Oscilloscopes reach 12-bits, finally

Not if the noise remains that bad. Did they have any comment as to why the SNR in your tests was only 55dB?

Part of what I do for a living is designing front end sections for ultrasound machines. There we'd never waste so many bits. We always make sure that the SNR is way above what the converter can do.

--
Regards, Joerg

http://www.analogconsultants.com/

. . e

I recently received the following in the email. Any rumors about what is different about Tektronix's revolutionary oscilloscope?

"An unexpected solution. One remarkable new scope.

This summer, Tektronix will unveil a revolutionary oscilloscope with functionality you've never seen before. And you're invited to the exclusive event. Help us launch the oscilloscope revolution.

Online Launch Event August 30, 2011"

HJoward

I can usually get a few weeks in advance "confidential" brochures for these sorts of new releases from my favorite instrument vendor sales reps.

But they know I'm down on Tek these days, so probably didn't bother to bug me about it.

Since it's only a few more days, I'll just wait and see.

--
_____________________
Mr.CRC
crobcBOGUS@REMOVETHISsbcglobal.net
SuSE 10.3 Linux 2.6.22.17

Well this is at 2Gs/s, so I give them some credit. But it is a bit disappointing.

It took me to the end of the session before I had a chance to get the numbers. I'll have to email the sales manager about this and see what they say.

It might be more important that higher resolution bench scopes become accepted and expected by the marketplace. Then once the big makers are competing with each other, hopefully the performance will converge toward goodness.

Just curious, but by how much is it cost effective to make the front end amp higher in SNR than the digitizer?

--
_____________________
Mr.CRC
crobcBOGUS@REMOVETHISsbcglobal.net
SuSE 10.3 Linux 2.6.22.17

Not really exciting. 8-bit scopes usually work well enough. The other issue is whether the noise and settling dynamics justify 12 bits.

John

You can usually move back and forth in time on a captured trace with a digital 'scope. 12 bits will let you zoom in and out on the y axis too.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Hi Mr.CRC,

Oh I remember the tds820 I used in 2000. 14 bit resolution and in comparison to the lecroy crap excelent low noise. No, I didn't quantified it but I remember the fantastic zoom function in Amplitude and time base

But not the lecroy 4 channel windows-stormy-box device I had to use in

2007. 7 kEUR for a thick line in each channel. Without digital filter and bandwith limitation there was nothing to see than noise.

But of course I do not need GSps.

Marte

Marte

Yes, I do.

I have been looking at some of the new offerings (not lecroy though). I do like the R&S RTO series although they are very expensive.

They will do "real time" signal averaging and DSP at the full speed of the digitizer, like 10GS/s. So for normal lower frequency use this can vastly reduce the displayed noise level. Also there are DSP low pass filters that can be used which look very useful too. So for low level work you could enable a 100kHz low pass filter and easily resolve sub-millivolt signals. One problem was they did not allow low enough settings to take full advantage of this, was limited to 5mV/div say. But I gather this is now fixed (possibly due to my feedback!).

It would be interesting to try this on a scope with 12 bit underlying resolution.

--

John Devereux

50KSa/s though. And only 50 ohm inputs, no attenuators.

Nice, a sampling scope though. (I didn't realise the TDS820 was a sampler).

--

John Devereux

Ok, yes 2GHz is tougher than what I have to do. But it isn't such a big deal to make a wideband amp for a TV distribution that goes clear to

800MHz and has to retail for $29.95 including enclosure, power cord and blister pack. So it ain't rocket science. But the oscilloscope folks must weave in DC-stability and a Hi-Z input which isn't trivial.

At several hundred MHz I'd think it would be ok to give up Hi-Z performance. Even in my wildest dreams I would have everything at 50ohms up there. In fact, my fast scope has only 50ohms inputs on all four channels.

I never found it to be a major cost factor, the AD section, FPGAs, DSPs and all that always dwarfed my BOM total. In the analog part one must not restrict the design to chips. Sometimes you can get enough dynamic range out of them so many of my front ends are largely discrete parts.

--
Regards, Joerg

http://www.analogconsultants.com/

Sorry, but it doesn't work that way. You can't get more bits out of a converter by just filtering the signal. Just imagine a pure DC signal sitting somewhere between two levels. Filtering won't help.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

.

LeCroy certainly knows more about making a data acquisition system than the other scopemakers.

But I think that we as designers and engineers often confuse scopes with data acquisition systems.

I suspect there is a real market for those of us who dwell in the confusion.

Tim.

t .

The tek guy gave be a brief overview. They are integrating a RF capture and analysis capability into a MSO. Fully time correlated signal analysis. Not being an RF engineer, my take away was that the scope could pick up RF signals and demodulate them, display them, (a hint at spectrum analysis but he didn't say those words), correlate them with digital processing that might be needed., If I remember correctly, he said something about having most of a SDR in within he scope... He said the cost would probably 'be expensive'....

John

Probably for 4G or whatever cell phones and similar complex modulation stuff. I bet it will cost in the 6 digits.

The old-line scope makers (Tek, Agilent) are being squeezed out of the low end by cheap asian scopes, and are battling for high end against newcomers like LeCroy and R&S.

John

Tell that to the guys that make sigma-delta ADCs :)

It's not pure DC, thats the point. It's DC with a lot of noise added, which dithers the signal over many lsbs. Of course there are inaccuracies which is why I would be interested in a 12 bit version.

--

John Devereux

Mr.CRC schrieb:

Hello,

the ideal 9 ADC has 54 dB of dynamic range. They should call it a 10 bit ADC, not a 12 bit.

Bye

Joerg schrieb:

Hello,

I would like to offer a "t" for free:

"Sometimes you can't get enough dynamic range out of them so many of my front ends are largely discrete parts."

Bye