(tr)uSDX power, efficiency and filter optimisation

Post Reply
Iw2fpe
 

Posts: 12
Joined: 20 Mar 2022, 17:06

Re: (tr)uSDX power, efficiency and filter optimisation

Post by Iw2fpe »

HI

I found some free time to optimize the 20 meter band inefficient (rw kit). The measurements made gave 50% efficiency with 0.35W at 5V, the second harmonic in the right position at 28200 but with -40 dbm attenuation. I replaced all the section caps and for each component I repeated the test with Nvna. Only with the cap C16 150pf I got a sharp curve with attenuation at -61 dbm. after the substitutions I got the following values:
20m band Efficiency 78% 0.53W at 5V (calibration of up to 12V still to be performed). it seems the right way to repeat on the other bands, 30,40,60 still not very efficient. 80 meters looks fine.

73 Iw2fpe
HA7MAC
 

Posts: 1
Joined: 31 May 2022, 10:02

Re: (tr)uSDX power, efficiency and filter optimisation

Post by HA7MAC »

Hello All,

I have problem with AliExpress Up Tech Store version.. First it was 0W power, but after change the BS170s the power increased about 0.4W. I started to play with filter optimization. The end is: power increased to 6-6.5 W but the efficiency is less than 55%. I If I'm checking the band specific LC then it shows good result, but If I'm checking the drain waveform it is near to sinus. My assumption is, the tuning process was wrong.
Could someone drop a link with E class tuning process or video?
Other way I would recommend to you nanovnasaver python app, it helps to handle nanovna from your PC. my measurement created with impedance matching on CH0 / S11.

Thanks!
73 DX HA7MAC Gabor

update1: Yesterday evening I went through the DL2MAN Manfred video tutorial on adjustment and followed. The curves are same and some improvement on the drain waveform, but really far from expected. Is my understanding correct: the second curve indicates incorrect capacitance on the "bandpass section"?
Attachments
3.png
3.png (13.03 KiB) Viewed 8950 times
3.5MHz-eclass.png
3.5MHz-eclass.png (15.66 KiB) Viewed 8967 times
Ohwenzelph
 

Posts: 207
Joined: 01 Jan 2022, 03:47

Re: (tr)uSDX power, efficiency and filter optimisation

Post by Ohwenzelph »

3.96MHz, 3.8W, 89%
3.96MHz, 3.8W, 89%
865590D0-72E5-482A-B2E3-269470FA26DD.jpeg (137.05 KiB) Viewed 8885 times
3.86MHz, 4.9W, 90%
3.86MHz, 4.9W, 90%
E0BD7897-6B5A-41EB-9DC0-15F0B8BC713B.jpeg (113.67 KiB) Viewed 8885 times
3.76MHz, 6.2W, 90%
3.76MHz, 6.2W, 90%
0DEFE022-D2D8-4D03-837F-F13B23989DC0.jpeg (129.99 KiB) Viewed 8885 times
Here is what I see on a scope across 80 meters

3.96Mhz, 3.8w, 89%
3.86Mhz, 4.9w, 90%
3.76Mhz, 6.2w, 90%
3.66Mhz, 7.3w, 86%
3.56Mhz, 7.2w, 76%
3.502Mhz, 6.2w, 70%
Ohwenzelph
 

Posts: 207
Joined: 01 Jan 2022, 03:47

Re: (tr)uSDX power, efficiency and filter optimisation

Post by Ohwenzelph »

3.66MHz, 7.3W, 86%
3.66MHz, 7.3W, 86%
F4D9E8FA-92BF-474A-BB55-B38385880C5D.jpeg (153.6 KiB) Viewed 8886 times
3.56MHz, 7.2W, 76%
3.56MHz, 7.2W, 76%
C82547CE-52C5-40A6-9E70-9434EA49E964.jpeg (170.58 KiB) Viewed 8886 times
3.502MHz, 6.2W, 70%
3.502MHz, 6.2W, 70%
354F646D-237C-49F7-8F45-4E5B6031C9F9.jpeg (94.64 KiB) Viewed 8886 times
At the lower frequency end the wave and efficiency are not as good, and so far, the peak power out and peak efficiency are not exactly together. Perhaps I need more capacitance on 53/52?
G3CWI
 

Posts: 12
Joined: 11 Jun 2022, 16:47

Re: (tr)uSDX power, efficiency and filter optimisation

Post by G3CWI »

My Up Tech built unit gives the following performance (12.0 Volts, 50 Ohm Load, internal metering):
MHz
3.560 5.8W 60%
5.351 3.3W 61%
7.027 3.1W 61%
10.11 6.3W 65%
14.06 3.5W 57%

Looks to be symptomatic of use of caps with the wrong delectric to me. I have emailed Up Tech for assistance.
G3CWI
 

Posts: 12
Joined: 11 Jun 2022, 16:47

Re: (tr)uSDX power, efficiency and filter optimisation

Post by G3CWI »

...interesting reply from Up Tech (immediate response - within seconds - excellent!).

"Hello,It can be corrected by setting Menu 8.6 RShunt to Value 22."

I think that I need to do a few measurements before changing the shunt value.
Last edited by G3CWI on 22 Jun 2022, 09:27, edited 1 time in total.
CR7BAQ
 

Posts: 9
Joined: 22 Mar 2022, 22:19

Re: (tr)uSDX power, efficiency and filter optimisation

Post by CR7BAQ »

Did it resolved the issue ?
G3CWI
 

Posts: 12
Joined: 11 Jun 2022, 16:47

Re: (tr)uSDX power, efficiency and filter optimisation

Post by G3CWI »

I have no doubt that changing the shunt value will improve what it says on the screen of the radio. However, that might not be the right approach. In the first instance I need to make a few measurements to see what the efficiency actually is.
G3CWI
 

Posts: 12
Joined: 11 Jun 2022, 16:47

Re: (tr)uSDX power, efficiency and filter optimisation

Post by G3CWI »

Okay. Measured the efficiency of Up Tech assembled unit using internal power meter to measure RF output power and bench PSU to measure DC input power.

MHz
3.560 59%
5.351 55%
7.027 55%
10.11 63%
14.06 51%

I have no easy way to measure the actual output power. Have people found the internal RF power meter to be reasonably accurate?

EDIT: Up Tech has pointed me to a thread here that indicates that the displayed efficiency is the efficiency of just the PA - not the whole unit. I will need to rework my calculations to subtract the 80mA that the unit draws on receive. I will be back!
G3CWI
 

Posts: 12
Joined: 11 Jun 2022, 16:47

Re: (tr)uSDX power, efficiency and filter optimisation

Post by G3CWI »

Reworked figures - now for PA efficiency only. RF output measured using tr uSDX power meter and DC input measured using bench power supply. RX current 80mA.

MHz
3.560 65%
5.351 66%
7.027 66%
10.11 69%
14.06 60%

EDIT: Using these measured values, I have found that for my radio a shunt value of 19 gives accurate readings for PA efficiency (assuming that the internal power meter is accurate).
DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Re: (tr)uSDX power, efficiency and filter optimisation

Post by DL2MAN »

When I designed the PCB for (tr)uSDX, I thought it was a good Idea to use a PCB Trace as RShunt for PWR measurement. It turned out, that PCB Manufacturing Tolerances are quite high, so this ends up in quite different results..... So I will not do that again in any future design.

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, so at least Efficiency is close to reality. PWR however is still 0,3-0,5W Off....

Unfortunately with chip crisis going on and on, there seem to be AtMegas out there, with tolerances higher then they used to be.
At this point, there´s not much we can do about that.

But the good news is: The rigs deliver Power and are efficient. It´s only an OnBoard measurement inaccuracy.

73 Manuel; DL2MAN
G3CWI
 

Posts: 12
Joined: 11 Jun 2022, 16:47

Re: (tr)uSDX power, efficiency and filter optimisation

Post by G3CWI »

Thanks Manuel. That clarifies the situation. Perfectly happy with the product itself. I will dive in with my VNA at some point and normalise the power output across the bands. Thanks for a super little radio.
yo2mke
 

Posts: 6
Joined: 03 May 2022, 12:03

Re: (tr)uSDX power, efficiency and filter optimisation

Post by yo2mke »

The SWR looks right for every band but the efficiency is very low (~20%) so I am trying to check the coils using the nanovna but the graph looks very different than what i have seen in the "tutorial". I do not have a notch, it just goes down and stays down...

I do not know if it makes a difference or not but the marker at 14000 is -4.25db (not ~0.somethign)
What am I doing wrong?
298930069_499878511946077_8760808103281767391_n.jpg
298930069_499878511946077_8760808103281767391_n.jpg (163.18 KiB) Viewed 8253 times
is this ok?
AC9YY
 

Posts: 133
Joined: 21 Mar 2022, 00:04
Location: Illinois, USA

Re: (tr)uSDX power, efficiency and filter optimisation

Post by AC9YY »

What band/frequency were you testing?
yo2mke
 

Posts: 6
Joined: 03 May 2022, 12:03

Re: (tr)uSDX power, efficiency and filter optimisation

Post by yo2mke »

AC9YY wrote: 12 Aug 2022, 16:52 What band/frequency were you testing?
20m 14Mhz
AC9YY
 

Posts: 133
Joined: 21 Mar 2022, 00:04
Location: Illinois, USA

Re: (tr)uSDX power, efficiency and filter optimisation

Post by AC9YY »

yo2mke wrote: 12 Aug 2022, 18:45
AC9YY wrote: 12 Aug 2022, 16:52 What band/frequency were you testing?
20m 14Mhz
Yes, that curve is not right. Is it connected correctly and are you sure you are on 20 meters?
yo2mke
 

Posts: 6
Joined: 03 May 2022, 12:03

Re: (tr)uSDX power, efficiency and filter optimisation

Post by yo2mke »

I did all the measurements again calibrating each band individually and played with coil 2 to achieve minimum value but still no notch....
My nano vna is SAA2N (the one with N connectors). Is it possible that this is because of the N to sma adaptors I am using? (My Open/Load/Short are for N connector and I calibrated everithing witout the sma adaptors...)
20m.jpg
20m.jpg (53.39 KiB) Viewed 8220 times
20M band
30m.jpg
30m.jpg (51.06 KiB) Viewed 8220 times
30M band
40m.jpg
40m.jpg (51.48 KiB) Viewed 8220 times
40M band
yo2mke
 

Posts: 6
Joined: 03 May 2022, 12:03

Re: (tr)uSDX power, efficiency and filter optimisation

Post by yo2mke »

same for rest o the bands... 80m band has a very little notch, it goes down and then slightly up
60m.jpg
60m.jpg (40.5 KiB) Viewed 8220 times
60M band
80m.jpg
80m.jpg (50.55 KiB) Viewed 8220 times
80M band
John
 

Posts: 49
Joined: 19 Feb 2022, 08:56

Re: (tr)uSDX power, efficiency and filter optimisation

Post by John »

You might want to take a photo that shows us how everything is cabled up, (coax cable,) along with the process you are going through to get those readings. Also a photo that shows the RF board up close.
John
*******
yo2mke
 

Posts: 6
Joined: 03 May 2022, 12:03

Re: (tr)uSDX power, efficiency and filter optimisation

Post by yo2mke »

299605063_3289358797998535_1314247682816383705_n.jpg
299605063_3289358797998535_1314247682816383705_n.jpg (157.71 KiB) Viewed 8151 times
From CH0 a cable goes to the pin header (GND & RF) and from CH1 a cable goes to the SMA connector
299248961_7865494280159141_934767356685194134_n.jpg
299248961_7865494280159141_934767356685194134_n.jpg (228.05 KiB) Viewed 8151 times
299215628_1222967545127663_3097890457214976977_n.jpg
299215628_1222967545127663_3097890457214976977_n.jpg (179.29 KiB) Viewed 8151 times
DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Re: (tr)uSDX power, efficiency and filter optimisation

Post by DL2MAN »

This is most likely a measurement problem.
If you use such long cables, try to calibrate them out.

So: Attach Short Open and Load to the ends of the cables you´re using while calibration.
For through calibration, connect both cables together....
Calibrate as you would measure.

73 Manuel; DL2MAN
John
 

Posts: 49
Joined: 19 Feb 2022, 08:56

Re: (tr)uSDX power, efficiency and filter optimisation

Post by John »

I see you are not using the pin header connection to the RF signal pin recommended in Manuel's video. Nor can the point where the connection is actually made be identified in the photo. It would be best to verify that your connection is made to that same trace in the circuit.
Unfortunately, I find that this sort of remote troubleshooting only works if one is extremely methodical about every detail. It can be a tedious process, but it does work.
Randal007
 

Posts: 8
Joined: 30 Jun 2022, 21:38

Re: (tr)uSDX power, efficiency and filter optimisation

Post by Randal007 »

I'm having some issues Tuning LX2 for 20m, my NanoVNA 2nd harmonic reading is 28.350 which is what i tuned it for, but I only get 2.5Watts out. I found if I remove a wrap I now get 3.5watts out. I also notice as i go down in frequency the power goes up still. Is my NanoVNA off or is something else happening? should i remove another wrap?

I am finding this on a few bands seems like many of coils have too much inductance.

I also always read 99% efficiency all the time. It did use to work but now it never does. I did check the 2 toroids multiple times and they seem good.

I believe my kit is Batch 3 from UpTech.
John
 

Posts: 49
Joined: 19 Feb 2022, 08:56

Re: (tr)uSDX power, efficiency and filter optimisation

Post by John »

To the best of my knowledge these parts in these circuits always put out more power at the bottom of the band and then taper off as you go higher in frequency. If you only use 1 part of a given band you can tune the power output for just that frequency range. But I tune them for the bottom of each band with the maximum power I figure is safe, (I don't let it exceed 6.5 watts,) and then I just accept whatever the output is at the high end of the band. Two or 3 watts is plenty for me. I'm not a QRP contester, and my goal is to run the lowest power I can get away with in any event. Sophisticated circuitry that the (tr)uSDX lacks is often added to RF stages of commercial gear in order to maintain uniform RF output. I suspect that is a feature that will be added to inexpensive digital rigs as time goes on, along with high heat and SWR protection.

As for me and for right now, I'm adding a small switching regulator to control the DC voltage applied to the FET output stage so I can dial in the power level I want and minimize power consumption, and of course comply with FCC rules. While it's another knob and a manual adjustment process, it's easy enough to implement, and I'm building the rig in a small metal box that has enough room for the modification, a high quality separate speaker to fix the speaker feedback problem, in case I want to use a speaker instead of earbuds, and a set of batteries and charge circuitry. Everything for communication but the antenna, a key and a separate mike will be in a solid metal box. It's a fun little project.
jil2
 

Posts: 1
Joined: 29 Sep 2022, 15:53

Re: (tr)uSDX power, efficiency and filter optimisation

Post by jil2 »

Hi to all,

Just want to share my (tr)uSDX experience.
First of all thanks to Manuel and Guido for the nice and tiny device!

I've ordered mine from AliExpress on 7th of June.
Classic band board was ordered on 31st of August.

Power was measured with Diamond SX-600 (not very accurate)
Here is the results of the test with R shunt = 22, Vcc=13.8:

Code: Select all

		Power Supply    Out	SDR meas.	Calculated		
Band	Freq    I (RX)	I (TX)	P meas.	P/W	Eff.	TX P/W	Eff.
80m	3560	79	652	~4.5	4.65	72.5	7.91	58.8%
40m	7030	80	527	~4.1	4.22	82.75	6.17	68.4%
20m	14060	83	567	~4.1	4.29	76.6	6.68	64.2%
15m	21060	86	658	~3.9	3.98	62.19	7.89	50.4%
10m	28060	88	678	~4.1	4.29	66.0	8.14	52.7%
Calculated values are not so perfect. It looks like R shunt should be equal to 17-18 in my case.

My observations:

1. The more longer PTT button is pressed the lower current, efficiency and power will become. Will stabilize after a few seconds.
2. After some time of testing all toroids became slightly warm (measured by my lips:)
3. After cooling down efficiency will be increased. Difference around 3-8%. The higher frequency the bigger difference. For red toroids the difference is 3-4%, for yellow toroids the difference is 5-8%.

Planning to buy new toroids and test again.

Will be very appreciated for some comments or suggestions.

Thanks! and 73
Attachments
my (tr)uSDX
my (tr)uSDX
truSDX.jpg (239.56 KiB) Viewed 7266 times
Post Reply

Users browsing this forum: No registered users and 1 guest