EEVblog #587 - Tektronix MDO3000 Mixed Domain Oscilloscope Teardown

What's inside Tektronix's new MDO3000 Mixed Domain Oscilloscope?
Dave does a teardown and looks at the circuitry and how they can get such a cheap 3GHz RF spectrum analyser to work.
Datasheets:

EEVblog Main Web Site:
EEVblog Amazon Store:
Donations:
Projects:
Electronics Info Wiki:

[profpep]

'Touch sensor' could be a condensation detector - I've seen similar things in old VCR's. Won't let the thing fire up if it has been brought in from a cold place and got condensation inside. Just a thought.

[redtails]

6:22 the cable management is fantastic. Even though this thing is never ever going to be repaired (at most, whole board replacement), I admire their dedication to making it easy to open

[tubical]

Refering to 05:10 -- As I worked as a PCB designer we had used the PCB itself to act as a RF shield for the underneath circuits. I used to have some ground planes within the PCB-layers to prevent RF electric fields gone straight through the PCB without beeing damped. But this do not work very well with magnetic fields. So one have to use other mechanical solutions go shield them (MuMetal works good but it´s not easy to deal with and it´s expensive as well)

[Almostanaccountname]

I'm glad I stumbled upon this channel.   I calibrate electronics for the AF and I use a lot of the equipment you use in these videos. It's awesome to learn more about it, you are very knowledgeable.

I've only been doing my job for about 1.5 years there is still so much to learn though... hope to be as knowledgeable of my job someday as you are.

[MundoDaEletrica]

Congratulations! Amazing video!
Really nice to have access to an equipment like this.
It is possible to feel how much you love electronics. You are an inspiration to me.

[retrotails]

yay, time for a long video again finally. I'm not super interested in the tech, I just kinda leave these videos up on my 2nd monitor while I do other stuff, he's like the bob ross of youtube.

[NerdNordic]

I really liked your explanation on the RF magic! 

[markbell]

I think your best one: Great quality video, audio and photos.  Great content and a butt-load of links.  Cheers, Mark

[DjResR]

That bodged SMD part is soldered between GND pin of the chip and via going somewhere. Thanks for teardown.

[ChipGuy]

The chip at 39:29 is a Vishay Siliconix DG411L Precision Monolithic Quad SPST Low-Voltage CMOS Analog Switches.
Nice teardown :)

[artforz]

47:50 those "360" signal switches in the spectrum analyser front-end appear to say 09E or O9E on them. Doesn't matter much, clearly just boring SPDT RF switches.
59:40 pretty sure it increases, not reduces SFDR.

My $0.02 on the "marginal" thermal performance: the ADCs are probably in the double digit W a pop. 2 minutes to critical junction temp with those tiny heatsinks and no forced airflow sounds about right.
As for the "last shutdown was due to overtemp" message... what's the alternative? "come back and find your scope shut itself off for no apparent reason"?

Otherwise great teardown, can't wait for the review :)

[fragalot]

So that's where my boombox from the 80s went too.. someone made it into an Osciliscope!

[oriole]

It seems pretty good for the price. The crippled spectral bandwidth is a bit of a shame since unlike time domain, you often want to see higher frequency components to things. The lack of simultaneous time domain and frequency domain display is also troublesome.. I suppose this scope can still do a regular FFT on each channel so that you can get a simultaneous display of both, albeit without the low noise floor of the spectral analyzer path? I was considering this scope but hmm not so sure. The MDO4000B seems like a better long-term investment. Looking forward to the review, thanks Dave!

[electronalchemy]

The "touch sensor" at 17:20 might be a moisture / condensation sensor. They seem increasingly popular in consumer gear now. 

[JohnFHendry]

16:45 absolutely amazing... looks like testing and quality control with the needs of the technician on a time budget were addressed as the end user won't be operating that switch. I just love a GOOD mystery! How cool... maybe it's a new form of advanced logic acupuncture using DNA/RNA phase timing loops and patterns where you just put your thumb on it and wait as preprogramed needle points preset using the test leads take effect. You'll probably have to update the firmware first... and that will cost a little extra.

[flaminggoats]

Could those RF switches at 48:28 be 09E not 360

[Petertronic]

I think that the pad at 17:00 is a condensation sensor too (I know someone already said that). Only one way to test it Dave ;)

[africanelectron]

Could that squiggly pad thing be either a test point to test the capacitance of the board or some sort of compensation device that measures the capacitance change of the board due to age, dust or heat ??

[rougenaxela]

Looks like a Rakon oscillator there at Y702. I've seen that interesting sort of hybrid oscillator package for some nice TCXOs before.

[maeanderdev]

Hey Dave,
i assume your Heat issue with the opened unit occured due to the fact that those little Heatsinks have a very low thermal mass causing the tempreature to reach its constant level very fast.
Also the thermal resistance to ambient is many times higher with the fan not blowing straight on it.
I finally assume the constant temperature of probably about 90°C with the fan blowing on it to rise by at least 5 times to something unreachable like 450°C, so if you now take a look at the thermal resistance and the thermal capacitance you get something like a RC-network in electronics where the rise can be assumed as e^x and therefore to complete my assumtion to let the temperature plausibly rise that far from ambient to the thershold within even less than a minute

BTW: great BOM-calculation at the end of your video I was very surprised, that such a spectrum analyser frontend would only add 50$ to the BOM...

greetings from Germany,
Mäander