EMRFD Message Archive 12213

Message Date From Subject
12213 2016-01-21 12:07:04 David J Nushardt What to use to sweep the 100 mhz Specan filter.
12214 2016-01-21 13:51:34 w7zoi Re: What to use to sweep the 100 mhz Specan filter.
Hello all,

This question seems to come up again and again.    There is an outstanding solution that is pretty simple.   It works well and if one has already built a power meter based upon the AD8307, is already half done.   This is the "SSG" that is found in the Files section of this Group.   It's the work of Bob Kopski (k3nhi) and John Lawson (k5irk.)    When in the files section, look under Bob's name.    Their name, SSG, stands for Simple Sweep Generator.

There are other solutions out there.    About any of them will work and can be implemented by experimenters.

The circuit by Kopski and Lawson is one of those gems that folks seem to have missed and that's really tragic.   It really is a very well done design that has been thoroughly tested and confirmed.

73, Wes
w7zoi

12215 2016-01-21 14:43:16 DuWayne Schmidlko... Re: What to use to sweep the 100 mhz Specan filter.
The easiest is with an SNA,  as you said the one from Midnight Solutions does not go high enough in frequency.
There are a couple different Chinese SNAs available on eBay. Difference is in the frequency range, but they use the same software.   I have two of them with different coverages.  One from .5 to 85 Mhz. the other from 35Mhz to over 4Ghz.
I posted a quick page on the first one on my blog, after I found links to the software and hardware.
You can down load the software at www.dl4jal.eu to take a look at it.

http://www.ebay.com/itm/35M-4-4G-USB-SMA-signal-source-signal-generator-simple-spectrum-analyzer-/171792212565?hash=item27ff9d2255:g:FA4AAOSwstxVV~NZ

http://www.ebay.com/itm/NWT70-frequency-sweep-meter-0-05-85-MHZ-with-0-to-50-db-attenuation-/121372362938?hash=item1c425b48ba:g:MboAAOSw-nZTq5V5

Another option is to use a noise generator and a TV dongle with SDR software.  There is a write up and links to several videos of the method at.
 
http://www.rtl-sdr.com/rtl-sdr-tutorial-measuring-filter-characteristics-and-antenna-vswr-with-an-rtl-sdr-and-noise-source/

http://www.ebay.com/itm/noise-source-Simple-spectrum-external-tracking-source-12v-power-supply-/181735993814?hash=item2a504f31d6:g:53AAAOSwstxVSMR2

It uses a cheap noise generator  and TV dongle along with  free  software.
There are a couple of different versions of spectrum analyzer software available that use the $12 tv dongles.
http://eartoearoak.com/software/rtlsdr-scanner     is one that I have tried.  Or you could use any one of the SDR software packages such as SDR#.  They give you limited bandwidth, only about 4Mhz max sweep , but should work for sweeping the bandpass filter.

DuWayne
KV4QB.blogspot.com





12216 2016-01-21 23:08:13 vasilyivanenko Re: What to use to sweep the 100 mhz Specan filter.
...."- a square wave signal has its energy distributed between the fundamental and harmonics which means that coupling the signal directly to the power meter we actually measure the whole available energy and not the fundamental's"......

Please also view crest factor:  http://ecmweb.com/power-quality/crest-factor-key-troubleshooting-parameter

....."- the previous lost energy could provide an explanation for the relatively high insertion loss that I measured since it would mean I did not measure only the insertion loss of the filter but also the loss of the harmonics' energy. I wonder how much of a difference a sine wave would make.".....
 
Beyond crest factor in your 50 ohm detector, and trigonometry based on a sine wave, basic filter characteristics / physics affect IL:

These include things like

-- Loss of filter components ; (resonator Q) As possible, ply air wound coils and air variable trimmers, or really high Q piston trimmers (or the like)
-- input and output port mismatch   (S11 and S22)
-- the 3 dB bandwidth you choose (while getting a single humped response)
-- proper tuning of each resonator with each port terminated in 50 ohms  ( Beyond tuning each resonator properly, I've helped builders perfect filters where they chose a trimmer cap that did not give enough C at full mesh in order to tune the resonator ). Also need to account for stray C.

A delicate dance ensues. We must balance the resonator coupling needed to get a single filter peak without too much coupling losses -- while getting a decent 3 dB, or half-power bandwidth --- all while reasonably matching the input and  output ports  --------- to ultimately get a reasonable insertion loss for our particular filter application or situation. Sometimes we'll sacrifice some IL to get steeper filter skirts with a little under-coupling and then just boost the filtered signal with a LNA.

I've got some related VHF filter experiments / notes in my archive from when I began studying VHF/UHF filtration:

QRPHB archive

http://qrp-popcorn.blogspot.ca/p/blog-page_9.html
[From menu]

Topics 2012-2014
VHF — Veronica
6. VHF Band-pass Filter Experiments

Sadly, the sweep images are missing. They were very nice when I achieved the best coupling and port matching for my situation . The Input / output match were tested with a return loss bridge and I got >= 20 dB return loss by tweaking the series input/output trimmer caps in the final filter. Now, at VHF, I usually match the ports by changing the input/output tap points on the inductors.

Best!

T/ V
12217 2016-01-22 18:56:45 David J Nushardt Re: What to use to sweep the 100 mhz Specan filter.

Hi Everyone, Many thanks to Wes and all who replied, I already learned a grate deal, When I Looked at Bob Kopski's work and saw Ramp generator A lightbulb went off , I have a pair of Tektronix RG501 ramp generators on a shelf above my bench, I completely forgot about, it seems a ramp generator is what's needed  to make something sweep, provided its varactor or voltage controlled.

Just before I made my post, I received a  very inexpensive Chinese  VCO , offhand I think it's 80 -180mhz. For about 10 bucks on ebay.

I'm thinking I can use my RG501 to drive the VCO and make it sweep from 80- 180 mhz. provided the level is enough.

Should work with mini circuit's POS - 200 as well.

I was confused by Bobs SSG Schematic  there is a quad op amp that doesn't seem to have a label. I take it that it is a LM324 , also there is a negative 2volt supply  that I dont see what's generating it or what it's used for. The ramp gen and the rf schematics are paged so I could be missing something.

Anyway I have a multitude of ideas  on rf sweeping now , but the question now is how to best
display the results.

I sure like the Midnight Solutions LCD screen  to display the results , pretty slick! and inexpensive but far beyond my ability to program, I do have a LCD screen, a Arduino UNO  and several DDS's.

The best solutions seem to be based on what one has on hand to work with , what's the cheapest, and simplest.

So no one answer for everyone.

The AD8906  for power meter is the obvious choice , but I do prefer the digital direct reading versions

I have a multitude of PIC Programming devices, but unfortunately not a pickit 3 yet.

I don't know which is easier to program the PIC or Arduino, I have yet to program either and usually get hung up with either.

Still wonder what the bottle neck is with the Midnight solutions SNA,  Seems like the SI DDS is capable of going higher?

73's
Dave

Sent from my Kindle fire HDX 8.9



12218 2016-01-23 06:32:44 AncelB Re: What to use to sweep the 100 mhz Specan filter.
VCO's can generate RF sweeps from any stable source of ramp, even the
lowly LM324 can probably adequately supply the sawtooth ramp wave form
once a stable voltage source is available.
YIG VCO's are a noted means of UHF freq. gen. with good spectral
qualities. Chinese VCO's (cheap) probably have a lot of harmonics and
are often not pure sine wave generators. They'd need characterizing with
a Spect. Analyzer for their lack of flatness and any significant RF
resonances due to use of lower quality components.

I work with PIC processors rather than arduino as I prefer the tighter
interfacing achievable with custom builds. Arduino is simpler and
bulkier from a modular & rapid prototyping of view due to the large
'prebuilt' module market. Arduino is not so great for signal integrities
(S/N ratio) due to the convoluted, elongated, mechanical path for
signals passing through add on modules.

I can help with any PIC programming/LCD interfacing etc as my interest
in the RF and EMRFD text is ramping up.

Ancel
12219 2016-01-23 10:21:05 vasilyivanenko Re: What to use to sweep the 100 mhz Specan filter.
hi Dave:  Do you own a ‘scope / DSO?   For a sweeper like Bobs (analog) – that’s “ the screen ”.
I’ve built Bob’s SSG – works as advertised. Precision calibration available.   Yes, it’s an LM324.
Bob’s  negative DC voltage generator = a multivibrator chopping an audio transformer
I’ve got it at the bottom of this page: DC-DC Converters

 

12220 2016-01-23 10:56:29 David J Nushardt Re: What to use to sweep the 100 mhz Specan filter.
Yes, I have 3 scopes but unfortunately there all analog no DSO's Tek 2246 and Tek 7603, Tek SC 502 there all 100mhz bandwith or less .

However If I understand correctly we will be using it in XY Mode so bandwith shouldn't matter.

Your. documentation is awesome.

Many thanks.
That really clears things up.

73
Dave

Sent from my Kindle fire HDX 8.9
12221 2016-01-23 12:48:39 kb1gmx Re: What to use to sweep the 100 mhz Specan filter.
Why not use the sepcan to adjust the filter!

Install the filter and put a fixed signal at the specan input and sweep it.  The 100mhz filter 
is then adjusted for min loss and best shape.


Allison
12222 2016-01-23 15:28:46 w7zoi Re: What to use to sweep the 100 mhz Specan filter.
Hi Allison and gang,

As is usually the case, you are exactly correct.    The spectrum analyzer is a self aligning instrument.    For this filter, there is no need to build a sweeper and certainly no need to order one from the other side of the world.

As you suggested, all that is needed is to put a signal into the front end and then adjust the VHF filter capacitors for maximum response.    The filter response you will see on the screen is not that of the VHF filter, at least with this spectrum analyzer.   Rather, we will see the shape of the 10 MHz filter that is in place at the time of measurement.    You could bypass the 10 MHz resolution filter and then see the shape of the 110 MHz filter.

There is a central theme here and that is to sit back and think about what is happening within the spectrum analyzer.    Then do what you can to take advantage of it with whatever you have around the shack at the time.    If you don't have a generator around for the purpose, build one.   It need not be elaborate.

The spectrum analyzer already includes a time base and that could be used with an external power meter and the VCO from the analyzer to generate a sweep of the VHF filter.    This is not needed to build the analyzer, but would still be a fun experiment.

Many thanks Allison for pointing out that the job can be done with what we already have on hand.   Thanks for all of your other insights that you post.

73, Wes
w7zoi

12225 2016-01-24 10:36:52 kb1gmx Re: What to use to sweep the 100 mhz Specan filter.
Back when I cheated.

Signal generator (really crappy one), 3 db pad, filter, 3db pad, and a simple diode detector.

The siggen had enough oomph to get down maybe 40 db from peak before I could not see 
it on the meter.  It was enough to start and then touch up in place.

My first specan was a simple osc swept using a 555 timer with a constant current source 
for linear sweep.  That drive the H input of the scope and the Opamp for the OSC Varactor.
IF was ua703s (three) with a diode detector on each output summed made a fair log
detector.  For the 10mhz IF filter a 30khzfrom a Aerotron VHF and a second filter for 
about 5khz using multiple of those filters cut open and hacked.  the HF filter was 
based on a VHF front end from a motrac (about 155mhz). 

The scope was a home brew solid state using a 3AP1 with maybe a 3mhz bandwdth.

It was an ugly affair but useful back around the mid 70s.

At that time the only instrument I had was a 350mhz frequency counter.


Allison
12226 2016-01-24 11:18:56 David J Nushardt Re: What to use to sweep the 100 mhz Specan filter.

Interesting, but getting back to your last reply,I wasn't sure if you were  awhere of  where I am at in the construction ,scheme of things.

Then when Wes chimed in , I thought about it awhile and saw , what you were getting at, yeah come to think of it the time base is a  ramp generator and the VCO is the RF generator.

So yeah I guess one could use elements of the kit to alighn  the filter, So why isn't this in the notes, or is this a brand new realization.

What really throws me for a loop is  self aligning really!  What's all this hoopla in the updates and EMRFD about coupling and filter shape?

Wes leads use down a path , then takes use in a circle.

I had two strokes for peat's sake , give me a break ,Yeah I want to learn something, but I just want to build the darned thing!

Hopefully were all learning something.

73's  Dave

Sent from my Kindle fire HDX 8.9



12234 2016-01-25 07:17:33 Adrian Scripcă Re: What to use to sweep the 100 mhz Specan filter.
Glad to see this thread taking off in a nice direction!

Wes, SSG seems like a nice tool but in this particular case we could only use its ramp generator, no? Besides that we would need to build a VCO capable to span the required spectrum. And since we are in the context of building the SA, can't we simply make use of the project's timebase and perhaps its 1st oscillator, granted it goes low enough.

Todd/Vasily, thanks for the suplements. It now makes me consider the filter matching. Whilst the powermeter has a 51 ohm resister on the input circuit, I just used the Si5351 CLOCK0 output as-is from the NT7S board and plugged it in the filter. Perhaps a 3dB resistive pad between the Si5351 output and filter would help presenting a better match.

I tried creating new pcb walls soldered on the base plate and re-measured the whole thing. While I did get about 10-12 dB improvement in out of band attenuation compared to the first "sweep", I couldn't really solder the lid in a satisfactory fashion so I ordered some 1mm copper sheet to build the walls and shield and lid. I have the feeling that this will do all the difference in the world and I can't wait to come back with a full report on that.

Farhan, while sweeperino is a nice tool for getting an accurate digital image of a filter's response, it takes quite a lot of time to run a sweep which makes it a bit harder to see the tuning effect in real time; for this task, a higher sweep rate is more beneficial, such as the analogue mentioned solutions. Having something like this would enable one to perform the insight gaining experiments that Wes mentioned above.

73 de Adrian YO6SSW

12235 2016-01-25 07:21:06 Adrian Scripcă Re: What to use to sweep the 100 mhz Specan filter.
Sorry, I kept the mail in draft for the weekend. It seems like I am repeating what others just said, hi