Low Power and low Efficiency on Uptech Classic Bands (tr)uSDX

[DL1DBY]

Hi all,
I have build a (tr)uSDX for the classic bands which I bought from Uptech/Sunny via Aliexpress. On most bands the power and efficiency is pretty low, down to 35%.

Here are my results:
80m: output 4.9 watts, efficiency 55%
40m: output 3 watts, efficiency 66%
20m: output 2.4 watts, efficiency 54%
15m: output 3.5 watts, efficiency 41%
10m: output 3 watts, efficiency 35%

All measurements were done externally, they and the calculations were done in the way Manuel DL2MAN suggested. The same measurements on my low bands (tr)uSDX had much higher and quite normal results.

I would have expected around 70% efficiency, maybe a little lower. While these values are not close to zero, which would clearly indicate a mistake while building the kit, they are pretty low.

I have already done the calibration process of the LX2 coils with a NanoVNA according to Manuel’s video. It displayed clearly visible minimums on all bands as it can be seen in the video. I also did the LX1 coils calibration as best as I could, peaking efficiency (but at these low levels).

The (tr)uSDX has the T2 mod with 3:21 turns.

Any ideas what could I try to get to output to around 4 to 5 watts on all bands and get efficiency higher?

Maybe experiment with a different number of turns on the LX1 coils or replace the MOSFET? Experiment with the number of turns on L15?

Thanks & vy 73 de Dieter DL1DBY

[DL1DBY]

Problem solved by increasing 8.2 PA Bias MAX from 128 to 160 for FDT86256 and retuning filters.

[vk2tty]

I got similar figures on my UpTech Classic Bands with PA Bias MAX=160, measured without tuning R Shunt:

• 80m 3.93W 72%
• 40m 3.52W 66%
• 20m 2.32W 59%
• 15m 4.35W 54%
• 10m 3.15W 32%

Tuning time!

[vk2tty]

[/quote]

In Sunnys Lot 3 (built with T2 Mod 3:21), we found that PWR measurement was also 0,3-0,5W Off from reality -depending on Band. This in combination with different RShunt Value of PCB Trace made me suggest to use RShunt Value 22…

Trying R Shunt=22, instead of the default 17…

• 80m 3.93W 95%
• 40m 3.52W 86%
• 20m 2.32W 80%
• 15m 4.35W 70%
• 10m 3.15W 47%

Reckon I'll need to learn how to measure output power with an oscilloscope. High-impediance probes on RF and GND while the unit is powered going to be OK?

Anyone got a set of expected figures for a recent UpTech Classic Bands kit? DL1DBY?

[vk2tty]

Reckon I'll need to learn how to measure output power with an oscilloscope. High-impediance probes on RF and GND while the unit is powered going to be OK?

I think not, given the signal across RF and GND haven't yet been amplified. The SMA socket at J12 is my only hope. If only I had a tee.

Anyone got a set of expected figures for a recent UpTech Classic Bands kit? DL1DBY?

These'd still be super handy.

[vk2tty]

Trying R Shunt=22, instead of the default 17…

• 80m 3.93W 95%
• 40m 3.52W 86%
• 20m 2.32W 80%
• 15m 4.35W 70%
• 10m 3.15W 47%

After tuning, still with R Shunt=22:

• 80m 4.10W 95% 80m
• 40m 3.35W 92% 40m
• 20m 2.15W 77% 20m
• 15m 4.05W 71% 15m
• 10m 4.15W 55% 10m

I'm not sure why I lost power at 40m but gained efficiency, nor why getting the notch right at 20m lost me both power and efficiency. (I can't transmit on 20m yet, but I'd like to know what's going on.)

Can we tune the other toroids for more power?

[bortek]

I have completed the build and tested to transmit but getting very low power out using LO bands board. The power is approx 0.5W on all bands. Have tried to set shunt to 22 as well as PA Bias MAX from 128 to 160. But no difference.

Here are some pics where you can see toroids before and after being glued to the pcb. Is there something visible that cold be of an issue?

PXL_20231216_102408380.jpg (207.6 KiB) Viewed 2511 times

PXL_20231219_184158031.jpg (219.12 KiB) Viewed 2511 times

[dl6sez]

Hello OM,

your plastic blobs, i think hot glue, are unnecessary and bad due to rf absorbing and change the behaviour of the filters....
I also see at T1 not insulated parts of wire, the 1 turn pass through, this should be done again.
I'm not sure if you did the connections and winding directiongs right on T1 and T2.
And please do your soldering again with good solder wire with abt. 3% flux inside. All your toroid windings look well done but not the soldering, it looks not really soldered through, maybe you need more temperature (10 to 15°C) too.
AND NO: PA bias is 128 for LO band with BS170s!!!
Please first test with 5V from USB, first all rx should work with good performance, than go to 5V tx not 12V!
Expect 0.2W to 0.3W (measure with external watt meter first, because you could have made errors there also) at 5V USB powered.
Than follows the notch and than power maximization on every band, see the videos.

73 de Chris

[bortek]

Thanks Chris for a quick reply. I did check soldering for continuity and they were all ok. Before I go ahead and remove the blobs and redo the soldering I need to ask a about toroids.

is the direction of winding important only on T1 and T2 or other toroids too? In the manual there was no specification of the winding directions or have I missed it?

Here is a more detailed pic of T1 and T2. The 1 turn pass through might be an old one which I replaced. this one should be looking better on the recent pic. Or shall I re do it ?

PXL_20231219_202759841.jpg (229.23 KiB) Viewed 2494 times

PXL_20231219_202722954.jpg (243.66 KiB) Viewed 2494 times

[dl6sez]

Hello OM,

T1 and T2 are wrong wound, start with the wire hole which lies off center end wind the wire immidiately into the toroid inner side and than wound.... than you will get exactly off the toroid at the other side to the off center wire hole from the inner side (none from outside).
You will ever measure conductivity have a look at the circuit, also without conductivity on T1 T2 toroids!
Lpf could be measured partly
Believe me your soldering flux and temperature is very bad aligned, no good flow and nearly every through contact whole is not through soldered!

73 de Chris

[bortek]

Well as we use to say re-do it and do it in the right way

As you understood this is my first transceiver build so please bear with me Reading up more on the toroids and from what I understand the winding direction for inductances is not important.

For the transformers on the other hand the direction of primary and secondary is important. Winding in the same direction makes the voltage in phase while winding in the opposite direction shifts the voltage 180 degrees.

Now I see that coils on my T2 are in opposite directions. Also that is what made the whole toroid being few degrees rotated relative to the rectangle on the PCB, where it is supposed to be sitting.

My T1 also has a wrong direction of winding relative to how it was supposed to be sitting on the PCB. I just made the wires go under the toroid to the other side to get them through the hole in the PCB. And I was wondering how come this could have been so misaligned. But now I know why.
But what makes me wonder is the other winding of 1 single turn. It is just a wire through the toroid and there is no actual direction in this case. So here it should not matter in which direction it is wound. Well except that it is not so convenient to solder it on the PCB.

Are these correct assumptions or am I off to the forest completely?

[dl6sez]

I think when you change direction on T1 7 turns it should work ok,
but why not simply do as Manuel shows in his videos and look there before doing things wrong?

73 de Chris good night

[bortek]

Hi again. Just an update!

As suggested I have removed the glue blobs. Rewinded T1 and T2 according to the official video. The direction of the winding, as it turns out, is actually called counter-clockwise Now T1/T2 are sitting properly directionwise on the PCB. I have also soldered through all toroids and transistors. Also set back PA Bias to 128 to match BS170.

But I do still get no more than 0.7W with 12V power. Something is still not right. If a lower RF shunt down to 3 then I get out 5W but I feel this is the wrong way of fixing the issue.

I am using a beta FW BTW.

Any more suggestions or clue are much appreciated?

[vk2tty]

Let's isolate the problem.

Do you have an RMS-capable multimeter or oscilloscope you can use to measure voltage across the antenna port while you transmit CW into a 5W 50Ω dummy load? Measure Vrms and Rload, and derive your output power Pout as Vrms*Vrms/Rload. Try each band. This'll be how you choose a new R Shunt value later when you finish troubleshooting.

If you measure adequate power with your test equipment but the radio thinks your power is terrible, perhaps your Power/SWR bridge needs work: T1–2, R1–6, C5–8, and D1–2. If the radio thinks your efficiency is terrible, it might also be the input voltage and current measurement: U1, C2–4, R7-8, and L15. You might also have blown your finals during your earlier testing: Q1–3, or Q4 or Q5 depending on your board.

If you're lucky at least one band is OK, and it'll turn the other band slots have poor joints on their odd and even inductors, more or fewer turns than intended, or something I haven't yet done to myself.

First, though: put R Shunt back to the 17–22 range for the time being and, while you're at it, double-check your LPF band setup setting matches your board? I hear that can make a big difference.

Good luck!

[dl6sez]

Hello OM Bortek,

what i don't undestand is you wrote you once have now seen 5W and think it was adjusted with R shunt. R shunt has nothing to do with the power measurement!! It is only the shunt resistance on the mainboard to measure current at tx. So adjusting R shunt doesn't change anything on displayed power!

So do you measure at CW mode or are you using LSB or USB?
Only CW makes here sense to find the problem and align your (tr)uSDX, in USB and LSB will be no power with no modulation input.....
I don't believe there is anything dead, maybe you killed FETs maybe 1 off 3 or so is possible during your tests, but please leave the suggestions from VK2TTY to check power and swr measurement components before you measure a power output with an extern power meter or oscilloscope at a dummy load beside.


73 de Chris

[vk2tty]

Success story related to re-soldering one of the components I listed:

In the past I had come across a post about issues with C7 on the PA board, so I reflowed the connections to that and I now have proper receive and transmit.

[bortek]

Thanks for all your suggestion guys! I am waiting for some of the orders to arrive soon, like dummy load and ext. power meter before I can continue to troubleshoot and do more testing. I can doublecheck other solderings to make sure nothing have been missed. So a new report will be made in about 2 weeks time. Until then 73 and Good New Year!

[bortek]

Hello


Here comes an update with the video showing my tests. The R-Shunt(18) and Bias(128) are back to their original settings as I did a factory reset of the radio. Now all seems to work much much better but still no 5W out. I have a dummy load of 50 Ohm connected and power meter which shows slightly higher value than the radio itself. I use 3S Lipo battery which was fully charged to 12.6V with lots of available current. Even with lower input voltage like 11.8 the output power was on the same level.

Any thoughts?

https://youtu.be/EsbV-d0rL9g

[dl6sez]

Hello Bortek,

and what?
I see a fully very good working (tr)uSDX regarding output power!
The absolute maximum is 5W and can be achieved sometimes but none of my (tr)uSDXs reaches it.
All is okay, your rig is not an Icom Yeaesu Kenwood. The trx uses a class-E PA series filter LPF which has the highest possible efficiency and is not comparable to other standard trx where you have a class AB1 PA with about 45 to 50% efficiency...
Read a little bit about it here and in the internet.

73 de Chris

[bortek]

Thanks. Will do. And yeas I think I am satisfied with the result. Now time to do a real test with a friend and see how long will it gi.

[KL7MJ]

Assuming you have the latest RF board with one pre-soldered PA transistor, not three BS170s, you will want to increase the PA bias to 160.

Mike, KL7MJ

[bortek]

I have a board with 3 BS170

[KL7MJ]

Then 128 is perfect!
Sorry, I thought that almost all classic boards used the SMD transistor that mine has. Raising the PA bias on my rig made a big improvement.

Mike, KL7MJ