EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: Legion on April 11, 2014, 04:21:09 pm
-
I've noticed most scopes top out in the hundreds of MHz and a few get into the single digit GHz range. So what do people use when they're trying to analyze a 20+GHz signal?
-
A sampling oscilloscope?
http://www.tek.com/oscilloscope/dsa8300-sampling-oscilloscope (http://www.tek.com/oscilloscope/dsa8300-sampling-oscilloscope)
-
Never heard of one of those before. Very cool. Though I guess that means high bandwidth applications are outside the range of the hobbyist.
-
People that analyze 20GHz signals don't use "most scopes" :)
There are DSOs available from several manufacturers that do 30, 60, even 100GHz. Prices are equally impressive of course...
Spectrum analyzers and old sampling scopes may be an option, depends on what you mean by "analyze" and what kind of signals.
-
Some use such a device to look what on I2C https://www.eevblog.com/forum/chat/is-this-overkill/ (https://www.eevblog.com/forum/chat/is-this-overkill/)
-
I used to work with Monolithic Microwave Integrated Circuits (MMICs), it was 25 years ago so the frequencies tended to top out at around 18GHz.
We never used scopes at all. Most measurements were done on vector network analysers - mainly the HP8510. These systems were very expensive at the time, like 50 times my annual salary as a young engineer so in todays equivalent £1M+!!
Now-a-days VNAs are much cheaper, only tens of thousands I think.
-
There are realtime scopes that extend that high, and more. For example, Tek's 70K series offers 4, 6, 8, 12.5, 16, 20, 23, 25 and 33GHz bandwidths. Agilent and LeCroy have models that run over 60GHz realtime. Tek will show 70GHz later this year.
Equivalent time sampling scopes, like the DSA8300 are a good choice when the signals are repetitive in nature. They sample at a lower rate, but with very precise timing with respect to the trigger, thus can build up a waveform over many triggers with very high bandwidth.
And yes, this type of performance is generally outside of the reach of most hobbyists. Although, these equivalent time sampling techniques have been around for many, many years - since the late sixties, and can likely be found in the used market.
-
Yeah, if you're doing something up there, you're better off with a spec or VNA. 'Tis how the engineers of olde did microwave work back when tubes were barely hitting VHF.
Tim
-
If you want to analyse the signal in the frequency domain, the modern $$$ lab method would be to use something like an Agilent PSA or PXA spectrum/signal analyser. These are available in models that go to >40GHz but they are very expensive.
However, if you just want to measure the power with reasonable accuracy then there are various USB power sensors available that work to >60GHz.
If you want to (critically) look at the spectral purity of the signal (eg close to carrier phase noise) then use something like an E5052 signal source analyser as these are available with options to cover way over 20GHz. These are also very expensive...
Generally, VNAs are used to analyser networks rather than signals but you can use some of them for measuring signal power fairly accurately.
If you want to measure this sort of signal at home then I guess there are old sampling scopes available but you can also buy something like an old 1970s or 1980s 22GHz HP spectrum analyser for anything from £500 to £2000 depending on the performance spec.
-
Just to be clear, I have no need for this type of equipment but I was curious to know what people use when they do.
-
Just to be clear, I have no need for this type of equipment but I was curious to know what people use when they do.
Most of the folks I know that are using these ultra wideband scopes aren't doing RF. They're either doing fast high speed serial data signaling, or high energy physics. Those that are doing RF with these scopes, are guys that are doing wideband modulation such as 500MS/s+ QAM type modulations for satcom, or wideband chirped radar applications, where even the best VSAs don't have the acquisition BW needed.
-
In them old days we used to "Rent" equipment. Test Equity I believe is still around. I remember we had a DSO for about a couple weeks or a month one time. I think it cost us about a thousand dollars to rent and a few hundred each way for shipping. This is over 20 years ago. We used the hell out of that thing cause we could not afford to rent it again! I remember getting that catalog and drooling over all the RF stuff.
-
Good example where fast realtime oscilloscopes are used: https://www.eevblog.com/forum/testgear/video-demo-experiments-with-the-world%27s-fastest-oscilloscope (https://www.eevblog.com/forum/testgear/video-demo-experiments-with-the-world%27s-fastest-oscilloscope)!-%28the- signal-path/
-
Good example where fast realtime oscilloscopes are used: https://www.eevblog.com/forum/testgear/video-demo-experiments-with-the-world%27s-fastest-oscilloscope (https://www.eevblog.com/forum/testgear/video-demo-experiments-with-the-world%27s-fastest-oscilloscope)!-%28the- signal-path/
Here is the link:
http://thesignalpath.com/blogs/2013/08/13/experiments-and-demo-of-an-agilent-dsa-x-96204q-160gss-62ghz-oscilloscope/ (http://thesignalpath.com/blogs/2013/08/13/experiments-and-demo-of-an-agilent-dsa-x-96204q-160gss-62ghz-oscilloscope/)
-
In them old days we used to "Rent" equipment. Test Equity I believe is still around. I remember we had a DSO for about a couple weeks or a month one time. I think it cost us about a thousand dollars to rent and a few hundred each way for shipping. This is over 20 years ago. We used the hell out of that thing cause we could not afford to rent it again! I remember getting that catalog and drooling over all the RF stuff.
http://www.trs-rentelco.com/ (http://www.trs-rentelco.com/) is still around, I had a bunch of friends that work there when I lived in TX and never took advantage of that at all. But now I live in Illinois.
-
In them old days we used to "Rent" equipment. Test Equity I believe is still around. I remember we had a DSO for about a couple weeks or a month one time. I think it cost us about a thousand dollars to rent and a few hundred each way for shipping. This is over 20 years ago. We used the hell out of that thing cause we could not afford to rent it again! I remember getting that catalog and drooling over all the RF stuff.
http://www.trs-rentelco.com/ (http://www.trs-rentelco.com/) is still around, I had a bunch of friends that work there when I lived in TX and never took advantage of that at all. But now I live in Illinois.
Where in TX were you? Not much in the Houston region...
-
Houston used to be a huge market with nasa and all there. Also the oil industry.
-
In them old days we used to "Rent" equipment. Test Equity I believe is still around. I remember we had a DSO for about a couple weeks or a month one time. I think it cost us about a thousand dollars to rent and a few hundred each way for shipping. This is over 20 years ago. We used the hell out of that thing cause we could not afford to rent it again! I remember getting that catalog and drooling over all the RF stuff.
http://www.trs-rentelco.com/ (http://www.trs-rentelco.com/) is still around, I had a bunch of friends that work there when I lived in TX and never took advantage of that at all. But now I live in Illinois.
Where in TX were you? Not much in the Houston region...
Southlake TX, so North of DFW (Dallas Forth Worth), trs-rentelco are located in Grapevine TX.
-
There is still a large rental market in the UK, and thanks to tax advantages, a large leasing market too.
-
Houston used to be a huge market with nasa and all there. Also the oil industry.
Not much in the oil industry. most of the actual hardware is no longer in Houston. Can't find any NASA auctions on GovLiquidation.
-
Old sampling oscilloscopes are one option. Mine goes to 14 GHz and just knowing that number, those who are familiar with them will identify it as some Tektronix variant using an S-4 sampling head.
With some design work you can design and build one good up to at least 4 GHz. I am tempted to do so using 300 MHz 2N3904 transistors in the sampling bridge just to see how well it would work.
There are USB based sampling oscilloscopes available including a recent Kickstarter project which failed to achieve funding.
http://www.fastsampling.com/index.php (http://www.fastsampling.com/index.php)
http://www.picotech.com/picoscope9000.html (http://www.picotech.com/picoscope9000.html)
-
I've noticed most scopes top out in the hundreds of MHz and a few get into the single digit GHz range. So what do people use when they're trying to analyze a 20+GHz signal?
This is a level of scopes that your average human will never touch nor be able to afford.
For starters, the probes alone in the 20GHz+ range can cost more than a new car, let alone the scope, which can cost the same as a house.
-
Also the oil industry.
I worked in the oil survey industry for over a decade, and I can't recall anyone who needed such a high end scope.
What aspect of the oil industry?
-
I've noticed most scopes top out in the hundreds of MHz and a few get into the single digit GHz range. So what do people use when they're trying to analyze a 20+GHz signal?
This is a level of scopes that your average human will never touch nor be able to afford.
For starters, the probes alone in the 20GHz+ range can cost more than a new car, let alone the scope, which can cost the same as a house.
Used sampling oscilloscopes can be had for less than the price of a good but inexpensive Chinese DSO if you have the experience to know what to look for at least in the US.
Low-z passive probes that work into the low GHz range are not difficult to fabricate but the usual solution is just to design in 50 ohm coaxial test points.
Performance verification is a significant problem though. For that a fast reference level pulse generator is needed and if you build that, how do you verify it? It is too bad that tunnel diodes are out of style.
-
In my experience, the main reason government/agency buy is to spend end of year budget fast...
-
Low-z passive probes that work into the low GHz range are not difficult to fabricate but the usual solution is just to design in 50 ohm coaxial test points.
We are talking 20GHz+ here though.
-
Have a look at the HP 54121T.
Four channel 20GHz sampling scope, with one channel including a TDR output. Inputs are 50ohm. It's excellent for characterizing the impedance tolerance of traces on PCBs, among other things.
For repetitive signals it's fine, and you can find them quite cheap on ebay sometimes. Also the manuals, though there are no schematics available, sadly.
I have two working sets, plus another mainframe for spares. That spares one has a dead power supply, that should be fixable even without schematics.
One complete unit was $600, the other was a couple of thousand. The mainframe with the dead supply was $75.
The only real downside with these is the scary +/-2V MAXIMUM rating on the inputs. Any more and (the manual insists) the inputs are destroyed. It means you have to be extremely rigorous with the anti-static measures. Can't even connect a piece of coax to the inputs without shorting the center conductor to the shield and ground first.
So far I haven't blown one. But this is why I got two. No spares available from HP/Agilent.
-
Also the oil industry.
I worked in the oil survey industry for over a decade, and I can't recall anyone who needed such a high end scope.
What aspect of the oil industry?
You went in a little different direction. I am just painting a picture that tech is leaving California and other traditional places and moving to Texas. Austin is now a hub of activity and growing
There is an extreme amount of communications equipment in the oil industry. Granted they are not installing test gear, but the oil business is surprisingly high tech nowadays. As far as NASA goes, they have a pretty good little setup over there at the Johnson Space Center. It is actually where they developed and tested many of the flight COMM and electronic systems and continue to this day. Texas is very high tech. There are a bunch of fabs there. I know Dallas has some. Austin also. I think Samsung may have something is Houston as well. If it were not so flat i might move there! (I need visible mountains nearby or I feel sad :'()
-
Also the oil industry.
I worked in the oil survey industry for over a decade, and I can't recall anyone who needed such a high end scope.
What aspect of the oil industry?
You went in a little different direction. I am just painting a picture that tech is leaving California and other traditional places and moving to Texas. Austin is now a hub of activity and growing
There is an extreme amount of communications equipment in the oil industry. Granted they are not installing test gear, but the oil business is surprisingly high tech nowadays. As far as NASA goes, they have a pretty good little setup over there at the Johnson Space Center. It is actually where they developed and tested many of the flight COMM and electronic systems and continue to this day. Texas is very high tech. There are a bunch of fabs there. I know Dallas has some. Austin also. I think Samsung may have something is Houston as well. If it were not so flat i might move there! (I need visible mountains nearby or I feel sad :'()
Texas Instruments used to have a large fab roughly 10 minutes where I live. Closed down recently I believe. Up in Austin and Dallas ther's quite a bit of semiconductor activity. We have HP in Houston as well. Tek might be letting me into their Houston cal lab to verify my 3478A. So yes, some stuff in Houston, but no where near the surplus stores like in California.
-
An article, HERE (http://www.electronicproducts.com/Test_and_Measurement/Benchtop_Rack_Mountable/Methods_used_to_boost_real-time_oscilloscope_bandwidth.aspx), describes techniques for high bandwidth acquisition.
-
Someone got experience with this USB sampling scope?
10 GHz USB scope (https://www.youtube.com/watch?v=4BdExgCdo2U#)