FYI - Another trunker/etrunker Windows port

[slicerwizard]

*shrug* Maybe the "Slicer Wizard" can answer that.

Well, I see his question about conversion utilities was answered by a helpful group member after only 30 minutes. Doesn't look dead to me.
Quiet? Certainly. Dead? Hardly.

 

[fmon]

Nice work Eric, my Motorola systems start immediately with wtrunker, however, neither etrnk will work with either normal or invert set.

 

[SCPD]

Doesn't look dead to me. Quiet? Certainly. Dead? Hardly.

Thanks Andrew. I'm not a member so I can't see what's going on over there. Someone who actively uses T88 should write up a review in the Wiki. Would be a nice thing to have (apologies to Eric for distracting this thread).

-rick

 

[EricCottrell]

Hello,

Occasionally I get an EDACS system that does not decode and find that I forgot to change the receiver from Super Narrowband FM to Narrowband FM.

I have a nearby SitePro EDACS system being built-out and tested. SitePro decoding and ESK will be included at a future time. If I can get a recording of a site when ESK is being sent it will help.

I found and fixed a few problems. I am in the process of getting LTRDump working under windows. I can decode LTR by feeding the discriminator audio into the computer but I am not totally happy with it. The lousy low frequency response of laptop sound inputs had me scratching my head for a bit. If a long series of ones or zeros occurs the signal tends to go to zero amplitude. So I need to do some more coding to fix it.

I am looking into the other problems mentioned on the thread. Thanks for the comments.

73 Eric

 

[SCPD]

I have a nearby SitePro EDACS system being built-out and tested. SitePro decoding and ESK will be included at a future time. If I can get a recording of a site when ESK is being sent it will help.

You got it.
-rick

 

[SCPD]

The lousy low frequency response of laptop sound inputs had me scratching my head for a bit. If a long series of ones or zeros occurs the signal tends to go to zero amplitude. So I need to do some more coding to fix it.

You now have a 5 meg WAV file in your inbox.
-rick

 

[EricCottrell]

You now have a 5 meg WAV file in your inbox.
-rick

Hello,

Thanks. I need to look over the EDACS code but one area I will look at is automatic ESK detection. I also want to get rid of the 4800 baud version as the only differences are the data rate of the control channel and inverted data.

There are a number of other areas I want to investigate. I need to read some more about the Windows MultiMedia API. A user problem with other windows based decoders seems to be the setting of audio levels. I want to see if I can put in code to adjust the audio level or to indicate when the audio level is too high or too low.

When I work with my AOR5000 or AOR 8600 discriminator output I have to put in a resistor to prevent loading down of the output. There is still signal on the output but I can tell it is loading down because the squelch control can not shut off the audio. So there is still some interfacing needed between the radio and the computer audio.

73 Eric

 

[blantonl]

Eric, all you need to do to enable ESK support is to XOR the EDACS command word (1st 8 bits) in msg 1a with "0A".

Also, does this new etrunker app support multiple sound cards? For instance, Can i run two separate versions of etrunker simultanously?

 

[SCPD]

XOR the EDACS command word (1st 8 bits) in msg 1a with "0A".

Don't confuse the guy. 0xA0 will work.

Also, does this new etrunker app support multiple sound cards? For instance, Can i run two separate versions of etrunker simultanously?

If you're running Windows XP - it will work (a feature of the OS). For Windows 9X and 2K, it probably won't work.

-rick

 

[pro92b]

Don't confuse the guy. 0xA0 will work.
-rick

That works if the system uses the default setting. It would be more flexible to have the program automatically calculate the key and apply it. If that is too difficult then the user should be able to manually override the default key if necessary.

 

[EricCottrell]

Hello,

If you are going from a data slicer to computer input be aware that you can not usually directly plug the discriminator tap into the audio input of a computer. The audio input will load down the discriminator output. It may not affect the sound output level very much but it can mangle the data enough that poor decoding results. In my case the AOR 5000 seems more sensitive to loading distorting the signal then my AOR 8600. This is the case with any decoding software that uses the audio inputs.

The classic solution is to use a buffer amplifier. I am going to experiment and see if a 2 or 3 resistor pad could also work. The idea is to match the impedances but the downside of a passive resistor network is the signal level will be lower.

You can use recording software like audacity to record the control channel and see what the data stream looks like. The signal should have sharp transitions from the positive level to the negative level. The positive and negative levels should not "drift" up and down very much. The signal should look like a stream of open squares and rectangles. Make sure the signal does not go all the way to maximum output (i.e. clipping).

My solution for the AOR 5000 is to build a buffered breakout box so I can also record the discriminator out on another computer while decoding with the main computer. I would rather not build a box that requires power from batteries or a serial port.

Multi-Net uses xor as well and I was able to figure out a way of determining it. I need to look at the bits to see how it is applied for ESK and if there is a pattern to test for.

First off I need to get the program to work. I have been cleaning up the sound code and fixing a few misunderstandings on how the windows console works. I also managed to introduce a few new bugs as well. The program ran all night on a 650 MHz Thinkpad 600x. I want to try it on a couple of Pentium 120s and 133s with Windows 95 to see if it works.

73 Eric

 

[blantonl]

Multi-Net uses xor as well and I was able to figure out a way of determining it. I need to look at the bits to see how it is applied for ESK and if there is a pattern to test for.

Take the idle command word that you see often in the EDACS data stream, XOR it with "0A" and use that as a pattern to test for. If you don't see the regular EDACS idle command word for 2 secs, XOR the stream, and look for the XOR'd idle command. Rinse and repeat as necessary.

 

[SCPD]

Take the idle command word that you see often in the EDACS data stream, XOR it with "0A" and use that as a pattern to test for. If you don't see the regular EDACS idle command word for 2 secs, XOR the stream, and look for the XOR'd idle command. Rinse and repeat as necessary.

I used a slightly different technique. In any case, be prepared for some funky decode results for that brief time period where the ESK mask is wrong. I'm note sure why Lindsay keeps referring to a value of "0A". If you treat the 28 bit EDACS code word as a 32 bit value and apply the XOR to the upper 8 bits then yes, that will work. It you're XOR'ing the upper 8 bits of the 28 bit code, then you'll want to use 0xA0. It's six for one, half-a-dozen for the other.

 

[DaveNF2G]

WTrunker and EWTrunker both fail on my XP Pro machine.

Using the Run command, the screen flashes black, then returns to the desktop.

Using the Command Prompt, I get an error message saying, "Failed to set video modes."

 

[jellotor]

Haven't had a problem yet with the 'new' trunker version for Windows. Working great decoding a Moto2 VHF system here since last Thursday.

 

[EricCottrell]

WTrunker and EWTrunker both fail on my XP Pro machine.

Using the Run command, the screen flashes black, then returns to the desktop.

Using the Command Prompt, I get an error message saying, "Failed to set video modes."

Hello,

The error message indicates it could not set the window to display 43 rows. Try setting the command prompt window to be 80 x 43. Right click on Command Prompt in the Toolbar and select properties. On the Layout tab, select 80 for width and 43 for height. Microsoft documentation indicates the setting of the screen size will fail if a window of the requested size can not be created due to factors like the font size being too big.

I did a test on my XP Pro machine and did not have any problem. I was able to test with a USB serial dongle as this laptop does not have any serial ports. I was able to run full screen or windowed.

I am almost to the point where I can start to add features. I got the beep to work and had to figure out an alternative to blinking. In the DOS version the program turned on and off the beep and used a duration that I figured as about 100 MS. The beep function in windows needs a duration specified so I use 100 MS. Different frequencies of beeps are used for talkgroup beeps and radio id beeps. In Win2K and XP the frequency of the beep can be specified but in earlier versions of windows the frequency of the beep can not be specified. One area I still need to test is when the emergency button is pressed.

Blinking is not supported directly in windows and is replaced by background intensity. This means that colors 8 to 16 can be specified for the background. As an experiment I turned on background intensity and flip the foreground intensity when blinking is specified. A default black background with light grey (code 7) text gets turned into a dark grey (code 8) background with bright white (code 15) text. It does stand out on the screen and will be included in the next release for people's opinions.

I am also looking into a problem where 44.1k sampling rate works but 22.05k sampling rate does not. I think it may be due to high frequency components from the radio being aliased into the frequency band of interest. With LTR I have the problem of 22.05k sampling rate works but 44.1k sampling rate does not. I want to merge the LTR stuff into a common codebase so I need to work it out. I am also not happy with the weak signal performance of the decoder. This may be due to using passive devices for matching to my radios. I sometimes deal with weak signals when DXing so weak signal performance is an area of interest. First I got it to work, now I am trying to get it to work right.

73 Eric

 

[pboy]

I am also not happy with the weak signal performance of the decoder. This may be due to using passive devices for matching to my radios. I sometimes deal with weak signals when DXing so weak signal performance is an area of interest.

Eric,

What does "using passive devices for matching to my radios" mean in relation to soundcard decoding?

Thanks.

 

[DaveNF2G]

Hello,

The error message indicates it could not set the window to display 43 rows. Try setting the command prompt window to be 80 x 43. Right click on Command Prompt in the Toolbar and select properties. On the Layout tab, select 80 for width and 43 for height. Microsoft documentation indicates the setting of the screen size will fail if a window of the requested size can not be created due to factors like the font size being too big.

A combination of Command Prompt settings has solved the problem.

In addition to your suggestion above, I found it necessary to visit the Options tab and select Window under Display Options. After that, all of the ported trunkers started up.

 

[EricCottrell]

Eric,

What does "using passive devices for matching to my radios" mean in relation to soundcard decoding?

Thanks.

Hello,

Using resistors and capacitors. Components that do not require power.

73 Eric

 

[pboy]

Hello,

Using resistors and capacitors. Components that do not require power.

73 Eric

Okay. In line with the discriminator tap.

Just didn't register that was what you were referring to.