G2 power supply - TAKE CARE!

[tony77]

Somewhere I thought I'd noted that the specification for DC power supply for the G2 was for 13.8v plus or minus 5 (five!) %, not the more usual 15%. Well, while transmitting a couple of days ago there was a power fault at my QTH that briefly dropped the DC voltage indicated on the shack power supply to 12.9v (I completely shut down the station within a few seconds, as soon as I noticed this).

This is a drop in the DC supply voltage of about 9.5%, for a few seconds while transmitting on 28 mHz. The result now seems to be that something in the transmit chain has been fried - the G2 is producing 28 mHz output regardless of what transmit VFO frequency is selected.

I haven't been able yet to access a suitable test rig to get hard data on this, but the external solid-state linear (that has automatic band switching) switches to 28 mHz regardless of the G2 VFO setting, and using another radio I have verified that the external linear is working perfectly.

Actual power output from the G2 is unaffected (monitored either side of the external linear with Wavenode sensors). A two-tone signal with the PureSignal AmpView window open produces a plot like I have never seen before - on any band other than 28 mHz there are now two separate, and slightly diverging amp output lines in the plot.

Receive is completely normal following a cold boot, BUT after transmitting even momentarily, on returning to RX the receive noise levels go 20-30 dBm higher.

Obviously a scope will reveal all, but in the mean time - BE VERY CAREFUL about your power supply. It appears the G2 is completely intolerant of a 10% supply voltage drop during transmit. Hopefully I'm not now the owner of a very expensive black doorstop...

UPDATE: I connected the G2 to a dummy load via a power meter, similarly connected my Xiegu X6100 to a dummy load, then put these two dummy loads right beside each other. With this setup I now know that the G2 transmit frequency is locked on 28.379 mHz. The TX frequency does not respond to any change whatsoever in the Thetis VFO setting, not even within the 10 meter band itself. Conclusion: at the very least, whatever G2 hardware is involved in setting the TX frequency is not functioning. Tx'ing a voice SSB signal on 28.379 produces output, but readability = zero.

Any constructive comments/info would be much appreciated, I'm not looking forward to the prospect of shipping a 3-month-old G2 back to India.

Tony, ZL3ES

[w-u-2-o]

It's not entirely clear from your post, but:

- Did the Pi crash during this power supply event?

or

- When you saw the voltage had dropped and you "completely shut down the station within a few seconds" did you follow the proper shutdown procedure so that the Pi was not just crashed off?

Because in either case it's possible the Pi file system is corrupted. That is a far more likely scenario than having any hardware actually be damaged.

[Trucker]

Tony,
Have you tried running the G2 from PiHPSDR in the radio? If it works properly from PiHPSDR, then it may be possible that P2APP has become corrupted in some way. Or,you need to reset your database in Thetis.
If you are using the G2 without the front panel, hook a keyboard, mouse and monitor up and see if it boots into PiHPSDR. Then close P2APP and start PiHPSDR. That should give you some indication where the problem lies. If it won't bootup to Raspbian Linux and/or P2APP, then it is possible your SD card's Raspbian Linux image has been corrupted in some way as Scott mentioned
James
WD5GWY

[tony77]

It's not entirely clear from your post, but:

- Did the Pi crash during this power supply event?

or

- When you saw the voltage had dropped and you "completely shut down the station within a few seconds" did you follow the proper shutdown procedure so that the Pi was not just crashed off?

Because in either case it's possible the Pi file system is corrupted. That is a far more likely scenario than having any hardware actually be damaged.

No, the Pi did not crash, and the G2 seemed to continue to function. The metered DC supply voltage is pretty much in my face at all times, and when this dropped I went QRT and closed down immediately to investigate. I'm almost paranoically careful about shutting down the Pi - when I start the G2 from cold, I give it about 10 seconds for the Pi to boot fully and get its system services running, then I open an SSH shell logging in as the "pi" root user. Then when it comes to shutdown I first execute a "sudo shutdown now" command in the shell, verify the G2 power switch has changed color to white, leave it 10 seconds, then power off completely.

BTW: the G2 power switch is the rig's worst feature by far in my book, with a truly horrible arrangement where you sort of half-lightly press it to invoke the Pi shutdown, then wait till it goes white to then give the power switch a full press for power off. WAY better to shell in to the Pi and invoke a proper Linux shutdown.

A while ago I made a backup of the Pi SD card, so I will try replacing the original with this copy and see if it makes any difference.
I should re-emphasise that the G2 + Thetis still runs perfectly in RX mode, as long as TX is not invoked on any band other than 10 meters. So it does really seem like there's something amiss with the TX band switching, and possibly with the Saturn TX pre-driver.

I emailed AL support. No response as yet (surprise!) but hey I'm a perennial optimist

Am already planning a station rebuild around using the G2 as purely a RX SDR, with some other radio for TX.

-Tony ZL3ES, ZL3LAL

[tony77]

Tony,
Have you tried running the G2 from PiHPSDR in the radio? If it works properly from PiHPSDR, then it may be possible that P2APP has become corrupted in some way. Or,you need to reset your database in Thetis.
If you are using the G2 without the front panel, hook a keyboard, mouse and monitor up and see if it boots into PiHPSDR. Then close P2APP and start PiHPSDR. That should give you some indication where the problem lies. If it won't bootup to Raspbian Linux and/or P2APP, then it is possible your SD card's Raspbian Linux image has been corrupted in some way as Scott mentioned
James
WD5GWY

Yep my G2 is the "blackface" version with no front controls. I haven't used PiHPSDR at all since I received the G2, but I'll give this a go. As I mentioned in my reply to Scott's post, one of the first things I did was to make a copy of the Pi SD card, and I regularly back up my Thetis database also to M-Disc DVD-R.

At least the G2 will make a nice SDR receive system, and I've already re-engineered and tested the power supply around it so that it is powered off a large AGM deep-cycle battery combined with an always-on float charger and a DC boost converter that produces a clean, steady 13.8v.

UPDATE: Email received from Doug. At his suggestion a Thetis database reset done. I've run out of time today to fully test things, but it looks like the database reset has fixed the problem. Am very happy indeed!

So over the few years that I've owned a 7000DLE-II and now the G2, this now makes it four times that I've emailed Support at AL with a <real> problem, and every time they have responded as quickly as could be expected given the time zones...and every time, their support has seen the issue fixed. Can't ask for more than that. Thanks everyone.

-Tony ZL3ES, ZL3LAL

[w-u-2-o]

So the problem had nothing to do with the G2 itself. It's seems simply a coincidence. It was difficult to imagine how running at 12.9V could cause any problems.

[Trucker]

Glad you got it sorted.
James
WD5GWY