TSP  

Nice job, Shahriar. A few comments: At about 12m you stated that the memory and sample rate drops when you add all channels. That's not true in this instrument. You retain full sample rate and memory depth on all channels. I like the idea of pluggable mixer modules. Of course, the MSO6 goes to 8GHz and thus could do this measurement without down-conversion. The ADC samples are not interleaved between the analog and RF paths - the same samples are available to both the analog path and DDC path simultaneously - this is why you can get Spectrum View up to the full BW of the scope. At around 22m you commented a few times about not being able to drop the spectrum time exactly where you want it. This only happens when you're using a mouse or touch screen to move the spectrum time. If you use the wave-inspector shuttle-jog control, you can put it anywhere. It's just a little bug in the UI when using the mouse to drag the spectrum time. One more thing that would be VERY interesting to do on this signal would be to double-tap the Frequency measurement badge that you have on, and turn on the Trend trace. Then, you'll see the Frequency vs Time trend of the hop (similar to the FvT trace on the MDO4k). Another handy feature of Spectrum View is that the pinch/pan/zoom all work on the Spectrum View display. Also, the Horizontal position and scale knobs directly control the center frequency and span. Pushing the Position knob will adjust the center frequency to the Reference Marker frequency. Another interesting point about having Spectrum View on multiple channels - you *can* have a different center frequency and a different spectrum time location on each channel - think of the possibilities... As for triggering on RF, MDO capability, etc. - all I can say is that this initial release of Spectrum View is just the tip of the iceberg....

From many here at Tek... thanks for your review ;)

I think it would be neat to do a video of a Tek product like this with W2AEW. He’s an engineer for Tek and lives in the NJ area. From your livestream last week I think you work at Bell Labs. At the lab in NJ? That would be a really cool crossover video to get some inside knowledge as you do a teardown.

Great video, really love this kind of content. It's like getting a tour around a strange city, that happens to be a few slabs of silicon :)

that's a 13000€ scope. Interesting that Tektronix doesn't do the whole "request quote" thing and just publishes their prices, gotta love some transparency. I never understood the principle of non-public pricing edit: derp, the 13k is the 4 channel version

"There's no such thing as digital circuits anymore of course everything is mixed signal and everything is BEAUTIFUL!" ha ha :))

Very grateful for this video ! Well produced !

I'm starting to believe that I could live with my right kidney only. I should be fine...🤤

Note the classy neutral colour scheme for the front panel compared to most manufacturers going for pinks and purples.

Momma was an ADC, Daddy was a DAC. I guess you could say I'm a bit mixed up.

Very cool instrument and analysis. Obscene amount of fast math computation. Thanks for your hard work. Watched at 1.5x haha. Some thoughts: -Product/market differentiation is an obvious business concern in terms of not wanting to enable every conceivable option on every instrument. Tek is very generous as a company, but they have to have a working business model of course. Enabling new firmware features on existing instruments over time, allows a lineup to remain competitive without manufacturing changes. Flexible hardware architecture goes a long way for that when it comes to initial product planning, also giving software/firmware guys more dev time after release. Higher initial costs vs higher long-term returns etc. Yay business things. -I've seen some weird BGA chips before that look like interposers from the top but actually have a die or two on the underside. No bypass caps in your case as you said, so probably not what's going on here. -I've always wondered if there are gov't restrictions for sale/manufacture of products that could function as an extremely wide bandwidth realtime spectrum analyzers, in terms of being used by bad actors to characterize complex radar signatures etc. Kind of like IR cameras being limited to 9fps for US export due to similar concerns of unwanted use. -Realtime wideband capture + multiple receivers for triangulation + "deep" analysis/characterization database + tactical interconnectivity = big current mil. interest for obvious reasons -w2aew's comment talked about frequency vs time trend, it would be interesting to see how "nicely" that's implemented. As a point of interest, one of the scopes I still own is an old 2GHz LeCroy WavePro from 2001, and with enabled options, it can do frequency waterfall display (time vs freq), 64Mpts single waveform capture + 2.5Mpts FFT on any part of captured waveform, which you can scroll/"play back" to see the same frequency hop changes you've shown on the 5-series. It even does vector demodulation with constellation/EVM etc displays. It's ultra awkward and quirky to use and requires high user expertise, and 90MHz cpu running VxWorks is comically underpowered for operating on 64Mpts waveforms haha, but just throwing this out there as a cheap used option if wideband analysis is needed for someone. These days I either use it for radio signal tracking, or for its acq. memory. It's networked, so I export to matlab and run various tools I've built over the years.

Great video, thank you very much.

Maybe the mystery chip is an alternate ADC position for the 4 channel version.

dream scope

at about 25min: some of your features in the FFT must come from the fact that your FFT window contains portions from before and from after the frequency shift. That said your actual switching between the frequencies might be faster that what you read from the transition from clean 150 MHz to clean 200 MHz. Is it possible to zoom into the waveform display once you did the single-shot capture? It would be interesting to see the transition between the two frequencies in the time domain.

Shahriar, if you still have a RTB2004 then you may want to have a play with it's FFT function - they've changed it's operation in a recent firmware release and you can do some of the same tricks as the Tek 5-series (e.g. take the FFT of a movable sub-set of a time domain capture). Like the Tek, moving the FFT window etc can be a bit finicky though. For someone who needs this functionality (on single channel only) it could be a lower priced option. Looking forward to seeing what new toys Tek enable in the future - also would be great if they could send a 6-series to you for a teardown!

Another unusual feature of this scope is that it can decode and trigger(!) on 100Mbit Ethernet. It can trigger on Ethernet frame content. I am however missing two features: * Ability to add custom frame decoding. At the moment it only decodes the Ethernet and TCP/IP headers. * Ability to save many decoded frames and timestamps for analysis.

Does the addition of "Spectrum View" effectively add the "RF Channel" capabilities from MDO scopes like the MDO4000C to this instrument? Thanks for the excellent videos Shahriar!

holy this is a great scope

Very well said

The middle chip is definitively exactly same chip as other two. All extra components and pads have exactly same placement.

it might have 8 channels and 2GHz bandwidth, but i still prefer my ancient 7704 :D

My first thought for the interposter chip was some kind of wiring matrix to connect the memory with the acquisition chips. The idea would be that if you added more memory you would also swap the interposter with one that interleaved the bits in a different pattern. Like RAID 0 with different strip sizes. A modern version of the classic plugboard. en.wikipedia.org/wiki/Plugboard

Hi, quick question, the analog inputs into the ADC doesn't seem to be length matched? It looks like some are longer than the others. Wouldn't it cause problems in synchronization between channels?

Thanks a lot for the video ! I am so lucky to get this scope at work😀. I'd like to know if there is a way to read the saved wafeworfm in csv or wfm offline with a pc. I know there is an official tool but not sure I can provide arguments to buy it. Does anybody know a tool that could help ?

Could the missing chip be a signal generator?

@The Signal Path Isnt Triggering by spectrum cant be done with triggering with rise/falltime?

Is that a PCI(e) slot on that first board in the upper right corner or some kind of custom interconnect for addons?

Now we play the 1080p waiting game :P

If they maid signal mixers to bring the measured signal into the middle of the bandwidth of the out of band oscilloscope. They would not sell the much higher bandwidth ghz oscilloscopes.

Wow

👍👍

Классный аппарат!

Come on, triggering on the SPI signal that you yourself switch the frequency with... The whole scenario only makes sense if you actually had triggers in the frequency domain. It's sad that this feature is missing, since it should be easy to implement (can't be too hard to look if bin with most power changed).

OMG, where did you get the money? I would be homeless to buy one if these. lets stay with 1Gsa/s, long live china !

God damn what a nice Car! I mean scope!

25:38 LeCroy WavePro can do it as well: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-0A6x1FyG5yY.html

That screen is bigger then my TV..

You did a Teardown on a *20 GRAND* scope ? jesus christ . The only way thats justified is if you work for tektronix, Please tell me "the signal path" YT channel is ran by tektronix

The real ADC comes from 6.25G*8bit. It is not real 12bit ADC. It combines two 3.125G*8 bit ADC to get 3.125G*12bit ADC. So the ADC bits drop down to 8 bit at 6.25 GSa/s. TEK is a liar and cheater!!!

You have too much gear.

This looks like small pc motherboards on one pcb. Why this crap is so expensive I don't know