Lost watts

[ea3ifh]

Hi,

Yesterday I've dug up my truSDX to test my recently installed OCFD antenna. I played with FT8 (audio cables) and didnt worked well - I guess because frequency drifting and lots of USB CAT disconnections / weird characters on screen / reboots... during TX.
TX was OK and low SWR. Feeded with 13.8V.

The fact is that, later I upgraded to 2x beta firmware (I wanted to try the "usb audio" thing) and I always like to try new firmwares and features.

So, later while trying to adjust frequency drifting.... I realized that suddently it has lost its power. Despite being powered by 13.8V or USB, doesn't matter, the power doesn't go up more than 0.5W. In fact my AT-200Pro Autotunner even doesn't realize that it's being fed with signal (so It means signal is too low).
Yesterday during FT8 testing - Autotuner was doing its job - so it was fed at least with 5W.
Have to say that FT8 testing was very short time so I don't thing (or I hope) that PA is burnt... It has no reason for it.

I've downgraded to 2.00i, same result.

Maybe during FT8 test transmission, something screwed up? Physically? How I could check it?

Thanks!!

[ea3ifh]

Hi,

The proof... https://imgur.com/a/TxBTDTD

Transmitting USB - 0.3W (13.8V input)


CW Transmitting - 0.0W (13.8V)

[dl6sez]

Hello OM,

i think you have blown your FETs....
Tuning antenna with an automatic tuner can kill a class E PA more easily than conventional class AB PA.
So it would be better to use a hand tuner and get around the tuning point first due to tuning without rf to highest rx S-meter levels. Than tune the rest fast with tx in CW dits.
To avoid this one can buy/build a resistive bridge tuning circuit like used in https://qrpguys.com/multi-tuner. There your PA FETs never see more than 1:2 SWR during tuning

73 de Chris

[ea3ifh]

Hello OM,

i think you have blown your FETs....
Tuning antenna with an automatic tuner can kill a class E PA more easily than conventional class AB PA.
So it would be better to use a hand tuner and get around the tuning point first due to tuning without rf to highest rx S-meter levels. Than tune the rest fast with tx in CW dits.
To avoid this one can buy/build a resistive bridge tuning circuit like used in https://qrpguys.com/multi-tuner. There your PA FETs never see more than 1:2 SWR during tuning

73 de Chris

Hi Chris,

That's interesting. Always learning (by trial error & screwing FETs)

I opened it and I see three BS170. Will try to change them. And lesson learnt.

Thanks,

[dl6sez]

Hello OM,

be careful not to buy chinese fake BS170s, many bad FETs are sold today.
I would suggest to get BS170 original OnSemi.
The problem with these FETs is, they are normally used to switch on/off simple things not rf as it us used in (tr)uSDX. So be careful buying real ones from well known suppliers, Mouser, Reichelt, Farnell or Digikey.
I checked it today Digikey offers original OnSemi.

73 de Chris

[ea3ifh]

Hello OM,

be careful not to buy chinese fake BS170s, many bad FETs are sold today.
I would suggest to get BS170 original OnSemi.
The problem with these FETs is, they are normally used to switch on/off simple things not rf as it us used in (tr)uSDX. So be careful buying real ones from well known suppliers, Mouser, Reichelt, Farnell or Digikey.
I checked it today Digikey offers original OnSemi.

73 de Chris

Hello,

Thanks, will take it into consideration. The problem is that these big suppliers (mouser, digikey...) usually you buy cheap components (say 2-3€) and shipment is extremely abusive (20-30€).

Also I'm recommended to put a Zener between to protect them in the future.


Until I replace these BS170 I have a nice (and expensive) SWL receiver


Thank you!!

73 de EA3IFH

[dl6sez]

The Zener will degrade output power and you have to retune your class E network,
also the efficiency goes down and can produce intermodulation!
Not sure if it really helps because it is not often an overvoltage which kills the FETs it is more often a high current due to reflected power.

73 de Chris

[ea3ifh]

Thanks Chris,

Always good to know it. Yes, next time I will be more careful. In fact I use tr(usdx) usually with a 20m monoband endfed, so no tuner is needed. Just this time wanted to test my "fixed" OCD dipole and... crash!

I'm used to my Icom 706MKIIG that is rock solid.

Best!

73 de EA3IFH

[QuantumRadio]

EA3IFH,
Good morning. I measured the Vp-p out of my (tr)uSDX into a dummy load using my PicoScope. When I clicked the Tune button in WSJT-X on FT8 mode I got 40.75Vp-p and 13.92Vrms.

I then computed the power output and got (13.92Vrms)ˆ2/50ohms = 3.88W. This matches up with the 3.87W that the (tr)uSDX displayed. My battery was feeding 12.2V to the (tr)uSDX.

Hopefully this provides a reference & best of luck in your hardware troubleshooting.

73,
Stephen

Test setup into dummy load

test_setup.jpg (144.52 KiB) Viewed 365 times

Vrms to Power

vrms_to_power.jpg (21.34 KiB) Viewed 365 times

Close-up of (tr)uSDX screen

close_up_of_screen.jpeg (41 KiB) Viewed 365 times

[QuantumRadio]

This is the scope output while WSJT-X was tuning.

PicoScope screen shot

FT8_truSDX_tune.jpeg (138.57 KiB) Viewed 362 times

[ea3ifh]

Thanks QuantumRadio!

Will investigate furher.

Best regards