apache-labs.com

Richie,

Will be nice to have the option in TCI spots to show the spotter callsign when holding "shift" key like we have on Thetis's spottig system.

ea3aqr wrote:Richie,

Will be nice to have the option in TCI spots to show the spotter callsign when holding "shift" key like we have on Thetis's spottig system.

That info is not included in the the spot message as part of the TCI protocol unfortunately. The additional text is normally for stuff like 'cq cq dx' etc.

I guess 3rd party logging/cluster software that supplies the TCI message to Thetis could include it in the additional argument, but some sort of common approach/format would have to be taken, or EE asked to add a SPOT_DETAILED message with extra arguments.

R.

In the F/W SET tab, the "Anan Options", which should be selected Dither or Random?
Right now both are selected.
thanks,
Don W9BHI

Ritchie, can you take a look at Auto AGC in rc2. The offset now seems to be way off, needs way more offset in db, not proportional.

NC3Z wrote:Ritchie, can you take a look at Auto AGC in rc2. The offset now seems to be way off, needs way more offset in db, not proportional.

I'm seeing the same thing.

I was just in 160 meters SSB and had the PS engaged and the correction bar was flashing on and off during transmission instead of being on steady.
The att. was at 11db.
I also got an SEQ error alert.
Here it is:

I've just installed rc-2b and having issues with RX2 using TCI client

Prior to up grade I was running two instances of JTDX - 1 for WSPR & the other for FT8 on 50Mhz.

Everything worked as expected. One client controlled RX1 rxtx & other client controlled RX2 rxtx.

Since the upgrade I can no longer change the RX2 VFO via the client 2. If I manually change the VFO B that is reported by the client.
If I TX the via RX2 client output is displayed on RX2 window - as expected.

RX1 client is behaving as previously. Able to change and read VFO via JTDX and TX goes to RX1 window.

I have tried upgrading the database file, restarting etc - no avail.

POST SCRIPT . I have gone back to RC1 -- Everything working correctly again. I'll wait for feedback before installing RC2 again

By the way the ability you created in the last upgrade to have two clients running RX1 & RX2 was for me was one of the the most useful changes made to Thetis in the last few years!

w9mdb wrote:rc2 seem to take less CPU...good job...
But the Random problem is still there. It's random though
Sometimes behaves OK and sometimes not.
Happens on both Tx and Rx
Here it is on Tx with random off

Having dither and random on improves your spur free dynamic range on receive. Both options directly control the dither and random pins on the ADC. Assuming you are in DUP, you are merely seeing the difference in performance on receive. Are you running an external coupler, are you perhaps overloading the receiver input occasionally? In addition, anything below -70dBc will rarely, if ever be seen on the air, only if you are S9+40 to someone.

The chances of this having anything to do with the latest release, or any release, is nil. If you think you might be overdriving the RX input, start a new topic.

At any rate, you want them both turned on. See here: viewtopic.php?f=6&t=2736

W9BHI wrote:In the F/W SET tab, the "Anan Options", which should be selected Dither or Random?
Right now both are selected.
thanks,
Don W9BHI

You want them both turned on. See here: viewtopic.php?f=6&t=2736

w2ner wrote:

NC3Z wrote:Ritchie, can you take a look at Auto AGC in rc2. The offset now seems to be way off, needs way more offset in db, not proportional.

I'm seeing the same thing.

Seems like it's OK here after adjusting the Auto AGC setting in the <DSP><AGC/ALC> menu. See attachment.

Paul, W9AC

Richie thanks for rc2-b,
now consumes less CPU power even when RX2 is enabled.
Subjectively, it seems a bit more reactive to me.

Re-enabled Auto-AGC after Thetis restart (with MIDI mapped knob) still refuses to work.

I am seeing the same auto AGC offset problem. Increased the offset as a workaround for now

NC3Z wrote:Ritchie, can you take a look at Auto AGC in rc2. The offset now seems to be way off, needs way more offset in db, not proportional.

On RX1 At 192KHz sample rate my dB Offset now has to be an addition 12dB, or in other terms 12dB of offset puts the Ga line equal to the noise floor line. And and number below 12dB offset puts the Ga line below the NF line.

On RX2 that offset dB is 24db for above.

Before RC2 the Offset was pretty close to what was expected.

I am seeing the same issue here with the Auto AGC. The offset is way off now, before seemed to be proportional.

Mike

I just can't explain the agc rc2 problem...

Video : https://youtu.be/T6Ab-cKXGKY

(watch it on x2 speed if I am too slow, just woke up )

RIchie.

Riche,

After watching the video, my Auto AGC result tracks with yours. In the earlier screen capture, I should have made Rx2 the same as Rx1. My point was that behavior seemed normal upon changing the Rx1 setting after the initial install.

Paul, W9AC

I can't see any differences in the offset. Both RX offset lines/values are identical for me. RX1 and RX2 are both configured identically.

Hi everybody,
did someone connect his RF-Kit RF2KS with Thetis via TCI?

Norbert dl8laq and I have had no success so far. After checking the Thetis settings and the Windows firewall, I ran a tcpdump on the Raspberry Pi in the RF2KS PA.

sudo tcpdump portrange 40010-40019 -w thetis-12.pcap

The IP of the PA is 192.168.6.146, that of the PC with Thetis is 192.168.6.116. The port set is 40011. I use the portrange because I have been surprised several times.

Switching from http to websocket works and Thetis transmits all data (ready;). However, a ping<>pong then starts in the websocket connection. At this point, further TCI communication begins with my ExperSDR3. I am not involved enough to determine if it is a problem on the Thetis or the RF-Kit side.

Settings:


PCAP File:
https://1drv.ms/u/s!Aqf-r97tkKLb-C_WzPGRO4ho0GEh?e=X8GH8d

73 Thomas, DF2LH

Does the PA implement multiple TCI client connections? So able to connect to Thetis + EESDR3 at the same time? If not I don't quite understand how it can work with two TCI servers at the same time (Thetis + EESDR3).

Thetis sends out ping frame every 20 seconds (websocket spec), and does not need a reply. However, it will reply with a pong frame if it receives a ping frame. If the frame send fails due to loss of connection for example, Thetis will close the connection.

Edit: I will be adding log windows for both TCI + CATtcpip next week so you can see inbound and outbound messages.

R.

ramdor wrote:I just can't explain the agc rc2 problem...

Video : https://youtu.be/T6Ab-cKXGKY

(watch it on x2 speed if I am too slow, just woke up )

RIchie.

For me, simply stated, before RC2b, a ZERO DB offset used to place the Gain Line at exactly the same level as the Noise Floor line on the panadapter, which made sense to me, The AGC gain tracked the noise floor with a 0 db offset.

Jeff
NJ2US

Hi Richie, I saw it immediately on RX1 , never tested RX2. To correct I had to increase the offset value from 20 to 43

EDIT; an offset value of 20db puts the AGC line on the noise floor

ramdor wrote:I just can't explain the agc rc2 problem...

Video : https://youtu.be/T6Ab-cKXGKY

(watch it on x2 speed if I am too slow, just woke up )

RIchie.

Tony EI7BMB wrote:Hi Richie, I saw it immediately on RX1 , never tested RX2. To correct I had to increase the offset value from 20 to 43

EDIT; an offset value of 20db puts the AGC line on the noise floor

ramdor wrote:I just can't explain the agc rc2 problem...

Video : https://youtu.be/T6Ab-cKXGKY

(watch it on x2 speed if I am too slow, just woke up )

RIchie.

25db for me, previously it was set to 0db

ramdor wrote:

ea3aqr wrote:Richie,

Will be nice to have the option in TCI spots to show the spotter callsign when holding "shift" key like we have on Thetis's spottig system.

That info is not included in the the spot message as part of the TCI protocol unfortunately. The additional text is normally for stuff like 'cq cq dx' etc.

I guess 3rd party logging/cluster software that supplies the TCI message to Thetis could include it in the additional argument, but some sort of common approach/format would have to be taken, or EE asked to add a SPOT_DETAILED message with extra arguments.

R.

spotTCIprotocol_message.png

What a pitty!

Could you show, at least, ARG5 on the selected spot when holding "shift" key?

w9ac wrote:Riche,

After watching the video, my Auto AGC result tracks with yours. In the earlier screen capture, I should have made Rx2 the same as Rx1. My point was that behavior seemed normal upon changing the Rx1 setting after the initial install.

Paul, W9AC

Before if you set the dB offset number for Auto AGC in the box for automatic noise floor compensation say 15, then the Green Ga line was 15dB above the noise floor on the spectrum. Now the numbers do not match and are not proportional.
You can raise the dB offset in the box to get the offset but it is around double the original number.

Cecil
W4CM

NJ2US wrote:

ramdor wrote:I just can't explain the agc rc2 problem...

Video : https://youtu.be/T6Ab-cKXGKY

(watch it on x2 speed if I am too slow, just woke up )

RIchie.

For me, simply stated, before RC2b, a ZERO DB offset used to place the Gain Line at exactly the same level as the Noise Floor line on the panadapter, which made sense to me, The AGC gain tracked the noise floor with a 0 db offset.

Jeff
NJ2US

With RC2B the noise floor is hit at zero offset and is shown 30dB above noise floor with 30 dB offset - but only if Display/RXn/BinWidth is set to the maximum left side. Interestingly regardless of Sample Rate setting with alters binwidth automaticly.

The effect of auto AGC on AF-Level seems correct in my opinion, but i think the positioning of the AGC line on display is off depending on binwidth setting.

73 Walter

NJ2US wrote:For me, simply stated, before RC2b, a ZERO DB offset used to place the Gain Line at exactly the same level as the Noise Floor line on the panadapter, which made sense to me, The AGC gain tracked the noise floor with a 0 db offset.

Jeff
NJ2US

I now see the point offered by Jeff and others. 20 dB of display correction is required for both Rx1 and Rx2 to equal the noise floor when connected to a dummy load. It does seem like a zero value should coincide with the noise floor. Apart from the display discrepancy, Auto AGC is functioning normally.

Paul, W9AC

In rc2-b, G will now move depending upon fft bin widths, just like it does when auto agc is off. It does this to maintain perceived level of audio. If you set minimum fft size and an autoagc offset of 0, G should sit approximately on noise floor in the spectrum view. If you then increase the fft size, G will be adjusted down. There is no way to keep G at noise floor even with a 0dB autoagc shift anymore as it moves depending on fft size.

R.

ramdor wrote:In rc2-b, G will now move depending upon fft bin widths, just like it does when auto agc is off. It does this to maintain perceived level of audio. If you set minimum fft size and an autoagc offset of 0, G should sit approximately on noise floor in the spectrum view. If you then increase the fft size, G will be adjusted down. There is no way to keep G at noise floor even with a 0dB autoagc shift anymore as it moves depending on fft size.

R.

Thank you Richie, your explanation makes sense now, we were all "holding it wrong" (a reference to the infamous Steve Job's antennagate iphone fiasco)

Jeff NJ2US

Richie,
Just checking, has this AGC auto change been applied in the download from where you posted above? Do we need to perform another download?

Thanks for everything and 73's
Joe
WD5Y

Joe wrote:Richie,
Just checking, has this AGC auto change been applied in the download from where you posted above? Do we need to perform another download?

Thanks for everything and 73's
Joe
WD5Y

the released rc2-b is the one to use atm. I am not working on it this weekend as I need a break Will look into anything that needs attention next week now.

Cheers, Richie.