[time-nuts] Leakage, tinySA

[Bob kb8tq]

Hi

> On Jan 16, 2021, at 5:32 AM, Hal Murray <hmurray at megapathdsl.net> wrote:
> 
> 
> My tinySA arrived a  few days ago.  I'm happy.
> 
> It comes in a nice box with an antenna that extends to a foot, a USB charging 
> cable (mini, not micro) and a couple of cables.
> 
> It doesn't come with a manual.  I haven't found a manual online.  Their web 
> site is pretty good, but sometimes I like a document organized to be read by 
> turning pages rather than following links.
> 
> The minimum receive bandwidth is 3.1 kHz.
> 
> The specs say the lower frequency limit is 100 kHz but the UI doesn't enforce 
> that.
> 
> My strongest signal is at 30 MHz.  It's -60 to -80 dBm as I wave it around.  
> At 100 kHz bandwidth, the noise floor is -110 dBm.  The 30 MHz blip is clean 
> enough to see the shape of the receive bandwidth filter.  (I'm assuming  the 
> raw signal is much narrower than 3 kHz.)
> 
> I haven't figured out where it is coming from.  At first, I thought it was 
> coming from low power do-it-yourself PCs.
>  https://www.mini-box.com/
> If I get near them, lots of spurs near 30 MHz come up out of the noise.  But 
> powering them off doesn't change the 30 MHz.
> 
> The 30 MHz is clean.  I'm guessing it's the 3rd harmonic of 10 MHz.  I only 
> have a few of them, but I don't like turning them off.  Or maybe the 2nd 
> harmonic from the Lucent boxes.  ??
> 
> The 10 MHz is stronger near coax carrying 10 MHz and next to a Z3801A.

First step is to pull the cable and put a 50 ohm load on the device ( 3801
in this case ). The radiation may not go down …. If it does go down, next
step is a clamp on ferrite on the coax. “Anything that will fit” is the first 
choice there ….

>  I'll 
> have to try some good cables.  Anybody have a favorite source?  Is there a 
> magic word?  Do I have to specify the type of coax?

If you don’t mind spending “many tens of dollars a foot” there are lots of choices 
out there. They often come with strange mating connectors so the cable
is only the start of the cash outlay.   

I would draw the line at LMR-400.

https://www.pasternack.com/50-ohm-low-loss-flexible-lmr400-pe-jacket-double-shielded-black-lmr-400-P.aspx <https://www.pasternack.com/50-ohm-low-loss-flexible-lmr400-pe-jacket-double-shielded-black-lmr-400-P.aspx>

Note that the term “flexible” is more in comparison to “rigid metal jacket” than 
what you would normally expect on a test cable. 

If you do want to go a bit more fancy than LMR400, stick with eBay and ten cents on 
the dollar pricing… Connectors may be a bit worn, but the price is right. 

https://www.ebay.com/itm/Megaphase-SF4-S1S5-96T-DC-to-4-GHz-50-96-SMA-to-SMA-R-Angle-RF-Test-Cable/231773017446?hash=item35f6bf9566:g:GU0AAOSwv-NWYJ3S <https://www.ebay.com/itm/Megaphase-SF4-S1S5-96T-DC-to-4-GHz-50-96-SMA-to-SMA-R-Angle-RF-Test-Cable/231773017446?hash=item35f6bf9566:g:GU0AAOSwv-NWYJ3S>

This is simply what is out there and seems to work pretty well. It is not the best of the
best by any means. It *is* something you *can* find on eBay. 

If your application is one that rigid cable ( = solid metal jacket) will work, that 
probably is a better bet. A thicker metal jacket is often a better jacket at 10 MHz. 

> 
> I have lots of Gigabit Ethernet.  I have a blob at 125 MHz, and spurs at 120 
> MHz and 130 MHz.  120 has a little brother 18 kHz higher.

Same trick with snap on ferrite cores applies here as well.  If the core is big
enough, putting one or more turns through the core may help. Core size
vs cable size is what will limit you there …

Fun !!!

Bob

> 
> The normal power up mode scans 0 to 350 MHz with enough attenuation so that 
> most of my signals are lost in the noise.  Sometimes it comes up without the 
> attenuation.  Then I can see all sorts of stuff.  The biggest ones are at 240 
> and 288 MHz.
> 
> I think there is a mode for overlaying displays.  I haven't investigated.
> 
> 
> Their web page has links to software to talk to it over the USB cable.  I 
> haven't investigated.  Plugging in the USB cable adds lots of cruft.  The 
> software would be unusable unless you are looking at a strong signal or at 
> something in a hole that the cruft misses.  It might work if you can freeze 
> the display, then plug in the USB cable to capture it.
> 
> 
> 
> -- 
> These are my opinions.  I hate spam.
> 
> 
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.