e 59 c0 sum 0
pi@raspberrypi:

************************************************************ Output end


I will be doing more testing and will let you know if I find anything funny.

Thanks for your work and looking forward to moving forward with this!

Let me know if I can help in any way.

Steve KA6PRW

On Saturday, April 7, 2018, 2:12:43 PM PDT, John Wiseman
<john.wiseman@cantab.net> wrote:


Steve,

I think this will fix it.

http://www.cantab.net/users/john.wiseman/Downloads/PITNCParamsApr18.zip

Could you test and let me know?

Thanks,
John G8BPQ
On 07/04/2018 18:34, John Wiseman wrote:
I think it is a bug in pitnc_get/set when the checksum (or in fact any other
character) is 0E. That character is used in i2c polling to indicate nothing
available, but is being being incorrectly ignored in get/set. I'll looking
into a fix.

73,
John
On 07/04/2018 17:45, ka2dew Tadd wrote:
I thought I saw a 1 in his PIC firmware version.  Could that be the deal?
Tadd --- Sent from Planet X
http://tarpn.net   

On Apr 7, 2018, at 12:43 PM, John Wiseman <john.wiseman@cantab.net> wrote:
Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For
some reason your TNC isn't sending the checksum, so the final sum is
non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but
it would be nice to know what is happening. Do you get the same problem if
you change some of the registers?

73,
John


On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone
into the ether.

Thanks John for your update.  The new version does print PIC version (1),
but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was
pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I
am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I
also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 =
0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d =
0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the
sources that were in the zip, but I am getting an 'undefined error:
FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW


Re: TNC-PI Checksum error

Steve (KA6PRW)
 

I have tried the new version (PITNCParamsApr18) and it looks good.
It now displays version of the pitnc_(g/s)etparams as 0.0.0.5

Below is the output from a couple of my tests:

************************************************************ Output start

pi@raspberrypi:~ $ ./pitnc_getparams 0 0
pitnc_getparams Version 0.0.0.5

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     14
8 1 0 40 a 0 0 0 0 e 4d c0 sum 0
pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2
pitnc_setparams Version 0.0.0.5
pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20
pitnc_setparams Version 0.0.0.5

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     14
8 1 14 40 a 0 0 0 0 e 59 c0 sum 0
pi@raspberrypi:

************************************************************ Output end


I will be doing more testing and will let you know if I find anything funny.

Thanks for your work and looking forward to moving forward with this!

Let me know if I can help in any way.

Steve KA6PRW







On Saturday, April 7, 2018, 2:12:43 PM PDT, John Wiseman <john.wiseman@...> wrote:


Steve,

I think this will fix it.

http://www.cantab.net/users/john.wiseman/Downloads/PITNCParamsApr18.zip

Could you test and let me know?

Thanks,
John G8BPQ

On 07/04/2018 18:34, John Wiseman wrote:
I think it is a bug in pitnc_get/set when the checksum (or in fact any other character) is 0E. That character is used in i2c polling to indicate nothing available, but is being being incorrectly ignored in get/set. I'll looking into a fix.

73,
John

On 07/04/2018 17:45, ka2dew Tadd wrote:
I thought I saw a 1 in his PIC firmware version.  Could that be the deal?

Tadd --- Sent from Planet X

On Apr 7, 2018, at 12:43 PM, John Wiseman <john.wiseman@...> wrote:

Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For some reason your TNC isn't sending the checksum, so the final sum is non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but it would be nice to know what is happening. Do you get the same problem if you change some of the registers?

73,
John



On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone into the ether.

Thanks John for your update.  The new version does print PIC version (1), but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 = 0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d = 0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the sources that were in the zip, but I am getting an 'undefined error: FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW




--
Steve Jones
KA6PRW


Re: TNC-PI Checksum error

Steve (KA6PRW)
 

Yes will test in morning.  
Have been out and about today.
Thanks for the efforts.

Steve KA6PRW

On ‎Saturday‎, ‎April‎ ‎7‎, ‎2018‎ ‎02‎:‎12‎:‎43‎ ‎PM‎ ‎PDT, John Wiseman <john.wiseman@...> wrote:


Steve,

I think this will fix it.

http://www.cantab.net/users/john.wiseman/Downloads/PITNCParamsApr18.zip

Could you test and let me know?

Thanks,
John G8BPQ

On 07/04/2018 18:34, John Wiseman wrote:
I think it is a bug in pitnc_get/set when the checksum (or in fact any other character) is 0E. That character is used in i2c polling to indicate nothing available, but is being being incorrectly ignored in get/set. I'll looking into a fix.

73,
John

On 07/04/2018 17:45, ka2dew Tadd wrote:
I thought I saw a 1 in his PIC firmware version.  Could that be the deal?

Tadd --- Sent from Planet X

On Apr 7, 2018, at 12:43 PM, John Wiseman <john.wiseman@...> wrote:

Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For some reason your TNC isn't sending the checksum, so the final sum is non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but it would be nice to know what is happening. Do you get the same problem if you change some of the registers?

73,
John



On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone into the ether.

Thanks John for your update.  The new version does print PIC version (1), but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 = 0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d = 0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the sources that were in the zip, but I am getting an 'undefined error: FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW




--
Steve Jones
KA6PRW


Re: TNC-PI Checksum error

John G8BPQ
 

Steve,

I think this will fix it.

http://www.cantab.net/users/john.wiseman/Downloads/PITNCParamsApr18.zip

Could you test and let me know?

Thanks,
John G8BPQ

On 07/04/2018 18:34, John Wiseman wrote:
I think it is a bug in pitnc_get/set when the checksum (or in fact any other character) is 0E. That character is used in i2c polling to indicate nothing available, but is being being incorrectly ignored in get/set. I'll looking into a fix.

73,
John

On 07/04/2018 17:45, ka2dew Tadd wrote:
I thought I saw a 1 in his PIC firmware version.  Could that be the deal?

Tadd --- Sent from Planet X

On Apr 7, 2018, at 12:43 PM, John Wiseman <john.wiseman@...> wrote:

Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For some reason your TNC isn't sending the checksum, so the final sum is non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but it would be nice to know what is happening. Do you get the same problem if you change some of the registers?

73,
John



On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone into the ether.

Thanks John for your update.  The new version does print PIC version (1), but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 = 0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d = 0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the sources that were in the zip, but I am getting an 'undefined error: FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW




Re: Stratum One NTP server using a RPi

 

Yes.


On Sat, Apr 7, 2018, 05:11 WA7SKG via Groups.Io <wa7skg=yahoo.com@groups.io> wrote:
Just looked up the Adafruit Ultimate GPS Hat. Neat device! Can it be
used with a TNC-Pi for APRS? Right now I'm using a USB GPS puck, but I'm
wondering if this might also be a viable option to free up a USB port.

Michael WA7SKG


John D Hays - K7VE wrote on 04/02/2018 10:50 PM:
> Yes, I have been running one for a couple of months.
>
> I used the Adafruit Ultimate GPS Hat.
>
> On Mon, Apr 2, 2018, 22:32 w4oa@... <mailto:w4oa@...>
> <w4oa@... <mailto:w4oa@...>> wrote:
>
>     Anyone using a RPi with a gps as a NTP server? Should work wonders
>     for FT8, even if is total overkill. :)
>
>     I am working on one using an Adafruit gps, got it working today.
>     Will be checking accuracy over the next few days. The biggest hassle
>     so far is the wonky serial port/bluetooth port on the Pi.
>
>     Mitch W4OA
>




Re: TNC-PI Checksum error

John G8BPQ
 

I think it is a bug in pitnc_get/set when the checksum (or in fact any other character) is 0E. That character is used in i2c polling to indicate nothing available, but is being being incorrectly ignored in get/set. I'll looking into a fix.

73,
John

On 07/04/2018 17:45, ka2dew Tadd wrote:
I thought I saw a 1 in his PIC firmware version.  Could that be the deal?

Tadd --- Sent from Planet X

On Apr 7, 2018, at 12:43 PM, John Wiseman <john.wiseman@...> wrote:

Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For some reason your TNC isn't sending the checksum, so the final sum is non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but it would be nice to know what is happening. Do you get the same problem if you change some of the registers?

73,
John



On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone into the ether.

Thanks John for your update.  The new version does print PIC version (1), but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 = 0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d = 0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the sources that were in the zip, but I am getting an 'undefined error: FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW



Re: TNC-PI Checksum error

Steve (KA6PRW)
 

Yes - the version reports as "1". 
I get the same response (checksum error = 0xe) if I read (all) or write any register.
Steve


On Saturday, April 7, 2018, 9:45:36 AM PDT, ka2dew Tadd <tadd@...> wrote:


I thought I saw a 1 in his PIC firmware version.  Could that be the deal?

Tadd --- Sent from Planet X

On Apr 7, 2018, at 12:43 PM, John Wiseman <john.wiseman@...> wrote:

Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For some reason your TNC isn't sending the checksum, so the final sum is non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but it would be nice to know what is happening. Do you get the same problem if you change some of the registers?

73,
John



On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone into the ether.

Thanks John for your update.  The new version does print PIC version (1), but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 = 0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d = 0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the sources that were in the zip, but I am getting an 'undefined error: FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW


--
Steve Jones
KA6PRW


Re: TNC-PI Checksum error

 

I thought I saw a 1 in his PIC firmware version.  Could that be the deal?

Tadd --- Sent from Planet X

On Apr 7, 2018, at 12:43 PM, John Wiseman <john.wiseman@...> wrote:

Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For some reason your TNC isn't sending the checksum, so the final sum is non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but it would be nice to know what is happening. Do you get the same problem if you change some of the registers?

73,
John



On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone into the ether.

Thanks John for your update.  The new version does print PIC version (1), but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 = 0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d = 0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the sources that were in the zip, but I am getting an 'undefined error: FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW


Re: TNC-PI Checksum error

John G8BPQ
 

Steve,

Something odd going on..

I get
    TNC Software Version           3
01 TXDelay - Zero means use ADC  50
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         3
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     34
8 3 32 40 a 3 0 0 0 22 52 c0 sum 0

As you can see, there is one more byte, the checksum, before the c0. For some reason your TNC isn't sending the checksum, so the final sum is non-zero.

I suppose as it seems be working perhaps it isn't worth worrying about, but it would be nice to know what is happening. Do you get the same problem if you change some of the registers?

73,
John



On 07/04/2018 14:35, Steve Jones via Groups.Io wrote:
Sorry if this is a double post - but my previous reply appears to have gone into the ether.

Thanks John for your update.  The new version does print PIC version (1), but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 = 0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d = 0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the sources that were in the zip, but I am getting an 'undefined error: FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW


Re: TNC-PI Checksum error

Steve (KA6PRW)
 

Sorry if this is a double post - but my previous reply appears to have gone into the ether.

Thanks John for your update.  The new version does print PIC version (1), but still pesky checksum error. Always '0xe' from what I am seeing.
           I believe the version of pitnc_get/set that I was using was pointed to by the TNC-PI assembly manual online?
                            I have tried to get the correct versions, so I am not clear what happened there.


Note that when I manually do the xor checksum on the data examples below, I also get a '0xe' as the result:
               8 xor 1 xor 14 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 59 = 0xe
               8 xor 1 xor 00 xor 40 xor a xor 0 xor 0 xor 0 xor 0 xor 4d = 0xe

I tried both 'get' and 'set' and through in a reset "15 2", for good luck.
After all these, the checksum error remained. 

EXAMPLE #1 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 1 20

   PIC Software Version           1
01 TXDelay - Zero means use ADC  20
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     89
8 1 14 40 a 0 0 0 0 59 c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #2 *****************************************

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

EXAMPLE #3 *****************************************

pi@raspberrypi:~ $ ./pitnc_setparams 0 0 15 2

pi@raspberrypi:~ $ ./pitnc_getparams 0 0

Using Serial port /dev/ttyAMA0

   TNC Software Version           1
01 TXDelay - Zero means use ADC   0
02 Persistance                   64
03 Slottime (in 10 mS)           10
04 TXTail                         0
05 Full Duplex - Not used         0
06 Our Channel (Hex)             00
07 I2C Address (0 = async) Hex   00
   ADC Value                     77
8 1 0 40 a 0 0 0 0 4d c0 sum e
** Checksum Error - Sum = e Should be Zero

Any suggestions? 
Quite willing to look at code if that might help.  I tried compiling the sources that were in the zip, but I am getting an 'undefined error: FIONBIO'.
               I will continue to look at that.

--
Steve Jones
KA6PRW


Re: Stratum One NTP server using a RPi

WA7SKG <wa7skg@...>
 

Just looked up the Adafruit Ultimate GPS Hat. Neat device! Can it be used with a TNC-Pi for APRS? Right now I'm using a USB GPS puck, but I'm wondering if this might also be a viable option to free up a USB port.

Michael WA7SKG


John D Hays - K7VE wrote on 04/02/2018 10:50 PM:

Yes, I have been running one for a couple of months.
I used the Adafruit Ultimate GPS Hat.
On Mon, Apr 2, 2018, 22:32 w4oa@arrl.net <mailto:w4oa@arrl.net> <w4oa@arrl.net <mailto:w4oa@arrl.net>> wrote:
Anyone using a RPi with a gps as a NTP server? Should work wonders
for FT8, even if is total overkill. :)
I am working on one using an Adafruit gps, got it working today.
Will be checking accuracy over the next few days. The biggest hassle
so far is the wonky serial port/bluetooth port on the Pi.
Mitch W4OA


Re: Stratum One NTP server using a RPi

WA7SKG <wa7skg@...>
 

Last year for Field Day, I set up an RPi and a wireless router for our logging computers to use. As there was no Internet access at the FD site, I connected a USB GPS puck to the RPi (~$35 Amazon device). The RPi served NTP, DHCP, and central file server for all computers on site. Worked great. Used the same setup for some remote FT8 ops just for time purposes.

73,
Michael WA7SKG


w4oa@arrl.net wrote on 04/02/2018 10:31 PM:

Anyone using a RPi with a gps as a NTP server? Should work wonders for FT8, even if is total overkill. :)
I am working on one using an Adafruit gps, got it working today. Will be checking accuracy over the next few days. The biggest hassle so far is the wonky serial port/bluetooth port on the Pi.
Mitch W4OA

6041 - 6060 of 13943