(tr)uSDX Forum

I have also recently experienced a transmit power failure. I used the radio for approx. 1 hr cw mode in a POTA activation. Lots of transmit time. The rig was working fine when I turned it off for the night. When I turned it on the next morning, no cw sidetone heard (in practice mode), and later discovered no power transmitted (back in transmit mode of course). The board does not appear to be damaged in any way, but this sounds like it may be a similar issue. The fact that it was working fine when I turned it off for the night is the most perplexing thing.

I appear to also have killed the FDT86256 mosfet in my high-band (tr)uSDX, with an SWR of 1:1.4 indicated on my Daiwa analogue meter.

I was characterising the efficiency of the unit, and it just stopped transmitting after a few PTTs of ~5 seconds.

Breaking out the trusty Rigol, I could see that I was getting the ~11V battery voltage on two of the BS170 pads, with no change in voltages on PTT press. I could see the overlaid RF ripple at less than 1v difference on both of those voltages with PTT, so I knew that portion of the systems was working. Checking the output on the antenna connector center pin, I was getting ~1v RMS for a few tens of milliwatts of output.

Comparing to the standard low-band unit I also built, that one is giving a nice 40v-50v RMS for the expected 3-4W output power.

Off to Digikey to order a replacement set. For me, a relatively easy fix once I get the spare part. Mildly frustrating, but all part of the kitbuilding game.

HB9HJF wrote: ↑06 Oct 2022, 10:19 that one is giving a nice 40v-50v RMS for the expected 3-4W output power.

40-50V RMS is 30-50W

I had a feeling I didn't have the probe 1x-10x slider set correctly, so that would instantly explain that.

Good catch.

Unlikely.

/10
4-5 V RMS gives us 0.3-0.5 W

But!
40-50 Vpp gives 4-6 W

<hands held up>

I'm in definite need of practice with my Rigol, and I also need to get more quailty sleep.
This is what I get for leaving it about two years since using the oscilloscope last.

Thank you for the pointing!

Right, digikey order arrived with the replacement FET (and a stack of capacitors to repair a mate's audio amplifier).

Some time with a hot air rework station and a few bits of kaptan tape later, the faulted FET was removed, and the replacement soldered into place. Quite an easy exchange at the end of it all, nice to practice the soldering with paste and hot air techniques

Reassembly, and the test shows there's now RF being ouptut with about the same listed power values as before the oopsie.

Hopefully it'll last a bit longer. If it does go again, at least I have nine replacements...
I'm putting it down to a random component failure.

Good afternoon.
Well I just joint the BS170 dead club. Build a standard kit (MALAHIT) from Amazon. Worked great for about three days. Checked the efficiency after build and it was around 72% on the bands with anywhere from 3 to about 6w output. I did not write it down unfortunately. Yesterday I was going to do the fine tuning of the coils and nothing. RX fine just no TX. Remove the three BS1701s and two are dead. Ordered new ones from Newark and they should be here on Tuesday. Manufacturer's part number is BS170-D27Z. I hope those will work.
I will update the results when I install the replacements.

James

I managed to blow another FDT86256 mosfet this morning in the high-config rig.

I was calculating the efficiency of the rig, with 13v supply (my portable lightweight battery which is a car battery jump starter), into a dummy load, with my Rigel monitoring the output voltages. PA bias max had been set at the 160, and I'm fairly sure my dummy load is working reasonably well. It's the same one DL2MAN uses on the bench, the 20W from QRP-Guys.

With key down after maybe 15 minutes of the activity, I watched the output voltage suddenly ramp down and die out, after I had the key down for about 15 sec, while I was measuring the output on 10m. The output voltage on the scpope was showing about 50 Vpp when this happened - a little bit over 6 watts.

Thankfully I had ordered other spares of that FET and in less than 5 min I had the dead one removed and a new one in place. I was able to complete the characterisation and get efficiencies in the region of 70-75%, with output powers in the 4.8W to 6W at 13.1V input, which is quite reasonable.

My limitation on the efficiency is the resolution of my DC inline power meter, which is 0.1A. Having a bit more resolution would be nice , but it'll only change the result by less than ~5% anyway so no really big deal.

Hopefully this FET relpacement will not need to be repeated.

(edit - wrong unit used, watts not amps..)

HB9HJF wrote: ↑19 Oct 2022, 22:17 I managed to blow another FDT86256 mosfet this morning in the high-config rig.

(...)13v supply (...)With key down after maybe 15 minutes of the activity, I watched the output voltage suddenly ramp down and die out, after I had the key down for about 15 sec, while I was measuring the output on 10m. The output voltage on the scpope was showing about 50 Vpp when this happened - a little bit over 6 amps.(...)

50Vpp = 6,25W Out.
13V*6A =78W in.....
So Efficiency is 8% and abt 72W are being disappated in your poor FET. This is why it dies....

If it was a misunderstanding and only the Current in the moment when it died: Use current limitation to prevent that. As soon as something goes wrong, your Power supply should cut off the Voltage and safe the FET. Especially while adjusting the Filters, this is important.
Max 1A, ans you should be safe.

73 Manuel; DL2MAN

Just an imagination, ...!
And maybe a possible 'solution'.

You can 'protect' your (tr)_u_sdx when 'testing SWR'!

I have a ZM-2 Z-Match. (The original one) When looking to the 'LED ' on the ZM-2, to 'monitor', how good or bad my SWR reflection is, i will not ruin my TRX.

When the ZM-2 is in the 'measuring-mode', the maximum SWR the TRX will see, is 1:2 !
Also if their is a short-circuit, or no antenna at all, ... !
( Z=0 Ohm, 50 Ohm, or nothing!)

The circuit inside the ZM-2 will 'protect' the TRX during the measurment! Also during SWR-mesurment, there will be a 6dB attenuation in TX and RX. So the maximum 'bad'-SWR is never worse then 1:2 for the TX-side.

The mesuring-circuit in the ZM-2 is a resistive-brigde, not a reflecto-meter.
(Like a "Weathstone-Brigde" for resistance)

When using the ZM-2 correctly for sending, I have to 'switch back' to 'normal operation'-mode!

By the moment, I never had any problems with 'blown-up' BS-170, ... !

72 de Markus ; db9pz

Hello,

I got the same problem here. Power and efficency are both zero.

I already checked the BS170 and switched them to new ones (the ones from the link to RS Components = ONSEMI). Still 0 PWR….

What else can I check? Any ideas?

Thx
73 OE3FSA

I can measure the filters and they look good. So am I right that the problem must be the FETs?

DL2MAN wrote: ↑20 Oct 2022, 10:44

HB9HJF wrote: ↑19 Oct 2022, 22:17 little bit over 6 amps.(...)

50Vpp = 6,25W Out.
13V*6A =78W in.....
So Efficiency is 8% and abt 72W are being disappated in your poor FET. This is why it dies....

If it was a misunderstanding and only the Current in the moment when it died: Use current limitation to prevent that. As soon as something goes wrong, your Power supply should cut off the Voltage and safe the FET. Especially while adjusting the Filters, this is important.
Max 1A, ans you should be safe.

73 Manuel; DL2MAN

My apologies for writing "amps" instead of "Watts"
This was the conversion from the Vpp to Watts - a correct conversion (6.25W RF output at the time) but an incorrect writing of the unit. I've edited my post for clarity and noted that fact.

The battery supply feeding the rig has an absolute maximum of ~30W output anyway and drops voltage before that point, I am quite confident that I'm not putting over 1A through the power input to the rig at any time. There's also a fast-blow 2A fuse in that spot in the powerpole breakout box. I don't have a current limited bench supply yet, i've found self-limiting battery supplies and low-value fuses to have been sufficient. One of those times when some of the crappy equipment actually is more useful for me.

The dead FET had no release of magic smoke, no case breakage (checked with a low power microscope), didn't smell burned at all.

I think I'll add a thin heatsink to the currently-installed FET at some point to help guard against short-term thermal excursions.

Thanks for the tips though.

I am now suspecting that the spares that I got from Digikey are not good quality.

I removed the non-functional FET, soldered on a replacement, put it together, attached dummy load and using the three battery pack for ~11V, I got CW output into the dummy then a quick ramp down to no watts output.

Another desolder, a replacement soldered, this time with a PiZero2 heatsink on the FET. It also gave up after less than 10 seconds total transmit time.

Another desolder, another replacement soldered in, this time with a tiny amount of non-conductive thermal paste under the FET and the same rPi heatsink on top. PA max bias set back down to 128, was putting ~3W into the dummy load with a quick test. This one died in the length of time it was taking my ATU-100 (QRP-built) to tune my 40m EFHW that I've recently built and was testing the rig with. 11 volts in from the triple battery, was listing about 3.5W of output just before it died.

I'm at a loss to explain why the FETs are failing in the manner they are. The internal power meter in the (tr)uSDX just starts at the expected output and ramps down to zero in about a second or so. Testing the failed FETs in my little component tester gives me about 25 ohms of resistance, where testing the device pre-installation did show it as a "N-E-MOS" with C=282pF and Vt of 3.2V.

I am now thinking of putting in x3 BS170 into that radio just so I can get something working, but I'll leave that for another day while I go and play with the low-band one I also built.

It's just very odd. I would love to know what I'm doing wronmg - if it's my actions that are killing it, but I'm currently at a loss to explain why.

Edit:- I've soldered in another one, this time taking great care to prep the solder pads before installation, and trying the use of a cooler soldering iron, as I think I may have overheated the FET when previously soldering them to the board. Looks okay so far at least, will not be powering from >12V nor will I be using the PA max bias of more than 128. Fingers crossed

Hello Guys.

I have the Same problem on my Classic Band (tr)usdx.
Power:0 and SWR: nan
Now i want to replace the PA-FET. I found this FDT86256
https://de.farnell.com/en-DE/on-semicon ... dp/3003915

Is this the right one? Or are there other versions of FDT86256 out there?

greets

Markus

If SWR shows nano, is the FDT86256 burnt? I ordered 10x from farnell.

I bought a prebuilt truSDX in April from one of the approved suppliers. At first glance it all seemed fine and I was able to download the software with my callsign. I've been busy so haven't been able to play radio much lately. I finally got some time this last week. I plugged it in and when I go to TX, the power meter says 0.00w. This is with a 13.8V power supply attached.

As far as troubleshooting, I've had a nearby ham listen on my frequency but he doesn't hear anything being transmitted. Also when transmitting, it doesn't move any needles on my inline Watt/SWR meter. I opened the case to see if any solder joints looked sketchy and they all appear fine. I'm not a hardware guy which is why I ordered the prebuilt version. But I'm stuck as to what to do.

Some folks mention testing the FET. I have a multimeter so I could always do some testing with that. But I need some help from the forum on how to do that. What circuit location(s) on the board should I look at? How do I test them? What am I looking for on the multimeter? Any other troubleshooting I should try?

73, Carl K7CGE

K7CGE wrote: ↑12 Jun 2023, 02:29 What circuit location(s) on the board should I look at? How do I test them? What am I looking for on the multimeter? Any other troubleshooting I should try?

With a multimeter, you can only measure the FETs outside of circuit, and then only if they have a really bad condition (Fully open or fully short). On an intact Fet, you should be able to determine breakthrough voltage of internal Diode.
Please provide some more information: What Version do you have (Lo/Classic/Hi Bands),
Does the radio receive ?
How much current does thr rig draw on RX and on TX ?
Does it maybe work when USB Powered ?

The reason I don´t reply by Mail to requests like this is:
Others in a similar situation can maybe find this and it does not end on individual level, while I would need to do this over and over and over again.

73 Manuel; DL2MAN

P.S. For all of you, that ask "how can we support the project":
Please provide assistance to those who seek help here.
Please provide "community support".

Hi Manuel,

Answers to your questions:

What Version do you have (Lo/Classic/Hi Bands): Lo bands
Does the radio receive? Yes, I'm able to receive and decode
How much current does the rig draw on RX and on TX ? RX draw is 1.2w. TX draw (holding PTT button on CW mode) also stays at 1.2w.
Does it maybe work when USB Powered? It functions the same when on USB and/or 12V power. No difference to the readings stays at "0.00W SWR:1.00" with any combination.

Thanks, Carl

When you put it to cw and press PTT: does it show P or T in Position of R ?

73 Manuel DL2MAN

It changes to 'T'.

OK, then most likely your PA Fets are defective.
Please organize 3xBS170 from a trusted source:
https://www.mouser.de/ProductDetail/ons ... rOUQ%3D%3D

Watch for the right pinout, when you order:


Order some more, to have spare. If you have a cheap component tester like this one:
https://www.amazon.com/-/de/dp/B08YNB7K ... 191&sr=8-4
you can sort them for equal Values (Vt, cg and rds as equal as possible).
If you solder in 3xBS170 with as equally as possible values, then you devide the load equally to all 3 FETs an they are much more forgiving in case of abuse. Imagine: 1 FET would have slightly different values than the others, this one has to carry more load and fails first. This causes a chain reactuion as the remaining two cannot handle the load and blow as well....

Unsoldering without hot-air soldering equipment is easiest done by:
-carefully cutting off the 3 legs of the Fets
- then unsolder each leg individually with a sucking-pump and a pair of tweezers. This way, you don´t damage PCB.

73 Manuel, DL2MAN

OK I'm back in business. I replaced the 3 FET's and am getting PWR readings ~5W. Just made a 20m FT8 contact about 1,000 miles away so I know it works!

I hope I don't ever need to replace those PA FET's again because I'm not sure the board would survive another butchering from my soldering iron. I think I'd be best in investing in a hot air gun if I need to do so.

Having a few other small issues with the radio with but I'll search the forums to see if there are any other mentions of those.

Thanks for your help!

73, Carl K7CGE