[App_rpt-users] Fwd: COR signal getting confused

Sat Sep 27 04:47:12 UTC 2014

The ctcss and carrier from are just bits so technically you could use either one and it would work the same. You do need to set the one you are not using to none. We always use carrierfrom even though it is ctcss that is driving it.

I tried what you are doing here. I listened to myself iaxrpt from a distant node. Then I key the local PTT on my handheld into my local node. I immediately unkey and then keyed again before I heard the courtesy tone and it worked fine. Tried it several times.

Why don't you eliminate the 1620 node in the test and iaxrpt directly to 1998 to see what happens there.

This is acting like what you would see with an rxdelay setting but it should be 0 by default unless you somehow got that into your config.   Specifically say   rxdelay=0  in the 1998 rpt.conf context to be sure.
73 Doug
WA3DSP
http://www.crompton.com/hamradio

From: kf7mbk at w7ara.net
Date: Fri, 26 Sep 2014 21:15:03 -0700
Subject: Re: [App_rpt-users] Fwd: COR signal getting confused
To: doug at crompton.com
CC: app_rpt-users at ohnosec.org

Doug, this behavior happens even if I manually toggle the COR independent of the radio and audio so unless I'm not understanding, I can't see this being the radio.
I tried changing unkeywait  to 1000 but made no difference other than extending the time the PTT is active for the courtesy tone, and thus extending the window in which the COR line going high will be ignored.  Seems that anything that extends the time that the PTT is active for the courtesy tone also extends the time that the COR is ignored.
Just to make sure I'm not doing anything stupid or have an unconventional setuo,. my test setup is:
HT -> DMK URI -> BBB (node 1998) -> ethernet -> Allstar on a VM (node 1620) -> IAX -> Zioper soft phone on laptop
HT has audio in/out connected to the URI.  PTT from the URI is not connected to the HT while testing.  Signal from receive LED on the HT goes to COR URI input.
I have the HT tuned to the weather frequency and I'm making/breaking the antenna connection to simulate picket fencing and I'm monitoring the audio received by the URI/BBB on the softphone.
Bottom line is if the COR line drops, the courtesy pause and tone is instantly triggered, triggering the URI PTT. This works.  But, if while the PTT is active, the COR line goes high again, the audio through the URI is not routed through until the COR line goes low* and high again after the PTT/courtesy tone is done transmitting.
* when going low after getting ignored going high, no courtesy tone is triggered.

Would it be useful/appropriate change carrierfrom to "no" and instead set ctcssfrom = usb and switch the radio output to the ctcss input on the URI and see if the problem still exists?
 Thanks for all the help!-Mike

On Fri, Sep 26, 2014 at 8:37 PM, Doug Crompton <doug at crompton.com> wrote:

Mike,

 You can try the changing the courtesy time wait then set back to duplex=1

The wait-times section
This section allows the wait times 
for voice responses and other audio telemetry events to be adjusted. All
 wait times are in milliseconds.

telemwait
 sets the time before sending the general audio telemetry responses used
 in the application which are not specifically defined below
idwait sets the time to wait before sending ID audio telemetry
unkeywait
 sets the amount of time to wait before sending the courtesy tone. This 
can be used to ensure breaking stations can get their turn
calltermwait sets the amount of time between an autopatch disconnect and the audio telemetry message indicating the call has been terminated

The unkeywait is the one to mess with. It is 400ms on the BBB by default. Make it longer like 1000 so it spans the dropout period you are seeing. Experiment with times.

If that doesn't work I really think you should try a different radio. My thoughts are that it is some kind of recovery issue with TX. 

73 Doug
WA3DSP
http://www.crompton.com/hamradio

From: kf7mbk at w7ara.net
Date: Fri, 26 Sep 2014 18:59:26 -0700
Subject: Re: [App_rpt-users] Fwd: COR signal getting confused
To: doug at crompton.com
CC: app_rpt-users at ohnosec.org

OK, so I did some more testing.
I have routed the signal out of the radio to an NPN and switched from usbinvert to usb as suggested.  This made little difference, but in this configuration I could make the RC work a little better bridging the dropouts.  But still would end up not detecting carrier after a quick dropout even though the COR signal into the dongle was high.
Duplex = 0 seems to eliminate the problem. Is there a setting to tell it how long to wait after COR dropped to start transmitting the courtesy tone? (though I suppose that would defeat the purpose of the duplex=1 setting)  Note that I have the PTT disconnected for testing, so all I see is the red LED come on as soon as the COR out of the radio drops.  
Actually as I'm typing this and playing with it, I just noticed it appears if the COR is restored ANY time the PTT is active for the courtesy tone, it will ignore the COR line until it's dropped and brought back up AFTER the PTT is done. This behavior seems VERY repeatable, even with a cleanly restored COR signal,  So it may have nothing to do with quick dropouts and everything to do with the courtesy tone..
Are there any settings that control this?
The LED on the radio (and hence my COR input)  is controlled by noise squelch opening.

On Fri, Sep 26, 2014 at 3:05 PM, Doug Crompton <doug at crompton.com> wrote:

Mike,

 Are you using PL or noise squelch or both? Like I said I have not seen this problem and I am using basically stock timing. Every time the PL or squelch goes untrue for a certain (short) duration you should hear a courtesy tone. Perhpas you should try duplex=0 just to see if there is an issue with TX recovery.  Do you have another radio you can try?  

73 Doug
WA3DSP
http://www.crompton.com/hamradio

From: kf7mbk at w7ara.net
Date: Fri, 26 Sep 2014 10:18:19 -0700
Subject: Re: [App_rpt-users] Fwd: COR signal getting confused
To: doug at crompton.com
CC: app_rpt-users at ohnosec.org

The scope image in the OP shows there doesn't appear to be a problem with pulling hi or low,  The transitions (without the RC) are very clean and go all the way to ground..  It seems like the URI is doing edge detection instead of level detection, and it's just missing fast edges.  Wasn't sure if this was something that could be tweaked in software or if I'll have to put some external debounce logic on the line.

On Fri, Sep 26, 2014 at 10:07 AM, Doug Crompton <doug at crompton.com> wrote:

Mike,

 Before you change any timing make sure as Jon pointed out that your COS signal is going hard high and low on signal and non-signal. You might need a pullup. The URI is kind of weak in that regard.  It should be TTL levels  <.5 volt low and >2.5 volt high and preferably rail to rail.

73 Doug
WA3DSP
http://www.crompton.com/hamradio

From: kf7mbk at w7ara.net
Date: Fri, 26 Sep 2014 09:56:21 -0700
Subject: Re: [App_rpt-users] Fwd: COR signal getting confused
To: doug at crompton.com
CC: app_rpt-users at ohnosec.org

Doug,
First: thanks for the excellent writeup on getting allstar running on the BBB.  I used it to get up and running very quickly and is the setup I am having this issue with.
I am running duplex =1 as well.  I have not changed any timing values, (I'm essentially running the as-installed setup from the install image on your site) but I will take a look at them and see if I can make some changes for the better.
Thanks!-Mike

On Fri, Sep 26, 2014 at 9:49 AM, Doug Crompton <doug at crompton.com> wrote:

From: doug at crompton.com
To: kf7mbk at w7ara.net
Subject: RE: [App_rpt-users] Fwd: COR signal getting confused
Date: Fri, 26 Sep 2014 12:46:14 -0400

That is very strange. Have you changed any of the default timing values? Are you using simplex/duplex? Courtesy tone? Timing values could effect that. 

I use simplex with courtesy tone (duplex=1 in rpt.conf)  on several nodes and I have never seen that problem. I have one user who has a signal like you describe. He walks in a park with a handheld and he is often in and out.

Also I assume you are not using any rxdelay?

73 Doug
WA3DSP
http://www.crompton.com/hamradio

From: kf7mbk at w7ara.net
Date: Fri, 26 Sep 2014 08:50:51 -0700
To: app_rpt-users at ohnosec.org
Subject: [App_rpt-users] Fwd: COR signal getting confused

Resending since my initial post has been pending moderation for 2 days now and got no reply from the list owner email.

I have modified an HT to bring out the receive LED signal and use it

as the COR input to a DMK URI. It is a digital signal as viewed on a

scope (i.e. very short slope in the transitions)  This seems like it

would work well and it does on strong signals. But on marginal signals

where the LED may blink on and off quickly (like during picket

fencing), it seems to confuse the URI (or the software) were it will

start off good, but then a quick dropout of the COR signal will cause

the software to no longer detect carrier, even though the signal on

the COR input has been restored.  Even if the input is solid high (or

low in my case, I'm using usbinvert) it will not detect a carrier

until there is another dropout of adequate time to "reset" it.

I've tried putting a simple RC on the input but it didn't really help

much.  Seems the URI (or software) needs a significant minimum time

between transitions to properly detect them, long enough that it would

delay the carrier detection too much without a circuit more complex

than an RC (i.e. something that would allow instant keyup, but delay

the key down.)

Linked is a scope image. 
https://lh5.googleusercontent.com/-CsJ9autUTAE/VCWK-uNJJpI/AAAAAAAAPSM/VsVvbPVvuEA/s800/20140923_184107.jpg
Carrier was detected until this signal

dropout event.  Ch A is my RC filtered signal sent to the URI, Ch B is

the signal out of the radio.  Carrier was not detected after the 5th

dropout even though the signal remained low.

Anyone ever experience this and/or have ideas on how to fix it?

Thanks.

Mike

KF7MBK

_______________________________________________
App_rpt-users mailing list
App_rpt-users at ohnosec.org
http://ohnosec.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://ohnosec.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the "Unsubscribe or edit options button"
You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.keekles.org/pipermail/app_rpt-users/attachments/20140927/69be9e4a/attachment.html>