High-resolution data acquisition. Comparing Tektronix TDS7254B and Agilent MSO8104A oscilloscopes

Sergei P. Skorobogatov
sps32 (at) cl.cam.ac.uk

Manufacturer: Tektronix and Agilent

Models and Date of production: TDS7254B (2005) and MSO8104A (2007)

This is another small project I have not finished yet, but I am working hard on it...

I'm working...

The aim of this project is to evaluate some of the high-end oscilloscopes available in our laboratory for suitability in my research applications, especially for low-noise high-resolution power analysis.

Very first impression from the Agilent MSO8104A oscilloscope was really good. The boot time is considerably shorter than for the Tektronix TDS7254B. The Agilent oscilloscope software (revision 05.20.0001) has some easy-to-use features such as minimisation button and touch-screen Zoom while the Tektronix software required several clicks and actions to get there. Data acquisition worked really fast even for 16MB of the acquisition memory, thus allowing acquire and average up to 4 trigger events per second while for Tektronix it took about 3 seconds to acquire each event.

Other advantages of the Agilent MSO8104A oscilloscope: I have a big list of them and I am going to place it here soon.

Talking about disadvantages of the Agilent oscilloscopes, and most of other oscilloscopes, I want to mention the customer robbering feeling for memory upgrades. When you buy the MSO8104A in a standard configuration (1MB of memory) you already get the oscilloscope hardware with 128MB of acquisition memory, but the Agilent software will not allow you to use all this memory until you pay tens of thousands dollars to Agilent for an enabling code that will tell the software to expand the use of the memory up to 128MB.
The other horrible thing with the Agilent MSO8104A oscillscope is the lack of the recovery CDs/DVDs. This is really horrible and puts a big cross on in-field applications for Agilent oscilloscopes. Just imagine - you are far away from the Agilent support team and your hard drive crashes (that actually happens very often especially when RoHS was introduced and became mandatory, but this is a different story and probably already a headache for their technical support team). For some long-term applications this might be very important issue and the buying decision will be made definitely in favour of Tektronix oscilloscopes which are fortunately supplied with the recovery CDs.
Other useful things which the Agilent oscillscope might benefit from is DVD Writer (their cost has dropped significantly these days and Tektronix does put CD-RW as a default driver). Relying on USB Flash drives as Agilent offers, is not a good practice as their failure rate is terribly high, especially for extensive data transfer use. Luckily, Agilent has a LAN port which can be used to trasfer large files. However, that requires some experience to set up, plus availability of a server to hold the files.

So far, I was able to spot only one problem in the Agilent software for MSO8104A oscilloscope. It is related to incorrect handling of the Math functions.


For example, I do some measurements with one of the channels (Channel 3 in the above picture) and define the following two Math functions: f3 = m3 - 3 (m3 is a copy of the earlier Channel 3 waveform), f4 = intg f3 (integrate). Once enabled, these functions work perfectly.


However, if I repeat acquisitions, sometime the function 'f4' shows incorrect result as in the picture above.

PM500-1A PM500-1A

Going to the Math menu, unchecking the 'Display On' box, then checking it again helps to restore the 'f4' waveform to the correct state as it the above picture.

I already notified the Agilent software development team and I hope the problem will be fixed soon.

Sergei Skorobogatov <Sergei.Skorobogatov (at) cl.cam.ac.uk> <Sergei.Skorobogatov (at) hushmail.com>
created 04-05-2007 -- last modified 04-05-2007 -- http://www.cl.cam.ac.uk/~sps32/