Gateware 20200420_71p0 Released

[Steve Haynal]

Hi Group,

Gateware 20200420_70p1 has been released to testing. Please see the link for release notes. I appreciate any testing and feedback. The 20200329_70p0 release has been moved to stable.

 

73,

Steve

kf7o

[VU3NXI Siddhalingesh Basawanal]

Hi Steve,

Just tried installing the firmware 7.1 and noticed that my Winkeyer from front jack is not functioning. I did also try with 6.9p3 and it had same behavior as well.

So got to rollback to 6.8 and all started working as intended. Not sure if I am missing something here.

73,

Sid

vu3nxi

[Steve Haynal]

Hi Sid,

Thanks for the report. How exactly is your Winkeyer not functoning with the newer firmware? Can you please provide more details? This will help with debug.

73,

Steve

kf7o

[VU3NXI Siddhalingesh Basawanal]

Hi Steve,

I have a very simple and straight forward setup. I use PowerSDR mRX software and a YCCC SO2R box which has a built-in Winkeyer. For keying I use N1MM software with virtual com-ports configured between SO2R and N1MM. The Winkeyer output of the SO2R box is injected directly to the front jack of HL2. It works fabulously well without any relay chattering with 6.8 firmware. The only downside I can see is the sidetone. 

With 7.1 /6.9p3 firmware's, while keying, the PowerSDR is put in TX (MOX is ON) but I see that there is no keying/output from HL2.

73,

Sid

vu3nxi

[Joe LB1HI]

Hi,

No 6 receivers version  ?  (Hermes-Lite 2.0 build5 and later, 6 RX, no TX)

[Matthew]

Hi Sid,

I have just tried PowerSDR and the HL2 with this gateware and a key connected. It works fine for me. Do you have a paddle you can connect directly to the HL2? Does the dit or dah key the HL2 and produce RF? I suspect this could be a 3.5mm jack, tip/ring/sleeve issue, or perhaps a setting in PowerSDR.

Does 69p1 work?

73 Matthew M5EVT.

[VU3NXI Siddhalingesh Basawanal]

Hi Matthew,

I just worked few stations (LZ1958GU, LX5WARD, HG20ME) with 6.8 firmware, running at 26 wpm and all went well.

I have tried with 69p3 and 70p1 with no luck. I'll try now with 69p1 and report back.

73,

Sid

vu3nxi

[Joe LB1HI]

Hi,

observations after v. 71p0 installation:

 

Quisk - RF slider in Quisk no reaction.
There are no LNA gain changes during TX.
Does not remember slider settings during TX. (changing the position of the slider with TX also causes a change in time RX)

Spark SDR RF gain during TX works. will remember the settings separately for RX and separately for TX.

 

PowerSDR OpenHPSDR  - AGC Gain no reaction. But S-ATT works the same as before. Eeverything indicates that it plays double role of RF gain and step attenuator. PowerSDR remembers the separate settings for TX and for RX.

 

In SDR Console V3 it is difficult to say because the slider operation is very strange already during RX. . Slider “Visual” behaves like RF gain. Causes a change in audibility levels.
However, the “LNA” slider does not react. it does not allow to change the signal level
It changes a bit but in exactly the opposite position.  Because in the minimum setting of -12dB it shows a bit gain increase.

 
Increasing the LNA "level" of the slider only causes ADC overload indication at some point, but moving slider does not correspond to a change in the received signal / LNA audible gain.

73, Joe, lb1hi

On Tuesday, April 21, 2020 at 6:48:47 AM UTC+2, Steve Haynal wrote:

[VU3NXI Siddhalingesh Basawanal]

Matthew,

Thank you for the 69p1 archive link. I could find all other firmware's under this archive folder.

Yes. I now confirm that 69p1 and 69p2 works as well.

I tried 69p3 and it did not work. Seems like the issue starts from here.

So, I have now reverted to 69p2 and all is well.

73,

Sid

vu3nxi

[Matthew]

That is good that you are progress in finding the issue. It will always be good to stay up to date with the latest gateware, so it would be good to find the root cause. Do you have a paddle to connect to the HL2 to test what happens with a dit and dah sent from the paddle?

73 Matthew M5EVT.

[VU3NXI Siddhalingesh Basawanal]

You are absolutely correct Matthew! It's always good to be with the latest gateware. I did simulate it without a paddle, by connecting a 3.5mm Stereo EP cable to HL2 front jack.

Very strange thing. Please see the below results while the PowerSDR is in OFF state.

With Gateware 68, 69p1/p2:

With PowerSDR switched ON
Tip to ground shorted - I hear the relay clicking sound and do get the RF power output.
The ring and ground - enables the PTT

With PowerSDR Switched OFF / Completely closed
Tip to ground - I hear the relay clicking sound and do get the RF power output (About 1 watt reduced output)
The ring and ground - No effect

-------------------------------------------------------------

With Gateware 69p3/p4/71:

With PowerSDR switched ON
Tip to ground shorted - I hear the relay clicking sound and do get the output, working well as intended.
The ring and ground - enables the PTT

With PowerSDR Switched OFF / Completely closed
Tip to ground - No effect
The ring and ground - No effect

Do you also observe this? Any guess what's going on?

Note: When I Tx with 68/69p1/p2 gateware, there is no relay chattering sound at all. It is kind of an AFSK mode and very smooth to operate.

73,

-de Sid VU3NXI

[Matthew]

Sticking with the 3.5 mm stereo test for now, with gateware 69p3/p4/71:

1) in PowerSDR settings, DSP-CW do you have Semi Break-In ticked? If so, set the delay to 350 ms.

2) With PowerSDR on and connected to the HL2, short tip to ground. Does the relay chatter disappear?

73 Matthew M5EVT.

[VU3NXI Siddhalingesh Basawanal]

Sorry, I forgot to mention:

1. With 68, 69p1/p2, I am able to key the HL2 from my YCCC SO2R Winkeyer (through N1mm) without any issues. 

2. However, with 69p3/p4/71, though I am able to trigger the HL2 manually through 3.5 mm stereo cable, my YCCC SO2R Winkeyer (through N1mm) is unable to key the radio at all, no output.

I have tried restarting HL2 post updating the gateware with no luck. Now I have reverted the gateware to 69p2 and enjoying CW. 

May be it is worth to compare and see if what is changed between 69p2 and 69p3? Just a guess..

73,

Sid

vu3nxi

[Steve Haynal]

Hi Joe,

Are you saying the LNA behaves differently in this release for you? I just tried and it behaved the same for me. Can you try power cycling your HL2 to see if that makes a difference? The LNA code was not changed, maybe timing on the external pins changed and affects some people.

Also, I did not release a 6 RX version as there is still more I want to do with the version 71. There will be a 6 RX version when it is done. For now you can use an older 6 RX version.

73,

Steve

kf7o

[Steve Haynal]

Hi Sid,

I've been following this thread and will look into a fix this weekend, but am not sure I understand the failure still. It appears you have verified that CW is keyed and RF output produced when you ground the tip of the 3.5mm cable with version 71. (When software is not running, you should not see output now, which may have changed recently.) But your Winkeyer no longer works. How exactly is your Winkeyer connected? Does it also key by grounding the tip of CN4, or by some other mechanism? Please describe.

73,

Steve

kf7o

[VU3NXI Siddhalingesh Basawanal]

Hi Steve,

I use YCCC SO2R box with a built-in Winkeyer, the output is directly connected to CN4.

I measured the Winkeyer output, it has some resistance while keying. 

Here is a small video clip of the same.

73,

-de Sid vu3nxi

[Matthew]

Sid,

Looking at this:

http://www.k1xm.org/SO2R/Downloads/SO2R_Box_Schematic.pdf

which connector/signals on this schematic are you connecting to the HL2?

Can you try making a cable that does not connect the ring from the yccc so2r box to the HL2?

73 Matthew M5EVT

[Steve Haynal]

Hi Sid,

I think Matthew is on to something about the ring. It is an undefined condition if both the the ring and tip are grounded during operation. Which should have precedence, CW or PTT? There are some changes in the gateware from 69p2 to 69p3 which might start favoring PTT. I will favor CW in the next gateware release as I suspect that is the more common use, but I still recommend only grounding the tip for CW or the ring for PTT and never both in normal operation. If the tip and ring are grounded and CW is favored, software could still receive a message that PTT was engaged, and that could lead to more unexpected  behavior.

73,

Steve

kf7o

[VU3NXI Siddhalingesh Basawanal]

Hi Steve,

Thanks for your reply.

I have connected from schematic, U3 pin 10 to HL2.

Today, I made a separate cable that does not connect/use the ring from the yccc so2r box to the HL2, only the tip and ground is used.

However, I still observe the same condition with 69p3/7.1.

73,

-de Sid vu3nxi

[VU3NXI Siddhalingesh Basawanal]

Just tried 7.1p1 with same behaviour. No change.

73,

-de Sid vu3nxi

[Steve Haynal]

Hi Sid,

This is quite baffling. As I understand, you are able to trigger CW manually by grounding the tip but the keyer is no longer working. Is that a correct understanding? Can you think of anything in you software setup which may be sending a PTT signal to the HL2 before the keyer grounds the tip? Does the signal from keyer look clean on a scope when connected to the HL2?

73,

Steve

kf7o

[VU3NXI Siddhalingesh Basawanal]

Hi Steve,

You nailed it !! 

What a guess! N1MM was sending PTT command. See the attachment. Unchecking those boxes worked like a charm.

Now I have uploaded 7.1p1 gateware and sending the PTT command as well from SO2R software to the HL2, CN4 ring terminal and all is working as intended.

I'll do some extensive testing tomorrow and will report back.

Thanks again to you and Matthew for the wonderful help!

73,

-de Sid VU3NXI

[Steve Haynal]

Hi Sid,

Glad it is working!

73,

Steve

kf7o

[neil whiting]

Hi Steve,

I am a little confused by this.

I would expect to ground the PTT (ring) first, then send CW by grounding the tip. When finished sending, unground the PTT.

If the HL2 is driving high power amplifiers or transverters this gives time for a sequencer to get the relays in the right state

before the RF comes along.

If the PTT is not used in CW mode then this sounds like semi-break-in mode. In this situation the relays will all be

clicking over during keying of the transmission and very likely something will get fried. It may be possible to design the

software and sequencer so that nothing untoward happens but it is placing quite stringent demands on them.

Am I missing something?

73,  Neil  G4BRK

[Steve Haynal]

Hi Neil,

This sequencing is done by the gateware. By setting the TX buffer latency, you set how early the relays are engaged before CW signal starts. Also there is hang time that is set to for transition from no breakin to semi-breakin. The gateware needs to manage this if one is to connect just an external simple key.

73,

Steve

kf7o

[James Ahlstrom]

Hello Neil,

This is from the protocol document, and may be helpful:

The Hermes-Lite2 has a 3.5mm phone jack at CN4. When the ring is grounded, C0[0] is sent as one. Software should interpret this as an external PTT switch. There is no internal keyer, so C0[1] is always sent as zero. When the tip is grounded, both C0[0] and C0[2] are sent as one. C0[2] follows the signal at the tip, and a shaped CW signal is generated by the Hermes-Lite2. C0[0] stays set during the entire time TX is on for CW, which includes any specified CW hang time. These bits are never set from software events.

So, when the tip (CW key) is grounded, the HL2 sends the PTT to the software and holds it for the CW hang time. It is up to the SDR software to interpret all this.

Jim

N2ADR

[neil whiting]

Hi Steve and James,

Thanks for the replies and further enlightenment.

I've been updating things so I can try the latest CW improvements. This lead to a big struggle to get the latest Quisk to work on

Windows 10 which is still not resolved, but no problem on Linux so I used that for tests.

I just loaded the latest gateware and see an improvement, the PTT derived from sending CW via the HL2 3.5mm jack tip

now seems to do semi break-in as expected instead of going to TX and never coming out again. Changing the hang time

seems to provide enough range for me to work with, though it will take some time to go back to RX when the gateware and PA

are both doing relay delays as part of the sequencing. I should be able to live with this once I have retrained my brain not to

hit the footswitch on CW :-)

(I hate semi break-in and also VOX - give me a footswitch any time. Properly engineered full break-in is a different story,

I love it but not really appropriate for VHF/microwave).

More experiments tomorrow and I will try SparkSDR too.

73,  Neil  G4BRK

[Max]

Hi all

I now use SDRC with keyer connected to HL2 jack tip. It works brilliantly with semi break-in as expected, hang time now set in the program.

So one last question? Why is it than in CW mode, the ring cannot be used as a manual PTT for CW operation, for a foot switch for example, as an alternative to semi break-in?

What I find at the moment is that if I ground the ring (in both Spark and SDRC), HL2 seems to go into TX mode, but no power out is then generated when keying the tip. Is this as intended? if so, why? It's not how I would expect it to be. Why does grounding the ring prevent CW keying the power out?

Also in SDRC there is a manual TX button, but in CW mode if I turn TX on using this, while in CW mode, when I key HL2 jack tip, no power out. I think I misunderstand reasoning behind the gateware logic regarding CW? it seems that gateware is programmed so that grounding the key jack tip (and this action ONLY, and in isolation) is the only action that will result in CW TX via the jack? Is that right? And why? Seems to remove the otherwise useful alternative of a manual TX switch, either by manual switching in the SDR program or by grounding ring via a manual switch?

73

Max

[Steve Haynal]

Hi Max,

The current gateware sequences PTT and CW separately, so if either PTT or CW occurs first, it will stay in that mode until TX ends. What I think you are asking for is that CW is always highest priority, so if the tip is grounded, everything else is dropped and we go into the CW sequence. This is a small change. I'm not sure how the change will interact with all the software out there. I can make this change as an experiment in the next testing gateware release and we can see how it works out.

73,

Steve

kf7o

[Max]

Hi Steve

I'm just looking for the ability to (optionally) manually hold the HL2 in TX mode and bypass the hang time completely. So yes, I think what you have described would achieve that? So manual TX sequence is as follows:

(1) Ring is grounded by (say) a foot-switch which mutes the receiver and switches HL2 into TX "standby".

(2) Any subsequent grounding of the tip transmits CW characters, and when tip is then left open by key-up, the HL2 is held in TX with the receiver muted until the footswitch is released and the receiver un-mutes.

(3) None of the above interferes with the current semi break-in operation if the tip is grounded in isolation, with no ring grounding. 

I think there are others who are interested in a manual CW mode to complement the existing semi break-in?

Truth be told 95% of the time I would use the usual semi break-in with adjustable hang-time, but last night I worked a very slow speed station. I therefore had hang time adjusted to maximum in SDRC (1000 ms) but it was at a speed that the slightly longer pause between words was just enough to keep dropping the HL2 back into receive exactly at the moment I was keying down for the start of the next word. it was very distracting, so it would have been nice to just use a totally manual TX mode to hold the HL2 in TX with receiver muted for the whole over.

Hope that explains it clearly?

73

Max

73

Max

[Pete Ferrand]