¿ªÔÆÌåÓý

sBitx No Power Out, blown Finals?


 

Hello,?
I think the transmitting transistors of my sBitx V3 (approx. 1 month old, delivered fully assembled) are defective,
there is no more transmitting power. When I measure with a qrp wattmeter, I can see the pointer on the meter twitching. The initial situation:?
?
I made FT8 with approx. 20w, 1-2 QSOs in 10 minutes, SWR under 1:1.2, from now on the transmission power was gone, the display shows an SWR of 1.0 and 0 watts.?
?
The first thing I did was to take a new SD card and set up the system again.?

I opened the device and there is no visible damage, to check the mosfets, I wanted to loosen both screws and found that the screws were loose and the mosfets were
not in contact with the heat sink (could be easily recognized by the gap in the thermal paste, about 1-2mm distance) The mosfets look visually ok, nothing burnt or apparently too hot.?
?
i didn't change the language either (i had already made that mistake on the first day :) )
?
How would you proceed? Write to Ashhar Farhan?, replace the two replacement mosfets yourself (btw, lead or lead-free solder?), check other components first?
?
Has anyone else had a similar problem or have any ideas??
?
Many thanks and regards Andy DO6DOC
?


 

Hi Andy,
?
You should check with HFSignals as there is a warranty for fully assembled kits. ?At a minimum I would expect a pair of replacement MOSFETs. ?

I would remove the two IRF510s and check with a transistor tester or using a DMM to verify they are good. ?You can search the web for how to test a MOSFET with a DMM.?

Remember that you must readjust the bias if you replace the IRF510s. ?I also suggest you replace both even if only one is bad. Try to use two from the same manufacturer batch.?

73
Evan
AC9TU


 

I would definitely contact HFSignals. My V3, bought complete at Dayton this year, failed on the second day of use with no power out. They shipped me a replacement board to swap in and I sent the failed one back. Worth a shot.

On Mon, Nov 4, 2024 at 11:55?AM Evan Hand via <elhandjr=[email protected]> wrote:
Hi Andy,
?
You should check with HFSignals as there is a warranty for fully assembled kits.? At a minimum I would expect a pair of replacement MOSFETs. ?

I would remove the two IRF510s and check with a transistor tester or using a DMM to verify they are good.? You can search the web for how to test a MOSFET with a DMM.?

Remember that you must readjust the bias if you replace the IRF510s.? I also suggest you replace both even if only one is bad. Try to use two from the same manufacturer batch.?

73
Evan
AC9TU


 

¿ªÔÆÌåÓý

When you finally put it back together again, you really might want to switch to aluminum nitride insulators. ?

On Nov 4, 2024, at 15:08, Dave Fugleberg via groups.io <dave.w0zf@...> wrote:

?
I would definitely contact HFSignals. My V3, bought complete at Dayton this year, failed on the second day of use with no power out. They shipped me a replacement board to swap in and I sent the failed one back. Worth a shot.

On Mon, Nov 4, 2024 at 11:55?AM Evan Hand via <elhandjr=[email protected]> wrote:
Hi Andy,
?
You should check with HFSignals as there is a warranty for fully assembled kits.? At a minimum I would expect a pair of replacement MOSFETs. ?

I would remove the two IRF510s and check with a transistor tester or using a DMM to verify they are good.? You can search the web for how to test a MOSFET with a DMM.?

Remember that you must readjust the bias if you replace the IRF510s.? I also suggest you replace both even if only one is bad. Try to use two from the same manufacturer batch.?

73
Evan
AC9TU


 

Of course you tried transmitting in another mode like CW. To make sure it's not a FT 8 issue??
Patrick?


 

I would at first try to measure bias current.. if there is no current the transistors are toast.
if it is under warranty then contact the seller.
I have no idea if they will then send you a replacement board.. or just a pair of MOSFETs?
Should be easy to repair in either way.
Dg9bfc sigi?

Am 05.11.2024 12:19 schrieb N8boapsc@...:

Of course you tried transmitting in another mode like CW. To make sure it's not a FT 8 issue??
Patrick?


 

Hi Andy,

Before doing anything, check if there is 12V in TX in the final stage. As you said the power

meter twitches then power transistors may be OK. worth checking.

Raj, vu2zap

On 04/11/2024 10:11 PM, Andreas Zahmel DO6DOC via groups.io wrote:
Hello,
I think the transmitting transistors of my sBitx V3 (approx. 1 month old, delivered fully assembled) are defective,
there is no more transmitting power. When I measure with a qrp wattmeter, I can see the pointer on the meter twitching. The initial situation:
I made FT8 with approx. 20w, 1-2 QSOs in 10 minutes, SWR under 1:1.2, from now on the transmission power was gone, the display shows an SWR of 1.0 and 0 watts.
The first thing I did was to take a new SD card and set up the system again.

I opened the device and there is no visible damage, to check the mosfets, I wanted to loosen both screws and found that the screws were loose and the mosfets were
not in contact with the heat sink (could be easily recognized by the gap in the thermal paste, about 1-2mm distance) The mosfets look visually ok, nothing burnt or apparently too hot.
i didn't change the language either (i had already made that mistake on the first day :) )
How would you proceed? Write to Ashhar Farhan?, replace the two replacement mosfets yourself (btw, lead or lead-free solder?), check other components first?
Has anyone else had a similar problem or have any ideas?
Many thanks and regards Andy DO6DOC


 

Thank you very much for your numerous answers!
?
support has contacted me and wanted a video of the error. So I reassembled the device, connected everything and connected my dipole.
I started the recording and the device is transmitting with normal power. The transmission power was not present before in the other operating modes either (SSB and CW tested)
I actually wanted to carry out the tests recommended by you (Voltage MOSFETs, Bias) but apparently this has been done. The only thing i did was to tighten the fastening of the two
mosfets by hand so that they have contact with the heat sink, the only thing i suspect now would possibly be a cold solder joint, i have no other idea.?
?
vy 73 de DO6DOC Andy


 

Ok, now it's getting weird.

I have successfully made a few tests in FT8 (SSB and CW also worked, FT 3 connections at first attempt)?
?
Then I switched off the TRX and a few hours later I did an update. I installed the 64 bit v3.1 (copied the finished image to SD) copied hw_settings.ini and sbitx.db and everything worked.
After an update (via git hub) to the current 4.12, the first cq call was with full power, a second one with 0 watts again. After a restart, power went out again with the first CQ call,
with the second no longer, I'll go back to 3.1 and test it longer with it.
?
73 de DO6DOC Andy
?


 

I have seen this.? ?Before I switched to AlN (aluninum?nitride) insulators, apparently my mosfets got pretty hot and it appeared there was "plastic flow" of the solder on at least ONE of the leads of one of my mosfets.? ?The power dropped by like 20dB.? ?Took me quite a while to figure it out (it was part of my months-long quest to fix one radio that I think I bought used, and probably contributed myself to its problems -- all of which are now solved and it is a delightful radio).? ??

LOOK at the solder -- if from the top it looks like it has "dropped a bit" then you have diagnosed it.? ?Otherwise, just reflow a bit more solder on the pins of the MOSFETS -- of course CAREFULLY so you don't overheat the transistor yourself, right?

After my various fixes, I have had no further issues.? ?A month ago, I bought ANOTHER "V2" radio but have not yet had the time to tear into it.? ?Too busy teaching and trying to deal?with pickup truck transmission failures (which are much, much more expensive than ham radios)

73
Gordon KX4Z


 

Andy,
This device should be treated like a male egg.
It's a computer-controlled radio device, and as such you can encounter countless problems that are thought to be minor errors. But you can learn to use it with a lot of patience. Some screws may come loose during shipping.
This was also discussed at the end of April 2023. So it's worth checking a few things before you use it.
What you should buy if you don't have one at home is a power supply with current limitation.
I use v2, which is sensitive to higher voltages, so I use it with 12.5V and 6.5A limitations.
The v3 can get more voltage, but don't overdo it.
When I used it from a battery once, I managed to destroy the final FETs.
The replacement parts still work today after the adjustment.
I hope to see you on the air: I have antennas for 7, 14, 21 MHz (a top-fed delta loop antenna installed at an angle).
I'm celebrating soon because I've been using the device for almost a year and have made more than 6600 QSOs with 161/142 dxcc/LoTW cfm certification.
--
Gyula HA3HZ


 

To me, it appears to be a heat related issue. This can be a crack, usually a dull connection, which is a poor solder job, or insufficient solder mainly around heavier components. In car audio work, toroids not glued down and large transistors will exhibit this.?

A few years down the road you might need to scrape off glue that turns conductive, causing lots of weird troubles, then glue the component again. (Obviously this is not yet your trouble.)

Scotty

On Tue, Nov 5, 2024, 2:20 PM Andreas Zahmel DO6DOC via <andreas.zahmel=[email protected]> wrote:
Ok, now it's getting weird.

I have successfully made a few tests in FT8 (SSB and CW also worked, FT 3 connections at first attempt)?
?
Then I switched off the TRX and a few hours later I did an update. I installed the 64 bit v3.1 (copied the finished image to SD) copied hw_settings.ini and sbitx.db and everything worked.
After an update (via git hub) to the current 4.12, the first cq call was with full power, a second one with 0 watts again. After a restart, power went out again with the first CQ call,
with the second no longer, I'll go back to 3.1 and test it longer with it.
?
73 de DO6DOC Andy
?


 

Hello a small update,
?
I am in contact with the support,? a few days after the purchase I upgraded the device to the 64bit version, so it ran for a good 2-3 weeks.
The support asked me to try again with the 32bit version, unfortunately I can't find a download anywhere,
except for a version 1 where I have problems with the update, hw-settings etc is all there, do any of you have a link??
?
Many thanks vy 73 Andy/DO6DOC
?


 

Here is some feedback,
in the last 3 weeks there was a lot of email traffic with the support of hfsignals. I have made several videos on request from the behavior of the
TRX to measurement setups to show in the video how the quiescent current is and what happens during transmission.
?
Today I was told to desolder the MOSFETs by chance, you don't know exactly where the error could be and install the supplied ones myself because I contacted them 3 days after the warranty expired.
?
The error was noticed for the first time 14 days after receiving the device, 1-2 times no transmission power at approx. 50 FT 8 QSOS in 4-5 days.
A week later I started up again and everything worked normally again, 4-5 days later the same error, out of 100 transmissions 2-3 times no power - of course I first redid the software,
opened the device at the weekend to look for faulty areas, reassembled everything and retested, the device then ran for about 1 week without problems and then the error
was there more often, at that point I turned to the forum here and at the same time to HFsignals
?
and now I'm told because I got in touch 3 days too late it's not a warranty case, that's kind of frustrating so next time, as soon as even the smallest error occurs in our
experimental hobby, write to support immediately and never try to solve it yourself.
?
Andy
?


 

¿ªÔÆÌåÓý

Sounds like an intermittent bad connection?

You might check and see if there¡¯s any possibility of looseness on the BNC pin-to-socket connection.

After that, I would look very carefully at the source connection of the power MOSFETS. ??


Make sure there is a tiny convex mound of solder on the source lead, and that it is not dropping into the printed circuit board concave ; if it is dropping in, I would very carefully add a tiny bit more lead tin solder.?

I had this happen to me and the output doesn¡¯t completely go away, but it drops by about 20 DB. One of the two mosfets was no longer making a good source connection so it was not really a push pull anymore. ? If you have a sensitive output monitor of some sort, you might be able to see that. ?

A tiny bit of solder and it was fixed. I would tell you that I switched my insulators to aluminum nitride, but unless you are really pushing higher power, I don¡¯t think that¡¯s a problem. ?

And I would mention that my issues were with the 40 W V2 version, so it¡¯s not guaranteed this applies ?to the current version?


Gordon kx4z?



On Nov 25, 2024, at 11:50, Andreas Zahmel DO6DOC via groups.io <andreas.zahmel@...> wrote:

?
Here is some feedback,
in the last 3 weeks there was a lot of email traffic with the support of hfsignals. I have made several videos on request from the behavior of the
TRX to measurement setups to show in the video how the quiescent current is and what happens during transmission.
?
Today I was told to desolder the MOSFETs by chance, you don't know exactly where the error could be and install the supplied ones myself because I contacted them 3 days after the warranty expired.
?
The error was noticed for the first time 14 days after receiving the device, 1-2 times no transmission power at approx. 50 FT 8 QSOS in 4-5 days.
A week later I started up again and everything worked normally again, 4-5 days later the same error, out of 100 transmissions 2-3 times no power - of course I first redid the software,
opened the device at the weekend to look for faulty areas, reassembled everything and retested, the device then ran for about 1 week without problems and then the error
was there more often, at that point I turned to the forum here and at the same time to HFsignals
?
and now I'm told because I got in touch 3 days too late it's not a warranty case, that's kind of frustrating so next time, as soon as even the smallest error occurs in our
experimental hobby, write to support immediately and never try to solve it yourself.
?
Andy
?