EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: KedasProbe on April 03, 2014, 08:11:41 pm
-
Any idea when we can see this in our scopes?
There is 10 bit vertical resolution, so the resolution is there now we also want to see it better.
I'm not saying they have to change the resolution of the scope screens, just add support to connect a hi-res monitor and see high res and not just a scaled up version of its own screen.
Some work that a student did, to show that it isn't really a big deal for developers to implement it and not very expensive, it just needs to be done.
http://web.mit.edu/6.111/www/f2013/projects/kramnik_Project_Final_Report.pdf (http://web.mit.edu/6.111/www/f2013/projects/kramnik_Project_Final_Report.pdf)
edit: Hameg entry level has DVI-out but the resolution is the same as the scope screen.
-
Any idea when we can see this in our scopes?
There is 10 bit vertical resolution, so the resolution is there now...
Where is the resolution, exactly?
I'm not trying to be facetious. I realize there are high-rez scopes that are relatively low BW and sample rate. Primarily instrumentation and DAQ applications (I'm thinking stuff like Yokogawa). And high-rez fast scopes that cost an arm and a leg (I'm thinking Teledyne LeCroy, for example). But in the general-purpose scope arena, they all seem to be 8-bit, and even then, usually don't have an ENOB of 8.
Having an HDMI output at 1080p on these will provide no real benefit.
...we also want to see it better.
Who is "we"?
[I'm not knocking you, I'm just curious.]
-
10 bit or even 12 bit from 8 bit ADC comes from processing multiple samples.
And yes for some settings it will just be 8 bit (256) shown on 1080.
(my current scope screen can and is showing 9 bit)
And on the horizontal resolution, nobody can be against 1920 points even if it's only 8 bit vertical.
"we" well if it becomes mainstream then all scope users, who can be against seeing more info of the signal on the screen.
-
At 1080p you will be bound by the refresh rate of whatever video controller is needed, say you can do 120Hz but the hardware can sample way more than it can be displayed.
I rather see more waveforms per second than high res, but that's me.
-
Nice find, but that's my 27" screen and setup so I know exactly how it looks like. ;D 8)
Yes they have but only the scope screen resolution.
I also suggested to Hameg to get their virtual sceen completely (no scrolling) on the DVI-out for their next design.
-
At 1080p you will be bound by the refresh rate of whatever video controller is needed, say you can do 120Hz but the hardware can sample way more than it can be displayed.
I rather see more waveforms per second than high res, but that's me.
Yes, refresh rate will stay the same (no point in going higher) but more info will be visible by each refresh.
-
10 bit or even 12 bit from 8 bit ADC comes from processing multiple samples.
Not really. Yes, you can process multiple 8-bit samples, to accrue additional bits of rez. Then, where do you go with it? The buffers in current scopes only hold 8-bit samples, so all you've accomplished is filtering, not rez extension. It's not like on my (old) LeCroys, where the inputs are 8-bits, but it has 16-bit internal sample buffers. That's a different kettle of fish, and you get real resolution enhancement there.
"we" well if it becomes mainstream then all scope users, who can be against seeing more info of the signal on the screen.
I'm not completely sure about that, however, that's not what you said. You said, "...we also want to see it better." As in, right now, current tense. Not some day in the future, "IF it becomes mainstream", after it becomes available and people eventually do (or don't) decide it has some value to them.
That's why I asked who this current "we" is you were referring to. Thanks for clarifying that they don't actually exist. Which probably helps explain why what you're asking for doesn't already exist either.
-
That's why I asked who this current "we" is you were referring to. Thanks for clarifying that they don't actually exist. Which probably helps explain why what you're asking for doesn't already exist either.
Most likely you also switched to 1080p TV instead of 480i so don't try to convince me you are the only one who likes to keep the screen resolution low.
About 8 bit storage only, hameg is giving me 9 bit now (well, since 2011) so ask them how they work, I don't see the problem. (not the same memory buffer)
-
60 Hz it's 16.66(6)ms per frame so I guess if you can update the buffer at 8.68us per pixel (up or down) without incurring delays, then wow!!!!
maybe a custom video that rasters vertically and then left to right without waits for refresh?
But that hardware doesn't exists, even an FPGA will have trouble keeping that signal synchronized. I guess you can delay the signal and always be a frame in the past, that might do it with some triple buffering on the display.
That will add some cost I'll say.
-
4 times 8 bits plus margin fits perfectly 1080 :-) Or do you got single channel DSOs? And the extra space would also help with the 16 ch logic analyzer option.
-
About the GPU power I'm not sure yet how much it would change:
If you have 300 waveform layers and you change that to 3 * 100 waveforms (3 after each other due to 1920)
Then didn't you require the same number of calculations.
-
yes, same number but the delay between them is shorter so it's achievable. But again, HDMI TX chips work like the old TV's scanning left to right from up down.
To update a scope you want them to work from up to down from left to right, but they don't make those and pixels are not addressable without waiting for the clock to tell you to turn the pixel on or or.
-
Doesn't that just mean you have to wait 1/60 sec before you can serialize your image to HDMI.
Brings me to thinking how much delay is there NOW between capturing and seeing it on the screen, and do we really care as long as it is below a certain max. like 1/10 sec.
There is an interesting test setup in something like that with a fast cam...
edit:
https://www.eevblog.com/forum/testgear/dso-display-delay-and-discontinuous-scope-behaviour/msg426331/#msg426331 (https://www.eevblog.com/forum/testgear/dso-display-delay-and-discontinuous-scope-behaviour/msg426331/#msg426331)
-
That's why I asked who this current "we" is you were referring to. Thanks for clarifying that they don't actually exist. Which probably helps explain why what you're asking for doesn't already exist either.
Most likely you also switched to 1080p TV instead of 480i...
So long ago that I can no longer remember. :)
...so don't try to convince me you are the only one who likes to keep the screen resolution low.
Well I certainly won't try to convince you of anything. Especially since I never claimed I "like to keep the screen resolution low". |O It would be great to have an adult conversation, respecting differing points of view. Without people trying to turn things into something confrontational, by putting words in others mouths. :(
Since you seem to be having difficulty comprehending my comments, I'll just clarify that I enjoy having larger screens, even at the current resolutions! I wish all my scopes had VGA or DVI/HDMI outputs. And I'd love to have higher resolution as well... not to see the signals in more detail :o, but to be able to see more channels of them on screen in a reasonable size. With instruments now supporting 4-analog real-time channels (or my Yokogawa, with 8-analogs), plus Math, plus References; along with 8 or 16-digital channels, and then Decoder channels (sheesh!)... current screens with 400p or 480p are rapidly becoming woefully inadequate (even 800x600 is strained, which is why R&S has moved to 1024x768 on their higher-end models).
About 8 bit storage only, hameg is giving me 9 bit now (well, since 2011) so ask them how they work, I don't see the problem. (not the same memory buffer)
You don't see the problem, because you don't want to see the problem. Of course it's possible for scope manufacturers to redesign their implementation, to provide buffers with more bits. I already commented on how my (old) LeCroy did exactly that... what? 20 years ago? And that was a full 16-bits, not the wimpy 9-bits from your Hameg. :P If they perceived a need (that people would pay for), they would. And some are testing the waters.
Teledyne LeCroy already has expensive hi-rez scopes with lots of bits of resolution. And 1280x800 screens. Not very affordable though, and not mainstream, because most folks don't need it, and can't cost justify it. Why don't you just buy one of those? Instead of posting hypotheticals here? I'm guessing the $$$ factors into that.
You originally asked the question of when we're going to see something you'd like to have happen, and why it hasn't already. I simply tried to respond to those inquiries with an intelligent appraisal of the reality of how things are. I'm sorry if you were offended by that.
-
OK, that makes sense, so we like to have higher resolution, which reason is more important than the other is personal taste.
About 8 bit storage only, hameg is giving me 9 bit now (well, since 2011) so ask them how they work, I don't see the problem. (not the same memory buffer)
You don't see the problem, because you don't want to see the problem. Of course it's possible for scope manufacturers to redesign their implementation, to provide buffers with more bits. I already commented on how my (old) LeCroy did exactly that... what? 20 years ago? And that was a full 16-bits, not the wimpy 9-bits from your Hameg. :P If they perceived a need (that people would pay for), they would. And some are testing the waters.
I could be wrong but I think you are overestimating the requirements to implement a high resolution image.
only your 1920 points need to be 10bits or more, your big sample memory is still 8 bit like in my hameg. (the same for Rigol)
But I do see that it's also possible (and likely more expensive) to pre-process everything to 16 bit and then do analysis on this set of data instead of on the 8 bit data set.
I assume the biggest reason for not implementing it yet is that they don't want to make 2 images, one low res for the scope and one high for the output. (scaling text to low resolution would make it look bad)
In this case as soon as the 1080p 8inch tablet screens become cheap enough to put in a scope we will also see the 1080p out connector.
You originally asked the question of when we're going to see something you'd like to have happen, and why it hasn't already. I simply tried to respond to those inquiries with an intelligent appraisal of the reality of how things are. I'm sorry if you were offended by that.
Yes, I asked 'when' because I'm sure it will come sooner or later so I wanted to check if someone knew something about current development plans of new scopes.
I'm checking to buy a second scope with more memory and higher waveform update speed.
I would be very annoyed if after I buy one now they release a new series soon with high resolution out, I can wait 1 or 2 years (it's only a second scope)
Compare it with the evolution of the nice screen of the Agilent 34460A DMM.
Do you really need it, no, but for the right price people will pay for it and Agilent took the next step.
-
I could be wrong but I think you are overestimating the requirements to implement a high resolution image.
You could be right! ;) There is more than one way to handle it, and some are more 'cost-effective' than others.
only your 1920 points need to be 10bits or more, your big sample memory is still 8 bit like in my hameg. (the same for Rigol)
OK, yes. I see what you're getting at. All the large sample memory could stay the same, but do processing (e.g. averaging, resolution enhancement) into a smaller pre-display memory buffer.
But I do see that it's also possible (and likely more expensive) to pre-process everything to 16 bit and then do analysis on this set of data instead of on the 8 bit data set.
That's what my old LeCroy does, but it came out of a nuclear physics lab. And it's screen is vector graphics, at I believe 1024x1024. Others (like my Tek 460A) offer rez-enhancement to 9, 10, 11, or 12 bits (at a corresponding loss of bandwidth), and is a digital scope, but it renders to an analog screen (at 640x480, with a VGA out). I don't know the internal memory architecture of this one though, if it even stores more than 8 bits.
The newer DSOs have LCD screens at typically 800x480 these days, and are carefully designed to optimize the speed at which the 700x400 trace area gets updated. The memory architecture of these is 8-bits throughout.
I assume the biggest reason for not implementing it yet is that they don't want to make 2 images, one low res for the scope and one high for the output. (scaling text to low resolution would make it look bad)
Yes, that would be a major factor. Currently the on-board screens have rez much lower than 1920x1080. And outboard displays don't seem to get used very much, if they're even available. On some scopes you can get the Option at a $$$, others (like Hameg) come standard, and others have no way to get anything at all out. So how much time and money do the manufs want to spend on something that they perceive many/most won't use?
A big issue though is the current push to optimize waveforms/sec, which Agilent seemed to pioneer, and others are now on the same bandwagon. With that as a primary goal, other considerations (like your hi-rez preferences) will fall by the wayside. They've already dropped previous functionality like Alternate timebases, and digital LPF/etc. filtering, because they hampered the wfms/s rates.
In this case as soon as the 1080p 8inch tablet screens become cheap enough to put in a scope we will also see the 1080p out connector.
I agree that if and when they upgrade to new screens in this class, that having a 1080p HDMI output will definitely follow in short order. At that point, all that will be required will be a relatively inexpensive video driver output... much as it already is in existing scopes with external display outs. When that's the case, why not?
Yes, I asked 'when' because I'm sure it will come sooner or later so I wanted to check if someone knew something about current development plans of new scopes.
I'm checking to buy a second scope with more memory and higher waveform update speed.
I would be very annoyed if after I buy one now they release a new series soon with high resolution out, I can wait 1 or 2 years (it's only a second scope)
Ah, thanks. I didn't understand the motivation behind your question. That's a tough one. I'm not 'someone who knows', because I'm not a designer working on new scopes. And likely anyone who really has a clue isn't going to say anything about it here. So all we can do is look at past time-frames, and current trends. And try to extrapolate. It seems unlikely (to me) that we'll see such improvements in the 1-2 year time frame you mentioned. That's like, what have they been working on already, and will be releasing in those 2 years? Development cycles are much longer than that. 5 years out though, I wouldn't bet against it.
Those changes could come first from innovators with lots of resources, selling expensive products. That would include manufs like R&S, Teledyne LeCroy, and Agilent. Whether anything they come up with is anywhere close to affordable by you or me though, is another story. Or they could come from the other end of the spectrum, from someone like Rigol who has made deep memory sampling (once far out of reach) very affordable. They could decide to leapfrog the competition, and do something similar in the display arena.
Compare it with the evolution of the nice screen of the Agilent 34460A DMM.
Do you really need it, no, but for the right price people will pay for it and Agilent took the next step.
I will agree with you on that. Manufs are always looking for 'product differentiation'. As technologies evolve, they try to gain an advantage by incorporating cool new stuff. Which is good, because that's how things advance. And I'll also agree that it's plausible we'll see higher-rez screens eventually, but probably (just my opinion, and possibly wrong) motivated more by a need for more real-estate, rather than displaying finer than 8-bit resolution signals (as the primary reason).
One thing slowing that down will be the need for faster graphics processors, since, just like on tablets, it takes more horsepower to update the screen, the more pixels it's packing. They're not going to degrade the screen speed and performance just to make for prettier images. :)
You can even see the influence of those considerations impacting things like the difference between the Rigol DS2000 and DS1000Z series. The 2000 has 14 horizontal divisions (700-pixels), with a collapsible side-menu. While the 1000Z has only 12 h-divisions, with 600 pixels, and a non-collapsible side-menu. They certainly could have done the same on the 1000Z as the 2000 (both 800x480), but with only 6/7 of the display area to update, they can use slower and cheaper parts. And thus make the scope more affordable. That's always going to be a big factor, especially on the low end.
-
I'm checking to buy a second scope with more memory and higher waveform update speed.
One big problem I see there is 'higher waveform update speed' is at odds with 'lets update 2M-pixels'.
But since you want to use a large external 1080p screen anyway, and are looking for more memory, you may want to consider something that doesn't already have it's own screen (and the constraints associated with that).
I.e., one of the high-horsepower USB scopes, like a pico-scope, or the cleverScope. They've got high-rez rendering, high-rez acquisition options (10, 12-bits), high-depth captures, multiple windows simultaneously on one screen, etc. And they've at least got the potential for high waveform update speeds inside the acquisition module, although the actual screen update rates will be more limited (due to the decoupling).
In which case, the answer to your question: "Oscilloscope with 1080p HDMI/DVI out, When?" would be... right now. :-/O
-
About graphics power
If you look at the 1080p game speed on battery powered devices I don't think graphic power is a problem.
We would just need a good 'game' engine to play the electronics games ;)
This APU would be overkill (4.5Watt), but even dual display would be an option, or even recording your scope screen as video file and watch again later.
http://www.amd.com/Documents/49282_G-Series_platform_brief.pdf (http://www.amd.com/Documents/49282_G-Series_platform_brief.pdf)
(maybe better wait for a DDR4 APU)
-
I.e., one of the high-horsepower USB scopes, like a pico-scope, or the cleverScope. They've got high-rez rendering, high-rez acquisition options (10, 12-bits), high-depth captures, multiple windows simultaneously on one screen, etc. And they've at least got the potential for high waveform update speeds inside the acquisition module, although the actual screen update rates will be more limited (due to the decoupling).
In which case, the answer to your question: "Oscilloscope with 1080p HDMI/DVI out, When?" would be... right now. :-/O
Yes, I haven't ruled out this option although I prefer a not connected to PC scope.
edit: 1280 *800 and HDMI (high resolution scope)
http://jaunty-electronics.com/blog/2013/10/look-inside-teledyne-lecroy-hdo4024/ (http://jaunty-electronics.com/blog/2013/10/look-inside-teledyne-lecroy-hdo4024/)
http://jaunty-electronics.com/blog/2014/03/apec-2014-lecroy-power-measurement-demo/ (http://jaunty-electronics.com/blog/2014/03/apec-2014-lecroy-power-measurement-demo/)