REL #18 Using a CMU200 in the 2020s

[@douro20]

In the early 1990s they built their digital hardware around Transputers. I read a bit about a system which used a 486 processor as the front-end and five Transputers which were used for processing the signals and presenting them to the 486 which ran the GUI. The 486 only ran the front-end OS and the GUI; everything else was done in the Transputers. There were other systems which used a different front-end processor but they also used Transputers.

[@mikesradiorepair]

I have a stack of 5 of these. I bought them when they first hit the market and no one realized they were good general communications test sets that could be used for amateur radio repair. I think the most I paid was $450 USD for one of them. The rest were around $350 USD. Within a few months the prices had sky rocketed to a couple of thousand dollars. They have their limitations but in general they are a great instrument that can be had for a tiny fraction of their original cost. I contacted my R&S rep and inquired about the original sales price of my units. The one I paid $450 for originally cost $189,000 new as optioned. I paid less than a penny on the dollar for what they cost new. The R&S CRTU is basically the exact same instrument with the exception of a different connector configuration. Those are the ones with 4 more N connectors running vertically up the right side on the front instead of the DB9 connectors. To expand the capabilities of the spectrum analyzer function there is a nice little program you can run on a computer to add some of the missing features you would normally find on a stand alone spectrum analyzer. One is a tracking generator function, a bit slow on the sweep speed but it works. Also implements things like automatically finding the 3dB down points (dBc). I don't want to post a link because my post may get rejected but if you search on KZbin for --- VMA Simple Spectrum Analyzer for CRTU - Implementing Tracking Generator Function --- you will find a guy that make a great little program you can download for free. You just need to connect the CMU200 / CRTU to a computer via RS232 or GPIB for automatic control. I still use my CMU200's from time to time but upgraded a couple years ago to a brand new R&S CMA180 Radio Test Set. Still enjoy using the CMU200's though. Almost forgot to add, the main bottom interconnect board is called a "Back Plane Board". 73's Mike KC3OSD WRCH578

[@rudiselectronicslab861]

Thanks, Mike. Yes, $350-450 is about the sweet price, I'd say. I have not tried the alternative code people created for the CMU, even if its on my agenda. >> The R&S CRTU is basically the exact same instrument with the exception of a different connector configuration. I think there is more to that… The CRTU is a full family of devices, focused on somehow different usage. These cover a wide range of protocol test scenarios beyond the CMU, including handover tests and multicell tests, where some scenarios include up to 8 CRTUs (where one acts as a master, and other units are conceived as slave units that add extra channels rather than being standalone protocol testers). Hardware-wise, the CRTU differs from the CMU in several ways, one of which is that (at least a part) of its system runs on Windows 2000 (whereas the CMU only runs on MS-DOS 6.22). The CRTU brochure, its specifications sheet, and a number of EEVBLOG threats offer more detailed information, and I also would like to refer to the IO Group threat "CMU200 CRTU differences".

[@frankfournier5183]

Hey Mike, I just recently bought CM 200. Can it measure frequency? Or did you have a beat frequency oscillator to measure freq.

[@derkarhu5079]

i think that one bit that not everyone understands about test equipment (i worked for Tektronix, Agilent, Nokia, and R&S), is that the test equipment performance needs to be ~10x *better* than the mobile phone or base station being tested! *10x BETTER* That's a 'non-trivial' task, especially in the early days of standards development, and when it's being done with components, not in ASICs that will be produced in 500 million quantities. These CMU's needed to be built to much higher standards than 'benchtop' test equipment, because they could not have *any* EMC issues internally, let alone affecting tested equipment, or a lab environment...imagine having to be 10-20x bettter than FCC requirements, at any point inside or outside the equipment!

[@alklapaxida850]

and its this idea that makes this equipment so great and i am honoured to have it.

[@raymondbruns3291]

Hoi Rudi, Thanks for this nice presentation. I also am a proud owner of a version 02 with B41 installed. One can Loose One self in all the options and different ways one can use this device. I'm looking forward to your next video.

[@rudiselectronicslab861]

Thanks! In the last week, I had a bit more time to explore the CMU200, and it indeed looks like having seemingly endless possibilities and use cases... Quite an intriguing and amazing machine, I start to like it more and more. Will try to cover some in about 4 upcoming episodes, focusing on general-purpose use (as opposed to specific standards like GSM, WCDMA). Let me know if you have specific interesting use cases as well!

[@vincentnonnenmacher9352]

Great ! Hope to see a serie episodes on this gear. And perhaps dig deeper on how to use them outside telecom test suite. E.g how to connect them to control them from outside, using their huge librarie of functions they seems to expose for custom tests requirements. Thanks

[@jeanaimarre8605]

Nice presentation as usual. Very informative and accurate. Thanks !

[@TMS5100]

@3:28 most people dont realize how huge that feature is. 50W directly means you dont have to worry about distortion from attenuators, or even having to mess with attenuators at all. and you dont have to worry about blowing up your input. these CMU200 are an incredible value for hobbyists. the one i have as configured cost almost US$200,000. I paid $800 on ebay. that is 99.6% loss of value.

[@TheGmr140]

Amazing collection of test gear 😊😊😊

[@MichaelKatzmann]

The hard drive can be replaced by a CF card and adapter. A CF card has the same interface as a PATA drive. The adapter is purely passive. Use Linux 'dd' to copy the first 4GB image from the 30GB drive to a file (or just enough to copy the partition table and only partition ~2GB). Then use 'dd' to copy that file to the CF card.

[@rudiselectronicslab861]

Michael, yes, you are right, the CompactFlash (CF) card interface is basically IDE/PATA, and I read of several people successfully running their CMU using a CF card and adapter to replace their HDD. Have considered both the CF card and the PATA drive routes when I wanted to upgrade my three CMUs but eventually chose the latter. But I've been quite a CF card adapt as almost all my earlier digital cameras (Nikon D70, D200, D300, D800E) had such cards. I wonder where all of them went, I cannot find many cards anymore, and the single one I found I now use in an adapter for the PCMCIA slot on the front panel of the CMU ;-) Nowadays, my main cameras use XQD (Nikon Z7, D500), and in the longer term, a move to CFexpress may be inevitable, but my workflow is not yet ready for that. Not to forget SD cards in my pocket cameras (Sony RX100M3, FujiFilm X100, and my drone). Too many standards ;-)

[@MichaelKatzmann]

@@rudiselectronicslab861 The CF card worked for me and was cheaper. It's also smaller so I just left the original drive in place and attached the CF adapter board with some of that very tacky clear adhesive strip. One thing to watch out for in the CMU200 is corrosion of the Murata monoblock filters in the RXTX board. All the RXTX boards (except for some very late production units) will suffer from this problem eventually. The RF absorber contacts the copper surface of these filters and corrodes them, increasing the insertion loss in the receive IF chain (mine was ~15dB). You see this in the spectrum analyzer and receive signal. There's a 836MHz IF filter and two 1843MHz filters (only one of which is in a cavity with absorber). Depending upon the frequency band covered the displayed level will differ from normal if you have this issue. I'd recommend inspecting, and if the absorber is touching the filter(s), isolating the filter from the absorber with some Kapton tape to prevent any further corrosion.

[@alklapaxida850]

i like the cmu200, nice gear indeed

[@AllElectronicsChannel]

Great channel!

[@derkarhu5079]

BTW, CMU200 was introduced in 1998 to 1999...I found a GSM brochure dated 10/99, and I recall presentations on it when I worked at Tek in '98/99 (Tek was the NA and Central America representative until about 2002)

[@rudiselectronicslab861]

Thanks, Der. For this type of equipment, it is not so easy to find information on the original introduction date. I tried to look at the dates mentioned on the earliest Spec Sheet or User Manual I can locate. But such documents are regularly updated, and I may not locate the oldest ones. For some years, I have found Test & Measurement catalogues, Distribution Catalogs, R&S News magazines ('NEWS' or 'NEUES'), and for some older years, R&S Mitteilungen. So I try to track down original introduction dates, by my data is sketchy and has considerable time gaps, so thanks for filling that in! Do you perhaps have a copy for me of the brochure you mentioned?

[@Janktzoni]

Dag Rudi, Thanks for these detailed explorations and explanations. Spectacular equipment! Do I understand it correctly that these SMU200 with the Audio analysis option are capable of analyzing plain audio frequencies (spectrum and THD on 20 to 20 kHz) or should that audio signal be modulated on some HF signal? I'm interested in frequencies from 0 Hz on to about 200kHz and these are generally not supported on HF class analyzers.

[@rudiselectronicslab861]

Thanks for the reply! Indeed, if the option Audio Analyzer/Generator CMU-B41 board is installed, the CMU can analyze audio devices (DUTs) by sending out a (stereo) signal via front panel connectors AF OUT / AUX 2 (output) and then receiving signals back via front panel connector connectors AF IN / AUX 1 (input). If the B41 board is not installed, these connectors are not connected to the inside of the device. It supports both single-tone and multitone (up to 20) signals. For more details, check the manual on Page 3.16 for a table with supported measurements, as well as Page 4.96 and further. Now, some confusion seems to have raised because some websites mention that you would need software option CMU-K293 would be required to the CMU for audio testing. This is not true; this is only required if you want the CMU also to modulate the audio signal to an RF carrier and then test a radio receiver. I plan to dedicate a video to audio measurements on the CMU. But before that, it is always good to keep in mind that, when all is said and done, do remember the CMU is intended for mobile telephone applications, not for measurement of Hi-Fi, high-end, or professional audio equipment, so it may not be a perfect match for the latter.

[@ronaldomartins499]

Friends. Hello My old CMU-200 doesn't start after turning it on. It's as if the boot is inoperative. Do you know where to start an error test? Do you have a script on how to recover this boot? Thank you

[@nameone7183]

Hi Rudi, Do you know if you can change the SMS message that CMU sends in a GSM test?

[@rudiselectronicslab861]

Good question. I recently successfully sent a number of GSM SMS messages from a CMU200 to a variety of mobile phones, but always used the default content (which reads something like "ROHDE & SCHWARZ - The quick brown fox jumps over the lazy dog"). Next time I am working on this, I will check whether/how that message can be changed ;-). I did note that for some phone brands/types, the CMU tended to crash just after sending the message :-| This is one of the very few CMU crashes I ever experienced.

[@nameone7183]

@@rudiselectronicslab861 Thanks very much. It would save money sending codes for testing alarms etc.

[@WilliamLagerberg]

Rudi ik heb een vraagje over de cmu200 ivm sinad meting, ik denk dat even email of telefonisch contact makkelijker is maar heb geen idee hoe ik je moet bereiken.

[@dragonlard4595]

About Software Options: There exists a Key-Generator, to activate every software option...

[@marcseclecticstuff9497]

I wasn't aware of that. Where can I find that?

[@dragonlard4595]

@@marcseclecticstuff9497 There still should be a thread about it, on EEVBlog Forums.

[@marcseclecticstuff9497]

@@dragonlard4595 Thanks for replying. I'll head over to Dave's place and check it out!

[@derkarhu5079]

Road ee and Schwarz Not Road en and Schwarz Sorry to be so picky, excellent presentation, but I worked with R&S, proudly, so feel a bit more pickia out that... BTW. TEK, in the 90's and early 00's was a great company to work for, and with, better than Agilent, by my estimation.

[@super-8]

Alter Schwede die Communication Tester von R&D each 100K€

[@vaualbus]

I got mine for 100. Just a bad keypad membrane and now it's working so well ❤️‍🩹!

[@rudiselectronicslab861]

Nice, that's a good price! Paid a bit more than that, but still was very happy with my price. In any case, a fair price for a CMU depends on, among others: (1) physical state, (2) functional/electrical state, (3) installed HW and SW options, (4) information on the history of the device, (5) information on possible repairs/tinkering, (6) warranty and (7) ability to see and test the device before purchase, and especially to run the 10+ self-test routines, including the ones with RF loop. Common errors seem to be faulty CPU, faulty or damaged screens, broken HDD, and - as you have experienced - the front panel buttons, where a mere clean is often sufficient.