Thetis 2.9 20 dB Attenuator

[w0ivj]

In Thetis 2.9 the 20 dB and 30 dB attenuator give the same attenuation on RX1. The attenuator works properly on RX2.

[w-u-2-o]

How did you test this?

[w0ivj]

Using Thetis v2.9.0 v64, I disconnected the antenna from Rx1 and Rx2 and set the attenuation to 0 dB. I increased both attenuators in 10 dB steps while monitoring the noise floor of the receivers. The noise floor should increase by 10 dB for each attenuator step change. The 20 dB attenuator on Rx1 increased the noise floor by 20 dB instead of 10 db which was the same for the 30 dB attenuator. This problem does not happen on Rx2 or Thetis v2.8.11. I am using Protocol 1 in both cases.

[w-u-2-o]

I can't reproduce your problem here with 2.9.0.6 (hardware is an 8000).

Assuming you are using 7000 it would be good if someone else could check this on that hardware.

Note also that there are physically two separate attenuators on the RF board: a 10dB attenuator and a 20dB attenuator. The 30dB attenuation setting simply turns on both of them.

I hate to say it, but you might try a fresh database to see if that fixes it.

Note for additional testers: just to reiterate, you can only see the noise floor rise as described if the antenna port is disconnected or, better yet, attached to a dummy load. It will not be exactly 10dB but it will be close enough. Otherwise the normal noise floor coming out of the antenna is much higher than the attenuator impact on the receiver noise figure.

[w0ivj]

Are there separate attenuators for each receiver?

[w-u-2-o]

Are there separate attenuators for each receiver?

This is an excellent question and one that made me look at the 7000 schematics. Again, I'm assuming you are using a 7000, is that correct?

If I'm reading the schematics correctly, the 7000 does not have a the separate 10 and 20dB attenuators that all the other ANAN models do. It relies solely on the 31dB step attenuators that are mounted on the Orion MKII SDR board itself, one for each ADC. Thus when you select ATT mode it merely adjusts the step attenuator as required.

Would you try your tests again with both RX1 and RX2 in S-ATT mode and report if the results are any different. That may help isolate the problem to either hardware or software.

Also, do you have RX1 assigned to ADC0 and RX2 assigned to ADC1? Look in Setup > General > ADC.

[w0ivj]

I am using an ANAN-7000DLE MKII running Protocol 1. I have one of the 7000s that have the FPGA timing problems that prevent me using Protocol 2.

Looking at the only block diagram that I have found (at the bottom of the Thetis User Manual edited by G8NJJ),
I have determined that there are separate 31 dB attenuators for each receiver. They are MiniCircuit DAT-31-SP+ attenuators.

I ran my previous tests using the S-ATT attenuator settings and the problem does NOT occur! The panadapter and the S-meter move in 1 dB steps together and the 20 db setting and the 30 db setting are correct. The 20 dB and 30 dB attenuator settings do not affect a change using the ATT tab. However, with a -70 dBm signal in Antenna 1, ADC R and ADC L both change with the attenuator switched between the -20 dB setting and the -30 dB setting using ATT. In fact ADC R and ADC L read the same when the attenuator is set by ATT or S-ATT. This makes me believe the problem is software and not hardware.

I have the RX1 assigned to ADC0 and RX2 assigned to ADC1 which is the default for Protocol 1.

If I set RX2 to ADC0 and apply a -70 dBm carrier into Antenna 1, I get -70 dBm on the panadapter and the S-meter on both receivers. If I change the attenuator on RX1 to -10 dB, RX1 panadapter and S-meter read -70 dBm, but RX2 reads -80 dBm until I change the RX2 attenuator to agree with RX1 attenuator, and then RX2 reads the same as RX1. This interaction between the attenuators does not seem correct to me.

I apologize for the length of this post.

[w-u-2-o]

Your understanding of the DAT-31 is correct. You should make a request to the factory for full schematics. They will provide them. They continue to ask that they not be posted publicly, unfortunately.

The problems with attenuator settings when RX1 and RX2 are both assigned to ADC0 are definitely associated with software bugs, and this has been known, although perhaps not widely known. It has been discussed elsewhere on this forum, though.

I'm not entirely sure what's going on when RX1 and RX2 are assigned to different ADC's. It would be good if someone else with a 7000 would duplicate your tests. However it does seem that S-ATT mode works perfectly fine, correct?

Again, it bears mentioning that testing by looking at the noise floor of a disconnected or, better yet, terminated antenna input is not a situation that holds true for normal operations. First, the normal noise level coming out of an antenna will be much greater than the system noise figure and therefore a 1:1 correspondence between attenuator setting and noise floor level will not be seen. Second, the software does account for the attenuator setting such that all signal levels are referenced to the rear panel connector. Thus, for example, a -70dBm signal will remain a -70dBm signal on the panadapter regardless of attenuator setting (leaving aside the aforementioned RX1/RX2/ADC0 bug, of course).

[oe3ide]

Just made a quick check on my 7000, Protocol 2 v2.1.18, Thetis 2.9.0.7 alpha

no antenna connected, USB, 2,7khz bw, auto-agc off
RX 1 DDC2 - ADC0
RX 2 DDC3 - ADC0


ATT:
0 db => -115 dbm
-10 db => -105 dBm
-20 db => -95 dBm
-30 db => -85 dBm

S-ATT
0 => -115 dBm
5 => -110 dBm
10 => -105 dBm
15 => -100 dBm
20 => -95 dBm
25 => -90 dBm
30 => -85 dBm

73 de Ernst

[w-u-2-o]

Thank you, Ernst!

Were the results the same for both RX1 and RX2?

Would you mind trying the case where RX2 is assigned to ADC1?

[oe3ide]

Thank you, Ernst!

Were the results the same for both RX1 and RX2?

Would you mind trying the case where RX2 is assigned to ADC1?

RX2 ADC0 identical to RX1 ADC0.

RX1 ADC0 (RX2 ADC1) => same results
RX2 ADC1 (RX1 ADC0) => same results

That means I get the expected values in all combinations.

73 de Ernst

[w-u-2-o]

Thank you, Ernst!

[w0ivj]

The difference may be Protocol 1 compared to Protocol 2. Both of you are running Protocol 2, whereas I am running Protocol 1. I am stuck on Protocol 1 because of the FPGA timing issue. Has this FPGA timing issue ever been corrected in Protocol 2?

Also, if I turn on power on the application with RX2 selected, RX2 S-meter reads -388 dBm. I am able to correct this by switching bands and then switching back. If I turn the application power on with RX2 not selected and then select RX2, the S-meter problem does not occur.

[w-u-2-o]

Which P2 versions have you tried?

[K4IBC]

I am confused. This whole ordeal with the hardware change and firmware. He says Mark II but does he have the Mark III? I thought all the 7000 prior to July 1st would run P1 and P2.
Website is still calling them Mark II.

[w-u-2-o]

How old is your ANAN, Tom? When did you get it? After 1 July 2022?

[w0ivj]

My 7000 was shipped on 03 April 2020, so the FPGA change does not effect me. It was shipped with PowerSDR and protocol 1. When I tried Thetis and Protocol 2 shortly after I received the radio, the radio would not work and the consensus at that time was that there was an FPGA timing issue in some radios. Later when a version of Thetis came out that would work with protocol 1, I switched to it. I am currently running Thetis v2.8.11 (02/08/21) with protocol 1. If protocol 2 has been updated since late 2020, I am not aware of it. If the FPGA timing problem has been fixed since late 2020, I will give that a try.

[w-u-2-o]

Hi Tom,

Yes, you are way, way behind the times, both software and firmware-wise.

The first thing I'd strongly recommend is to upgrade to Thetis 2.9.0.6. There are a lot of new features, bug fixes and other enhancements in 2.9.0.6. You can download the installer for that here:

https://github.com/ramdor/Thetis-2.9.0/releases/tag/v2.9.0.6

I recommend you install the 64 bit (x64) version.

The second thing you should do after getting that running the way you like it is to again try P2 firmware, specifically P2 firmware revision 2.1.18. Rick, N1GP, put in a ton of time making P2 work better. So much better that it's very rare to find a 7000 or 8000 that won't run 2.1.18. Not guaranteeing it will work on your hardware, of course, but there's a good chance it will. If you don't mind a little work with HPSDR Bootloader to try it, that is. And of course be prepared to use Bootloader to go back to P1 if there is a problem.

You can find 2.1.18 at this link:

https://community.apache-labs.com/viewtopic.php?f=32&t=3192

[w0ivj]

Well, I downloaded Thetis 2.9.06 and P2 2.1.18 and the attenuator issue is still there and now PureSignal will not work. I did find on my computer a Protocol 1: metis_Orion_MkII_v2.7_K5SO_5-4.rbf which is the latest P1 that I have. At this time, I think I will continue to use Thetis with P1, unless someone has further ideas. Thanks everyone for all the help. 73, Tom

[w0ivj]

Also, A special thanks to you Scott!!