I Like ‘Scopes

Well, the title makes the point. I like oscilloscopes, own them, sometimes repair them, and have been thrilled to use them in my work. 'Scopes are the ideal window to the fast-moving world of electrical signals. They are to engineers what the telescope is to the astronomer, the x-ray machine to a doctor, or the occasional spoonful to a chef, the tractor to a farmer, leaf blower to a gardener. They are just beautiful and useful instruments.

I even tried to make an oscilloscope. I obtained a revered Tektronix microchannel CRT (inside job) and started to make surrounding circuitry. My first section was a trigger circuit. After a year of weekend design and lab construction, I gave up. I could not make circuits that even came near the performance of the then champion 7000 series 'scope from whence the CRT came.

I then studied the manuals of the 7000 series and saw everything was stacked against me. It had custom ICs, not discretes as I was using, and that IC process was a 15GHz type. They even had a very fast high-voltage 50V process for driving the CRT deflection plates. The selector switches were actually connected to tiny cams and even tinier leaf switches assembled on the circuit board to minimize inductances.

The boards used microchannel striplines as interconnect and had stubs and matching sections to keep frequency response amazingly flat. The level of sophistication gave my ego a good spanking.

So, now I am a minor collector of oscilloscopes. Not modern ones — they disappoint. The transition to digital 'scopes has seen the loss of decent triggering, fat noisy waveforms, and yes, even analog purity less than the 2465 series. Of course, digital 'scopes are the thing to use to make numerical measurements, but they suck at being a general-purpose lab 'scope compared to the analog ones.

Our man Steve Wong got involved with a settling-time measurement project and discovered that the digital 'scopes in the lab had worse overload recovery and settling details than the 2465. Yes, the POS 1GHz 'scope (I won't tell you its name; I'll just say it makes me cross to use it) was beaten by the out-of-production 2465. The POS 'scope even had a lousy input SWR at the 50Ω setting compared to the 2465.

I'm not as bad as some though. Jim Williams had about 40 'scopes, most of them real antiques. Thirty years ago, I would have been impressed with them, but with today's standards, I reject the old stuff's inferior performance. I'm not nearly as promiscuous as Jim either; he had more than one of a couple of models.

No, I collect only a few historically important ones. I couldn't pass up the chance to get a 7104 4-bay mainframe. Hey, this is THE fastest analog 'scope ever with 1GHz bandwidth. It's got that cool microchannel plate CRT that allows really fast writing rates, and it can use various vertical plug-ins. We keep a couple of 7104s in the Milpitas labs, mainly for the high sensitivity vertical plug-in (10μV/div) and the fast-settling vertical plug-in (1mV settling from 10V step in 400ns.). And yes, I own those plug-ins.

I have a Tektronix 465 'scope, too. This was the first transistorized portable. Well, the input devices are miniature Nuvistor tubes for safety against overload, but the rest are discrete transistors. Pretty nice overall, with 100MHz bandwidth. I need to fix one input channel, though.

Naturally, I have a 2465B. I don't know why, but I passed up a 2465A with a bad channel at the local electronics flea market, for only $100. That would have been a great parts 'scope. I'll have to start acquiring beater 2465's for parts. Tektronix stopped making them in 2002, and they're all old now. Nothing lasts forever.

So, I'll get to the point of all this. When you see these analog 'scopes in the lab not being used, please turn them off. The CRT's don't last forever, and neither will the rest of them. These machines are superb and deserve a break. Please don't run the trace very bright. They'll do it, but you can burn features into the phosphors over time and use up the cathode's emission life.

I don't mind what you do with the digital 'scopes.

What 'scopes do you use and why do you like them?

14 comments on “I Like ‘Scopes

  1. Scott Elder
    August 26, 2013

    I've always wondered why one finds HP/Agilent scopes favored by the test engineers on the test floor, but mostly only Tek scopes found in the design labs.

  2. Barry Harvey
    August 27, 2013

    Yeah, I've noticed that too.  I always wondered why they were bought in the first place; HP analog 'scopes didn't trigger well, have much bandwidth, nor flexibility.  Ironically, my favorite early digital 'scope was the HP54100.  Like Tektronix, it had flexible plug-in's.  It had a great menu system and did a random but slow sample, and could display 1GHz signals.

  3. BillWM
    August 27, 2013

    I remember using the old DVST storage scopes back in my early days — keeping the CRT from “Blooming” was always an art with those — I even worked production doing old aircraft Navigation radars that Used DVST rather than Digital — And I had a 465 with the TDR addon in the oil field for finding cable breaks/shorts.   Some interesting art in those old scopes.

  4. Barry Harvey
    August 27, 2013

    Still have a 7000-series storage mainframe in the lab.  I use it to make x-y plots mostly.  You can do this with some digital scopes, but when high-slope regions of display occur the sample pixels are sparse and pretty ugly.  Not so with the analog storage; nice continuous lines throughout.  Of course, phosphor storage is pretty cludgy.

  5. PZman
    August 27, 2013

    I also grew up with the 465 with the TDR option. Also have on my bench two Sencore scopes from the 1970's. Not a lab scope but great for troubleshooting, especially tube amps.

  6. Davidled
    August 27, 2013

    Generally speaking, analog scope is really heavy because the old components of electronic might add the weight of scope and PCB board is really big. But when I used the digital scope, it is really not heavy. Also, when checking component inside board, the component is getting small and better performance compared with old components used in the analog scope. The light weight scope is better.

  7. Barry Harvey
    August 27, 2013

    I dunno…I seldom lift my 'scopes.  If I were a travelling repair kind of guy, weight would matter.  Come to think of it, my Tesla S is damn heavy too but handles really well and is fast.  Reminds me of the 7104.

  8. Davidled
    August 27, 2013

    Weight comparison is based on the almost same functionality between analog scope and digital scope.

    Well, I did carry scope and test the device from place to place. Then, I realized that old scope is more weight than that of new scope (digital scope).  I knew that some engineer measure the frequency response by Ipod.  It seem like smartphone is one- size- fits- all.

  9. jbike
    August 27, 2013

    Well I must admit that after reading your blog I just about bid on a Tektronix 2465B oscilliscope on ebay. I may end up with one yet! I could not agree with you more about the analog scopes.

  10. RedDerek
    August 27, 2013

    I did enjoy the old Tek scopes with the phosphor screens. I remeber taking pictures with the polaroid camera as well. With phosphor screens going out, it might be an interesting project to tap into the X and Y drive of the screen and interpret onto an LCD screen. This would solve the bad phosphor issues.

    I have traveled with the Fluke handheld scope a few times and we found that was the best scope to use when making power measurements across a 12 inch long, 1 inch diameter flash lamp. We used one channel for the differential across the lamp and the other for a current donut. Then let the scope do the math to calucalte the joules per pulse.

    So, the analogs are great, but the digitals do have a good spot in my tool chest. I certainly agree that an analog could still capture somethings the digital cannot. Thus the idea of the screen conversion circuit.

  11. D Feucht
    August 28, 2013


    Tek's triggering improved significantly over the older 545-era scheme of differentiated-edge pulse synchronization when it was replaced by ECL arm and trigger RS flops and an input gate with hysteresis. The hysteresis provides some delay time for flop setup when holdoff releases (holdoff is the inhibition of a new sweep until the trace has settled on the left side of the screen) and when the new trigger edge (from the trigger level/slope comparator) arrives. (Details on this are in my book on waveform-processing circuits at ) The key problem is in deciding whether or not to run a new sweep when the holdoff releases simultaneous with a new trigger edge. Equivocation results in trace jitter on the screen.

    When I was working in the Tek portable scopes group in the late '70s on the trigger IC (the Tek M202) for the 465, 475 replacements, it was able to reliably trigger (with acceptably low trace jitter) to over 350 MHz with what I recall were 7 GHz NPN BJTs. (The layout was in rubylith on large light tables!)

    If you really want to get into fast 'scope details, get ahold of John Addis, who designed the 7104 vertical amplifier front end, or Audio Precision Inc. analog founder Bruce Hofer, who designed its fast time-base. Addis also invented the very fast overload-recovering vertical amplifier scheme used in Tek 'scopes nowadays. Both are up in the Beaverton area.

  12. Navelpluis
    August 28, 2013

    We also love our couple of analog scopes we still have around. Both are HP's 1742A I think they are born in the '70 😉 These are the 'low power' versions with the neon bulb functions illumination, a brilliant concept. We bought them 25 years ago as surplus from the navy. They still go 100% okay. Triggering I always found fantastic, no problems at all. I also had a 2465 from Tek with that special tube, but they made the failure to skip flash in their design to hold the calibration. What a failure !! Due to time limitatins I sold it away to a good friend: He managed to go through the tedious and flakey calibration process. Tek certainly did not make a good job here. The previous than 24xx series older Tek designs I did not like too much: Lots of contact problems. Due to the custom chips often unrepairable (here in Europe), and the modular design with too tiny buttons…

    Nowadays we certainly have digital scopes. I have an old LeCroy LC584AL. This one is a digital scope with amazing analog front end. Wow, what a beast, only from the fan noise you recognize it from 50 feet distance 😉 But I had amazing engineering results with this one. It calculates your brain out: Jitter from here to the moon. Beautiful. They still had their own OS, so no free storage of your measurement results at the NSA…

    Another oldie I bought 10 years ago is a HP54124T, you know, the large digital scopes with a test set attached to it with a flat cable. We used this one for trace TDR measurement. Wonderful piece of equipment, would not miss that experience. Last year I saw one at the HAM fest in Friederichshafen (Germany) for 230 EURO's (about $300) At the end of the day I saw it still sitting on the stand: Nobody bought it, unbelivable…

    Don't start this topic with spectrum analyzers. I know that I have a little fetish in this respect 😎  All of the old equipment still is around in our lab. Most of the time this equipment is not in use, but from time to time I take a day to play with some of them again. Lots of fun !

  13. Barry Harvey
    August 28, 2013

    I'm not gonna do this HP vs. Tek thing.  Not gonna.

  14. David Maciel Silva
    August 31, 2013


    Technological advancement has made him more viable to purchase digital oscilloscopes with even more features some engineers still prefer to use the analog oscilloscope, I personally do not use much, because they think they're a little slow …

    I would like to know from our readers, what type of oscilloscope and best for your applications and why?

Leave a Reply