¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. QRPLabs
  2. Topics

FSKCW on U3S QRSS 30m in preparation for New Year's Eve

 

My results after epoxying a #4-40 aluminum spacer to the top of the 27 mHz crystal on my U3S' Si5351a board. Here I'm running 2W TPO (w/ 5W PA kit) with a 13.6V linear power supply on 30 meters, and the unit's been transmitting every 10 minutes for about an hour. Also, this is local reception on a SDRplay RSP-1 in digital mode with AGC shut off. I think it's looking better than than without the 'thermal damper" added.

--Al
WD4AH
 

Dear AL,

That is much better and looks clean.? I am curious if your 12VDC supply on your amplifier is regulated.?? Also checking the? SWR of the input to the amplifier.? There is something in your setup that is still giving you some ripple from somewhere but the spacer on the crystal case has definitely made a difference.?? you signal is definitely very readable now.?? So will be looking for you on my
30M grabber and so will the others on the frequency.?

Good Job!
73 Dave WA5DJJ
 

I'm not saying that the thermal damping concept is imaginary, but I have to say that for me it has not brought the stability people have suggested it should, at least in remedying drift.? Wandering drift seems to be a mystery which can completely remedy itself without explanation.

Though certainly no expert, I suspect that your squiggly problem has to do with power.? Possibly something as simple as adding a filter capacitor would help.? I'll leave this to the more technically advanced.

I have not seen my signals picked up on any grabber (and probably never will, given the appalling propagation lately), so I can't judge what mine looks like.? I should really learn how to analyze my power situation with my 'scope in any case.
 

Dave and TrueBlue,
Thank you for your comments! Regrading the question on the power supply used with my U3S installation, the external supply is regulated. Earlier I mentioned it was homebrew, it was originally a Clegg model 011, probably rated at 7A. I've reconfigured it using the same transformer and fresh caps, but using a regulator board designed by KD1JV,??Current limiting is set up for about 3A service. Inside my U3S chassis is a small LM7805 based regulator board that's fed by 13.6V from the external supply. On top of that there's the 3.3V regulator on the Si5351A synthesiser board. So that's 3 levels of linear regulation.

When I run the U3S barefoot I'm seeing about the same ripple as what I get running with the QRP Labs' 5W PA. One thing I have failed to mention is I'm using a homebrewed Low Pass Filter built on plain perf board and it really shouldn't be a source of the problem, but it's a variable nonetheless.

I will try poking around with an oscilloscope on the power rails and maybe track down my squigglies.

It seems to me I might get better results using?the TCXO option on the synthesiser board or maybe try out the OCXO kit.

Thanks again!

--Al
 

As the problem is likely thermal, externally regulated power to the PA is also inside of the equation. In the U3S the finals are below both the MCU crystal and the synth crystal, so even slight variations in the raw input will make different power dissipation. Convection inside the case and draft will make this worse. Thermal damping will help, but eliminating all other variables will also make a difference. I would recommend all supply lines having local & individual (linear) regulators inside the case, that would mean 9V for the PA if you feed the box with ~12V. One additional reg for +Heat if using the OCXO. Thermally damping or heat spreading the finals will probably improve this as well.
The Fox TCXO has discrete steps in its frequency compensation and will likely be visible at 10m and above, say within 5Hz or so.
Qrp-labs OCXO works really well in my tests and I strongly recommend it if you want excellent stability in the U3S/ProgRock/VFO.


Virus-free.

 

On Tue, Dec 19, 2017 at 02:02 am, Daniel Ekman SA2KNG wrote:
.....In the U3S the finals are below both the MCU crystal and the synth crystal.......Convection inside the case and draft will make this worse.....
?
Good points! Maybe an intermediate approach would be to try a either a small shield above the final(s) or foam box covering the Si5351a board? I was thinking, some of the small ripple in my signal might be has from the microprocessor getting into the synth board given their proximity. Your point about a separate supply for the PA is not out of the question.?

I do use an improvised heat sink on the one BS170 FET in my rig...and it goes something like this:

However, if working with a QRP Labs sized enclosure, this design won't work.

Thanks!

--Al
 

Over the past couple of days I've tried a few things to help clean up my signal. First, I added a small cardboard "Quonset hut" over the Si5351a daughter-board in an attempt to dampen thermal issues. I also improved some of the internal wiring by shortening power supply cables and using miniature coax for the main RF out. I've also added some shielding to the underside of the synth board in hopes it might keep any hash being radiated from the main board out.

Pictured below is a locally received trace using Argo of my U3S (barefoot) running on a 12V battery.?I'm still a bit ragged looking, but I think this is going to rule out questions of the regulated supply. Ok, I still have the LM7805 regulator board installed.

In probing around yesterday with my oscilloscope I was surprised seeing so much RF on the 5V supply during transmit. I did what I could to keep it out of the power supply board, but monitoring the 5V line going to the GPS module it's rather pronounced(sorry no pix).

I'll probably keep trying a couple more things before the big event on NYE, but I think will run it with low power, 200mW.

Cheers!!

--Al
 

¿ªÔÆÌåÓý

Hi Al,

The edges look to me as being too rapid to be thermal drift, More like as you say RF / Noise induced on the supply rails.
(Assuming the receiver audio is clean & without earth loops...)

Have you seen Han's & Peters High precision OCXO U3S write up in the hints & kinks:-



(In particular, decoupling capacitors around the Voltage Regulators, mounted as close as you can to the device legs.
As well as hopefully decoupling RF, and general hash they prevent parasitic oscillation that affects regulators under some load conditions.)
Pick your favourite semiconductor manufactures data sheet, and check out their recommendations for values.
I also added an additional reservoir cap locally to the U3 for good measure.

Alan

On 20/12/2017 20:00, Al Holt wrote:

Over the past couple of days I've tried a few things to help clean up my signal. First, I added a small cardboard "Quonset hut" over the Si5351a daughter-board in an attempt to dampen thermal issues. I also improved some of the internal wiring by shortening power supply cables and using miniature coax for the main RF out. I've also added some shielding to the underside of the synth board in hopes it might keep any hash being radiated from the main board out.

Pictured below is a locally received trace using Argo of my U3S (barefoot) running on a 12V battery.?I'm still a bit ragged looking, but I think this is going to rule out questions of the regulated supply. Ok, I still have the LM7805 regulator board installed.

In probing around yesterday with my oscilloscope I was surprised seeing so much RF on the 5V supply during transmit. I did what I could to keep it out of the power supply board, but monitoring the 5V line going to the GPS module it's rather pronounced(sorry no pix).

I'll probably keep trying a couple more things before the big event on NYE, but I think will run it with low power, 200mW.

Cheers!!

--Al
 

Alan, that's a good article and I'll work towards that goal, thanks!

--Al
 

Dear Al,
Just for giggles.?? unplug your GPS module and see if that changes anything.?? Also put some .01MFD and .001 Capacitors on the input and output side of your 5 V DC regulator card.
Check the solder connections on the inductor on the final amplifier.?? That may have a resistive component that is causing the power to fluctuate.? Also resolder? the ground end of the bypass
capacitors (.1) around the U3S board.?? It may be a bypass issue somewhere on the board.?

Good luck.?? You will find it.??
73 Dave WA5DJJ

1 - 10 of 10