EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: mrpackethead on September 13, 2014, 05:18:36 am
-
Hi, due to some changes in what i'm building, i need some better tools.. I'm often working with Ethernet Phys, transformers and stuff, that is running signals that may be up to 200MHz.. I current have a Rigol DS1102E 100Mhz Scope but its simply not enough.
I quite like that scope its done plenty of useful work for me. Any suggestions on what might be useful?
-
If you haven't seen these already, here's some reading for you:
https://www.eevblog.com/forum/testgear/review-siglent-sds2304-a-comparison-of-features-with-rigol-ds2000-series/ (https://www.eevblog.com/forum/testgear/review-siglent-sds2304-a-comparison-of-features-with-rigol-ds2000-series/)
https://www.eevblog.com/forum/testgear/siglent-sds2204-mso-review/ (https://www.eevblog.com/forum/testgear/siglent-sds2204-mso-review/)
https://www.eevblog.com/forum/testgear/choosing-a-four-channel-oscilloscope/ (https://www.eevblog.com/forum/testgear/choosing-a-four-channel-oscilloscope/)
Now you wouldn't call me biased would you? ;)
-
For ethernet a spectrum analyser may be a better tool. There is not much to see on an oscilloscope because the signals are coded using various levels. On a spectrum analyser you can see the frequency spectrum and determine whether you have too much damping, phase noise, etc.
-
How much do you have to spend?
High bandwidth = high $$$$, it's almost a direct correlation.
I'd at least check here if you have some coin:
http://stores.ebay.com/keysight (http://stores.ebay.com/keysight)
I'd probably be shooting for a 300MHz or 500Mhz scope here.
-
I wouldn't recommend this one:
http://www.ebay.com.au/itm/Tektronix-TDS-5054-500MHz-4-Channel-5GS-s-Digital-Phosphor-Oscilloscope-DPO-/360974897202 (http://www.ebay.com.au/itm/Tektronix-TDS-5054-500MHz-4-Channel-5GS-s-Digital-Phosphor-Oscilloscope-DPO-/360974897202)
-
For ethernet a spectrum analyser may be a better tool. There is not much to see on an oscilloscope because the signals are coded using various levels. On a spectrum analyser you can see the frequency spectrum and determine whether you have too much damping, phase noise, etc.
Yes..
How much do you have to spend?
High bandwidth = high $$$$, it's almost a direct correlation.
I'd at least check here if you have some coin:
http://stores.ebay.com/keysight (http://stores.ebay.com/keysight)
I'd probably be shooting for a 300MHz or 500Mhz scope here.
Any comments on the Rigol DS2000A. Its affordable, especially if you buy the 70Mhz one and well do some trickery to it.
-
Any comments on the Rigol DS2000A. Its affordable, especially if you buy the 70Mhz one and well do some trickery to it.
Excellent value scope, but only 200MHz.
-
You can get decent 4 channel 500MHz used scopes for under $1K
http://www.ebay.com.au/itm/Tektronix-TDS-744A-4-Ch-Color-Digital-Oscilloscope-500MHz-2-GSa-s-GUARANTEED-/111389782591 (http://www.ebay.com.au/itm/Tektronix-TDS-744A-4-Ch-Color-Digital-Oscilloscope-500MHz-2-GSa-s-GUARANTEED-/111389782591)
Not all the bells and whistles of today's scopes, but if performance is what you need on limited $ then it's not a bad option.
-
I wouldn't recommend this one:
http://www.ebay.com.au/itm/Tektronix-TDS-5054-500MHz-4-Channel-5GS-s-Digital-Phosphor-Oscilloscope-DPO-/360974897202 (http://www.ebay.com.au/itm/Tektronix-TDS-5054-500MHz-4-Channel-5GS-s-Digital-Phosphor-Oscilloscope-DPO-/360974897202)
Its just sad to see them like that. They should end their days with a switch that won't work, and you swearing at them.
-
Even better:
http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814 (http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814)
That's 400Kwfps, 4 channel, 500MHz, 1M sample memory, colour intensity display
https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/Tektronix-TDS784C_Datasheet.pdf (https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/Tektronix-TDS784C_Datasheet.pdf)
Dare I say that's impossible to beat for performance for under $1K?
Can anyone find anything better?
-
Any comments on the Rigol DS2000A. Its affordable, especially if you buy the 70Mhz one and well do some trickery to it.
Excellent value scope, but only 200MHz.
DS2302A-S 300MHz = DS2000A + software? and 300Mhz?
-
Even better:
http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814 (http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814)
That's 400Kwfps, 4 channel, 500MHz, 1M sample memory, colour intensity display
https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/Tektronix-TDS784C_Datasheet.pdf (https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/Tektronix-TDS784C_Datasheet.pdf)
Dare I say that's impossible to beat for performance for under $1K?
Can anyone find anything better?
Do you know where to buy 3.5" floppies any longer?
-
Do you know where to buy 3.5" floppies any longer?
The average stationery shop never throws anything away. They'll have some somewhere. :)
-
Do you know where to buy 3.5" floppies any longer?
You can buy them, and maybe a USB replacement one.
But even if not working, I wouldn't quibble when you get all that for under $1K
-
We have one of these scopes (the 754) at work, and it is a very useful bit of kit still. Annoying single set of vertical controls however.
-
Even better:
http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814 (http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814)
That's 400Kwfps, 4 channel, 500MHz, 1M sample memory, colour intensity display
https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/Tektronix-TDS784C_Datasheet.pdf (https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/Tektronix-TDS784C_Datasheet.pdf)
Dare I say that's impossible to beat for performance for under $1K?
Can anyone find anything better?
Well, for slightly over $1k you can get this:
http://www.ebay.co.uk/itm/Lecroy-LT374L-WaveRunner-500MHz-4GS-s-Digital-Oscilloscope-DSO-/111368356482?pt=BI_Oscilloscopes&hash=item19ee124282 (http://www.ebay.co.uk/itm/Lecroy-LT374L-WaveRunner-500MHz-4GS-s-Digital-Oscilloscope-DSO-/111368356482?pt=BI_Oscilloscopes&hash=item19ee124282)
500MHz 4Ch, 4GSa/s, 4M per channel, 8.4" LCD (see the spec file I enclosed). Hardware is made by Iwatsu and the software and technology comes from LeCroy.
This one even comes with Wave Analyzer and HDS options. There are many other useful options (i.e. Jitter Analysis, Power Analysis, Advanced Trigger) available as well.
-
Well, for slightly over $1k you can get this:
http://www.ebay.co.uk/itm/Lecroy-LT374L-WaveRunner-500MHz-4GS-s-Digital-Oscilloscope-DSO-/111368356482?pt=BI_Oscilloscopes&hash=item19ee124282 (http://www.ebay.co.uk/itm/Lecroy-LT374L-WaveRunner-500MHz-4GS-s-Digital-Oscilloscope-DSO-/111368356482?pt=BI_Oscilloscopes&hash=item19ee124282)
Yep, great value, but I believe they are a PITA to use?
I think I'd rather the Tek as my everyday use scope.
-
Well, for slightly over $1k you can get this:
http://www.ebay.co.uk/itm/Lecroy-LT374L-WaveRunner-500MHz-4GS-s-Digital-Oscilloscope-DSO-/111368356482?pt=BI_Oscilloscopes&hash=item19ee124282 (http://www.ebay.co.uk/itm/Lecroy-LT374L-WaveRunner-500MHz-4GS-s-Digital-Oscilloscope-DSO-/111368356482?pt=BI_Oscilloscopes&hash=item19ee124282)
Yep, great value, but I believe they are a PITA to use?
Not really, at least not more than any other complex test instrument from any manufacturer. A few things are different with LeCroy (i.e. horizontal controls left and vertical right) but after a short time that's a nobrainer. Other than that it's very easy to use thanks to a simple menu structure.
In addition, if you know how to use this scope then you also already know how to drive a 9300 Series, LC Series, WaveRunner 1 and WavePro 900 Series scope, because unlike other manufacturers LeCroy doesn't re-invent the UI everytime they come up with a new scope. In fact, they only changed the UI twice, once when going from the 9400 Series to the 9300 Series in the early 90's, and then in 2001 when they introduced their Windows scopes (XStream) with MAUI touch interface.
think I'd rather the Tek as my everyday use scope.
Not sure I would. The Tek UI wasn't great when it came out, it's only useable because there is not that much functionality in these scopes. Then there's the small screen (doesn't this one use a mono CRT with LCD shutter?). It not only looks like a scope from the 90's, it feels and operates like one as well.
-
Not so keen on buying a 2nd hand scope.. If theres a problem its a PITA.
I have a budget of up to $4000 USD. Though i'd only spend that if i had to, to do the jobs which i need to. The Rigol DS2000A running at 300Mhz, likely would seem to do that.
-
The Rigol DS2000A running at 300Mhz, likely would seem to do that.
While the DS2000A has some excess bandwidth headroom (you only hit -3db attenuation @ ~350-400MHz), if you are interested in details of your waveform at those speeds you will be lacking the samples (at "only" 2GSa/s). I'd consider that if you're getting a faster scope, you are probably interested in viewing the integrity of the signal at those higher speeds and most likely would benefit from a higher sampling rate.
-
Maybe the Rigol DS4000 Series.
-
Then there's the small screen (doesn't this one use a mono CRT with LCD shutter?). It not only looks like a scope from the 90's, it feels and operates like one as well.
I have a Tektronix TDS744A (4ch. 500MHz long memory and FFT) with a TFT screen for sale. ;D
Still 200MHz is more than enough for ethernet. Most of the signal content is below 100MHz.
-
Not so keen on buying a 2nd hand scope.. If theres a problem its a PITA.
That really depends on the what you buy. With some luck you can save enough with a second hand scope to have enough budget left for any repairs or a second used scope as spare. For example, LeCroy offers full support for 7 years after a scope has been out of production, and offers repair even for the 9300 Series scopes which is now older than 15 years. They even offer additional warranty for repaired devices.
Maybe the Rigol DS4000 Series.
The DG4000 seems to be a rather buggy scope, much more so than the DS2000 (which has seen more bug fixes).
-
Not so keen on buying a 2nd hand scope.. If theres a problem its a PITA.
True, but odds of failure are relatively small if you get a more recent one.
I have a budget of up to $4000 USD.
That changes the game entirely.
-
Not so keen on buying a 2nd hand scope.. If theres a problem its a PITA.
2nd hand is not an option, at our company.
I have a budget of up to $4000 USD.
That changes the game entirely.
Well, it just means that we have to work harder with a bit more paperwork. I could spend more, but would have to be able to justify it.
-
2nd hand is not an option, at our company.
I could spend more, but would have to be able to justify it.
Look at it another way.
Let the boss justify not spending money on top grade equipment.
Higher quality production, better test results and happier worker.
Everybody wins.
EDIT
Siglent Hat OFF
-
2nd hand is not an option, at our company.
I could spend more, but would have to be able to justify it.
Look at it another way.
Let the boss justify not spending money on top grade equipment.
Higher quality production, better test results and happier worker.
Everybody wins.
That kind of sales line won't work. Facts work.
-
2nd hand is not an option, at our company.
I could spend more, but would have to be able to justify it.
Look at it another way.
Let the boss justify not spending money on top grade equipment.
Higher quality production, better test results and happier worker.
Not really. Cash spend on test equipment cannot be spend on a raise!
-
2nd hand is not an option, at our company.
I could spend more, but would have to be able to justify it.
Look at it another way.
Let the boss justify not spending money on top grade equipment.
Higher quality production, better test results and happier worker.
Not really. Cash spend on test equipment cannot be spend on a raise!
Boss gets bigger bonus for spending less.
Siglent hat OFF
-
Do not forget the cost of the probes. High impedance passive probes have quite a bit of loading and will not deliver full bandwidth from a twisted pair Ethernet transmission line.
A Tektronix TDS500 or TDS700 series would be my choice as well although they are getting long in tooth so I would get a return warranty. Some of the TDS400 series are 400 MHz and might work.
-
2nd hand is not an option, at our company.
Well that's different again!
In that case shoot for something a bit better than the Rigol 2000
-
Some of the TDS400 series are 400 MHz and might work.
I'm not sure if spending money on an old boat anchor with paltry 100MS/s max real time sample rate and lowly 120kpts sample memory is a sane investment. Such a scope is absolutely useless for looking at stuff like Ethernet, no matter the analog bandwidth.
-
Some of the TDS400 series are 400 MHz and might work.
I'm not sure if spending money on an old boat anchor with paltry 100MS/s max real time sample rate and lowly 120kpts sample memory is a sane investment. Such a scope is absolutely useless for looking at stuff like Ethernet, no matter the analog bandwidth.
This depends on what needs to be measured. If only eye diagrams are necessary to check phase and amplitude distortion which is most of what is needed at the physical link layer, then sample rate and record length are largely irrelevant except insofar as it will take longer to generate the display. Even an old but fast analog storage oscilloscope could do this although I would not recommend it.
On the other hand, no amount of sample rate and record length will make up for low bandwidth.
-
This depends on what needs to be measured. If only eye diagrams are necessary to check phase and amplitude distortion which is most of what is needed at the physical link layer, then sample rate and record length are largely irrelevant except insofar as it will take longer to generate the display.
Even for simple eye diagrams a TDS460 is still a very poor scope. It also would have to be used in equivalent time sampling mode, which means for non-repetitive signals (like Ethernet) valuable signal information is lost.
Even an old but fast analog storage oscilloscope could do this although I would not recommend it.
Well, at least a decent analog scope wouldn't suffer from the drawbacks of ETS, although I agree that it's not a viable solution for the OP.
On the other hand, no amount of sample rate and record length will make up for low bandwidth.
While this is certainly true, the opposite is, too (no amount of analog bandwidth can compensate for insufficient sample rate and memory). A TDS460 seems to sell for somewhere between $200-$350 on ebay (the majority seem to sell closer towards the upper end), which is silly money for an antique boat anchor with such poor specifications. For the OP's use case this wouldn't be a sane investment, it would be money flushed down the drain.
The prices are even more ridiculous when considering that (with a bit of patience) the same money can give you 1GHz/2GHz sample rate and if I remember right 1Mpts sample memory:
http://www.ebay.com/itm/LECROY-9374M-1GHZ-4CH-2GS-S-OSCILLOSCOPE-PP093-2GS-S-ADAPTER-amp-MANUALS-/151363188746?pt=BI_Oscilloscopes&hash=item233df2f80a&nma=true&si=AtmF14dZIgHkPnYTheDg0mBaHOs%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557#ht_1752wt_905 (http://www.ebay.com/itm/LECROY-9374M-1GHZ-4CH-2GS-S-OSCILLOSCOPE-PP093-2GS-S-ADAPTER-amp-MANUALS-/151363188746?pt=BI_Oscilloscopes&hash=item233df2f80a&nma=true&si=AtmF14dZIgHkPnYTheDg0mBaHOs%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557#ht_1752wt_905)
I guess the main reason people pay so much for these Tek scopes is because it has "Tektronix" written on it, and I bet that many people are not really aware of its limitations and expect it to be useable for non-repetitive signals over 20MHz.
-
This depends on what needs to be measured. If only eye diagrams are necessary to check phase and amplitude distortion which is most of what is needed at the physical link layer, then sample rate and record length are largely irrelevant except insofar as it will take longer to generate the display.
Even for simple eye diagrams a TDS460 is still a very poor scope. It also would have to be used in equivalent time sampling mode, which means for non-repetitive signals (like Ethernet) valuable signal information is lost.
Even an old but fast analog storage oscilloscope could do this although I would not recommend it.
Well, at least a decent analog scope wouldn't suffer from the drawbacks of ETS, although I agree that it's not a viable solution for the OP.
I have made eye diagrams with analog oscilloscopes, a Tektronix 7834 (400 MHz analog storage), and using a analog sampling oscilloscope (1 GHz to 14 GHz at 50ksamples/second) on clocked signals up to 100 MHz. All were completely usable but only the sampling oscilloscope had the resolution to see pattern dependent jitter produced by different logic families and analog storage was needed to display rare events.
Long ago I did the same thing with a low end TDS400 series DSO but not at 100 MHz. :)
I actually bought the 7834 recently as the least expensive way to accomplish this with 4 channel 400 MHz performance and I still feel justified based on checking the recent prices of similarly performing DSOs but this is not a path I would recommend unless ones likes diagnosing and restoring old equipment. I was not familiar enough at the time with the various options for old DSOs to go that route.
None of my old DSOs support variable persistence so they are not very useful for eye diagrams and their waveform acquisition rates are too slow for hunting rare glitches unless I plan on a long coffee break. My 7834 is pretty good at that though at 50k waveforms/second.
On the other hand, no amount of sample rate and record length will make up for low bandwidth.
While this is certainly true, the opposite is, too (no amount of analog bandwidth can compensate for insufficient sample rate and memory). A TDS460 seems to sell for somewhere between $200-$350 on ebay (the majority seem to sell closer towards the upper end), which is silly money for an antique boat anchor with such poor specifications. For the OP's use case this wouldn't be a sane investment, it would be money flushed down the drain.
The prices are even more ridiculous when considering that (with a bit of patience) the same money can give you 1GHz/2GHz sample rate and if I remember right 1Mpts sample memory:
http://www.ebay.com/itm/LECROY-9374M-1GHZ-4CH-2GS-S-OSCILLOSCOPE-PP093-2GS-S-ADAPTER-amp-MANUALS-/151363188746?pt=BI_Oscilloscopes&hash=item233df2f80a&nma=true&si=AtmF14dZIgHkPnYTheDg0mBaHOs%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557#ht_1752wt_905 (http://www.ebay.com/itm/LECROY-9374M-1GHZ-4CH-2GS-S-OSCILLOSCOPE-PP093-2GS-S-ADAPTER-amp-MANUALS-/151363188746?pt=BI_Oscilloscopes&hash=item233df2f80a&nma=true&si=AtmF14dZIgHkPnYTheDg0mBaHOs%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557#ht_1752wt_905)
I agree there are higher performance options for the money but disagree that sample rate, equivalent time sampling, or record length matter for eye pattern generation which is one of the better examples of a repetitive signal suitable for equivalent time sampling. Higher sample rates and higher waveform acquisition rates are certainly nice for generating the display more quickly. Segmented memory can help a lot here despite its limitations.
I guess the main reason people pay so much for these Tek scopes is because it has "Tektronix" written on it, and I bet that many people are not really aware of its limitations and expect it to be useable for non-repetitive signals over 20MHz.
I am more familiar with the old Tektronix oscilloscopes so I let others make the LeCroy and Keystone/Agilent/HP recommendations. I think a lot of the goodwill Tektronix has is from their historic and long expired policy of excellent service documentation. Agilent/HP enjoys similar goodwill for the same reason.
I did a quick search yesterday and found lots of suitable used Tektronix and Agilent/HP oscilloscopes within the specified budget.
-
I agree there are higher performance options for the money but disagree that sample rate, equivalent time sampling, or record length matter for eye pattern generation which is one of the better examples of a repetitive signal suitable for equivalent time sampling. Higher sample rates and higher waveform acquisition rates are certainly nice for generating the display more quickly.
Higher sampling rates also help with capturing signal glitches: with equivalent-time, you only capture random points of glitches that may not necessarily register as obvious issues on their own and you end up in a similar scenario as hunting a one-per-million glitch with a 1k waveforms/sec scope. With a full-rate oscilloscope, once you capture the glitch, you have the whole thing instead of only a few uncorrelated outliers telling you the plot MIGHT need to run for a few more hours to see how many more seemingly out-of-place points crop up to connect the dots.
Can equivalent-time work? Sure. But a full-rate scope makes it one less thing to worry about.
-
I agree there are higher performance options for the money but disagree that sample rate, equivalent time sampling, or record length matter for eye pattern generation which is one of the better examples of a repetitive signal suitable for equivalent time sampling. Higher sample rates and higher waveform acquisition rates are certainly nice for generating the display more quickly.
Higher sampling rates also help with capturing signal glitches: with equivalent-time, you only capture random points of glitches that may not necessarily register as obvious issues on their own and you end up in a similar scenario as hunting a one-per-million glitch with a 1k waveforms/sec scope. With a full-rate oscilloscope, once you capture the glitch, you have the whole thing instead of only a few uncorrelated outliers telling you the plot MIGHT need to run for a few more hours to see how many more seemingly out-of-place points crop up to connect the dots.
Can equivalent-time work? Sure. But a full-rate scope makes it one less thing to worry about.
I do not disagree with this although I did not cover it in any detail. High sample rates and high waveform acquisition rates make all of the difference minimizing the time taken looking for glitches. This was one of the reasons I compromised with an inexpensive old analog storage oscilloscope for my glitch hunting.
Lots of modern DSOs support various real time glitch trigger modes but the couple of times I have gone glitch hunting with one, I started off not knowing what I was looking for so I ended up using infinite persistence and relying on raw waveform acquisition rate to narrow the search.
-
I agree there are higher performance options for the money but disagree that sample rate, equivalent time sampling, or record length matter for eye pattern generation which is one of the better examples of a repetitive signal suitable for equivalent time sampling. Higher sample rates and higher waveform acquisition rates are certainly nice for generating the display more quickly.
Higher sampling rates also help with capturing signal glitches: with equivalent-time, you only capture random points of glitches that may not necessarily register as obvious issues on their own and you end up in a similar scenario as hunting a one-per-million glitch with a 1k waveforms/sec scope.
That's valid if you can define what you are looking for and develop a trigger for it. If not then it is necessary (and sufficient) to use eye pattern displays - and ETS is sufficient for that.
Evidence? Testing of communication systems of all types is based on eye diagrams + ETS.
With a full-rate oscilloscope, once you capture the glitch, you have the whole thing instead of only a few uncorrelated outliers telling you the plot MIGHT need to run for a few more hours to see how many more seemingly out-of-place points crop up to connect the dots.
Can equivalent-time work? Sure. But a full-rate scope makes it one less thing to worry about.
If you can afford it, sure. But at higher bit rates a full-rate scope even if it is available is exorbitently expensive.
-
Evidence? Testing of communication systems of all types is based on eye diagrams + ETS.
That used to be the case when real time DSOs were slow but not anymore, unless we're talking about signals in excess of 65GHz (which is the case for a few communication systems, but certainly not all of them). Everything below is covered by real-time scopes these days, and soon this will be true for signals up to 100GHz.
Can equivalent-time work? Sure. But a full-rate scope makes it one less thing to worry about.
If you can afford it, sure. But at higher bit rates a full-rate scope even if it is available is exorbitently expensive.
As I said, real-time scopes are available for up to 65GHz bandwidth and 160GSa/s, so for the majority of cases availability of a real time scope shouldn't be a concern. On the other side, the few sampling scopes that are still made (like the LeCroy WaveExpert 100H and the Agilent Keysight 86100DCA) are those with bandwidths in excess of 65GHz where real time scopes are not yet available, and the price of these sampling scopes very well fit the description of "exorbitantly expensive".
If you want to look at signals from common modern communication systems and buses like SAS/SATA, 10Gbps Ethernet and such then these days you do that with a real time oscilloscope.
-
Evidence? Testing of communication systems of all types is based on eye diagrams + ETS.
That used to be the case when real time DSOs were slow but not anymore, unless we're talking about signals in excess of 65GHz (which is the case for a few communication systems, but certainly not all of them). Everything below is covered by real-time scopes these days, and soon this will be true for signals up to 100GHz.
Can equivalent-time work? Sure. But a full-rate scope makes it one less thing to worry about.
If you can afford it, sure. But at higher bit rates a full-rate scope even if it is available is exorbitently expensive.
As I said, real-time scopes are available for up to 65GHz bandwidth and 160GSa/s, so for the majority of cases availability of a real time scope shouldn't be a concern. On the other side, the few sampling scopes that are still made (like the LeCroy WaveExpert 100H and the Agilent Keysight 86100DCA) are those with bandwidths in excess of 65GHz where real time scopes are not yet available, and the price of these sampling scopes very well fit the description of "exorbitantly expensive".
If you want to look at signals from common modern communication systems and buses like SAS/SATA, 10Gbps Ethernet and such then these days you do that with a real time oscilloscope.
Since I haven't checked, I'll bow to your more up-to-date knowledge w.r.t. max real-time sampling rate. I'm astounded that you could directly digitise a 60GHz radio signal. They almost cease to be electronic (by definition) and become infra-red; IIRC infra-red starts at 300GHz :)
Nonetheless, I wonder to what degree they are exorbitently expensive! I'm also curious about the circumstances in which (presuming you've verified the eye diagram is good) it is insufficient to debug digital signals in the digital domain.
-
If you're in the US, and official company-endorsed refurbs are acceptable, Rigol has a couple of refurb DS4054 (4-channel 500 MHz) available for US$2.99k, over 50% off:
http://www.rigolna.com/clearance/#a_DS4054 (http://www.rigolna.com/clearance/#a_DS4054)
[ Some reasonably sweet deals for other, cheaper stuff there as well. E.g., a DS2202 (now an obsolete model, technically, but I very sound scope) for sub-$900. That's a fair bit less than I paid for mine (yes, I know, I should have just bought a DS2072 and hacked it, but I didn't know that back then :palm: ) ]
Even better:
http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814 (http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814)
That's 400Kwfps, 4 channel, 500MHz, 1M sample memory, colour intensity display
https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/Tektronix-TDS784C_Datasheet.pdf (https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/Tektronix-TDS784C_Datasheet.pdf)
Dare I say that's impossible to beat for performance for under $1K?
Can anyone find anything better?
So, when I saw that the OP wasn't interested in this... I kinda just went ahead and bought it ;D :-+ It should complement my Rigol DS2202 nicely; more channels (reason for buying), more bandwidth, less features. To have it all in one scope would be even more expensive than the cost of both combined. Does anyone have any feel for when the TDS754C was released, and what its original list price was?
-
Does anyone have any feel for when the TDS754C was released, and what its original list price was?
It was first in the 1997/1998 catalog at a price of $19,500 without any options.
-
Even better:
http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814 (http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814)
So, when I saw that the OP wasn't interested in this... I kinda just went ahead and bought it ;D :-+ It should complement my Rigol
The screen on that TDS754C looks a bit funny to me :(
-
Since I haven't checked, I'll bow to your more up-to-date knowledge w.r.t. max real-time sampling rate.
Well, you don't need to just trust me on this, see for yourself:
http://cdn.teledynelecroy.com/files/pdf/labmaster_10_zi_datasheet.pdf (http://cdn.teledynelecroy.com/files/pdf/labmaster_10_zi_datasheet.pdf)
This is the spec sheet for the LeCroy LabMaster 10Zi, which offers up to 40 (yes, forty!) 65GHz channels at 160GSa/s. If you're fine with 33Ghz you can even have 80 channels!
LeCroy was actually the first reaching 65GHz, but eventually Agilent Keysight did catch up (almost, they're only at 63GHz, and this is in 2Ch mode only ;) ):
http://www.keysight.com/en/pd-2392275-pn-DSAZ634A/infiniium-oscilloscope-63-ghz?nid=-32899.1083518&cc=US&lc=eng (http://www.keysight.com/en/pd-2392275-pn-DSAZ634A/infiniium-oscilloscope-63-ghz?nid=-32899.1083518&cc=US&lc=eng)
And Tek? Well, probably thanks to the very successful DMS (Danaher Management System) they apparently managed to be stuck at 33GHz and 100GSa/s for their real time scopes:
http://www.tek.com/oscilloscope/dpo70000-mso70000 (http://www.tek.com/oscilloscope/dpo70000-mso70000)
As to what is technically possible, there is still this:
http://teledynelecroy.com/100ghz/ (http://teledynelecroy.com/100ghz/)
I'm astounded that you could directly digitise a 60GHz radio signal. They almost cease to be electronic (by definition) and become infra-red; IIRC infra-red starts at 300GHz :)
Well, they don't digitize these frequencies directly. Instead, they split the signal into various frequencies which are then down-converted and digitized. LeCroy calls it DBI (Digital Bandwidth Interleaving) and has apparently some patents on that.
Nonetheless, I wonder to what degree they are exorbitently expensive!
I don't know the starting price for the LeCroy LabMaster but the Agilent 63GHz DSAZ634A is listed as starting at $453k.
I also vaguely remember that the LeCroy WaveExpert sampling scope did cost a similar amount when fitted with four channels (although the basic scope frames are relative cheap, the money is in the front end modules).
I'm also curious about the circumstances in which (presuming you've verified the eye diagram is good) it is insufficient to debug digital signals in the digital domain.
There's a nice document by HP (yes, it's older, but even more valid these days when real time sampling rates have gone through the roof since the time it was written!):
http://www.hpl.hp.com/hpjournal/96dec/dec96a1a.pdf (http://www.hpl.hp.com/hpjournal/96dec/dec96a1a.pdf)
But even all this aside, a quick look at the market should show you that sample scopes have no real future. The Agilent/Keysight 86100A Sampling Scope is an old design that (f I remember right) did come from HP before it became Agilent. The LeCroy WaveExpert is similarly old, I think did originally come out around 2002. Despite the fact that the amount and presence of high performance communication systems and buses has certainly not decreased within the last decade, these scopes are essentially niche products that very likely will be taken off the market rather sooner than later. Nowadays, most comms systems and buses are worked on with real time scopes.
-
Even better:
http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814 (http://www.ebay.com.au/itm/TEKTRONIX-TDS754C-COLOR-4-CH-500-MHz-2-GS-s-DIGITIZING-OSCILLOSCOPE-/350952354814)
So, when I saw that the OP wasn't interested in this... I kinda just went ahead and bought it ;D :-+ It should complement my Rigol
The screen on that TDS754C looks a bit funny to me :(
Yeah, I noticed that; might be fun to try and fix, and it has a VGA out as a safety net. Might even rip out the CRT and slot a little TFT in there :P No worries either way :)
-
The screen on that TDS754C looks a bit funny to me :(
All of the Tektronix liquid crystal shutter displays look funny in photographs to me. I have never used one in person though.
-
The screen on that TDS754C looks a bit funny to me :(
All of the Tektronix liquid crystal shutter displays look funny in photographs to me. I have never used one in person though.
Thanks for the hint that this screen has a "liquid crystal shutter";
1. Wow, I had never heard of liquid crystal shutters before. Really curious idea for making a colour display, but a really great way to avoid the resolution limitations and complexities of three electron guns and a precisely lined-up shadow mask.
2. It's absolutely plausible that an interaction between a shutter (whether mechanical or electronic) with the liquid crystal shutter on the screen could cause colour variations over the screen.
3. Furthermore, if you look at the pictures on the eBay site, that's exactly what seems to be happening. The "Tek" is blue in one picture, the other picture it's almost invisible.
I'll update this post when I get it with my impressions, and a long-ish exposure photo to remove these effects.
-
1. Wow, I had never heard of liquid crystal shutters before. Really curious idea for making a colour display, but a really great way to avoid the resolution limitations and complexities of three electron guns and a precisely lined-up shadow mask.
Tektronix originally used the liquid crystal shutter on the 5116 mainframe oscilloscope when operating as a DSO to display blue-green, orange, and neutral.
Apparently they continued to use it in the later TDS oscilloscopes because it was the best solution for high resolution color.
2. It's absolutely plausible that an interaction between a shutter (whether mechanical or electronic) with the liquid crystal shutter on the screen could cause colour variations over the screen.
3. Furthermore, if you look at the pictures on the eBay site, that's exactly what seems to be happening. The "Tek" is blue in one picture, the other picture it's almost invisible.
The oscilloscope is completely functional without the liquid crystal shutter. They have a couple of age induced failure modes.
-
1. Wow, I had never heard of liquid crystal shutters before. Really curious idea for making a colour display, but a really great way to avoid the resolution limitations and complexities of three electron guns and a precisely lined-up shadow mask.
Tektronix originally used the liquid crystal shutter on the 5116 mainframe oscilloscope when operating as a DSO to display blue-green, orange, and neutral.
Apparently they continued to use it in the later TDS oscilloscopes because it was the best solution for high resolution color.
2. It's absolutely plausible that an interaction between a shutter (whether mechanical or electronic) with the liquid crystal shutter on the screen could cause colour variations over the screen.
3. Furthermore, if you look at the pictures on the eBay site, that's exactly what seems to be happening. The "Tek" is blue in one picture, the other picture it's almost invisible.
The oscilloscope is completely functional without the liquid crystal shutter. They have a couple of age induced failure modes.
Just to clarify, I was indicating that I think the funny colours in the picture are purely an artifact of the camera used to take the picture, and not visible to the human eye. It is good to know that it fails gracefully, though.
Looking at the manual, it mentions that you can choose colours for the waves, and it's a full RGB selection. Pretty impressive stuff, I can't quite wrap my head around how they managed that when the only way to completely block light is with 90 degree polarisation. Do they have a two-layered arrangement? That's all I can think of. Also, can you buy the shutters as stand-alone parts? I can think of some project ideas... turn any white 7-segment display into a colour one...
-
Actually it works quite simple. There is 5 bit video coming from the acquisition and text display hardware (the 5th bit says it is a pixel from the acquisition hardware or not). At this point there is the ram-dac which creates the signal for the VGA output. The 5 bit video data also gets stored in a dual port memory. The dual port memory feeds a ram-dac running at 3 times the display rate (180Hz). The 5 bit video data is complemented by 2 or 3 bits telling which color (R, G or B) is displayed. The ram-dac converts that input data together with the intensity settings into the signal which goes to the CRT. On the CRT you get interlaced R, G and B frames at 180Hz but if you move your head a little you can see the image getting seperated.
Regarding the funny display: I have seen displays with decoloration in some areas. Due to the age finding a good one can be very hard. Lot's have failed already and replacements are few.
@David Hess: be happy you've never seen one. These displays are quite dim and they give me a headache. I always converted these scopes to TFT.
-
Actually it works quite simple. There is 5 bit video coming from the acquisition and text display hardware (the 5th bit says it is a pixel from the acquisition hardware or not). At this point there is the ram-dac which creates the signal for the VGA output. The 5 bit video data also gets stored in a dual port memory. The dual port memory feeds a ram-dac running at 3 times the display rate (180Hz). The 5 bit video data is complemented by 2 or 3 bits telling which color (R, G or B) is displayed. The ram-dac converts that input data together with the intensity settings into the signal which goes to the CRT. On the CRT you get interlaced R, G and B frames at 180Hz but if you move your head a little you can see the image getting seperated.
@David Hess: be happy you've never seen one. These displays are quite dim and they give me a headache. I always converted these scopes to TFT.
I understand the driving-the-CRT part; what I'm curious about is the construction of this LCD shutter than can turn red, green or blue on demand.
Also, how elegant is the convert-to-TFT job?
-
Some pictures:
(http://imagizer.imageshack.us/v2/1024x768q90/854/img0482xd.jpg) (http://imageshack.com/f/nqimg0482xdj)
(http://imagizer.imageshack.us/v2/1024x768q90/513/img0433p.jpg) (http://imageshack.com/f/e9img0433pj)
(http://imagizer.imageshack.us/v2/1024x768q90/140/img0478o.jpg) (http://imageshack.com/f/3wimg0478oj)
(http://imagizer.imageshack.us/v2/1024x768q90/10/img0430px.jpg) (http://imageshack.com/f/0aimg0430pxj)
-
Some pictures:
Er, why have you posted those pictures?
-
Some pictures:
Er, why have you posted those pictures?
Because rs20 wanted to know how elegant the TFT solution is.
-
Some pictures:
Er, why have you posted those pictures?
Because rs20 wanted to know how elegant the TFT solution is.
I was curious too. Thanks for posting them. :-+
-
Some pictures:
Er, why have you posted those pictures?
Because rs20 wanted to know how elegant the TFT solution is.
Thanks; now we know what we are looking for!
-
Since I haven't checked, I'll bow to your more up-to-date knowledge w.r.t. max real-time sampling rate. I'm astounded that you could directly digitise a 60GHz radio signal. They almost cease to be electronic (by definition) and become infra-red; IIRC infra-red starts at 300GHz :)
Nonetheless, I wonder to what degree they are exorbitently expensive! I'm also curious about the circumstances in which (presuming you've verified the eye diagram is good) it is insufficient to debug digital signals in the digital domain.
I was going through the list of Rigol's Canadian distributors (looking for local prices on the DS1054Z) and noticed that TMetrix's parametric oscilloscope search had sliders going up to about 500k$ and after raising the low end to 300k$, I was left with this:
http://tmetrix.com/labmaster-10-65zi.html (http://tmetrix.com/labmaster-10-65zi.html)
Around 450k$ for 65GHz / 160GSPS.
-
Thanks, particularly Wuerstchenhund and DanielS, for the comments about the current state of the art scopes.
I like it when I learn something new (even if I could have found it out with a little research!). This forum appears to have a good SNR; long may it continue.
-
I wouldn't recommend this one:
http://www.ebay.com.au/itm/Tektronix-TDS-5054-500MHz-4-Channel-5GS-s-Digital-Phosphor-Oscilloscope-DPO-/360974897202 (http://www.ebay.com.au/itm/Tektronix-TDS-5054-500MHz-4-Channel-5GS-s-Digital-Phosphor-Oscilloscope-DPO-/360974897202)
Just a simple misunderstanding. He said "trusty scope" so they brought him a "rusty scope".
Here's another example of overly hopeful ebay pricing:
http://www.ebay.com/itm/HP-16488B-Standard-Inductor-0-28-H-and-8pF-/261506907064?pt=LH_DefaultDomain_0&hash=item3ce306bbb8 (http://www.ebay.com/itm/HP-16488B-Standard-Inductor-0-28-H-and-8pF-/261506907064?pt=LH_DefaultDomain_0&hash=item3ce306bbb8)
Anyone want a badly smashed up 'standard inductor' for any price?