Least expensive but reliable used-market scope with at least 1 GHz?

[EV]

....  It seems that the 7T11 and 7S11 are not both needed, seems like 7S11 should do it, yes?  Then comes the question of what are the tradeoffs between S-6 and say S-4, 3, 2, etc?  And finally, which probe?

You need both plugins.
S-2 has 50 Ohm GR874 connector and needs no probe. Risetime is 50 ps.
S-3 has probe connected to it. Rise time is 350..500 ps dependind on the used attenuator.
S-4 has 50 Ohm SMA connector and need no probe. Rise time is 25 ps.
S-6 is for 7S12 TDR. I have not used it with 7S11.

[Howardlong]

As a matter of interest, what is the technology behind these old sampling scopes?

I assume there's a sample and hold, but what technologies were used to achieve such a tiny sample time?

Regarding the timing relative to trigger, how did that work? I realise you can do random or sequential, but at some stage there needs to be very low jitter picosecond accuracy timing involved, how did they achieve that?

[Berni]

One important question to ask if is you are looking to get a proper real time scope or is a sampling scope fine?

[nctnico]

As a matter of interest, what is the technology behind these old sampling scopes?

I assume there's a sample and hold, but what technologies were used to achieve such a tiny sample time?

Regarding the timing relative to trigger, how did that work? I realise you can do random or sequential, but at some stage there needs to be very low jitter picosecond accuracy timing involved, how did they achieve that?

Usually a diode sampler. That is the easy part. Triggering and time correlation are a different story.

[Electro Fan]

....  It seems that the 7T11 and 7S11 are not both needed, seems like 7S11 should do it, yes?  Then comes the question of what are the tradeoffs between S-6 and say S-4, 3, 2, etc?  And finally, which probe?

You need both plugins.
S-2 has 50 Ohm GR874 connector and needs no probe. Risetime is 50 ps.
S-3 has probe connected to it. Rise time is 350..500 ps dependind on the used attenuator.
S-4 has 50 Ohm SMA connector and need no probe. Rise time is 25 ps.
S-6 is for 7S12 TDR. I have not used it with 7S11.

Thanks - some questions:

If the S-2 has a risetime of 50 ps (plenty fast in my book), what is the downside to a S-2?
 - if the benefit of a S-4 is 25ps I can forgo that; if the benefit of the s-4 is the SMA connector I can get around that on the S-2 with a GR874 to BNC adapter... I think; not to say the S-2 wouldn't be preferred if one can be found...

When you say no probe is needed where is the signal coming from?   (I need an overview of the 7000 sampling system architecture).

The original goal was to be able to measure risetime on a 1GHZ or faster signal...

Thanks

[Electro Fan]

One important question to ask if is you are looking to get a proper real time scope or is a sampling scope fine?

Good question.  As mentioned, the original goal was to measure risetime on 1 GHz or faster signals.  What are the differences between a proper real time scope and a sampling scope?

Your question caused me to go look for/find this:
http://electronicdesign.com/test-amp-measurement/what-s-difference-between-real-time-and-sampling-oscilloscopes
-seems like for higher bandwidth at lower cost for repetitive signals that are not difficult to trigger the sampling approach will be better, including less jitter?
- seems like for more challenging trigger requirements and deeper analysis (but maybe for relatively slower signals and with higher cost) the real time scope would be better?

FWIW, I have noticed on my 2467B that under some circumstances a pulse generator signal with a risetime of perhaps 700ps or less will drift; it seems to happen more with certain cables or cable lengths.  I'm not sure if this drift is considered jitter, or if it's something else; perhaps the drift would be less with a better/faster real-time scope or perhaps it would be less (more stable) with a sampling scope?


- feel free to add further clarifications     Thx

[Electro Fan]

A couple of these comments from Howardlong reflect some of my thinking:

I do admit to a strange fascination of seeing changes at 10ps/div.

It's been pretty educational looking at this stuff in the time domain, usually I use an 8753A VNA for matching stuff, and will continue to do so, but seeing the impedance bumps in the time domain give you a different perspective.

The current objective is to be able to attain better risetime measurements (ie, to be able to measure risetime for 1 GHZ signals, or faster if affordable) but anything that gives better measurement of signal performance in general along with other improved visibility into waveform and signal attributes would be a positive.  The challenge is to do it at a good cost since it's strictly an educational hobby with no chance of revenue generation.

So far, the idea of adding some plug-ins to a Tek 7000 seems like the path of least resistance....
The one thing I most miss on the 7000 is cursors (go figure) along with the general lack of digital readout measurements but if I can get from the performance of a 500MHz 7904/7A29 to some of the performance and "observable measurements" normally associated with a 1 GHz or faster scope then the sampling plug-in approach seems like a reasonable path.

I'm definitely up for leaning more about the sampling configs of the 7000.

[EV]

When you say no probe is needed where is the signal coming from?   (I need an overview of the 7000 sampling system architecture).

As I told they have 50 Ohm input connector. You can connect the signal to it with RG58 cable. If the voltage is too big (max value is 5 V) you need also attenuators.

You can find free manuals for 7T11, 7S11, S-2 and S-4 by google. You can get more info from them.

[rx8pilot]

I have a TDS754C and really like it overall. It's old, big, and does not have some of the bells and whistles the newer models have. It does, however, work every day with no issues discovered yet. if the CRT every dies, it has VGA to extend its life a little if I need. I am using it as a step to a new and modern $15k+ scope when I have the cash to pay for it.

[Electro Fan]

When you say no probe is needed where is the signal coming from?   (I need an overview of the 7000 sampling system architecture).

As I told they have 50 Ohm input connector. You can connect the signal to it with RG58 cable. If the voltage is too big (max value is 5 V) you need also attenuators.

You can find free manuals for 7T11, 7S11, S-2 and S-4 by google. You can get more info from them.

Ok, I understand that the 50 Ohm connector (or GR874) can be used via a coax cable rather than a probe (I thought when you mentioned that no probe was needed that maybe the signal was routed internally from another plug-in) - now I understand.  Thx

If a probe were to used is there any particular Tektronix probe (from the 7000 era or more recent) that is recommended for use with a S-4?

[Electro Fan]

I have a TDS754C and really like it overall. It's old, big, and does not have some of the bells and whistles the newer models have. It does, however, work every day with no issues discovered yet. if the CRT every dies, it has VGA to extend its life a little if I need. I am using it as a step to a new and modern $15k+ scope when I have the cash to pay for it.

These look like fine scopes but just to confirm the TDS754C is 500MHz?  Need to go to 784 for 1 GHz?  Thx

[rx8pilot]

I have a TDS754C and really like it overall. It's old, big, and does not have some of the bells and whistles the newer models have. It does, however, work every day with no issues discovered yet. if the CRT every dies, it has VGA to extend its life a little if I need. I am using it as a step to a new and modern $15k+ scope when I have the cash to pay for it.

These look like fine scopes but just to confirm the TDS754C is 500MHz?  Need to go to 784 for 1 GHz?  Thx

That is correct - but it is the same box minus some filter caps and software keys.

[nctnico]

I think you can uphack the TDS754 to a TDS784. This info is somewhere on the forum but you'll need to calibrate the scope afterwards for frequencies >500MHz.

[rx8pilot]

I think you can uphack the TDS754 to a TDS784. This info is somewhere on the forum but you'll need to calibrate the scope afterwards for frequencies >500MHz.

Yep, I have seen it done. I am too chicken to mess with it. One small mistake will send me scrambling to fix a 15 year old scope. I may try it when I order my new scope - just for fun.

[Electro Fan]

On the Tek 7000 series...
The Tek S-2 spec says:

Operating Input Voltage Range:  +1V to -1V P-P signals.  Step signals greater than 200mV may require more than 1 dot to display 100% of the step transition

Maximum Input Voltage:  +/-5VDC or sine wave of 10V P-P up to 10MHz

Does that mean for anything above 10MHz the limit is +1V to -1V P-P ?

Likewise on the S-4 the spec says Signal Voltage Maximum Operating is 1VPP, but on the front of the unit it says +/- 5V Max. 

http://bee.mif.pg.gda.pl/ciasteczkowypotwor/Tek/S4_In.pdf

Under what conditions is it safe to go to +/- 5V?  Thx

[OldSchoolTechCorner]

The TDS784 is a good one. Only issues is uses tantalum capacitors and they are prone to fail and burn hole though board.  A tantalum capacitor is actually much like a bomb or thermite. It has an oxidizer (manganese dioxide) and fuel (tantalum). You'll only need to ignite it. Well the electricity is also there so just start the countdown clock. The reason why I replace them as they age. The older series Tektronix DSO's you have to replace every SMD electrolytic and though hole and neutralize and clean the boards.

[Electro Fan]

 a little bit off topic but saw this discussion/comparison of sampling scopes:

https://books.google.com/books?id=JP3sCAAAQBAJ&pg=PA64&lpg=PA64&dq=tekronix+s-4+vs+s-6&source=bl&ots=YVKKgsCnqU&sig=R0GUfGIQ6S92AR8dzeUOTF7AjCM&hl=en&sa=X&ved=0CD4Q6AEwBWoVChMIp-bu9-royAIVEN1jCh0wZgAQ#v=onepage&q=tekronix%20s-4%20vs%20s-6&f=false

[edavid]

S-2 has 50 Ohm GR874 connector and needs no probe. Risetime is 50 ps.

If the S-2 has a risetime of 50 ps (plenty fast in my book), what is the downside to a S-2?

I believe the S-2 actually has a 75ps risetime (4.6GHz bandwidth).  Downside is the GR connector... GR adapters tend to add aberrations.

[Electro Fan]

S-2 has 50 Ohm GR874 connector and needs no probe. Risetime is 50 ps.

If the S-2 has a risetime of 50 ps (plenty fast in my book), what is the downside to a S-2?

I believe the S-2 actually has a 75ps risetime (4.6GHz bandwidth).  Downside is the GR connector... GR adapters tend to add aberrations.

Yes, I think 75ps is the spec. 

How big a deal is the GR aberration?  On 10 typical measurements, how many are likely to be impacted by an aberration?  Kind of a subjective/random question I realize - just trying to get some impression of the impact.  Thx

[Rupunzell]

Start here at the Picosecond Pulse Labs app notes to get an idea of what is involved.
http://picosecond.surtos.com/techinfo/content.asp?spid=19


Bernice

[TSL]

A HP 54720A or D is a reasonable scope. Plugin modules to 2Ghz, color screen etc. i have one with 54712A and 54721A modules and that's fine for me. $1000Us will buy you a fine specimen thus...

http://www.ebay.com/itm/HP-HEWLETT-PACKARD-MODEL-54720A-OSCILLOSCOPE-W-TWO-54721A-AMPLIFIERS-1159-/401018882278

Yes its big and a little noisy but its exceptionally capable

rf-loop had high praise for them too...

https://www.eevblog.com/forum/chat/is-a-hp-agilent-54721a-a-decent-oscilloscope/

More info on it here...

http://www.keysight.com/main/facet.jspx?&cc=AU&lc=eng&k=54720A&sm=g

[Electro Fan]

Start here at the Picosecond Pulse Labs app notes to get an idea of what is involved.
http://picosecond.surtos.com/techinfo/content.asp?spid=19

Bernice

Thanks for the link to the info.  Picosecond Pulse Labs looks like an intense / focused outfit; looks like they are now part of Tektronix?
http://www.tek.com/document/news-release/tektronix-acquires-picosecond-pulse-labs

Privately-held Picosecond Pulse Labs, based in Boulder, Colorado, offers products that include ultra-high-speed pattern generators, the world's fastest pulse generators and highest bandwidth sampler modules. The company recently introduced the PatternPro line that includes multi-channel 32 Gb/s data generators and analyzers for 100G/400G applications.

[Electro Fan]

A HP 54720A or D is a reasonable scope. Plugin modules to 2Ghz, color screen etc. i have one with 54712A and 54721A modules and that's fine for me. $1000Us will buy you a fine specimen thus...

http://www.ebay.com/itm/HP-HEWLETT-PACKARD-MODEL-54720A-OSCILLOSCOPE-W-TWO-54721A-AMPLIFIERS-1159-/401018882278

Yes its big and a little noisy but its exceptionally capable

rf-loop had high praise for them too...

https://www.eevblog.com/forum/chat/is-a-hp-agilent-54721a-a-decent-oscilloscope/

More info on it here...

http://www.keysight.com/main/facet.jspx?&cc=AU&lc=eng&k=54720A&sm=g

I'm sure it's high performance but it looks a bit esoteric.  I briefly had a HP scope - it had some good attributes but I think I'm with whoever said "Tektronix for oscilloscopes, HP for everything else" - I'm kind of a Tek scope fan.  And a closet Keysight fan.  Not to say I wouldn't take a LeCroy or HP or another scope - almost any oscilloscope looks good to me.

Having said that I always admire your posts and I know rf-loop also has a lot knowledge and experience - and he definitely knows what he likes:

Congratulations, you have get really high-end real tool, not toy (what still may be reasonable for some use becouse price also is low - but any these cheap scope can not beat this machine as real tool. It hjave been some kind of "state of art".

and he also knows what he doesn't like..... a "cross word puzzles" UI   

What about Tektronix some UI...  just like cross word puzzles

- that comment by rf-loop was very FUNNY!

[Wuerstchenhund]

A HP 54720A or D is a reasonable scope. Plugin modules to 2Ghz, color screen etc. i have one with 54712A and 54721A modules and that's fine for me. $1000Us will buy you a fine specimen thus...

http://www.ebay.com/itm/HP-HEWLETT-PACKARD-MODEL-54720A-OSCILLOSCOPE-W-TWO-54721A-AMPLIFIERS-1159-/401018882278

Yes its big and a little noisy but its exceptionally capable

That's only true for its bandwidth and modularity (which includes optical inputs). As a scope per se, the 54720 is actually pretty basic. Trigger, math and measurement capabilities are somewhat limited for a scope in this class (it does have statistics, though), and processing is painfully slow. FFT on this thing is a real dog even though it only uses a few thousand points.

We had 54720As and Ds when they were current (~1995?), and they were great for the things we needed them to do (TDR, jitter measurements), but as general purpose scopes they weren't great even at those days, especially when considering the input limitations of the various modules. Many people also didn't like the typical HP single knob user interface (also found on the older 545xx Series scopes), although I have to say I did really liked it.

Don't get me wrong, the 54720 can be exceptionally great value for money if you need very high bandwidth, the available optical inputs and the modularity, but outside some niche areas it's not much better than any other digital scope from the early '90s. The same money ($1000 for an untested 54720D) can get you much more versatile scopes that satisfy the 1GHz bandwidth requirement and make better general purpose scopes.

I briefly had a HP scope - it had some good attributes but I think I'm with whoever said "Tektronix for oscilloscopes, HP for everything else" - I'm kind of a Tek scope fan.  And a closet Keysight fan.  Not to say I wouldn't take a LeCroy or HP or another scope - almost any oscilloscope looks good to me.

The thing with that old saying "Tektronix for oscilloscopes, HP for everything else" is that it was true when analog scopes were still current, and Tek was without doubt a leader in analog scopes who has pushed technology forward. However, that isn't equally true in regard to digital scopes, in fact Tek has been pretty much a follower when it comes to DSOs, and pushing DSO technology was left to mostly LeCroy and (to some extend) Agilent/Keysight. Plus many of the old Tek's suffered from a capacitor plague, and the UI was pretty horrendous even by those days' standards.

If you want an old digital scope I'd pretty much focus on LeCroy and HP/Agilent.

I don't know if the reason you started this thread is for curiosity or if you're thinking on buying a scope, but if its the latter then you should first get your requirements right (i.e. what you want to do with the scope, why you think you need 1GHz bandwidth etc), then sort out your budget, then find out what it buys you and then do some thorough research (i.e. read the manuals for each scope to get an impression on how they work and what they can do, talk to users that do use one of the scopes for similar things you want to do, find out what the weak points are etc), and then decide based on your findings.

[Electro Fan]

I briefly had a HP scope - it had some good attributes but I think I'm with whoever said "Tektronix for oscilloscopes, HP for everything else" - I'm kind of a Tek scope fan.  And a closet Keysight fan.  Not to say I wouldn't take a LeCroy or HP or another scope - almost any oscilloscope looks good to me.

The thing with that old saying "Tektronix for oscilloscopes, HP for everything else" is that it was true when analog scopes were still current, and Tek was without doubt a leader in analog scopes who has pushed technology forward. However, that isn't equally true in regard to digital scopes, in fact Tek has been pretty much a follower when it comes to DSOs, and pushing DSO technology was left to mostly LeCroy and (to some extend) Agilent/Keysight. Plus many of the old Tek's suffered from a capacitor plague, and the UI was pretty horrendous even by those days' standards.

I don't know if the reason you started this thread is for curiosity or if you're thinking on buying a scope, but if its the latter then you should first get your requirements right (i.e. what you want to do with the scope, why you think you need 1 GHz bandwidth etc), then sort out your budget, then find out what it buys you and then do some thorough research (i.e. read the manuals for each scope to get an impression on how they work and what they can do, talk to users that do use one of the scopes for similar things you want to do, find out what the weak points are etc), and then decide based on your findings.

I agree that the Tek/HP saying had it's best merit back in the analog days.  While I still think highly of Tek scopes (trust me, you don't have to tell me about why LeCroy is better, I read your posts) it's clear that Tek's heydays didn't fully traverse the A to D revolution; the dominance they enjoyed in the analog days declined but I still respect them and have a fondness for their MSO/MDO gear.  I know, I know LeCroy is better - I can see you reaching for the keyboard .  The only reason for mentioning the Tek/HP saying was that my experience with the HP analog scope seemed to confirm the saying.  In the digital era it's a different situation, I get it.

The reason for the thread is pretty much what I mentioned to dr.diesel on the first page of the thread - what started this inquiry was simply a desire to achieve better rise time measurements (but so as not close off other possible improvements I didn't want to constrain a wider discussion of what folks think might be a really good 1GHz scope in general - ie, a scope with other functionality and performance improvements beyond just improved risetimes, while trying to keep the price down).  Said differently, for this endeavor maximizing risetime performance is the base case and everything else is the value-added case.

I have 4 scopes.  (According to the recent thread underway I'm 16 scopes away from par if par is 20; so I have a long way to go to get to par much less a birdie or an eagle, in golf-speak).  Most of what I do (which is just trying to learn through small projects and experimenting) can be done with 100MHz or less - so that can be done on my Rigol (70MHz) or Tek 2247A (100MHz).  I also have a 2465B which is spec'd at 400MHz but does somewhat better.  Lastly, I have a Tek 7000 which is spec'd to 500MHz but my amplifier plugins are only spec'd for 200MHz; I'm on the lookout for a 500MHz 7A29.  In the meantime I became fascinated with the Jim Williams generator (now I have two of those, not much in test equipment land seems to hold steady at quantity 1).  The JW generators showed my scopes couldn't keep up with the fast pulse gens.  So, I started thinking how much bandwidth do I need and what could I maybe afford to see hundreds of picoseconds risetimes, or better yet, maybe tens of picoseconds risetimes.  (The 2467B is spec'd at 875ps and does somewhat better - but going faster seems to be an addiction for no good reason.) 

So, this all led to researching 1GHz scopes (an arbitrary number but a notable threshold, perhaps).  That led to the suggestions here to investigate the Tek 7000 sampling plug-ins.  Somehow even though I had the Tek 7000 I wasn't at all on top of the sampling capabilities.  I've found a 7S11 with a S-1 (only good to 350ps but that's 525ps faster than the 2467B); so I think for now I just need a 7T11 and I'm sub-500ps risetime ready.  With a S-4 I'd be down to 25ps.  There is no good reason to do this except that a) it's fascinating to think about looking at signals operating in the trillionths of a second, and b) as Howardlong mentioned (I'm paraphrasing) the more you can visualize signals the more insight/learning you can gain.  There is no payoff other than curiosity addressed.

I appreciate all the advice here - some of the recommended scopes would add more functionality but this thread helped me realize that I'm sitting on just a few incremental pieces short of what I need for pretty fast risetime viewing with the Tek 7000.