EEV Blog: 199; "Tektronix MDO 4104-6, mixed domain oscilloscope Review"

[Lawsen]

Thanks for the comprehensive review on magnificent and innovative integration of a spectrum analyzer frequency domain along with the time domain oscilloscope functions into one instrument.  I just started working on a temporary, on call, as needed HP giant Eagle Jet plotter repair technician.  With all of those years stuck at school, I still do not have enough service requests to earn a good living, oh. . .  The American economy collapsed and still have not gotten better.  I would not need a Tektronix MDO 4104-6. 
Why:
1. It is too heavy and large for me to pack it to the plotter repair place.  It is light weight that it is four instruments as one.
2. It is too expensive, at $28,000.00 USD and additional $2,000.00 USD for module software plug ins, no.  I am usually paid about $150.00 USD per repair.  I only make about 3 to 8 repairs per year.
3.  Plotters are not RF circuits, just basic switching power supply and stepper motor computers and digital signals, a basic four channel digital oscilloscope would be more than enough for around $315.00 USD from a Hong Kong store.  I use a Rigol 1052E or Atten with 7 inch display, both are inexpensive and displays what I need. 
4. Plotters are really mechanical intensive machines, some control and power electronics. 
5. I do not think; I can be an electronics tech until retirement, because the jobs are all in Asia, not in the U.S.A.  I would not invest that much into an oscilloscope.  If I could earn that much money, that it would be to volcanology school for one summer. 
6. Electronics has been a tiny side income, not a primary income with so much outsourcing to other countries.  Electronics is a declining industry in the 2000s like the specialty garment industry was in the 1980s.  The clothes industry went over seas for cheaper skilled labor.
7. The oscilloscope software have too much lag time and not instant response.  It needs a more higher performance processor.
8.  The USB memory device did not works at first or second try.
9. It is not clear how well it works with the PC. 

Here is another way to look at the MDO 4104-6 mixed domain oscilloscope:
10. It does save bench space that it is four instruments in one.  Yes, it will weigh a tad more, but it is four instruments into one and the controlling, displaying, and storage software are nicely integrated together as one. 

The four threaded tapped holes in the rear of the Tektronix MDO 4104-6 are for mounting the oscilloscope onto a swivel articulating arm to save work table top or bench top space.  You will not lose space on your shelf for this $28,000.00 USD mixed domain oscilloscope.  The large 10.5 cm diagonal LCD display will not likely survive a drop test upon concrete floor, because the large surface area of the glass LCD display.  The smaller the surface area, the more likely it will survive the drop test unbroken.  The Fluke 1,2,3 industrial strength Scope meter with 20 MHz bandwidth are more likely to make it through a drop test. 

I cannot afford an Agilent X.  The Agilent 3062 series in the 60 MHz second hand market is all I can afford in the Agilent brand, which is similar to the Rigol 1052E series at 60 MHz.  The Agilent 3062 is made in China, too.

http://www.newark.com/tektronix/mdo4104-6/oscilloscope-mixed-domain-21-chan/dp/48T2687?Ntt=TEKTRONIX+MDO+4104-6

My favorite Tektronix is an older Tektronix analog 400 series for $100.00 USD.  It is a tad heavy to carry.  I do not think the exotic, modern oscilloscopes like the ones you reviewed are for me.  I view your video as for entertainment than for practical use.  The Agilent X series and the Tektronix MDO oscilloscopes are not for use everyday.  It is like a Ferrari, Tesla S, or the Rinspeed Squba exotic cars.  This review is not for me, more for entertainment.  The MDO 4104-6 is for those who need extremely four instruments integrated into one instrument and one visual displaying tool.  It does have a market, but specialized.  Will it revolutionized the oscilloscope market?  I doubt it, because it is not always needed by a broad customer base.  The design is sure to win some innovation awards for integrating four instruments as one.  I doubt it would be a common concept with all oscilloscopes in the future.

Conclusion:
Tektronix MDO 4104-6 is an innovation, not a revolution.  It is definitely a new inventive concept.  It is not for me.

[alm]

What's wrong with the current thread on the same topic?

This scope is not targeted at field service, scopes in this price class never are. If a cheap scope satisfies your requirements, good for you. Not sure why you need to write a long post to explain this to the world. It's not like Agilent and Tek market these scopes as 'the ultimate plotter repair tool'.

[Computeruser]

I have a TEK 455A2B2 that I purchased new many years ago. It remains in excellent condition and well calibrated as well. But it is a big heavy machine even though "portable". It is worth more to me than the $100 I could get, so I hang on to it when I need an analogue scope.

I also just get a TDS2024C. It is fine machine for me and does what I want (although it is not everyone's cup of tea). But it is lightweight and very portable. And it does more than my 455A2B2. The only downfall I can see for you (from your description) is that it is probably a bit expensive for you (I don't really know, though) because you can only get them new at this point.

... C

[Wartex]

The "OoOoOoOoOhhhhhhhhhh!" whine at 00:17:40 about horiz scale knob is absolutely fucking priceless.

[Computeruser]

Did you not stick around for the RF section?  It was even better than the Horizontal Scale knob. ... C

[Computeruser]

Now I very much know this post is from another era altogether, but the insides of a TEK 7A19 1 GHz Analogue Amplifier are interesting.

7A19_Delay_Line shows the delay is a formed brass tube. The input is the front panel BNC connector and the output is the AC/DC switch.

7A19_Inputs shows the output of the delay line into the white AC/DC/OFF (Precharge) switch. Then its output goes to the black Attenuator cam switch. Turns and joins are via brass coax couplings.

7A19_Amplifier: All done, and now on to the main amplifier and out the backplane to the rest of the oscilloscope.

Still some interesting high speed signal design.   ... C

[vk6zgo]

 I expected it to be a very good Oscilloscope with a so-so Spectrum Analyser tacked on.
 Instead it turned out to be quite a good Spectrum Analyser & a so-so Oscilloscope.

I had a fault where it would have been very useful:-

A PIC told a PLL chip (via a ribbon cable) what frequency it should be on at "turn-on" of a transmitter.
Instead of going to that frequency,the PLL went to a higher default frequency,after producing,(according to our Spec An),a series of discrete frequencies between the requested & default frequencies.

It turned out to be a faulty connector between the PIC & PLL,causing the frequency to go directly to the high default freq.
Our Spec An couldn't sample fast enough & rendered the rapid change as a series of peaks.

The MDO 4104-6 would have been able to trigger on the data ex the PIC,show the PLL control voltage,& the output RF spectra.(& in this case,show that the data wasn't reaching the PLL).
It took us quite some time to find the problem,which would have been immediately obvious.

Then again,Tech's manhours are a lot cheaper than top level test equipment!

It's not all that expensive,though,compared to the new price of Tek 545Bs & 7000 series in their day!

VK6ZGO

[chscholz]

Hmm, it should be really easy to trouble-shoot this kind of issue with a traditional scope:

1) measure frequency
2) track frequency over time
3) increase the number of cycles you measure
any change in frequency at the output of your PLL will become quite obvious.

Use a 12 bit scope if you need better dynamic range.


Chris

[...]
It took us quite some time to find the problem,which would have been immediately obvious.

[...]

Then again,Tech's manhours are a lot cheaper than top level test equipment!

VK6ZGO

[Lawsen]

A viewing the long video about this MDO oscilloscope, that it is a multiple scopes in one instrument.  I asked myself, if it was priced the same price as a Rigol 1052e, would I get one?  The answer is still no, because I do not need the spectrum analyzer.  I would rather have an Agilent DSOX2002A, a 70 MHz, 1 Gsamples/s for $1,200.00 USD.  I can use this DSOX, because it is simple to use, no confusing menus and sluggish, less than responsive on screen command and intuitive.  For those who works with digitally controlled radio frequency products like digital controlled VFO in a radio or stereo, wi-fi sets, and RF amplifiers will see an use in this.  Products development career.  If you are just a service tech on call, hardly any revenue, the answer is no for the MDO, even if it was priced closer to the Rigol 1052e.  I would purchase a DSOX2002A and not look at the MDO.  When I repair a plotter; I like to take the instrument out of the box and just start using it, not focus on the instrument, but on the repair and go home, when done.  It should not be like a research project or home work at the university complication.  MDO feels too much like an university experience and the long video lecture/presentation from Dave Jones.  I was even assigned to repair a bubble hockey game.  Agilent DSOX2002A feels like an intuitive, automatic oscilloscope, as easy to use as multimeter.

[joelby]

Actually, once you've owned a spectrum analyser, you'll wonder how you ever lived without one. It's a frequency counter! It's a power level meter! It's a filter tester! It's a frequency sweep generator! It's a VSWR meter! If it supports vector measurements, it's an impedance matching network designer and cable length/break tester!

All electronic equipment is a bit quirky. Having used Tektronix stuff a bit in the past, I find that you do soon learn the menu structure and don't need to think about it much, and while lag is a bit annoying, once you've got things set up, the measurement functions are fine.

Clearly, a $30000 tool has not really been designed with someone who makes hardly any revenue in mind!

[vk6zgo]

Hmm, it should be really easy to trouble-shoot this kind of issue with a traditional scope:

1) measure frequency
2) track frequency over time
3) increase the number of cycles you measure
any change in frequency at the output of your PLL will become quite obvious.

Use a 12 bit scope if you need better dynamic range.


Chris

[...]
It took us quite some time to find the problem,which would have been immediately obvious.

[...]

Then again,Tech's manhours are a lot cheaper than top level test equipment!

VK6ZGO

Yes,you are quite right!

Unfortunately,the employer did not supply any Oscilloscope at all,so our investigations were mainly limited to the frequency domain.

Once we realised the Spec An was lying to us,we could ignore the supposed extra RF carriers & concentrate on the change of frequency.

If we had an Oscilloscope of any kind,we could have looked at the input port on the PLL whilst watching the freq change,& discover that the data to the PLL was non-existent.

In the end,the problem was found by "wiggling" the header connector wiring in the device.

To use your rather more elegant method,we would have needed a 500MHz 'scope!

Later,we began to bring our own equipment into work to make up for the shortcomings in test gear.

VK6ZGO

[sqkybeaver]

I will be getting my hands on one of the new scopes for a couple weeks, are there any requests specific to the performance of the RF section?

[Lawsen]

The Tektronix MDO is not really for me.  Please, do not take Dave Jones review as a must have item.  It was an excellent introduction and review and asked yourself, is this the right instrument for you?  The review model is not the only model.  There are four models of the Tektronix MDO.  The difference from the cheapest model to the most expensive model is 3 GHz and $10,000.00 difference. 

Here are price examples and various models available in the U.S.A. by Testequity. 

Tektronix MDO-4054-6 is a 500 MHz oscilloscope, logic analyzer 4+16 channels with 3 GHz RF spectrum analyzer for $19,900.00 USD.  That is $10,000.00 USD less than the one Dave Jones reviewed.  If your signals are less than 500 MHz and RF less than 3 GHz, this one should works.  I have not went to RF engineering school, so you will need to asked an RF engineer.

Tektronix MDO-4054-6 is the same as the previous model, but with a 6 GHz RF channel for $22,900.00 USD. It is still less than the one Dave Jones reviewed, but think about it, for $1,000.00 USD more, the spectrum analyzer is a 6 GHz, twice the bandwidth than the previous one as Dave Jones mentioned this in his conclusion of the review.  If you do not have analog signals more than 500 MHz for the oscilloscope, than this is probably the MDO oscilloscope to buy.

Tektronix MDO-4104-3 is a 1 GHz oscilloscope with 4+16 channels logic analyzer and 3 GHz spectrum analyzer for $24,900.00 USD.

Tektronix MDO-1404-6 is the exactly the same one that Dave Jones reviewed and it is 1 GHz oscilloscope with 4+16 channels logic analyzer and 6 GHz spectrum analyzer for $28,400.00 USD. 

These prices do not include the plug ins and extras.

Surf here:

http://www.testequity.com/products/4206/?src=home

I liked their price for the Agilent DSOX2002A for the side job in repairing plotters.  I use spectrum analyzer, but I rather have the cheap vintage HP-3569A with a limit of 20 kHz audio spectrum analyzer for geologic seismic wave viewing, the vintage, obsolete battery powered HP-3569A.  I like it because I have output the wave signals from a seismograph in the outdoors, "bush," or rural places to this small, battery powered, and light weight spectrum analyzer.  If you are helping out as a technician as a side job, then the Tektronix MDO is not likely for you.  The Tektronix MDO is for computer controlled RF signal products development workshop, which businesses like that are few and far in between.  Many dot com businesses closed as venture capital money disappeared with bank failures and now the government and house owners are litigating some large banks and mortgage firms.  If they fail, the FDIC will have a tough time to continue insuring them and the banks could not pay the plaintiffs the money from the litigation over house loan schemes and failed loans and foreclosed houses.  Most of the world's money is lost in the U.S.A. over houses.  We have ghost neighborhoods and rampant poverty and no hope for few years.  The U.S. and NATO lost money over the many wars being fought and U.S. Homeland Defense security is a large expensive that burdens the economy, too.  We, the U.S. will not likely to recover fully for many years to come and I am getting too old to enjoy a recovery in a decade from now.  I feel depressed at the moment. 

If you needed an MDO for just the project, then rent one.  It is still too expensive for many product development places to own one with venture money unavailable.  We are not in the dot com boom.  President Obama's economic analysis thinks the U.S.A. economy will be in a semi-slump state until 2017.  Is this the time to be buying test gear, based upon desire?  If you do not work around RF and just a service things, then the Agilent DSOX2002A, Rigol 1052E, or Atten 7 inch display oscilloscope, or an old $100.00 Tektronix analog oscilloscope will likely be adequate for you.  Dave Jones's reviews are good for educational and entertainment.  For many who do not have the opportunity to attend an university, then this is a good place.  I liked his lecture on how LCR meters work and the mathematical relationship between resistance, capacitance, and inductance could be solved by the knowing just two out of three different quantities.  frequency=1/(2pi*R*C). 

http://answers.yahoo.com/question/index?qid=20110406043237AASsMOt

[Spiro]

Where are high quality pics of FR part? ARJ22A12 is 5GHz relay. 09V is analog switch (something like AS169-73LF). At 20:37 it can be seen 2 different path signal can take. Left is normal line and right is "something". At 20:39 are those varicap diodes at lover right corner? 311/nt6 must be amp. There are 5-6 of them. SIM762H is mixer maybe. ADF4350 is VCO and only component I could find datasheet.
And now interesting stuff. What are white ceramic stuff with brown strip? Attenuators? Lot of them are with 4 pins but there are few with 6 pins. One other interesting thing is E5 at 20:33. What is it?

[eb4fbz]

Where are high quality pics of FR part? ARJ22A12 is 5GHz relay. 09V is analog switch (something like AS169-73LF).

09V is MASW-007107 (2mm 8-PDFN)

At 20:37 it can be seen 2 different path signal can take. Left is normal line and right is "something"

Left line is direct, right line is via a PI attenuator. It's just a bypassable PI att.

At 20:39 are those varicap diodes at lover right corner?

Just SMD RF inductors.

What are white ceramic stuff with brown strip? Attenuators?

LTCC Filters. 4 pins are low pass filters. 6 pins are band pass filters. Mostly from Minicircuits.

One other interesting thing is E5 at 20:33. What is it?

Minicircuits MAR-5 MMIC amplifier

[Zad]

I must admit, I was expecting to see some much more exotic hardware under there. There's nothing that an amateur radio enthusiast couldn't get hold of - in terms of long-term repairability this is probably a very good thing - just very well engineered. I'm not sure I would have placed the output so close to the input attenuators, but it isn't like they are short of screening.

[tom66]

I find it odd that it's made in China. Tek's margins on those units is probably very large ($1,000 - 3,000 worth of components, if that, the most expensive parts probably being the custom ASICs.) I mean aren't their margins large enough to get them made in the USA, like the Tek scopes of old?

My HP 54501A (circa '93) was made in the USA with big parts from Japan (CRT, and CRT drive board) and Malaysia (ASICs/CPLDs.) It cost $3,100 new I believe. Even with inflation that's only $4,700 or so.

Everything is moving to Asia. I'd better learn Chinese before it is too late.

[mikeselectricstuff]

I find it odd that it's made in China. Tek's margins on those units is probably very large ($1,000 - 3,000 worth of components, if that, the most expensive parts probably being the custom ASICs.) I mean aren't their margins large enough to get them made in the USA, like the Tek scopes of old?

They will already have manufacturing set up for their higher-volume/lower cost products, so it probably wouldn't make sense to split between sites - I'd expect they have their own staff out there to keep an eye on quality...

[tinhead]

here we go, there is the missing "10GS/s ADC". There is ADCMP562 comparator, traces (red lines marked) are going from the RF PCB region where the mysterious ADC should be located (blue makred area) this comparator. I bet TEK is using it as "sampling head".