Tektronix 2901 Time Mark Generator Tear Down & Inspection

[BravoV]

Sharing few shots of this Tek vintage scope calibration tool that just received recently. Did a brief external inspection, it looks fine and well preserved. Powered up fine too, its just I haven't hook it to the scope to see it's output.

Planning to check it's signal and an internal inspection tear down to see what and how it look inside now, wishing no bad surprise, and hopefully I can manage to do it at this weekend, will post the innards photo shoots once its available.

This little thing is quite heavy for it's size. 

Top front panel view, serial B031078 with few stickers overlapped with years printed on them, calibration date maybe ?








Front panel, with a date 1981 written at yellow sticker, probably that is the last traceable calibration date conducted. The main crystal oven led at the right side was on once its plugged at the mains, even the unit wasn't turned on. It blinked on and off periodically, looks like its maintaining the oven temperature, measured with kill-a-watt meter, while the unit was off, when that led was on, the oven it self consumes approx. 14 watt. 




Rear panel, with switch and bnc for external reference clock.




Top and bottom view.

[tekfan]

Nice one! Looks like it's in good cosmetic condition. These things are really handy for calibration of oscilloscope time bases. I'm surprised there is actually a proper ovenized oscillator in there.

[alm]

I believe this model was between the 184 (tubes) and TG501 (compact TM500 module). Appears to share some mechanical parts with the TM500 series. It will be interesting to see the insides. Too bad time mark generators are kind of a one trick pony, not much else you can use them for apart from calibrating analog scopes.

[BravoV]

Nice one! Looks like it's in good cosmetic condition. These things are really handy for calibration of oscilloscope time bases. I'm surprised there is actually a proper ovenized oscillator in there.

Thanks, yeah, external condition is still very well preserved as you can see, thats one of the reason I grabbed it, quite a gamble though without knowing how it looks like inside. 

I was also quite surprised its still working very well, did a brief 15 minutes test with my frequency counter showed the decades old ovenized 10 Mhz oscillator inside is still at pretty good accuracy, will post the result once I opened it up.

I believe this model was between the 184 (tubes) and TG501 (compact TM500 module). Appears to share some mechanical parts with the TM500 series. It will be interesting to see the insides. Too bad time mark generators are kind of a one trick pony, not much else you can use them for apart from calibrating analog scopes.

Actually it was kinda impulse buy even I knew it does only that. 

Stay tuned, definitely will post it's internal parts shots here.

[elementip]

I have one of these 2901's - neat units. Ideal for scope calibration. One of the cool features is you can press in just about as many of the marker selector switches as you want at the same time - it will output markerss for all of them.

How long did you let it warm up? I think Tektronix recommended at least an hour and I seem to recall they also recommend leaving it on all the time if possible.

[BravoV]

I have one of these 2901's - neat units. Ideal for scope calibration. One of the cool features is you can press in just about as many of the marker selector switches as you want at the same time - it will output markerss for all of them.

Thanks, I didn't know that !

Btw, do you have the manual ? This one is bought from a surplus tech shop which has nothing else with it, just the unit. 

How long did you let it warm up? I think Tektronix recommended at least an hour and I seem to recall they also recommend leaving it on all the time if possible.

Not sure, I guess its an hour like other classic ovenized crystal does at measurement instruments.

Btw, just curious, does your oven led blink ? I mean like turned on maybe for 10 or 15  seconds, then off for another 10 or 15 seconds or so and so on.

[BravoV]

The internal photo shots.

It has two compartments top & bottom which has it's own pcb, and isolated by aluminium plate.

I leave all pictures at their thumbnail state to save bandwidth when viewing this thread, the file name of the picture should be self explanatory.

Interesting spot :

"10. Dual Coils & Old School Transistors.JPG" shows there are two coils windings but using different materials, wonder why used two different materials for the coils ?

[BravoV]

..continued..

These are more details close up shots at some parts, few interesting spots which are :

- "16. Bulova Ovenised  Crystal.JPG" shows the toaster of this unit, actually the whole unit feels very warm especially at the rear after the unit was turned on just for few minutes. Time to find for an external 10 Mhz maybe ?
 - "17. Shiny Coils.JPG" , the silvery wires still so shiny after all this years, are they silver ? or other metal ?
- "18. Solder flux residue.JPG", the Semtech rectifier is the only semiconductor that is soldered while others like transistors and ics are socketed. Also at the inset, the solder points have the flux residue while others are clean, just curious if its manually replaced ?
- "19. Trim Cap Winged Bug.JPG" , funny looking trim cap, looks like a bug with folded wings. 

[BravoV]

Interesting spot :

- "20. Very Clean Untarnished PCB.JPG" shows the pcb is still very shiny and almost untarnished at all.


Last, wonder what happened to Ray and where is he now ? 

[codeboy2k]

Great pictures, thanks!

The socketed transistors blew me away. 

The coils look almost like nickel plating. Silver is a little duller even when it's first applied, and then tarnishes very quickly.

I used to build much larger silver and nickel plated coils for high power RF equipment, so I've seen them both after a few years.

[david77]

What a brilliant piece of kit. I love that PCB, you don't get stuff like that anymore.
Hard to believe that people actually bothered to but simple transistors in sockets.

[tekfan]

Very nice! The OCXO is made by Bulova the famous company that made the wristwatch with the tuning fork oscillator. Nice old HF design for the 500MHz multiplier. Looks like Tektronix had a habit of putting as many components in sockets so they can easily be replaced if they should ever fail.

If you're interested you might want to see how Tektronix made the circuit boards back then:
http://www.classictek.org/index.php?option=com_content&view=article&id=185:tektronix-circuit-boards&catid=88:tektronix-videos&Itemid=80

[alm]

Great pictures. I wonder why the sockets for small signal transistors. Was this a holdover from the tube days which would have to replaced regularly? I remember that some manuals mention that it's not recommended to periodically remove and check semiconductors, suggesting that it is recommended for some equipment. It may also be because they considered semiconductors expensive, just like how people often used sockets for expensive DIP parts. Not much choice now most expensive parts are BGA, however.

I think the sockets cause more issues due to bad connections than they solve by making part replacement easier.

I believe the OCXO was not very accurate. A decent modern day XO will have a similar performance, but you don't exactly need 0.01 ppm performance for calibrating an analog scope.

[BravoV]

I used to build much larger silver and nickel plated coils for high power RF equipment, so I've seen them both after a few years.

Why used two different wires type instead of just using the copper enameled for both ?

If you're interested you might want to see how Tektronix made the circuit boards back then:
http://www.classictek.org/index.php?option=com_content&view=article&id=185:tektronix-circuit-boards&catid=88:tektronix-videos&Itemid=80

Thanks, that is a great website, bookmarked.

Its so amazing how complex they were made back there.

I believe the OCXO was not very accurate. A decent modern day XO will have a similar performance, but you don't exactly need 0.01 ppm performance for calibrating an analog scope.

Agreed, beside personally I don't need that kind of accuracy either. Anyway, looking at the results, what do you think of that decades old ocxo's performance ?

[alm]

I don't see any reason to doubt the accuracy based on your results, but an analog scope isn't exactly a hard test for an OCXO.  My comment was based on the 20ppm accuracy spec after warm-up, which is close to the performance of a normal XO.

[JoeyP]

I have one of these too! Thought I was the only one still hanging on to one. They're not needed for digital scope calibration, but they do still have an interesting use. The time marks are essentially impulses, and so are very rich with harmonics. You can use them as a comb generator for RF work.

[BravoV]

I don't see any reason to doubt the accuracy based on your results, but an analog scope isn't exactly a hard test for an OCXO.  My comment was based on the 20ppm accuracy spec after warm-up, which is close to the performance of a normal XO.

You're right actually, brought it to my friend's counter with gps reference, this one its still within 15ppm after about 30 minutes of warming up.

I have one of these too! Thought I was the only one still hanging on to one. They're not needed for digital scope calibration, but they do still have an interesting use. The time marks are essentially impulses, and so are very rich with harmonics. You can use them as a comb generator for RF work.

Joey, any chance you could share the manual ?

[JoeyP]

I don't have the manual in electronic form. I don't have a scanner, but if there was something specific you needed out of it, I could take a hi-res photo of the page. If you want an entire manual, check this out:

http://www.ebay.com/itm/TEK-2901-TIME-MARK-GENERATOR-SERVICE-AND-OPS-MANUAL-/370221905595?pt=BI_Books_Manuals&hash=item5632f1f6bb

I have some of their CD manuals for other instruments, and they do a really good job.

[alm]

The time marks are essentially impulses, and so are very rich with harmonics. You can use them as a comb generator for RF work.

The OCXO reference may actually be important for this application.

You can probably get the manual cheaper if you buy it directly from his website. Especially if you opt to download it instead of having a CD shipped. The 2901A not listed on his website for some reason, but you'll probably get a quick response if you send an email. He's a very reliable source and has a good reputation within the Tek/HP repair communities.

[BravoV]

I don't have the manual in electronic form. I don't have a scanner, but if there was something specific you needed out of it, I could take a hi-res photo of the page. If you want an entire manual, check this out:

http://www.ebay.com/itm/TEK-2901-TIME-MARK-GENERATOR-SERVICE-AND-OPS-MANUAL-/370221905595?pt=BI_Books_Manuals&hash=item5632f1f6bb

I have some of their CD manuals for other instruments, and they do a really good job.

Thanks for pointing that out, actually I purchased the physical manual just last week, its just it hasn't arrive yet and prolly will take few weeks. 
Bought it from ebay seller that looks like he has lots of this 2901 manuals, and still offering them like ebay deal no 130601052761.

Just curious about it's full technical specification, if its not troubling you too much, appreciate if you can take the shoot on that particular tech spec page.

You can probably get the manual cheaper if you buy it directly from his website. Especially if you opt to download it instead of having a CD shipped. The 2901A not listed on his website for some reason, but you'll probably get a quick response if you send an email. He's a very reliable source and has a good reputation within the Tek/HP repair communities.

Thanks, yes, I'm aware of that Dave's offering.

The time marks are essentially impulses, and so are very rich with harmonics. You can use them as a comb generator for RF work.

The OCXO reference may actually be important for this application.

Gents, mind explain what is this all about ?

[JoeyP]

The time marks are essentially impulses, and so are very rich with harmonics. You can use them as a comb generator for RF work.

The OCXO reference may actually be important for this application.

Gents, mind explain what is this all about ?

Because the time marks are narrow, with a relatively fast rise time, they are rich in harmonics similar to a square wave. The difference is that an impulse has a much lower level for its fundamental and near neighbors. This means the harmonics are relatively much flatter in amplitude compared to a square wave, and that allows them to be useful over a wider range for a given full-scale amplitude.  Hence the term "comb generator".

One common use for a comb generator is to test the frequency span accuracy for RF spectrum analyzers (analog hardware-based analyzers, not talking digital FFT analyzers here). For example, if you set your time mark generator to 1us, it will produce many 1MHz harmonics. So feed that into your spectrum analyzer and set it to 1MHz/div, and you should see one harmonic per division. You can instantly view the linearity and overall accuracy of the analyzer's sweep.

[BravoV]

Joey, thanks for the info, it might be useful someday.

[ciccio]

If you're interested you might want to see how Tektronix made the circuit boards back then:
http://www.classictek.org/index.php?option=com_content&view=article&id=185:tektronix-circuit-boards&catid=88:tektronix-videos&Itemid=80

Great memories.
I was very good in designing and drafting PCBs with the adhesive pads and tape.
I used a slightly different method: a sheet for pads only, two sheets (component side and solder side) for tracks only, then with the reprocamera we made three negatives (pads + component side tracks, pads + solder side tracks and pads only for the solder resist).
I worked at a 2 to 1 size, not 4 to 1 as Tek did: film, pads and tape was too expensive..
I still have some Mentor Graphics and Mecanorma supplies and tools in a closet: if my computer fails, I can still design boards!

Please note that nobody, in the board fabrication plants, wears gloves or eye protections or masks: workplace safety was less important, in those days.

[BravoV]

Please note that nobody, in the board fabrication plants, wears gloves or eye protections or masks: workplace safety was less important, in those days.

Yeah, just watch that pretty lady at the solder reflow station, her face was so close to that molten solder when she adjusted the components placement at the pcb, a single splash of that hot molten metal by accident will ruin her pretty face and eyes, quite scary.

[SoundTech-LG]

Nice one! Looks like it's in good cosmetic condition. These things are really handy for calibration of oscilloscope time bases. I'm surprised there is actually a proper ovenized oscillator in there.

Thanks, yeah, external condition is still very well preserved as you can see, thats one of the reason I grabbed it, quite a gamble though without knowing how it looks like inside. 

I was also quite surprised its still working very well, did a brief 15 minutes test with my frequency counter showed the decades old ovenized 10 Mhz oscillator inside is still at pretty good accuracy, will post the result once I opened it up.

I believe this model was between the 184 (tubes) and TG501 (compact TM500 module). Appears to share some mechanical parts with the TM500 series. It will be interesting to see the insides. Too bad time mark generators are kind of a one trick pony, not much else you can use them for apart from calibrating analog scopes.

Actually it was kinda impulse buy even I knew it does only that. 

Stay tuned, definitely will post it's internal parts shots here.

Actually there is at least ONE other trick that these ponies can accomplish, and that is audio loudspeaker impulse testing.

"Loudspeakers[edit]An application that demonstrates this idea was the development of impulse response loudspeaker testing in the 1970s. Loudspeakers suffer from phase inaccuracy, a defect unlike other measured properties such as frequency response. Phase inaccuracy is caused by (slightly) delayed frequencies/octaves that are mainly the result of passive cross overs (especially higher order filters) but are also caused by resonance, energy storage in the cone, the internal volume, or the enclosure panels vibrating.[citation needed] Measuring the impulse response, which is a direct plot of this "time-smearing," provided a tool for use in reducing resonances by the use of improved materials for cones and enclosures, as well as changes to the speaker crossover. The need to limit input amplitude to maintain the linearity of the system led to the use of inputs such as pseudo-random maximum length sequences, and to the use of computer processing to derive the impulse response.[2]"   http://en.wikipedia.org/wiki/Impulse_response#Loudspeakers

Another benefit of the Tektronix 180A, is a space heater, with 54 tube heaters glowing in the fan cooled cabinet, it can provide some comfort on a chilly evening. Almost every time constant cap had changed value, and were changed out. The electrolytics are still good after 54 years. They don't make them like this anymore!