Stable Candidate 20201212_72p8 gateware released

[Steve Haynal]

Hi Group,

Please see the 20201212_72p8 stable candidate release. It will be come the next stable release in a few days if there are no problems found. It includes all supported variants since it is a stable candidate release.

Makerfabs has sold out and is preparing for the next build of 100 Hermes-Lites. Over 300 units have been sold just this year. I would like to make this release the factory image installed in this next build. I'd appreciate any feedback over the next couple of days.

73,

Steve

kf7o

[Wolfgang DL9UFB]

Hallo Steve,

i have tested on my HL2beta2.
The transmit function not working, also
PTTout for PA.

73, Wolfgang

[Steve Haynal]

Hi Wolfgang,

Someone else said that the previous hl2b2 gateware did not work properly, so I made sure this release did. I just had a FT8 QSO with my hl2b2 and the 20201212_72p8 gateware. Did you install the specific files for the hl2b2 found here:

https://github.com/softerhardware/Hermes-Lite2/tree/master/gateware/bitfiles/testing/20201212_72p8/variants/hl2b2_main

73,

Steve

kf7o

[Wolfgang DL9UFB]

Hello Steve,

this is the version I tested,
RX ok, TX without success (PowerSDR_OpenHPSDR 3.4.9).
Currently I am using v3_dhcp_191119_PureSignal_beta2.jic
as the last "working" version.

73, Wolfgang

[ron.ni...@gmail.com]

Hi Steve,

No problems with installing 20201212_72p8 on my HL2v9(2nd batch) using hermeslite.py from the Raspberry Pi command-line.  72.8 seems to run just fine Rx and Tx.  And unplugging and replugging the ethernet cable no longer requires a power cycle for its IP address to work.

But, in hermeslite.py, , per the hex string example in the jupyter notebook, I tried:

hl.set_eeprom_mac("13:de")

And got:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "hermeslite.py", line 466, in set_eeprom_mac
    b1 = int(mac[1])
ValueError: invalid literal for int() with base 10: 'de'

Thanks and 73,

Ron

n6ywu

[Steve Haynal]

Hi Wolfgang,

I don't recognize that gateware file. Have you modified your Hermes-Lite 2.0 beta 2 in any way? Is "Hermes-Lite 2.0 beta2" printed on the PCB? I suggest you:

* Undo any mods you may have made.

* Make sure your hl2b2 has the standard mods as described here: https://github.com/softerhardware/Hermes-Lite2/wiki/Build-Notes-2.0beta2#issues-and-required-modifications-6

I have tested TX on my hl2b2 and it is working. How exactly are you testing TX?

73,

Steve

kf7o

[Steve Haynal]

Hi Ron,

I just updated github with a fix.

73,

Steve

kf7o

[Steve Haynal]

Hi Group,

I have made this release the stable release. Makerfabs will use it in the next production run. Also, the hl2factorytest repository is updated to this version:

https://github.com/softerhardware/hl2factorytest

This means you can update a Raspberry Pi factory test image to the latest gateware just by double clicking on the Update desktop icon. This is how Makerfabs pulls the latest gateware.

73,

Steve

kf7o

[Wolfgang DL9UFB]

Hello Steve,

first, sorry for the confusion with the
Gateware, the file was from Taka San.
I have only one modification in the receiver to
improve the decoupling of the sampler signal
in PureSignal (Predistortion).
I have tested many firmware files.
The last working gateware for HL2b2
is the version 20200119_69p0 release on the
the Softerhardware /Hermes-Lite2 Github
as well  the older version hl2b2_20190616,
which works very well with PureSignal and PA control.
All newer versions do not work.
I found the reason for this.
The Pin42 FPGA TP1 enable IC U2 +VOP and
enable IC8 +VPA  (+2,5V TP1) is not activated.
Also not enabled is Pin28 FPGA (TR) for external
PTT (+3V) Pin1 DB18.
I hope this helps.
I don't understand why your HL2b2 works with this gateware.
Was there a modification in the power supply section?

73, Wolfgang

Translated with www.DeepL.com/Translator (free version)

[Steve Haynal]

Hi Wolfgang,

There were not power supply modifications. These pins and functions have been stable for quite some time.

What software are you using? Do you have the PA enabled? Do you have "Disable T/R switch in low power mode" set to True or False? You could see the behavior your report if these have changed values due to a different MAC in the more recent gateware. For example, your PowerSDR settings are saved based on the MAC and might be cleared if the HL2 reports a different MAC address.

73,

Steve

kf7o

[Wolfgang DL9UFB]

Hello Steve,

I have been using the HL2b2 board since the beginning.
It always has the same MAC-Address .
I have tested with OpenHPSDR 3.4.7 and 3.4.9, the Hermeslite Mod 3.5.0_Beta10,
Spark-SDR 2.0, SDR-Console and the Kiss console from OpenHPSDR group.
In all cases the same result, no TX.
None of the newer versions work, the
older versions work with all five programs.
Is maybe by mistake a wrong version for
HL2b2 on the Github?
Merry Christmas anyway!

73, Wolfgang

Translated with www.DeepL.com/Translator (free version)

[Steve Haynal]

Hi Wolfgang and Group,

We can verify the exact gateware you used with MD5SUM. On my Linux machine:

shaynal@sonata:~/shaynal/Hermes-Lite2/gateware/bitfiles/stable/20201212_72p8/variants/hl2b2_main$ md5sum *

671bea99962a9b4ea4264098435cc952  hl2b2_main.jic

f10b813c0e1312ed65a88da0f4447969  hl2b2_main.rbf

On Windows you can compute a MD5SUM by following:

https://onthefencedevelopment.com/2017/08/15/windows-10-builtin-md5-checksum-calculator/

It would be helpful if you can share what your PA (enabled/disabled) and TR switch settings are during your tests. There are directly related to the problems you see. Does your HL2 work in low power mode?

Are there any other reports from HL2beta2 users? I think only about a dozen HL2beta2 were every built.

73,

Steve

kf7o

 

[Wolfgang DL9UFB]

Hello Steve,

the MD5 checksum is identical.
The transistors of the "power amplifier" Q3,Q4.
are not populated in my case.
I use the output RF1 (Low Power TX) followed by a
BLF188 LDMOS amplifier, so must be
the operating voltage +VOP in case of transmission
but it doesn't, because the Pin26
still has low potential like in receive mode.
During the tests I always used the same operating software
and only changed the gateware,
of course restarted HL2.
With the old gateware everything is ok, with the new gateware
no enable of the required power supply
and output TR switch.

73, Wolfgang

Translated with www.DeepL.com/Translator (free version)

[Steve Haynal]

Hi Wolfgang,

I did not know that Q3,Q4 are not populated for you. Do you have the PA temperature sensor Q6 installed? Are you reading valid temperatures? If the temperature sensor is not installed or providing invalid readings, then it may be that your HL2 is not transmitting because the gateware thinks the temperature is above 55C. A recent feature of the gateware is to disable TX if the temperature is high.

Also, since this sounds like a custom build, do you have 10K R117 installed? This is the TX inhibit input. It must be pulled high to *not* disable TX. This feature may have been enabled in hl2b2 gateware recently.

73,

Steve

kf7o

[Wolfgang DL9UFB]

Hi Steve,

your last posting brings us closer to solving the problem.
Of course, because of the missing final stage, I did not assemble
IC U13 and U18 including the peripheral components.
A quick test on the I2C bus +3,3V at SCL U13 caused TR to work.
TX inhibit is ok.
I will install U13 with the pullup and
pulldown resistors and hope to simulate
a cool PA for the gateware .
The necessary components should still be in the tinker box.

What do you think about?

73, Wolfgang


Translated with www.DeepL.com/Translator (free version)

[Steve Haynal]

Hi Wolfgang,

That is good to know. Another option is to try and make the gateware detect if U13 is present. The resistors R106 and R107 should still be installed. I can try to take a look at this for the next stable release.

Installing U13 with the AIN pins tied to an appropriate constant voltage should also work.

73,

Steve

kf7o