Beta firmware CSV file problems?

Michael Cooper

unread,

Oct 8, 2013, 3:03:59 PM10/8/13

to flysig...@googlegroups.com

Hi All—

Some early versions of the “talking” firmware had an issue where the FlySight couldn’t keep up and would write a truncated line to the log file. I haven’t noticed any of these in the last couple of months, jumping with the latest beta firmware, but was wondering if anyone out there has had continuing problems with this.

Michael

spikeandjackie

unread,

Oct 8, 2013, 3:16:56 PM10/8/13

to flysig...@googlegroups.com

Hi Michael

Tony jumped here at the weekend and wanted to see his track but it wouldn't load into Paralog. I'm fairly sure that it was still running the original version so I updated it with the version from Github. I haven't heard of any others and there are a few using it here. I can't confirm myself as I always use the nav software. Do you think you will get a chance to look at the spoken distance option?

Regards
Spike

--
You received this message because you are subscribed to the Google Groups "FlySight Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flysight-dev...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Luke Hederman

unread,

Oct 8, 2013, 5:25:22 PM10/8/13

to flysig...@googlegroups.com

Hi Folks,

I was considering adding the Nav functions to the speaking software, but as it's a bit of effort to do that I was waiting to see if there were any further revisions first.
If the speaking version is already nearly maxing out the CPU, adding the Trig calculations for the Nav functions may push it over the edge as they are fairly processor intensive.

Michael, have you any thoughts on if there is enough processor power for both? Or can you give me any data on how the frequency and duration of audio playback commands affects the CPU?ï¿½ Perhaps I can alternate between the calculation and indication between log writes...ï¿½ Is there a new hardware revision on the horizon? (adding more cycles perhaps?)

Also, any thoughts on a way to make it easier to add the Nav functions to newer versions of the main software?ï¿½ (I haven't looked at any of the code since the last Nav version.)

On 08/10/2013 20:16, spikeandjackie wrote:

Hi Michael

Tony jumped here at the weekend and wanted to see his track but it wouldn't load into Paralog.ï¿½ I'm fairly sure that it was still running the original version so I updated it with the version from Github.ï¿½ I haven't heard of any others and there are a few using it here.ï¿½ I can't confirm myself as I always use the nav software.ï¿½ Do you think you will get a chance to look at the spoken distance option?

Regards
Spike

On Oct 8, 2013 8:03 PM, "Michael Cooper" <mic...@flysight.ca> wrote:

Hi Allï¿½

ï¿½

Some early versions of the ï¿½talkingï¿½ firmware had an issue where the FlySight couldnï¿½t keep up and would write a truncated line to the log file. I havenï¿½t noticed any of these in the last couple of months, jumping with the latest beta firmware, but was wondering if anyone out there has had continuing problems with this.

ï¿½

Michael

--
You received this message because you are subscribed to the Google Groups "FlySight Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flysight-dev...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Tom van Dijck

unread,

Oct 8, 2013, 5:29:48 PM10/8/13

to flysig...@googlegroups.com

what trig functions do you use? would a simple look-up table not be sufficient? It doesn't have to be extremely accurate right?

Tom.

On Tue, Oct 8, 2013 at 2:25 PM, Luke Hederman <luke.h...@gmail.com> wrote:

Hi Folks,

I was considering adding the Nav functions to the speaking software, but as it's a bit of effort to do that I was waiting to see if there were any further revisions first.
If the speaking version is already nearly maxing out the CPU, adding the Trig calculations for the Nav functions may push it over the edge as they are fairly processor intensive.

Michael, have you any thoughts on if there is enough processor power for both? Or can you give me any data on how the frequency and duration of audio playback commands affects the CPU? Perhaps I can alternate between the calculation and indication between log writes... Is there a new hardware revision on the horizon? (adding more cycles perhaps?)

Also, any thoughts on a way to make it easier to add the Nav functions to newer versions of the main software? (I haven't looked at any of the code since the last Nav version.)

On 08/10/2013 20:16, spikeandjackie wrote:

Hi Michael

Tony jumped here at the weekend and wanted to see his track but it wouldn't load into Paralog. I'm fairly sure that it was still running the original version so I updated it with the version from Github. I haven't heard of any others and there are a few using it here. I can't confirm myself as I always use the nav software. Do you think you will get a chance to look at the spoken distance option?

Regards
Spike

On Oct 8, 2013 8:03 PM, "Michael Cooper" <mic...@flysight.ca> wrote:

Hi All—

Some early versions of the “talking” firmware had an issue where the FlySight couldn’t keep up and would write a truncated line to the log file. I haven’t noticed any of these in the last couple of months, jumping with the latest beta firmware, but was wondering if anyone out there has had continuing problems with this.

Michael

--
You received this message because you are subscribed to the Google Groups "FlySight Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flysight-dev...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Luke Hederman

unread,

Oct 8, 2013, 5:38:36 PM10/8/13

to flysig...@googlegroups.com

You can see them in calcBearing and calcDistanceRad in this file :
https://github.com/Luke900/flysight/blob/master/FlySight/FlySight/Nav.c

You are right it doesn't need to be very accurate and at the time I did consider lookup tables but it was just easier to do it this way as I was re-using calculation functions I found online.
Perhaps it would help to use them now, but rather than dump what's working I thought I'd try and get a feel for how the talking is affecting things first.

On 08/10/2013 22:29, Tom van Dijck wrote:

what trig functions do you use? would a simple look-up table not be sufficient? It doesn't have to be extremely accurate right?

Tom.

On Tue, Oct 8, 2013 at 2:25 PM, Luke Hederman <luke.h...@gmail.com> wrote:

Hi Folks,

I was considering adding the Nav functions to the speaking software, but as it's a bit of effort to do that I was waiting to see if there were any further revisions first.
If the speaking version is already nearly maxing out the CPU, adding the Trig calculations for the Nav functions may push it over the edge as they are fairly processor intensive.

Michael, have you any thoughts on if there is enough processor power for both? Or can you give me any data on how the frequency and duration of audio playback commands affects the CPU?ï¿½ Perhaps I can alternate between the calculation and indication between log writes...ï¿½ Is there a new hardware revision on the horizon? (adding more cycles perhaps?)

Also, any thoughts on a way to make it easier to add the Nav functions to newer versions of the main software?ï¿½ (I haven't looked at any of the code since the last Nav version.)

On 08/10/2013 20:16, spikeandjackie wrote:

Hi Michael

Tony jumped here at the weekend and wanted to see his track but it wouldn't load into Paralog.ï¿½ I'm fairly sure that it was still running the original version so I updated it with the version from Github.ï¿½ I haven't heard of any others and there are a few using it here.ï¿½ I can't confirm myself as I always use the nav software.ï¿½ Do you think you will get a chance to look at the spoken distance option?

Regards
Spike

On Oct 8, 2013 8:03 PM, "Michael Cooper" <mic...@flysight.ca> wrote:

Hi Allï¿½

ï¿½

Some early versions of the ï¿½talkingï¿½ firmware had an issue where the FlySight couldnï¿½t keep up and would write a truncated line to the log file. I havenï¿½t noticed any of these in the last couple of months, jumping with the latest beta firmware, but was wondering if anyone out there has had continuing problems with this.

ï¿½

Michael

--
You received this message because you are subscribed to the Google Groups "FlySight Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flysight-dev...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Michael Cooper

unread,

Oct 9, 2013, 10:33:10 AM10/9/13

to flysig...@googlegroups.com

Hi Luke—

I don’t think processor power will be a concern. Speech has tight timing because it’s filling a 512 byte buffer 60 times per second, so a delay of just a few milliseconds results in a buffer collision. This is going to be the main timing constraint, I think—the “Tone_Task” function needs to be called perhaps every couple of milliseconds, so you can’t have another function blocking longer than that. However, that gives you a few thousand clock cycles to work with. In the worst case, you can do the same thing I’ve done with writing to the log—use a state machine to “stage” the calculation, so that only a little bit is done each time the task is called.

I’m hoping to take another look at things in the next week or so. I’d like to nail down the timing on the spoken indications so that we know logging is completely reliable, and at the same time I may be able to get a better idea of how much processor power is available.

Unfortunately, there are no “easy” changes which can be made to the hardware. The AT90USB646 is pretty much maxed out for the voltage it’s running on. I’d like to move to an ARM processor in a future hardware revision. This will give a lot more horsepower to work with, but it will also require a new round of FCC/CE testing, since it’s essentially an all-new hardware design. In the meantime, I’m trying to squeeze out everything I can from the hardware we’ve got in the field. :-)

Michael

From: flysig...@googlegroups.com [mailto:flysig...@googlegroups.com] On Behalf Of Luke Hederman
Sent: October-08-13 3:25 PM
To: flysig...@googlegroups.com
Subject: Re: [flysight-devs:34] Beta firmware CSV file problems?

Hi Folks,

I was considering adding the Nav functions to the speaking software, but as it's a bit of effort to do that I was waiting to see if there were any further revisions first.
If the speaking version is already nearly maxing out the CPU, adding the Trig calculations for the Nav functions may push it over the edge as they are fairly processor intensive.

Michael, have you any thoughts on if there is enough processor power for both? Or can you give me any data on how the frequency and duration of audio playback commands affects the CPU? Perhaps I can alternate between the calculation and indication between log writes... Is there a new hardware revision on the horizon? (adding more cycles perhaps?)

Also, any thoughts on a way to make it easier to add the Nav functions to newer versions of the main software? (I haven't looked at any of the code since the last Nav version.)

On 08/10/2013 20:16, spikeandjackie wrote:

Hi Michael
Tony jumped here at the weekend and wanted to see his track but it wouldn't load into Paralog. I'm fairly sure that it was still running the original version so I updated it with the version from Github. I haven't heard of any others and there are a few using it here. I can't confirm myself as I always use the nav software. Do you think you will get a chance to look at the spoken distance option?
Regards
Spike

On Oct 8, 2013 8:03 PM, "Michael Cooper" <mic...@flysight.ca> wrote:

Hi All—

Some early versions of the “talking” firmware had an issue where the FlySight couldn’t keep up and would write a truncated line to the log file. I haven’t noticed any of these in the last couple of months, jumping with the latest beta firmware, but was wondering if anyone out there has had continuing problems with this.

Michael

--
You received this message because you are subscribed to the Google Groups "FlySight Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flysight-dev...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Michael Cooper

unread,

Oct 9, 2013, 10:39:26 AM10/9/13

to flysig...@googlegroups.com

Hi again—

This should be doable. We’re getting close on SRAM—3274 bytes of 4096 used in the latest build. We don’t want to go too much further than that, since it starts to eat into stack space. However, we’re only at 33952 bytes of 65536 in program memory, so there is lots of space for data tables provided they use the PROGMEM attribute so that they’re stored in program memory instead of SRAM.

Michael

From: flysig...@googlegroups.com [mailto:flysig...@googlegroups.com] On Behalf Of Luke Hederman

Sent: October-08-13 3:39 PM
To: flysig...@googlegroups.com

Subject: Re: [flysight-devs:36] Beta firmware CSV file problems?

You can see them in calcBearing and calcDistanceRad in this file :
https://github.com/Luke900/flysight/blob/master/FlySight/FlySight/Nav.c

You are right it doesn't need to be very accurate and at the time I did consider lookup tables but it was just easier to do it this way as I was re-using calculation functions I found online.
Perhaps it would help to use them now, but rather than dump what's working I thought I'd try and get a feel for how the talking is affecting things first.

On 08/10/2013 22:29, Tom van Dijck wrote:

what trig functions do you use? would a simple look-up table not be sufficient? It doesn't have to be extremely accurate right?

Tom.

On Tue, Oct 8, 2013 at 2:25 PM, Luke Hederman <luke.h...@gmail.com> wrote:

Hi Folks,

I was considering adding the Nav functions to the speaking software, but as it's a bit of effort to do that I was waiting to see if there were any further revisions first.
If the speaking version is already nearly maxing out the CPU, adding the Trig calculations for the Nav functions may push it over the edge as they are fairly processor intensive.

Michael, have you any thoughts on if there is enough processor power for both? Or can you give me any data on how the frequency and duration of audio playback commands affects the CPU? Perhaps I can alternate between the calculation and indication between log writes... Is there a new hardware revision on the horizon? (adding more cycles perhaps?)

Also, any thoughts on a way to make it easier to add the Nav functions to newer versions of the main software? (I haven't looked at any of the code since the last Nav version.)

On 08/10/2013 20:16, spikeandjackie wrote:

Hi Michael
Tony jumped here at the weekend and wanted to see his track but it wouldn't load into Paralog. I'm fairly sure that it was still running the original version so I updated it with the version from Github. I haven't heard of any others and there are a few using it here. I can't confirm myself as I always use the nav software. Do you think you will get a chance to look at the spoken distance option?
Regards
Spike

On Oct 8, 2013 8:03 PM, "Michael Cooper" <mic...@flysight.ca> wrote:

Hi All—

Some early versions of the “talking” firmware had an issue where the FlySight couldn’t keep up and would write a truncated line to the log file. I haven’t noticed any of these in the last couple of months, jumping with the latest beta firmware, but was wondering if anyone out there has had continuing problems with this.

Michael

--
You received this message because you are subscribed to the Google Groups "FlySight Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flysight-dev...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Reply all

Reply to author

Forward