clean rf sampler

[ab7r]

I have a clean rf RFS-2K to use with my kpa500 and 7000dle skiing. It has an adjustable knob but cannot find anything about it.

Any experience with this for pure signal?

Thanks
Greg
AB7R

[w-u-2-o]

I had one. The adjustable nature of it makes it very risky to use for PureSignal. The adjustment range allows for power levels well in excess of what is safe for the ANAN hardware. You need a safe way to set it initially before connecting it to the ANAN. Then you need to make sure the adjustment never gets accidentally moved, and it is very easy to move.

Not recommended!

[ab7r]

Was only one I could find. Are there others?

[ab7r]

Belay my last. Found the vk-amps one and just ordered it.

[w-u-2-o]

VK has a good reputation as a quality unit.

[K1LSB]

I built my own ugly but effective pure resistive 18k-ohm sampler, usable everywhere from 40 watts to 1.5k watts.

Mark

Capture.jpg (561.25 KiB) Viewed 11736 times

[Gibo]

Bird Sampler works fine with adjustable attenuation from -40 to -70 dB
https://birdrf.com/Products/Test%20and% ... plers.aspx

[w-u-2-o]

Bird Sampler works fine with adjustable attenuation from -40 to -70 dB
https://birdrf.com/Products/Test%20and% ... plers.aspx

It better work good for $500!

[Gibo]

I was given two from a maritime receiving station that shut down and did not realize they were so expensive!!

[w9ac]

I built my own ugly but effective pure resistive 18k-ohm sampler, usable everywhere from 40 watts to 1.5k watts.

Mark

Mark,

At some point, you may want to consider replacing that large string of resistors with two 1KV SMD capacitors to yield a -50 dB coupling coefficient with a flat sampling passband. PS works in the same 40W to 1.5KW range so there's no need to bypass when using typical exciter-only power. Unlike resistors, the caps can't heat at any SWR being a near pure reactance. The cap V rating is sufficient if accidentally terminated into a high SWR at legal limit.

Paul, W9AC

[K1LSB]

Paul,

Thank you for that advice, I have thought about that but I'm not nearly savvy enough to know how to design such a circuit.

Mark

[w9ac]

Paul,

Thank you for that advice, I have thought about that but I'm not nearly savvy enough to know how to design such a circuit.

Mark

I've linked to a project on Oshpark in case you want to try one day. Oshpark receives all proceeds.

The parts BOM shows the required SMD caps. Additionally, small aluminum spacers, SO-239, coaxial "T," and BNC jacks are required. You get three boards for $8.65 and free U.S. shipping.

The circuit is open, so some means should be used to keep it out of harm's way or place it in a small plastic enclosure.

Paul, W9AC

https://oshpark.com/shared_projects/5RHcs2HH

[dj1yr]

With capacitive and resistive coupling, the influence of the terminating impedance on the attenuation must be taken into account.
In my opinion, a loose coupling in conjunction with a resistive ATT is the better solution.
I used the solutions up to 4m, with an attenuation of 50db +/-0.5db

[w9ac]

The termination Z at the PureSignal sampling port is at or near 50+j0. The PureSignal load does not (or should not) appreciably change as a function of frequency, even with changes in 50-ohm sampling cable length and nothing else is in the sample signal path. The capacitive sampler performs no differently than with my Xtronic or homebrewed toroid sampler.

Before construction, I ran a LTSpice response simulation taking the sample load into consideration. The worst-case response is on 160m but still, the in-band response is flat and that's what matters.

You are absolutely correct that the end-to-end passband response is flatter with loose inductive coupling, like a toroid. But that precision isn't required in this application like it is with a directional coupler. In fact, it matters little for PureSignal sampling because we care about flatness and group delay within the SSB transmit passband, not from say...1 through 200 MHz, for example. We're not dealing with sharp filter edges, so group delay isn't a concern.

Certainly, some folks will pay a premium just to have "the best." I'm guilty from time-to-time. So, for anyone wanting to spend additional money on a super nice sampler or commercial DC, I say go for it but at least try and consider the benefits and deficiencies of alternatives.

Paul, W9AC

[K1LSB]

Thank you very much, Paul!

Mark

[n1gp]

@W9AC

Hi Paul,

I saw this post and am interested in your capacitive sampler.
I was wondering how the attenuation was related to the capacitance values.

Ideally I'd like to build one that worked with 100 as well as 1000 watts.

So I thought I'd try to simulate your schematic. I've never run LTSpice before
but I have some idea of what I'm trying to accomplish and I got so far... but

I can't read the details of your signal generator but thought I got close.
When I RUN the simulation I just get a window with some ascii results, but no
graph like in your simulation.

I'll keep poking around and maybe find a way to get a graph, but if you can throw
some clues my way I'd appreciate it, but not a big deal.

Is there a file I can load in to LTspice?

Here's a screenshot of your spice simulation and my attempt...

cap_samp.png (758.8 KiB) Viewed 11323 times

[w9ac]

@W9AC

I saw this post and am interested in your capacitive sampler.
I was wondering how the attenuation was related to the capacitance values.

To the eye, C2 and C3 appear to be in series when looking at the PC board image although they're not. They're electrically in parallel and laid out that way for board symmetry. Schematically, C2/C3 are represented as one capacitor.

Nothing special is going on. The caps create a voltage divider with a resistive termination. The simple LTSpice simulation was done on the fly; I saw no need to save the file after the screen capture.

Paul, W9AC

[n1gp]

@w9ac

Paul, tnx for the suggestion. I hadn't heard of a capacitive solution
for an RF sampler before. I built one up and switch between a 6.8pf
and 22pf cap for less attenuation which works great with 5-10 watts.

[img]

cap_sampler.jpg

[/img]

[w9ac]

Nicely built using the recycled Pomona enclosure! Be mindful that the In/Out BNC jacks are good up to about 1200W into a 50-ohm resistive termination. The connector's 500V rating is a limit during high SWR conditions.

Paul, W9AC

[N4QNT]

The famous Coupler is finally back in stock. Dave started building them again. https://www.hamradio.com/detail.cfm?pid ... 8OSoATSwpv