New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

FIRMWARE TOPICS ONLY--non-firmware topics will be MOVED
Forum rules
Until such time as the New Protocol firmware goes into general release, all discussion will be concentrated here.
kd4jm
Posts: 22
Joined: Fri Oct 25, 2019 12:31 am

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby kd4jm » Tue Mar 10, 2020 11:34 pm

n1gp wrote:Tnx guys for your testing a great feedback.

pre9 works flawlessly on my ANAN-10. I can't make PS misbehave!

This next test image I added a reset to the rx0 & rx1 fifos on a transition of PTT with the
thought that rx0 & rx1 may get out of sync on PTT transitions.

Purely experimental, but looking forward to see if this may make PS more stable:

https://www.dropbox.com/s/9b9b2kdnh1wmyhh/Hermes_Protocol_2_v10.8_pre10.rbf?dl=0

-Rick / N1GP


Hi Rick,
pre10 PS still working in/out for me.
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Wed Mar 11, 2020 6:14 am

Hi Rick & all others,

Many thanks for digging into this so hard and especially for not giving up!

Unfortunately no fix here as well - jumps in and out.

I wrote it before, but it's interesting to see that in all cases when PS-A stops the "GetPk" value stops reading - and a second later it jumps to half the value and stays there.
When working correctly GetPk shows changing values of about 0,29x - when stopping it jumps to 0,14x (half the value) and freezes.
Releasing and pushing the ptt starts a new game.

Rick, if you can't produce this failure at your end it's probably very hard to analyze this in depth. How can you do that?
If I would live in the US I would send you my ANAN 10 for testing - all I can offer is a little non-professional movie via my dropbox.
I could try to show you what's happening.

73,

Andy
Last edited by dl6eat on Wed Mar 11, 2020 11:30 am, edited 1 time in total.
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Wed Mar 11, 2020 11:06 am

Hi Rick,

Thanks to our time difference I already captured a little movie - it's 3 minutes long. If you find some minutes please watch it:

https://www.dropbox.com/s/hn8ovs34nmxsy2y/ANAN%2010%20problem_DL6EAT.mp4?dl=0
(Movie removed - see updated version a bit further down)

Please note all numbers in the Linearity menu (in the box "Calibration Information") get frozen when PS stops.
PS stops unexpected, I would say random - you never know what's next.... :roll:

Thanks and 73,

Andy
Last edited by dl6eat on Fri Mar 13, 2020 12:50 pm, edited 1 time in total.
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
User avatar
Arno Imig
Posts: 11
Joined: Sun Oct 20, 2019 6:53 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby Arno Imig » Wed Mar 11, 2020 5:06 pm

Hello Rick,
also I have made a YouTube video about the problem P.S. jumps in / out.
It shows the same reaction as with video of Andy.
(The video is without sound!!)

https://www.youtube.com/watch?v=5gdQyzIeA1U

Always pay attention to the mouse pointer.

I wonder why you can't reproduce the problem at home. How exactly does your setup look like? All those who have the problem with P.S. in / out use a big PA and a sampler to feed the HF signal back into the ANAN / Hermes.

Rick, do you also have an external PA with a sampler in use?

https://www.qrz.com/db/DL5AI
73. DL5AI
Arno
Rig: Hermes Board
User avatar
n1gp
Posts: 175
Joined: Sun Apr 09, 2017 6:34 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby n1gp » Thu Mar 12, 2020 12:31 am

Hello Andy Arno and ALL,

Tnx so much for the videos. I have watched them very closely to make sure I understand
your settings and operation. You both have the same symptoms with the peaks going to 1/2.
Seems maybe 1 out of 4 transmissions you lose the PS. Interesting that it is on a PTT transition.
Once you engage PTT either it is on and STAYS ON or it is OFF and STAYS OFF.

What sample rate are you using? My testing with 100%PS is with 192K.
My setup is a bare ANAN-10 and no coupler, so internal RF. Very much different I am guessing
that you two with the Big PA and couplers. I am on a poor man's budget :)

But I am able to reproduce your PS problems if I put my sample rate to 384K or higher. So I will
go on this for now and try to get PS working better on at least 384K. I know it works well on the
Orion's.

-Rick / N1GP
User avatar
n1gp
Posts: 175
Joined: Sun Apr 09, 2017 6:34 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby n1gp » Thu Mar 12, 2020 1:27 am

So I have been playing with the 384K and greater sampling rates just because I can reproduce the
PS problem dropping out with 1/2 the peak.

EDIT!

I can also reproduce at 192K, just not so easily.

Working on it...

Tnx for your testing!

-Rick / N1GP
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Thu Mar 12, 2020 8:04 am

Good morning Rick and all others,

Well, I'm using 192K as well - and get the failure very frequently - a demonstration could be done right away and would take me under 1 minute.
This failure is not happening from time to time - it is present all the time.

Since the ANAN-10 is only capable to produce 10W (with its internal little PA) most of us use bigger high gain amplifiers (either LD-MOS, VMOS or tubes) and use low drive power of just a couple watts (1-3W). BTW, in Europe it's legal to manufacture and offer amps with super high gain, such as 50 dB or more (think there are some FCC limitations in the US, correct?).

The ANAN-10 can not handle Pure Signal without modification - or let's better say not really as it should.
The reason is the build-in relay K19 which switches the RX to ground when transmitting. (big isolation!)

Since your ANAN shows a much better and more stable behavior than ours I'm guessing Pure Signal stays alive due to internal cross talk.
When using the ANAN (not modified) you need quite a bit of feedback-power to get Pure Signal running.
Also its dangerous using bigger amps and not modifying the ANAN - imagine K19 is not grounding correctly due to dirty contacts. In this case you are blowing up the internal att or even worst the ADC.

An easy fix is removing the cable from internal plug J29 and feeding the ANAN by external RX and TX cables…..the rest can easily be programmed within Thetis. Pictures and test results are provided @ http://www.qrz.com/db/DL6EAT (scroll down a bit…)

Coming to the point: My guess is your ANAN stays alive due to undefined crosstalk while ours get a specified external low power signal of just about 1 mW (0 dBm) via an external fully sensitive RX-port.
That power level is relatively low but at least according specs. And this level will never produce disturbing unwanted crosstalk.
Just my guess but that could be the reason for our differences.

Is it possible that the current F/W misses to monitor the RX port correctly and all the time? Forgive all my stupid ideas….I'm not a programmer at all: May be there should be somehow a watchdog function ensuring the feedback signal will be monitored continuously? It appears to me that the signal gets lost - and when its lost it stays lost until ptt is released. Or its not even recognized from the beginning….as demonstrated in my little movie. The behavior is at least not consistent - sometimes it works for a minute or two - then it gets lost again and needs to be triggered by 2-tone tests…..strange!

BTW, I can confirm that many, many folks here in Europe use the Hermes board, the Hermes-copy of Gerd Loch or the original ANAN-10 of Apache Labs (like I do).
I have never heard one of them not facing this issue - all the very same. But all use external amps - no QRP. The failure can always be demonstrated right away. The only way getting around this is using "Single Calibration" - not good but Thetis becomes usable on Hermes.
And to be honest Rick once you start using Thetis you don't wonna miss it again :-)

Your help and your efforts are more than appreciated Rick! Think once this is covered the person who finally solves this gets a Golden Star on the "Walk of Hamradio Fame" :-) Thanks so much and please keep fighting…..

73,

Andy
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
User avatar
n1gp
Posts: 175
Joined: Sun Apr 09, 2017 6:34 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby n1gp » Fri Mar 13, 2020 1:47 am

Hey Andy and ALL,

I should have described my setup more accurately. My ANAN-10 has been modified via K19
so that it doesn't ground and the signal from the SWR network same as the ANAN-200/7000
Tx signal is fed back for PS running 'barefoot'.

Never-the-less, I am able to reproduce your problem with PS going away and the peak going to
1/2 the value. Only to re-appear on another PTT transition, usually 1 out of 4 times.

I find that this happens MUCH more frequently after my ANAN-10 warms up.

I understand that this is a problem for MANY Hermes - ANAN-10/100 users.

There is a fundamental difference in the firmware for the ANAN-10/100 in that there are only
2 receivers in which rx1 is switched out during xmit + PS. There are differences in the circuitry,
FPGA, on-board regulators, potential Fuse problems, and I wonder what differences exist in the code in Thetis
for the ANAN-10/100 -vs- Orion based boards?

I know I cannot troubleshoot this problem on a platform such as you have with high power and couplers,
but I feel I can at least reproduce the symptoms on my 10 watt setup.

I have merged in the FW code from the (PS working) Orion Protocol 2 to this Hermes code, which the Orion (ANAN-200/7000) seems somewhat stable in regards to PS. But this makes no difference on the Hermes it seems. What is the problem? :(

One thing I'll to try is to rule out the GigE code. I have compiled the same pre10 image based on 100T to work on 1000T code.
I wonder if you could test this. I have and what I found is that it works much better, but not perfect.

https://www.dropbox.com/s/h35057milmy7ab4/Hermes_Protocol_2_v10.8_pre12_100T.rbf?dl=0

Tnx!

-Rick / N1GP
kd4jm
Posts: 22
Joined: Fri Oct 25, 2019 12:31 am

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby kd4jm » Fri Mar 13, 2020 2:44 am

n1gp wrote:Hey Andy and ALL,

I should have described my setup more accurately. My ANAN-10 has been modified via K19
so that it doesn't ground and the signal from the SWR network same as the ANAN-200/7000
Tx signal is fed back for PS running 'barefoot'.

Never-the-less, I am able to reproduce your problem with PS going away and the peak going to
1/2 the value. Only to re-appear on another PTT transition, usually 1 out of 4 times.

I find that this happens MUCH more frequently after my ANAN-10 warms up.

I understand that this is a problem for MANY Hermes - ANAN-10/100 users.

There is a fundamental difference in the firmware for the ANAN-10/100 in that there are only
2 receivers which rx1 is switched out during xmit + PS. There are differences in the circuitry
FPGA, on-board regulators, potential Fuse problems, and I wonder what differences exist in the code in Thetis
for the ANAN-10/100 -vs- Orion based boards?

I know I cannot troubleshoot this problem on a platform such as you have with high power and couplers,
but I feel I can at least reproduce the symptoms on my 10 watt setup.

I have merged in the FW code from the (PS working) Orion Protocol 2 to this Hermes code, which seems somewhat stable in regards to PS.
This makes no difference on the Hermes it seems. What is the problem? :(

One thing to try is to rule out the GigE code. I have compiled the same pre10 image to work based on 100T from the 1000T code.
I wonder if you could test this. I have and what I found is that it works much better, but not perfect.

https://www.dropbox.com/s/9591225xqcy2zb2/Hermes_Protocol_2_v10.8_pre12_100T.rbf?dl=0

Tnx!

-Rick / N1GP


Hi Rick
I tried the 10.8pre12 and PS still working the same for me with Thetis 2.7.0 a2 Anan100, I tried 192 and 384 sample rate made no differences.

Thanks for not giving up!

Jorge kd4jm
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Fri Mar 13, 2020 6:08 am

Hi Rick and all,

Many thanks for digging into this.
Well, if your ANAN 10 is modified in regards of K19 there is physically no difference to our setup, or?
Feedback is feedback no matter where it comes from - the level should be good but the rest should be pretty similar.
Sorry about my long lesson about these ANAN-10 specifics - I should have asked you before.
Many over here do not know that and run PS anyway - but wonder about poor results.

I will be out for the next 6 hours - but have your new F/W already on my NUC and will report my results a bit later today.

Many thanks - I'm glad we are talking the same issue and we are on one page.
I'm a rf geek - never did programming. Really poor not to be able to add anything to this here…..

73,

Andy
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Fri Mar 13, 2020 12:37 pm

Hi Rick - hi all together,

Find another 3,5 minutes movie - nothing has really changed (unfortunately):

https://www.dropbox.com/s/cl3mzkp38wasdr1/ANAN%2010%20problem_pre12_DL6EAT.mp4?dl=0

Many thanks for all your work - don't hesitate to send me any new version or trials for testing - I'm more than happy to do at least my part: Testing!

73,

Andy

PS: I deleted the previous movie in order to gain some space at my Dropbox
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
User avatar
n1gp
Posts: 175
Joined: Sun Apr 09, 2017 6:34 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby n1gp » Fri Mar 13, 2020 7:54 pm

Andy, Jorge, Arno, ALL,

Tnx for the new video Andy. My testing shows very similar results with pre12 but
more like 1 out of 10 PTTs and only after my ANAN-10 has warmed up.

I am not feeding back that much power from my K19 SWR bridge modification so
when I transmit with PS set to auto-attenuate my S-ATT goes to 0. I notice yours
goes between 14 & 15, just to note a difference.

Is there a certain version of Protocol2 firmware that works for you guys? If so I'd like
to know which version so I may do a comparison of the source code. Also has Thetis
versions made any difference?

BTW, I had added a bootloader switch to my ANAN-10 quite some time ago. I had to replace it
today since I wore it out :)

-Rick / N1GP
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Fri Mar 13, 2020 8:26 pm

Hi Rick,

As far as I know no Thetis version works on Hermes boards...the only work-around is "Single Calibration" instead of Pure Signal in auto-mode (PS-A).......very annoying.

My hope was that there might be an easy fix just ensuring that the feedback signal will be monitored and calculated all the time while transmitting. Especially because it works fine - if it works. Its just that stupid instability....

Since the results are displayed in numbers my guess was that this can be found and traced. Again - just my ideas.....

Take your time Rick.....may be it is possible at the end.

If your s-att setting is next to zero it may result in a different behavior. I use 60 dB externally....at 1.5 KWout So I feed the ANAN with 1,5 mW and face 15 dB additional internal additional attenuation (by the s-att) ..... so after that internal step att we get about 0,04 mW or -13,5 dBm.......
If yours show 0 dB s-att your feedback is at around that level.....

@Arno: Do know about any Thetis version which works with the ANAN 10 regarding PS-A?

73,

Andy
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
User avatar
Arno Imig
Posts: 11
Joined: Sun Oct 20, 2019 6:53 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby Arno Imig » Fri Mar 13, 2020 8:51 pm

Hello, Andy and everybody,
I don't know of any Thetis version that runs stable on ANAN10 / Hermes,
and I have tested many versions.
And all the people I know have the same problem.

If I switch to protocol 1, Power SDR runs 100% stable.

73, DL5AI
Arno
Rig: Hermes Board
kd4jm
Posts: 22
Joined: Fri Oct 25, 2019 12:31 am

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby kd4jm » Sat Mar 14, 2020 1:27 am

Hi Rick,

Here is a video on how my Anan 100 PS with pre12 is working for me, if I switch to portocol 1 it work every time.

https://photos.google.com/photo/AF1QipM ... f9wDuyqXo3

73,
Jorge kd4jm
User avatar
n1gp
Posts: 175
Joined: Sun Apr 09, 2017 6:34 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby n1gp » Sat Mar 14, 2020 1:28 am

Hello All,

I went through this Apache-Labs website searching for clues. Mainly trying to see if ANYONE had success
with Hermes / ANAN-10/100 and Protocol2 / Thetis.

I came across this series of posts where it appears Rob W1AEX claimed running Hermes_Protocol_2_v10.3 : "I also ran Pure Signal and it is correcting reliably and behaving perfectly. I am running it with RX1 and RX2 set for a 192K sampling rate as you suggested it should it be at this time and so far everything has been running smoothly for the past hour." Hopefully I didn't mis-read that post.

https://apache-labs.com/community/viewtopic.php?f=32&t=2392

This made me investigate version 10.3 more closely. First I tried it as-is from the github TAPR site and it would not work at all
with pure signal. I deleted my database and restarted but still no PS. This was with the latest 2.7.0-a2 and earlier versions.

So I decided to try merging in the Hermes_Protocol_2_v10.8_pre12_100T changes I made. I was careful not to change
any ADC or DAC timings, only update protocol2 changes which is why I believe PS was not working at all.

The below image is the result. NOTE that it reports as version 10.3 and I have not tested anything more than xmt & rcv & PS
at this point. But wanted to throw this out there since I am seeing consistent PS behaviour on my ANAN-10 but only @ 192K BW.
If I go to 384K I lose PS.

https://www.dropbox.com/s/49s4fctuor4ytel/Hermes_Protocol_2_v10.8_pre13_100T.rbf?dl=0

Same FW but at Gige rate (1000T)
https://www.dropbox.com/s/iip3lr0hiy0cuyv/Hermes_Protocol_2_v10.8_pre13_1000T.rbf?dl=0

EDIT
After playing with these images for some time unfortunately I still see the PS dropouts.

-Rick / N1GP
kd4jm
Posts: 22
Joined: Fri Oct 25, 2019 12:31 am

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby kd4jm » Sat Mar 14, 2020 2:25 am

Rick,

Should I delete my database before I try pre13 100T ?

Jorge
User avatar
n1gp
Posts: 175
Joined: Sun Apr 09, 2017 6:34 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby n1gp » Sat Mar 14, 2020 2:30 am

Jorge,

You might try it before, then delete, then after...
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Sat Mar 14, 2020 12:22 pm

Hi Rick and team,

Tried both versions Pre13 - no significant changes…..sri to say.

What came in my mind - and this can be damp stupid: "Open HPSDR mx" runs perfectly regarding PS-A on Hermes, ANAN-10 and Co.....
The routine for PS may be very different but may be not the way it is monitored and/or enabled?
See, my point is that PS works to perfection even on our ANAN-10 under Thetis if it works.
The point is that it falls to zero right at the beginning or it doesn't even start.....
Once it works it appears that is stays alive all the time until ptt is released again.

So may be I'm wrong but we realized that the internal decision whether PS will be active or not is always made at the beginning of every single transmission.

May be you find a minute and can monitor and check this on your ANAN and can agree or disagree on that.
If we are correct it is obviously more a problem of starting that specific routine accordingly.

Again Rick - may be all wrong but these are just my/our observations……

Pse do NOT give up on this - all of us have time - we are so happy that somebody out there is looking after our little ill patient.
Any and every job on this is very much appreciated! Thank you.

73,

Andy
Last edited by dl6eat on Sat Mar 14, 2020 1:26 pm, edited 1 time in total.
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
kd4jm
Posts: 22
Joined: Fri Oct 25, 2019 12:31 am

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby kd4jm » Sat Mar 14, 2020 1:12 pm

Hi Rick,

My results are the same as Andy on both pre13 verison.

73,

Jorge
User avatar
Arno Imig
Posts: 11
Joined: Sun Oct 20, 2019 6:53 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby Arno Imig » Sat Mar 14, 2020 3:29 pm

Hello Rick + all
with me, there's no change with P.S.

I just think out loud, so it can all be stupid. ;)

If P.S. works or not is decided in the first sesond after the PTT is pressed.
If P.S. works (SetPk=GetPk) I can talk for one hour without letting go of the PTT and it probably works 100% the whole time without problems.
And exactly the other way round. If PTT is pressed and P.S. does not work (GetPk half of SetPK) the state will not change until I release PTT and press PTT again.
After pressing PTT again the decision if P.S. is running or not is made in the first second. If PTT is pressed the whole time the situation never changes.

So the system have to make a calculation at the beginning of a run (PTT), which is either right or wrong. What could that be? Does it have something to do with the delay calculation? In our setups the delay over the RF-feedback line is much longer than in your setup. We have large filters (LPF) in our PA, so the delay is longer.

And by the way.... even if only a few people write here, there are hundreds of ANAN10 / Hermes users waiting for a solution.
Pse not give up - we all have time - we are so happy that someone out there is taking care of our little sick patient :-)
We are very thankful for any work in this area! Thank you very much !!!

73,DL5AI
Arno
Rig: Hermes Board
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Sun Mar 15, 2020 12:26 pm

Hi all,

Just a quick one:
I played around again with version Pre13 and changed all the timing values in the Linearization menu.
I could not find any setting which improves things.

Looks like all the same - like when using my ANAN-7000.
The only difference the 7000 works stable - the ANAN-10 not …..

All, have a great remaining we,

73,

Andy
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
K1LSB
Posts: 638
Joined: Wed Feb 05, 2020 5:25 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby K1LSB » Mon Mar 16, 2020 1:10 pm

Gents,

I was seeing the same symptoms as you guys are reporting (difficulty getting and keeping a PS "lock"), and I'm getting much better results now so I want to report my observations in the hope that it may benefit some of you.

DISCLAIMER: I'm running Thetis 2.6.9 on a 7000DLE MKII but I'm assuming the PS algorithm is basically the same on all installations.

1) To enable PS to get a strong lock on the feedback I found that I really needed to push the mic level as far towards max (Zero on the mic meter) as I reasonably can. I observed my PS locking much more quickly and strongly on a full-strength mic level. That's the single most critical variable I've observed regarding reliable PS functioning.

2) If PS is unable to obtain a reliable lock initially, it will probably continue to struggle until PS is disengaged then re-engaged. So I disable the "DUP" and "PS-A" buttons then make a short transmission in non-PS mode (with some audio), then re-enable the "DUP" and "PS-A" buttons and make another transmission, and PS invariably successfully engages (if you are running something less than 100% mic audio you will probably want to use the two-tone to engage PS, because PS needs a strong source of multiple frequencies).

Just letting you know that is what works for me. I've only had my radio for a couple of weeks and I struggled for several days trying to get PS to operate reliably. I finally stumbled on the fact that my mic level needed to be really pushing near 100% to make PS happy.

And it's not just a matter of how much the total transmitted power is, it's more a matter of what percentage of the total power is actually mic audio. For example, I've found that PS is generally not as quick to acquire and maintain a lock when I'm transmitting in AM mode even with strong mic audio, because at least 25% of the total transmitted power is just carrier, which doesn't contribute anything as far as giving PS the multiple strong signals it needs to run it's calculations.

Hope that helps,

Mark
User avatar
w-u-2-o
Posts: 5525
Joined: Fri Mar 10, 2017 1:47 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby w-u-2-o » Mon Mar 16, 2020 4:13 pm

Mark,

Good points, but you may be missing a critical bit of info: you need to hit 0dB on the ALC (or, better yet, if you are using Thetis, the ALC Group) meter. PS only kicks off a measurement when it sees 0dB ALC.

Since the ALC algorithm is a look-ahead, soft limiter, not at all like the nasty hard clipper that the old, legacy Flex PowerSDR had, it is completely safe to set levels in your audio chain in order to achieve ALC peaks of +3dB or so (read this on the ALC COMP meter or, in Thetis on the ALC Group meter). Just this tiny bit of over-drive will make your PS performance nearly 100% reliable.

Achieving the requisite audio levels can be done a number of ways. A good starting place are the "tacked" audio setup guides posted here: viewforum.php?f=6

Also, turning DUP off and on accomplishes nothing. I recommend just leaving it on. Turning PS-A off and on will reset the PS algorithm.

73,

Scott
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Mon Mar 16, 2020 4:47 pm

Hi Mark and all others,

Your comments are very much appreciated.
May be I miss understood your note - you mentioned you use an ANAN 7000 DLE?.... different story….
This topic is related to the ANAN-10 / Hermes F/W only…...so my 7000 (I own one as well) runs next to perfect.

But I agree on the point that the percentage of modulation changes the chances to see PS-A running.
The higher the better - but at the end I will not accept to run my station with any kind of over-modulation.
A 2-tone test is mostly good - but this is max power and no SSB.

My modulation has to be perfect - it is my business card….it has to show dynamic and quality.
Since this no issue on my 7000 it should work on the ANAN-10 as well…..at least one day….my hope.

73,

Andy
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
K1LSB
Posts: 638
Joined: Wed Feb 05, 2020 5:25 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby K1LSB » Mon Mar 16, 2020 7:33 pm

Hi Andy,

Yes, I do use a 7000 MKII, I just wanted to post my observations regarding mic drive power and also the point of disabling then re-enabling PS to get it to acquire a lock-on, for whatever those observations might have been worth.

Regards,

Mark
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Mon Mar 16, 2020 8:01 pm

Understood Mark.....

Well, I have never had that on my 7000 but the percentage of modulation is a factor on the ANAN-10 / Hermes.
I went back in my lab and did some analyzis. Even when modulating with a generator at full audio power right from the beginning PS-A fails frequently.

As Arno stated before it looks like the decision is made in the first few milliseconds if PS-A works or not.
I couldn't produce stable operation this way or the other up to now.....

Again, my 7000 works fine today but on a totally different Firmware.
Interesting - on Open HPSDR mx (protocol 1) all works fine on Hermes as well......
I am not a programming guy but there should be at least a difference in enabling the Pure Signal process. If it starts working fine the procedure stays alive.....all fine. (Prot 2)
But if not there is no second trial......it stays off.
"Get Pk" drops and gets frozen at half the value....the green "correcting" and "corrected" identicators (lower right corner) stop flashing.

Once ptt is pushed again - next game begins (its like Russian Roulette) .... not predictable at all.

It would probably help (as an interims solution) to recall the last cal values if "Get Pk" fails and the correcting procedure stops. Similar to Single Cal but Pure Signal would at least be active until the next transmission begins … and so on....

Right now it hops on and off - no serious operation possible.

73,

Andy
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
User avatar
n1gp
Posts: 175
Joined: Sun Apr 09, 2017 6:34 pm

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby n1gp » Tue Mar 17, 2020 6:09 pm

Hi All,

Here's an update. I actually found a bug. I wasn't using the proper muxing bit to switch between
the ADC and DAC.

Though this tests well on my ANAN-10 I still wonder about the signal delay you guys have when going through
the LPF stages of your amps. I cannot simulate that here. It would be interesting to see if you bypass the amp
what happens. But hey maybe this will work this time :shock:

https://www.dropbox.com/s/wpeor3gnvj7j4jx/Hermes_Protocol_2_v10.8_pre14.rbf?dl=0

-Rick / N1GP
User avatar
dl6eat
Posts: 45
Joined: Wed Feb 14, 2018 1:31 pm
Location: Ellerhoop - JO43VR
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby dl6eat » Tue Mar 17, 2020 8:27 pm

Thanks Rick.......I will test it first thing tomorrow morning and will report my results.....
Timing doesn't matter....I tried everything but will certainly do it again.
Also changed them on my 7000 - no changes at all. The 7000 always works (PS-A) as far as I could test it.

Typical delay values are between 0 and 200.

Thanks so much!

73,

Andy
No knobs, Pure Signal and Thetis (open source)
ANAN 7000, ANAN 10, Intel NUC 8i7 32 GB + 16 GB Optane
N8RWS
Posts: 13
Joined: Sat Feb 24, 2018 9:32 pm
Location: Otsego
Contact:

Re: New -10/-100 16-BIT Hermes Protocol 2 Firmware Topic

Postby N8RWS » Wed Mar 18, 2020 12:08 am

Hi,

Many thanks for the work on this.

Just tried pre14 with my ANAN 100 using crosstalk for PS

PS will lock in and correct but as I am transmitting it will stop correcting, then start up again. Feedback is blinking the entire time. Also receiving SEQ errors !

Thanks

Jeff

Return to “Protocol 2 Firmware (all radios)”