The AM Forum

Well, the holiday season is getting near. Time to think about purchasing a Christmas gift for the shack. What is a good spectrum analyzer for the shack? Something affordable and can measure up to 30MHz or so.

At least on paper looks like a fun toy and might be of some help searching the neighborhood for RFI.

Craig,

Up to 30 Mhz wouldn't be all that useful. You need to see the harmonics when testing a transmitter.

The old HP 141T series are now pretty inexpensive. I have one with plug in RF sections up to 1.5 Ghz (I have the highest frequency module up to 20 or so Ghz but it needs some work and I have not had a use for it.)

The  module I use the most is the 110 Mhz module for HF stuff. I bought a 141T with three modules and the B series IF unit a couple of years ago for $600 on EBAY with a new CRT. I have a couple of mainframes and spare modules also. I picked up a tracking generator also.

There is a 110 Mhz 141T tested unit on EBAY buy it now for $429.

The more modern HP portable units are still $2500 or more used but are coming down. You might find one that can be repaired for less.

Soon the HPSDR guys will offer a 1/2 watt transceiver module. The spectrum display will be as accurate as any HP device and usable through 6 meters.
And you will end up with one heck of a rig. I bet it will sell for under $400.
Even a used SDR1000 would make a great spectrum analyzer. The bandwidth will be limited but still usable.

Coming out December/January, Flex 1500 for $579:
http://www.flex-radio.com/Products.aspx?topic=F1.5k_features

(http://www.flex-radio.com/Data/Image/F1500_dimetric_view_small.png)

If you are looking for a panadapter to look at signals on the band - then you don't need a Spectrum Analyzer.

If you are looking for a true piece of test equipment to test transmitters , oscillators and design filters and RF equipment, then you need a Spectrum Analyzer. The device above would be worthless for what I use my spectrum analyzer for.

Pat
N4LTA

Craig,

I use a Tektronix 7L13 that plugs into my Tektronix 7000 series storage mainframes.  Equipment like this can be found at reasonable prices since it is considered vintage at this point in time.  It is very useful for determining signal purity in its wideband mode and in the narrow sweep modes it makes alignment of the vintage phasing type gear an easy task.

The "front end" of a spectrum analyzer is easily damaged so whichever one you acquire you need to be very careful when testing transmitters/amplifiers. 

Rodger WQ9E

Some of the sdr stuff is more accurite than the old spectrum analizers.
They are not the toy screens like what is on the icom 756 pro stuff.

http://www.rfspace.com/SpectraVue.html
http://www.software-radio-laboratory.com/

Look at the displays, just like lab equipment, maybe better, better then my equipment...


I just sold a nice tektronics digital spec an that went to 23 GHz for $400.00.
I listed it here and did not get any bites.

I still have the nice HP one I use to test LO's and so on, it also goes up to 23 GHz but had more options and resolution than the tektronics. You can watch warm up drift with a span of 200 Hz across the screen, dual markers, etc.
Very handy to test freq range, output and stability of homebrew LO's, along with transmitter stuff, and locating the correct satellite when pointing dish antenna's.
The sdr stuff makes it look very dated with its green screen though....

Brett

HP8566 is still one heck of an SA and now you can replace the DRT with an LCD screen. We had one done at Tucker that works very well.

RE: Flex series being used as a spectrum ana.  The input is supposed to be calibrated for 0 dBm.  Has anyone ever checked to see how accurate it is?  It could double as a spec ana and also a darn good RX. 

I agree that some of the older SA's probably wouldn't be as good spec-wise as the newer SA's.  I used to repair the 8566's and have a comment -- it's a darn good SA but I wouldn't want to one as they are a good hernia inducer and if anything should go bad in 'em they are no longer supported by Agilent (in their end-of-life phase).  the Yig tuned filter alone is around $2K.  A better choice might be a 8568 but they are still way too much for the average ham use.

Al

Also, I dont think the flex 1500 is going to be a great receiver, you only get 48 KHz, and the specs will likely be even lower than the flex 3000. The sdr-iq might be better, at lower cost, but without any transmitter....

Brett

As far as the accuricy of the flex goes, all the lab tests say how good it is, it gives an S meter reading and a dbm readout.
The panadaptor is calibrated in dbm.

Brett

HPSDR with power SDR tracks an HP8640B line on line. And you can see phase noise on the display right down to -140 dBM.

How accurate is the internal attenuator? How wide is the bandwidth - can I put a 50 Mhz signal on it and see the 20th harmonic? Can I use it with a tracking generator? How well does it work with my microwave local oscillator when I set up my microstrip filters?

Probably a good panadapter or even a great one.

A spectrum analyzer it is not.

I only threw it up there based on the low price. The 5000 probably would run rings around it as a test instrument.

Don't confuse frequency coverage with whether something is a spectrum analyzer or not. I've used "real" spectrum analyzers made by Hewlett Packard that only worked up to 50 or 100 kHz.

If someone only needs HF coverage, then some of the SDRs are probably fine. Since they are FFT based, they will have far better close-in resolution than any swept-tuner analyzer like the 141. Since FFT analyzers aren't swept, a tracking generator is not needed - any sweeping generator will do (in many cases you could probably sweep it manually).

It all depends on your requirements.

Not confusing anything. I just don't think a $400 SDR is going to substitute for a decent spectrum analyzer.

My bet is when it comes out - it will be worth about what you would expect a $400 receiver to be.

It's a receiver. Does the manufacturer have the software to make it function as a tracking generator or a sweeper? Does the manufacturer even care to write that software.

If I was going to spend $400 for a digital FFT SA , I'd build Scotty's SA or build a the popular VNA kit.

If I was going to buy a SDR receiver - I wouldn't buy a Spectrum Analyser.

That's just me - my opinion.... probably slightly molded from listing to crappy sounding IPOD music and lousy "CD" quality music.

Well, if you look at the fr space web page, nasa actualy uses the sdr-iq to track stuff, maybe bacause you can also record the entire spectrum for hours, along with watching it.
They list it as a spectrum analyser.
It goes 500 CYCLES to above 30 mhz I think, an mamazing little box for $400.00.
The hp 8592l I have is $13,000.00 used....and it does not make a good receiver....

Brett

Dont forget about USB based Spec Ans and Oscopes.  Search Ebay for USB Spectrum analyzer.  There are several models out now that are affordable and have multiple channels and up to 200hz

You can make or buy a coupler to intercept your Coax to provide a BNC right to the USB analyzer. Plug and play.

I have a Big 100lb Spec An. I use it less and less.

Clark

So many choices and so little time to try them all. Gotta do more research! The SDR-IQ seems quite a nifty little device. I don't think it is repairable, sort of a throw away item, but for the price not bad.

I doubt a spectrum analyzer in my shack would see much duty in HBing or trouble shooting. It's primary use would be looking at the HF bands. Would be nice to stay away from surface mount models, not repairable here. Lots of RFI in the neighborhood! Something like the SDR-IQ is a plus because it is portable and could be taken in the car to sniff about. Although, things would really have to be bad before knocking on doors.

Time to surf the web and research.

Tnx for all the tips.

Craig

At one point I had four 141T systems here just to keep one or two working. After awhile I found most of the common failure modes, fixed everything and sold off 2. With a much later SA now on the bench I moved a 141T into the shack and the other sits on a shelf; both have the 110 and 1215 mHz plug-ins. For now the active one is used in the SSB/CW shack and mostly on VHF/UHF where it has been a big help finding "rotten signals" and making sure mine are OK.  Once the AM shack gets reorganized I'll put the other 141T in there. They are an OK SA for that use where their dynamic range and other vintage design shortcomings arent an issue.

Carl
KM1H

The sdr-iq is a very nifty little box.
I think its quite rugged, I never blew mine up with 700 watts carrier floating around in the shack.
Its a bit clunky as a receiver, mostly because it did not have band buttons, or band stacking registers, you always had to adjust the frequency in steps.

With the flex stuff, you can click on 80, 40, 160, and its there at the AM window, if that is where you left it...

As a receiver, its quite good, but no sync detector when I had it, they have new software out now I think, but i dont know what improvements it has.

I sold mine for $300.00, but should have kept it as it was so small and powered by the usb port, handy for portable stuff... its like pack of smokes size....

Brett

Future HPSDR boards will include a SA/VNA so it will have a tracking source function.
Flex stuff is yesterday. DDS is dirty, RX dynamic range is only around 97 db. I'm surprised Flex continues to use the Tayloe driven by a DDS. Way too much hardware for marginal performance. Heck Power SDR has been modified to deal with HPSDR so it would be easy to modify for new Flex hardware.
I bet Covington could put them out of business if he offered a transceiver. Rob Sherwood would have a ball testing that kind of performance.

Yes, but a lot of stuff has been 'in the works' for years and still no product other than a bare receiver with no front end.
If you want a rig to use today, or last year, or next year, the flex 5000 is about it.

3rd on the Sherwood list is not bad, and the the transmit audio is better than the two above it...

Brett

Well that is why I have a Viking 2 and TR7A.
HPSDR also has an exciter so I am on with 300 watts PEP or 50 watts am (modified MRI Amp) . IP3 is better than -50 dB out of the exciter and dynamic range about 120 dB. I'll wait. Reference/ GPS interface shoud be in the mail in a week or so. Preselector is slated for the end of the year.  Sure it is a bit slow but a lot cheaper than a Flex. 

Is it that much cheaper?
$1500 for a flex 3000, $2600 for a 5000, the icom 756 pro cost more...

Yes, its an old design, I wish they would change, but they seem to be branching out into govt work or something, they built a small box with about 13 seperate receivers in it, and I think its the direct sample method.

I had a laugh looking at the chassis on the hpsdr web page, very experimental looking, I think it looks cooler with the boards all hanging out in the open. That looks very ham radio.
The flex you want to hide someplace as its nothing to look at, but its very nice inside.
The entire box is a heat sink with a big (6 inch?) fan...

Brett

I think it was a flex plan to use hams as testers so they could get into the mil. market.
HPSDR all interconnects are external out the back. I put a switch and neon bulb on the front panel but the chassis is turned around backwards right now.
I mounted Lamda linear supplies in my chassis.
Direct sample adding receivers is just more softtware.
It seems a waste to use a Tayloe/DDS when it could be replaced with an A/D and FPGA doing DDC

Pretty much right Francois.
..and wait 'till the beast comes along that runs 24 bit with no overload until 100 dbm with a clock running 10Ghz.  It's only a matter of time.

Phil's board is now 16 bit, 9 to 10dbm, 125Mhz.
Persus is 14 bit.

These h'ain't no sound card limited ADC's.

But I don't look forward to having to endure that horrible music they will be playing over and over and over and over in every retail outlet for the next 6 weeks.

Bah humbug!

it's meant to be listened to after drinking the boozed-up egg nog.

Well Don, I was doing just fine until being reminded about the music.

Thanks ;D

I find that having to listen to that crap on the music radio stations I usually like to be even worse; especially without a remote to kill the audio.

The alternative of listening to talk or sports radio all that time is almost as bad ::)

Carl
KM1H

What's wrong with talk radio?  I MUCH prefer listening to the talk radio that happens from 1.8 mhz to 30 mhz to ANYTHING below 1.8, tho :)  That talk radio is just MUSH.

--Shaen

I agree Shane.  I like talk radio. I am listening to talk radio right now on 40 meters. 


Clark

Rick,
QSR1 and HPSDR use the same A/D but HPSDR used the preamp that came later. This dropped the MDS lower than QSR1 but also the saturation point by 20 dB. HPSDR has a relay controlled 20 db pad at the input to shift the operating range equal to QSR1. Perseus came first and the 14 bit A/D was the hot rod part then.
LT has a 170 MHz 16 bit A/D and hope to make a 250 MHz part in the future. The 170 MHz part is pin for pin with the  130 MHz part but then the FPGA needs to be faster to support the data flow. I can just imagine the parts the mil guy are using in those nice gold plated hybrids.

Frank,

Thanks for info. I wonder when Phil will 'finish' (if ever) his QS1R with the promised RFFE (front end board with two stage preamp, bandpass filters, and attenuators along with add'l control buttons on some version of SDRMax)?

No definite replies to several others' queries about this on the Yahoo board. He's probably bogged down just keeping up with QS1R orders.

Rick,
The problem is cost. HPSDR at TAPR is having a real problem keeping the cost down for two boards for the TX LP and RX HP. All those little relays run into some real money. Then there is the 30 or so inductors required. I think it will cost as much as the RX and TX boards combined. I was hoping for a tighter preselector but that would have even cost more. 

Other than being a real boatanchor and a cause of hernias, anyone played with a HP 8568B?  Still window shopping.

Craig,

Used them and the sister unit 8566 for many years. These units were probably the peak in analog based swept-tuned spectrum analyzers.

I love my HP8590 series.

Try this one:  http://www.science-workshop.com/  the poor mans spectrum analyzer. 

Kit built and reasonably priced. Seems like a good deal for a brand new spec analyzer although it won't have the bells and whistles a HP or Rhoades  & Schwarts will.

Not affiliated or anything etc etc.