Schlumberger 4002 signal generator

While working up some extra circuits for the spectrum analyzer, I managed to pick up an old signal generator from eBay.

I heard a lot of positive things about the German (actually French origin, please look at comments below. Thanks to Rohit for pointing this out!) brand “Schlumberger” before, even though there is no relation to any personal experiences with their equipment. Seems like they also ran some kind of subcompany outfit called “Solartron” or “Enertec” which would today sound more than fishy, what with all those copycat-brands out there. But when an auction came up for a reasonable price I decided to go for it after some short research on the net.

Fig. 1: Schlumberger 4002 signal generator, already opened up.

Fig. 1: Schlumberger 4002 signal generator.
Don’t mind the tearing effect on the LED displays, not visible to the naked eye. I already pulled all the side panels.

What I got was a Schlumberger 4002 signal generator. It ranges from 0.1 to 2160 MHz with 10-20 Hz tuning accuracy, selectable output amplitude from -138.9 dBm up to +13 dBm in 0.1 dB steps, auto-sweeping and several extras like an OCXO for stability, 20 dB of linear attenuation range without using the step attenuator, an internal modulator and IEC bus remote control. If you looked at the photo closely, you will have noticed that the frequency range is written as “0.1…1000/2160 MHz” on the front panel. The reason for this is the optional doubler module included in this instrument. If the module is installed and detected, the software switches over to extended range without any further changes. Else, 1000 MHz is as far as it goes. More detailed specs will follow as soon as I can decypher the bad scan of a manual page that cropped up on Google. Judging from the inventory labels on the backside, the device must have been used in the manufacturer’s own lab. Unfortunately I have not yet managed to find any service info even though the manuals seem to be sold sometimes, for rather terrible prices.

Upon arrival, the first thing I did was open it up to check for transport damage or any other things demanding immediate attention. Finding none, I was absolutely stunned at the old-fashioned, but VERY clean way this device has been built, similar to HP/Agilent products. Since there are few useful pictures on the net, I will start to provide some along with any technical info I manage to develop for this fine piece of tech. Here are some quick shots done while inspecting, more will follow later.

IMG_1461_web

Fig. 2a: Front panel, even the LED dot matrix displays are shielded.

IMG_1446_web

Fig. 2b: Top of frame. Modulator (right), attenuator (top left) and double (bottom center) modules.

IMG_1480_web

Fig. 2c: Modules can be easily pulled – oscillator unit in this case.

IMG_1447_web

Fig. 2d: Backside ports for OCXO unit: External source in and standard outputs.

IMG_1452_web

Fig. 2e: Right side of frame. Filter/AGC (topmost), multi-oscillator (middle) and OCXO (bottom) units.

IMG_1451_web

Fig. 2f: Left side of frame. Logic cards, controllers, bus connectors, etc.

IMG_1450_web

Fig. 2g: Left side of frame. Power supply can be pulled as one block.

IMG_1448_web

Fig. 2h: Backside ports for external modulator signals and IEC bus connection.

After the checkup, I tried turning it on. It booted fine right away, with the sweet clicking sound of relays inside. I then hooked up the output to my HP analyzer and ran some tests. First up was the selftest, executed by pressing the conveniently labelled key at the bottom of the front panel. Pressing it again after reading the error message brings up the next, and the next, and so on until the IEC bus address is displayed last. The errors coming up in my case are:

  1. AGC DOUBLER (intermittent, might have been a bad connection)
  2. AGC OUTPUT STAGE
  3. MOD-GENERATOR

So there is something wrong that even the device itself can detect. It would be nice to find out how it determines the faults to get more detailed hints. Several signal lines labelled “SELFTEST” can be found on the doubler and filter/AGC modules, so these probably lead to the respective points of interest.

Moving on, next up is selecting arbitrary frequencies and checking the freq/amplitude values on the analyzer. Turns out that the generator seems to contain multiple circuits for different ranges. The main divisions I could make out so far are 0.1 to 479 MHz, 480 to 1079 MHz and 1080 to 2159.9999 MHz. At these points, relays click inside and the output behavior  changes dramatically. In the first region, I get LOTS of spurs and harmonics, and the fundamental is always off by a certain amount. Funny enough, that offset is almost “binary”, since it results in 127 MHz instead of 100, 255 MHz instead of 200, and so on. In the second range, the frequency is almost right and the spurs vanish. Instead, the level drops and moves about. This might fit the AGC (auto gain control) error. The third range sometimes delivers something, but never quite right. This might be a problem with the doubler block which sits next to the attenuator. Have a look at the next pictures. The last one shows a complete recorded sweep over the whole band.

Fig. 3a: Lots of harmonics/spurs below 480 MHz. Not even the main peak is right.

Fig. 3a: Lots of harmonics/spurs below 480 MHz. Not even the main peak is right.

Fig. 3b: Clean signal from 480 MHz all the way up to...

Fig. 3b: Clean signal from 480 MHz all the way up to…

Fig. 3c: ...1079 MHz.

Fig. 3c: …1079 MHz.

Fig. 3d: At 1080 MHz, it's...gone?!

Fig. 3d: At 1080 MHz, it’s…gone?!

Fig. 3e: Fullrange sweep from 0.1 MHz to 2160 MHz. Obviously, there is something above 1080, but not much.

Fig. 3e: Fullrange sweep from 0.1 MHz to 2160 MHz (1700 MHz on the display right, actually). Obviously, there is something going on  above 1080 – but the gain is totally off.

For measurements, I have already disconnected the doubler block and the attenuator from the generator and checked the signal from the generator directly. The harmonics can be seen there, too. A first visual inspection of the blocks did not yet bring any revelations, except that someone seems to have done some soldering work in what I believe is the Filter/AGC block.

Well, seems like I have some work to do. Next up is the removal and inspection of all the RF blocks to get a better feel for how this works, and developing a crude block schematic from what I find. I will probably dedicate a separate post for each module as I work my way up the line.

If you happen to have a copy or scan of the original service manual you don’t mind sharing or even know where one can be obtained – I’d be glad to hear from you in the comments below 🙂

13 comments

  1. Ton says:

    Hello 4002 technicians, some very good info on this site about the repair of the Schlumberger 4002. I also own such a generator and it works fine except for the internal modulator, when I run the diagnostics with the frontpanel button it gives an error message; error; mod-generator and indeed, modsignal Bu78 is absent. When I apply external audio then I can modulate both AM and FM, I noticed also some leakage from electrolytic condensators but upon replacement I still got the same problem. The real problem is of course that I do not have schematics and I have to find an extender board to puton the board were Bu78 signal comes from. Does anyone in the group owns a service manual and can perhaps sent me the schematics of the mod generator ??

    Happy signal hunting !!

  2. Markus says:

    Hi Mario,
    hab gesehen daß du dich mit den Stabilocks sehr gut auskennst.
    Hast du Erfahrung was defekt sein könnte wenn im Selbsttest Duplex failed angezeigt wird?
    Was mir auffällt, im Analyzer Mode kann man ja normalerweise Filterkurven, Antennen etc. durchwobbeln, hier kommt nur noch undefiniertes Zick Zack im Plot.
    Ach ja noch was: Kennst du die Tastenkominationen für die versteckten Diagnose/Kalibrier Menüs?

    Gruß,
    Markus

    • µ says:

      Hi Markus,
      leider habe ich selbst keinen Stabilock, insofern nicht ganz so gut.
      Der Duplexer ist üblicherweise die Sende-Empfangs-Umschaltung. Bei der möglichen RF-Leistung gehe ich davon aus, dass im Stabilock dafür ein elektromechanischer Koaxschalter verbaut ist, irgendwo kurz hinter dem RF Ein-/Ausgang. Die hängen mit zunehmendem Alter auch mal fest, insofern würde ich da erstmal prüfen ob am Schaltelement Spannung ankommt und ob es dann auch tatsächlich was tut. Allerdings kenne ich von meinem Synthesizer die Eigenart, dass die Fehlermeldungen manchmal etwas neben das eigentliche Problem zeigen, weil das Gerät mit begrenzten Möglichkeiten versucht, einen logischen Überblick über seine Module zu bekommen. Es kann also gut sein, dass da was in der Umschaltung hinüber ist und dadurch auch die Störung im Wobbelbetrieb ausgelöst wird, weil er den Ausgang nicht mehr richtig konfiguriert bekommt. Klingt für mich jedenfalls so. Und da die Stabilocks grundsätzlich solide sind, würde ich Fehler definitiv ausgehend vom RF-Anschluss suchen.
      Die Tastenkombos kenne ich leider auch nicht. Service Manuals sind für diese Geräte irgendwie extrem dünn gesät. Rateversuche waren bisher nicht von Erfolg gekrönt :/
      Gruß Mario

  3. Victor says:

    Pretty good and insightful article, I like it 🙂

    I’ve always worked in electronics, just never in RF until less than a year ago. In my work we have this beast, a STABILOCK 4032, totally abandoned with no interest to repair or sell it as is (we mostly use General Dynamics equipment). It doesn’t measure or generate anything and the on-screen indicators dance like Christmas lights. Seeing what you did motivates me to try once again, to convince my boss not to let it rot away. Can you recommend a course of action to check and possibly fix it?

    http://i.imgur.com/sLdQf1k.jpg

    Curiously, my first approach to Schlumberger was repairing fuel pumps at service stations. They were reliable, robust machines, it was a pleasure working with them. I see that apparently extends to all products of the brand.

    Regards,
    Victor

    • µ says:

      Thanks 🙂 I’ll have more on the 4002 soon, but it takes a lot of time writing all of that up.

      Regarding your StabiLock – maybe you are lucky and the fault is not even anything RF related. If the controls misbehave, my first thought usually goes in the direction of power supply (bad electrolytic capacitors or the like, very likely at this age) or bad/loose connections. A clean supply is essential for the RF circuits.
      I’m not quite sure if I get you right, do you say that the screen output is disturbed, or rather that the measurement data shown on screen is bad? Because if that picture is of your set, at least the microcomputer and the display output seem to be alright. In that case, you should have access to some kind of self-test function that can point you in the right direction. Take the output with a grain of salt though. I suspect that error chains are still possible, meaning that you have to find the earliest error in the chain first.

      In any case, my recommendation is:

      • Check the power supply first (Absolut values of voltages and quality/noise level, also measure/replace critical electrolytics if possible).
      • Check logic boards, which are probably mostly plugged add-in cards, for bad caps. In my 4002 they were already clearly visible from the leakage and corrosion.
      • While checking, remove everything that has a connector, check for dirt or damage and reseat the connections – just to rule that one out. Also check any RF cabling for damaged crimpings or broken solder joints! Especially semi-rigid cables seem to break a lot if mounted under mechanical stress.
      • Only if faults persist after this should you begin checking the RF sections, starting from the inputs/outputs since damage is less likely to occur as you get further away from the “user interface”. At this point, the selftest output comes into play.

      You are fortunate that the StabiLock series seems to excite more interest than the 4002, there are copies of the service manual to the 4031 floating around. I gather the 4032 is very similar. You’d first try to identify all the modules and their connections from the block schematic in the first file, just so you know your way around inside.
      http://www.download-service-manuals.com/download.php?file=Schlumberger-5324.pdf
      http://www.download-service-manuals.com/download.php?file=Schlumberger-5325.pdf

      As for tools: In the ideal case, you have access to a proper spectrum analyzer and a RF signal generator so you can inject signals into each module’s ports and check the output.

      Best regards and good luck,
      Mario

      • Victor says:

        Wow that was fast!!

        I’m quite sure there are bad capacitors, for example the CRT image “wobbles” back and forth somewhat, and the supply make a subtle but weird hiss sound, both classic faults. Apart from that, the dust, and the weird readings (that’s what I’ve been referring, English isn’t my native language sorry) the rest is in good shape. I hope the faults relate to aged components only and nothing too difficult.

        Yes, this is my set. Well not mine, they are fixated on not letting it go but at least we have equipment for a quick test. Maybe if they see it’s “unrecoverable” I could buy it for a fair sum… only thing I know is that some guy blasted >100W of RF thru the GEN port, and the next day there was one less technician, a ~20kg paperweight, and a new General Dynamics R2600 in place. That was roughly 8 years ago.

        Thanks for your help 🙂 and kind regards from Chile.

        PS. AFAIK, StabiLocks were manufactured by Wavetek too. Maybe they have some kind of manuals for the 4002??

        • µ says:

          No wonder. I get the e-mail alert, too 😉
          Don’t worry about your English, it’s pretty much perfect. My question was solely about the content.

          It seems like you have experience in fault finding, that makes things easier. In your specific case, you *might* also want to give the RF frontend a look right from the start. That some-guy-phenomenon can be a nasty one, true. Judging from the 4031 manual, chances are pretty good that the synthesizer stages are not damaged. There are plenty of amplifiers in between, hopefully providing enough isolation. I don’t know if a quick test is possible, but you can try.

          The critical parts are the mechanical attenuator, the output unit and the metering and counter circuits (#229 031, p66 in 2nd PDF). If the N-connector was the victim, the damage might be small. Most of the energy should have dissipated in the load resistor, and the meters should be able to take a little excess voltage. The direct input is more fragile because mechanical attenuators are usually not built to dissipate much power. This might be the cause for the missing generator output.
          If the MOD GEN output was hit, there might be substantial damage in the MOD GEN A block, which you would probably be able to see right away when you inspect that module. At >100W, the low-power output circuitry should be shiny metal vapor all over the insides by now.

          I would start on the RF part by disconnecting the RF attenuator from the OUTPUT UNIT (#230 032) and check the generator signal there. If there is none, go one step further back and see if anything at all comes out of the UHF SYNTH block. The same goes for the RF POWER METER (#229 031) and RF COUNTER (#237 032), by feeding their inputs with a generator and checking the output lines or the device display. From what you find, the search will then go further along the chain or else into detail in the block where a fault is expected. I am not yet sure how the whole modulator/demodulator chain works, hard to read from the partially overlapping pages.

          Thanks for the tip about Wavetek, but no luck so far. Fortunately, the manuals are unnecessary unless something important breaks beyond recognition. Except for that worst case, I like the challenge of slowly reverse-engineering the whole thing to learn from it.

          Best regards,
          Mario

  4. Rohit says:

    Hi Mario!

    I stumbled across your blog while browsing DangerousPrototypes. Just a small correction – Schlumberger is *not* a German company; it’s actually a French (now French-American) Oilfield Services Company. Schlumberger makes almost all of its acquisition and test & measurement equipment. It used to own Daystrom and Fairchild Semiconductor (but has now sold them off). While I don’t work in the instrumentation/manufacturing division, as a Field Engineer involved in oil exploration, I use a lot of in-house-manufactured electronics. I can attest to their quality and robustness.

    Cheers!
    -Rohit

    • µ says:

      Hi Rohit,
      thanks a lot for the correction! I was actually under the impression that the original company was no longer existing, having split up somewhere in the middle of the timeline. My understanding is that the manufacturing labels on the backside set me up on the wrong track, going after some german branch or something like that. Maybe it was just too much of a good fit, since most of the RF-related companies are concentrated around Munich in Germany, and that’s where the labels point.

      Good to know that they are still in this business, though! And I totally agree, the build quality of what I have seen is right up there with the other big names. Some day I need to get my hands on one of their StabiLock transceiver testing devices.

      Very best regards,
      Mario

  5. µ says:

    You should check the 80 MHz output from the FM mod on BU 116, but that is probably okay. The frequency locking mechanism in the VCO block is somewhat strange, my understanding is that it upmixes the 10.7 with the second signal from the VCO (BU 115) and the 80 MHz standard signal. The oscillators are then partly locked to that and partly to each other – need to do some more analysis on that. However, if any of the signals is missing, lock is gone or sporadic.

    The 10.7 MHz on BU 114 is essential for the VCO, it goes from the OCXO to the FM modulator (same jack number there) into the frequency/phase modulator unit and out again on BU 104 to the VCO block, now carrying the modulation. Seems to me like the corresponding unit in your OCXO (probably a PLL) is damaged, you should remove and check that block first. You can also remove all other blocks and still operate the device with the OCXO lid open, does no damage. I did not open mine yet since the outputs were okay, but I can do this tomorrow if you need any advice on that.

    EDIT: Comment out of order, read below.

    • µ says:

      I have opened my OCXO now, a lot of PCBs in there. The 10 MHz TTL comes directly from a logic gate next to the oscillator, as does the one on the back. The 10.7 MHz are produced by mixing and filtering 10.625 MHz with 65-85 kHz. The wide-range signal on BU 115 is generated by a stack of VCOs in the chamber right behind the SMC jack. The control/locking circuit is next to it, the PCB with lots of ICs on it.
      I would start off with checking the internal 10 MHz references leaving the PCB in the chamber next to the backside BNC jacks. Next is the 100 kHz/10.625 MHz source in the bottom chamber directly beneath that, followed by the 65-85 kHz source (VCO/divider in the center of the bottom compartment, gray wire). Also look for any bad connections, there are some that might get loosened by shock.

      EDIT: Comment out of order, read below.

      • µ says:

        If you prefer german, we could just switch languages.

        No signal at BU104 is normal since there is no signal on BU114. As the 80 MHz is fine, start with the OCXO block.

        I put some comments in the following pictures for orientation (click for large view):
        OCXO bottom side
        OCXO bottom side

        EDIT: I have removed a string of comments by a reader who I believe is trying to sell his (still broken) 4002 generator at a substantial financial gain, while he claimed to only have a private interest in the device during a prolonged mail exchange – especially when I asked specifically. In summary, the exchange and circumstances make me doubt this. While I have no problem with commercial ambitions whatsoever, I strongly recommend to truthfully state your intentions. Who knows, you might still get an answer…
        I have left some of my own responses intact for the readers’ benefit. ~µ

  6. µ says:

    Hi Franz,
    first of all, check the output from the OCXO unit at the bottom! There are three labelled outputs, and each should supply approximately what is written next to it. If these are okay, check the FM modulator block on the top left next. Inside towards the back is a PLL/VCXO that multiplies the 10 MHz TTL to 80 MHz. Check the output from this stage, because all the VCOs of the generator are locked to this 80 MHz signal. Because of this error, there should also be two LEDs (red and yellow) lit up if you open the right side panel.

    The connectors are almost all some kind of SMC plug, I suggest building a suitable adaptor for testing. A SMB coupling works for most of the connectors, but some have a backset center-pin and therefore do not. Not reliable.

    Don’t take the AGC output stage error too seriously, the error also appears if the signal from the VCOs is bad. The IEC addr. is no warning at all.

    Any chance you bought yours on German ebay a short while ago? If so, it’s probably the same seller as mine 😉

    Cheers, Mario

    EDIT: Comment out of order, read below.

Leave a Reply

Your email address will not be published. Required fields are marked *

*