Advantest R9211 spectrum analyzer

[Jan]

I have a Advantest R9211E spectrum analyzer which is very good for audio work. It's about 20 years old now and i'ts not so well documented on the net so I thought I'd collect some info here.

Some background info about the instrument:
- Measurement ranges adjustable in around 100 (1dB!) steps from -60dBv (1mV) to 200V
- 16 bit resolution, 2 channels
- FFTs from mHz to 100kHz
- Background noise < 140dBV (2KHz range)
- Steep input antialiasing filters
- GPIB, thermal printer(option)
- Some versions have a built in signal generator
- Robust protected inputs, DC and AC coupled


- I have disassembled the instrument and swapped power supply, replaced most ecaps and some other parts so I have some knowledge (and pictures) on the interiors. The manual recomended life of the power supply is five years, most likely for the e-caps....

If you are interested in the instrument and / or have questions or information to share please post!

[Jan]

Some useful documentation for the R9211A R9211B R9211C R9211A R9211E

Product datasheet (140Kb)
Operation Guidebook (2.6Mb)
R9211A and E operating manual (13Mb)
R9211B and C operating manual (17Mb)
GPIB handbook (4Mb)
GPIB command matrix (1.5Mb)

Power supply datasheet

- I am looking for service manuals
- Calibration manuals / information

If you have more PLEASE send it to us!

Conversion of R9211 floppies to generic formats:
* AR9211 A utility that reads R9211 floppies and converts data to either STAR or UFF format (readable by Matlab?). Written in VB6, tested on Win XP. Written by http://www.ssinst.com
* ADVANT One older DOS version and a newer one in PERL by Lennart Karssen. The Perl version should run on any system supporting the PERL language, like UNIX / Linux, Win32 etc.

[Jan]

Like many other instruments from this time, the R9211 has a GPIB interface that can be connected to a pen plotter.

If you have a computer with a GPIB interface and the right software you can dump the plot to the computer instead. I tested a large number of PC software for grabbing GPIB screen printer dumps. A summary is available here:

/downloads/R9211/R9211_gpib_screendump_test.htm

[Jan]

- I have a problem with the built-in screen disturbing the measurements; There is a peak in the FFT around 15.75KHz

I have replaced all e-caps I can find and shielded as well as I can but the peak is still there.

Now I am thinking of replacing the screen with an LCD or similar.
So I am looking for a converter for the screen signal. It seems to be 15.6KHz or 15.75KHz scan rate, 16MHz signal. It's apparently similar to a CGA signal.

There is a connector on the back of the instrument for the screen signal, so it should be possible to test without ripping out the screen. Any help appreciated!

I will post here any findings....

[ghg]

Hi from Vienna

I respond regarding your pm via www.diyaudio.com

I have a few R9211C schematics, think of the 100kHz antialiasing input filter;
have to dig them out of 35 years of tech papers, no pdf's at that time ... but no way of getting a service manual.

My R9211C gives the 15.75kHz peak, too.
20 years ago, I put a switch to the V+ tube power supply and things went fine.
Additionally, if I remember correct, an IBM CGA monitor worked at the video printer out,
but the composite video was of low quality. I just used it just to lengthen the lifespan of the display tube.
Btw, you can buy the tube as a spare for about $200, eg. at www.helmut-singer.de , a wellknown and trustful company.

Today, I use the Prologix GPIB2LAN adapter and 7470.exe to get screen outputs via the COPY button; any way to upload example images here within the post?

further comments
date stamp gives the only Y2k bug I've ever seen;
7470.exe works fine, but there is a grid within the harmonics readout window, of course not at the tube display. John Miles could not solve this issue, he stated, he just gets the data that way from the GPIB, curious, cause a real world HP 7470 plots the harmonics window without a grid, just like at the tube display.

BR

Gary








I hate sounding amps, except these are Marshalls,driven by Eric C. or John F., period.

[Jan]

Hi and thanks for the feedback!

Thers is no way of uploading files here. You can emil them to me (preferred) or link to external images.

I was considering adding a scren switch like you suggest, and or using another screen. If the signal is CGA, there are CGA-VGA converters for sub $100. I may try one.

I am using an NI GPIB interface and it works good enough for screen dumps (for those who did not try, a screen dump takes about 30s so it's no screen substitute).

I would like to zero offset for the most sensitive ranges but am reluctant to fiddle with the pots without a manual. I think there are four pots per ADC board.

If you find the schematics, please send them...

Hopefully some other Advantest R9211 owners will find their way and we can exchange whatever is available. Advantest R9211C, R9211E, can't hurt to lure some search results here

[Jan]

A program in Basic / Perl for converting binary data written to the R9211 floppy drive; dump the data to the Advantest R9211 floppy, read and convert the data file on a PC.

http://tiki.karssen.org/tiki-index.php?page=AdvanTest

[Jan]

Just measured on the screen output, on the back of the Advantest R9211E screen output:

- vsynk 60Hz, continous pulse train 0V-5V app. 0.4mS long
- hsynk 15.7kHz continous pulse train 0-5V, app 6uS long
- video, spread pulses, most <1uS, 0-5V
- One klock output 16MHz

(The connector pinout is in one of the manuals).

This looks like CGA, without color information. As far as I understand, the problem converting this to VGA is that one needs to buffer one "CGA line" and display it twice, at double frequency, in VGA mode.

There are CGA converters out there. I think I will try one, with the assumtion that disabling / replacing the internal CRT will solve the 15.7KHz problem (and 60Hz)... I will report here when there are results to show.

[Jan]

Here are some screen dumps from a R9211C, sent by Gary. First one with the built in signal generator running at 1KHz

[img]/images/r9211/R9211C_01.gif[/img]

This plot is for a signal from a Kron Hite 4400A signal generator, combined with a -60dB active notch filter so the analyzer can utilize a high sensitivity input range. Because of the notch, the analyzer does not see the full 1kHz signal so the THD value is not valid, It shoudld probably be around -108dB as the signal is really 3.16dBv and 2:nd harmonic is -105dBv
[img]/images/r9211/R9211C_02.gif[/img]

Last an image of the analyzer with inputs grounded and CRT screen turned off, showing a nice < -135dBv background noise at 100kHz, except a peak near 100hHz. If one limits the bandwith, one gets even lower background noise. The specification is -145dBv at 2KHz bandwidth.
[img]/images/r9211/R9211C_03.gif[/img]

Thanks Gary!

[vr2whf]

My R9211C recently encountered the power on self test error. Failed item: "ANALOG_BOARD", Error code: 003f.
Any idea? Anyone has the service manual to share?

Jan, you have the power supply document? Can you share because i suspected the aging power supply. Thanks!