Tektronix 2465B oscilloscope teardown

[BravoV]

OK, I had the video of Exerciser 02 for safety's sake. The video quality was so bad that it's a miracle that I was still able to write all the data to the hex editor ... <snip>

But the device did not start now, Trigger'D flickered happily, the device complained about corrupt calibration values.....<snip>

... looking at the solder joints of the NVRAM under the microscope and discovered that I had severed a trace directly at the solder pad while soldering out ...<snip>

<snip>... everything work again.

Bravo  ... reading your "adventure" above, that you are saved by the Exer 02 video really excites me.

Thanks for sharing, this will be handy for others, or at least motivate them. 

[Alfons]

Thanks for the feedback.

Well, I wanted to be sure. That's why the video. And in my first tests with the device, I had the impression, that the calibration values are still OK. It had only 1200 hours behind and it has a high serial number, probably one of the last devices?.

Of course, I downloaded some bin-images here because I thought I had not transferred my data correctly. But the mistake remained. The chip was OK, could be described and read clean. Actually, it was logical, that I had caused a demange. Under the magnifying glass was nothing to see, only under the Microscope it was visible. And of course I could have measured all the traces and find it ...
I can only advise anyone to watch the video synonymous, if it is OK. I had extra download a video-cut-program to improve the image quality and then it went. My God, two days went for it, just because I did not work properly. Well, I learned something else for that.

[Alfons]

Still looking for the cause of the flicker. And came across other things ...

I programmed the NVRAM with a bin file from another 2467B to see, if it changed a bit because of the flickering, so it might be affected by the cal values. The device started normally, but reported errors because of the installed option 9. Also in the lower readout points appeared between displayed values. Nothing changed on the flickering. So again programmed the original firmware, that I had previously saved as a video and copied.

The device started normally, but at the end it showed Fail 4 10. After I pressed A / B-trig everything went normal. What happened now? Another NVRAM programmed, the same. Something was wrong with the Cal values. But why did it work without errors, before I played the other bin file.

It occurred to me then, that in the meantime I was running Cal 08 and that, after I played the original bin file. So I ran through CAL 08 again, without setting anything and then the error disappeared. As it stands, the software actually stores CAL 08 values, even though they are not set by software, but by trimmers.

You always learn something new about the device. It is only important that the error message Fail 4 does not necessarily indicate a defective NVRAM or incorrect (Cal 1) calibration values.

[Alfons]

Ok, even if I give the solo entertainer here:

The flicker occurs only in connection with the Owon AG1012 and only in the range between 5-10Mhz. If I press the x10Mag button, the flicker disappears too. The signal representation is otherwise first class. The CRT is extremely bright, so I almost completely reduce the intensity for normal signals. Only with critical signals do I turn up a little more. There is a screen panel installed, which increases the contrast. It is not this blue that you usually see on the other devices in the series.
Actually I wanted to sell the device again, but I will keep it. You should not give that up. A fantastic device. 

Maybe I'll change the Elkos in the power supply. I have all lying here (Panasonic FR).

[tggzzz]

Are you aware that the CRT beam is used for both the readout and the trace. If it it drawing the readout, it isn't drawing the trace. This timesharing leads to flicker.

Test. Get a display with flicker. Rotate the readout knob to the centre, so that there is no readout. Does the flicker disappear?

[Alfons]

The flicker is independent of the position of the readout knob. But others also wrote that, who have this flicker too. It has nothing to do with the readout, it also occurs when the readout is turned off. It has a very slow frequency, anything around the 2-3 Hz, it looks like waves of a modulation.

[Koreth]

Hello all,

I have recently come into possession of 4 Tek 2465Bs. I purchased them on the cheap as they were all listed as "for parts", "not working", "untested", etc. I'm well aware of the fact that all of them will need repair work to be functional, and some parts may be hard to come by. I'm okay with that; the whole point of the purchase was to begin a restoration on a complex device, which, after completing, I would have a useful and capable tool,  and understand it's function, capabilities and limitations better than the average bear. I'm pretty sure that between the 4 scopes, there's enough parts to get at least one of them operational, if not two. If I can get 3 or all 4 fully functional, freaking awesome , but I'll be happy with two.

I wasn't sure if I should post here in this thread, start my own thread here in the Test Gear section of the forums, or perhaps one in the Repair section. But, this thread seems to have become the de facto thread for pictures of 2465B teardowns, so I figured the photos I took during the teardown I did of the first scope would be appropriate here, at least.

So here they are. The first two arrived earlier in the week, the 2nd two arrived later in the week.

B0152xx - 2465B CT
B0120xx - 2465B
B0123xx - 2465B CT
B0595xx - 2465B CT



I'll be honest, I purchased B120xx to serve as a parts mule for the others. As I go through these scopes, I'll be moving all the broken bits and problems to that one. Then, if it is still in salvageable shape at that point, I may work on restoring it. Or I may part it out to have spares on hand to keep the others alive, or sell off the parts to recoup my costs. We'll see. It seems a shame to treat what was once a nice instrument as scrap, but if its still good or easily-restoreable parts can serve to keep its brothers functional and useful, that strikes me as less of a shameful waste than it ending up in a landfill or dissolved in hazardous chemicals to extract the precious metal content.

I was hoping to put this all in a single post to this thread, as it's been my experience on other forums that such is considered better etiquette. However, it looks like there are limitations on total attachment size, and I didn't want to shrink my photos down too small, lest details be lost. So, I'll be following this post with a number of other posts, each containing a logical group of photos, shrunk to keep it under the attachment limit.

[Koreth]

B0152xx  I opened up first; I have not even attempted to power it up. When I received it, I heard something rattling around inside, and I did not want to take the chance that whatever the something was, that it was something conductive, or that it might land across a couple points that should not be conducted across.

First up is pictures of the casing. The case was a right mofo to get off. I don't know how clear it is in the photos, but the case has taken a few good, hard smacks, including one to the right handle mount, collapsing that inward causing it to rub against the chassis. I ended up bending a bit of the rear panel trying to pry the collapsed parts past where they were hanging up. I'll have to bend that back later to the case will mount properly. You can see in one of the pictures that one of the screws for the bag on the top had been lost and replaced with another screw that's way too long. I had to screw that out to avoid it hanging up on the chassis or damaging the components as I pulled the case off.

[Koreth]

After disassembling the scope enough to get at the power supply and seeing its condition, I knew that this project was going to take a while. With that in mind, I focused on getting the case back into shape that it could slide on the chassis easier without scraping on the chassis or being a risk to the components therein. Some work with a 45oz deadblow hammer and some flat boards did the job. Turns out, a piece of scrap 1x6 board from the local big box home improvement store fits inside the case just perfectly.

There is definitely a visceral satisfaction that comes from bending things to your will by pounding them with a hammer.

It's not perfect, but it will do. Hopefully the pictures show the reduced degree of protrusion on the inside of the case. In the future, I'd like to get another board and cut a 3 inch hole it it so as to give some room for the mounting boss to be hammerd out, while properly flattening the rest of the case, but this will do for now. It lets the case slide back onto the chassis, keeping dust and debris out of the unit while I'm not working on it, hopefully reducing the amount of self-inflicted work I will have to do in order to get B0152xx operational again.

[med6753]

Yes, this thread has become a "catch all" for the entire 2465 series. So go ahead and post your adventures of repair or questions here. You'll get the quickest response because a lot of folks monitor this thread. 

[Koreth]

med6753: noted (and exemplified by your reply not a minute after my own, while I was preparing the next batch of pictures!)

With the case off, let's examine the A1 board,  in all of it's fully-optioned glory (note the extra cables draped across it going to various headers on the board). One of the cables routes right by the much-discussed unobtanium U800. I'm slightly worried that they might impede airflow, but if the scope still works after I get the power supply repaired and U800 doesn't show any signs of thermal stress, then I intend to leave it the hell alone. I suspect more U800s have been destroyed by monkeying with them when it wasn't necessary.

When I posted this picture elsewhere, I was told that this board was a thing of beauty. I kinda agree. When I was a child, all those through-hole doodads were inscrutable deep magic. Now that I've grown and understand that the actual magic is the smoke hidden inside each doodad, I can really appreciate this board. This board was not trivial to lay out and took some serious engineering to get right. And with the invariable parasitic impedance that comes with through-hole components, the fact that it sucessfully operates at 400Mhz is quite impressive. Mad props to the Tektronix engineers for making it it all work without wholly sacrificing the ability to effect adjustments or repairs in the field.

There will be close-ups in the next post.

[Koreth]

Not much to say here other than remark on the date codes. The date codes on the various 3rd party ICs look to be from late '88, and the Tek-made heat-sinked opamps have date codes from the last week of February, and the 1st week of March, of '89. The latter is the latest datecode I've seen so far. So I'm going to guess that this scope was  built in early to early-mid March of '89.

I don't see any damage to any of the parts of the board, or broken components, thank-god. Some of the parts on that board are definitely unobtanium.

There's only a few electrolytic capacitors on the board, but given their age, I intend to at least check them and verify they are still within spec. But that can come later.

[Koreth]

Next up, we have pictures of the CRT-side of the chassis. If you look you can see scrapes on the chassis where the case refused to come easily. But, looking closely, I an see that none of the parts going down into the tube's drive mechanism itself are damaged.


After that, we can see the top plate, which protects the power supply, and the various option boards. In these pictures you can see scrape marks from that way-too-long screw in the second post in this series. I'm glad I noticed that before it did some serious damage.


In the next photo, I have a confession to make. You can see from the disconnected leads going to the word-recognizer trigger board that I had in fact begun taking this scope apart before taking pics. I put it back together again so I could properly document the teardown with pictures, but I couldn't remember which order the leads went on, because stupid me hadn't taken pictures before disassembly, and there were no helpful markings stamped into the metal plate. Oh well, I'll look it up in the service manual later.

[Koreth]

Nothing terribly exciting here except some more pics of the the top plate.

I do note a couple things.  There's various marking stamped on the plate denoting connections, A-E for example on the CRT side of the plate. I have headers C, D, and E, populated. For comparison, those were not present when I opened up the plain 2465B, B0120xx. I wonder, are those cables supposed to carry signals from the option boards down to the main board?

Another stamped marking to note is Pin 1, in various places, where a ribbon cable connects either to the option board, or the A5 board down just the side of the chassis from the marking. Another bit of attention to detail that tells you that Tek build this thing intending for it to last and be servicable, not be some hermetically sealed throwaway device. Such design came with a pricetag of more than my parents' combined monthly salary in the 80s, but, hey, you get what you pay for.

[Koreth]

Onto the A5 board.

One thing I note here, is that unlike most of the teardown pictures I've seen of 2465Bs, is that the A5 board is 1) all through-hole in design, and 2) the infamous leaky SMD capacitors, with their board-and-component-destroying corrosive electrolytic aren't present. In their place are a few regular through-hole axial electrolytics. So, I think I dodged a bullet there. However, due to the scope's age, I do intend to test those electrolytics to ensure they have not drifted out-of-spec over the past 30 years.



Another thing I note is that I don't have the famous Dallas battery-backed SRAM, as seen on the later, SMD-style A5 boards. This thing looks like a plain battery, not a RAM-chip or ROM chip of any kind, especially given it's only 4 pins on the board. Apologies for the blurry photo there, but it looks like there's some dried flux residue on those solder joints. I don't know if that's normally there, or if that means that this board has been worked on before, and the battery has already been replaced once. It's hard to tell, given the lack of date codes on the label. I'll have to compare the A5 boards in my other pre-B5xxxx scopes.

[med6753]

Some of your scopes have the Counter/Trigger option which accounts for the additional ribbon cables you see.

You are wise to leave U800 alone. The additional ribbon cables over it from the options will NOT affect it. I have a 2465 DMS which is fully optioned with the C/T as well as the DMM option and I've never had any issues with U800. The key to U800 longevity is adequate airflow which means keeping the fan operational. Playing around with U800's mounting screws and attempting to add additional heatsinking is just asking for trouble....although they are some who will disagree with me. 

[Koreth]

The more astute of you might have thought that maybe something was missing in those pictures of the A5 board. You would be correct. As I mentioned before, when I first received the scope, I heard something rattling around inside the case. When I stated to remove the case, I found what it was: the 10.000Mhz crystal from the A5 board. Apparently, when the side of the case took the hard smack that collapsed the handle-mounting boss inward, that impact also knocked the A5's crystal free from the board. Well, damn. That's gonna be an important thing to fix if I want the scope to work at all. Microprocessors have a hard time working without a clock to drive them, after all. I forgot to take pictures of this, or of the crystal itself. Sorry.

Fortunately, I dodged another bullet here in a way. Take a careful look at the solder joints on the back of the board where the crystal was. The joints are still there, less two tiny holes in the middle where the crystal's leads were. I do believe I somehow managed to escape damage to the board, pads and traces. I'll probably want to replace the crystal with a new one, but damn! That could have been much, much worse, with potentially the A5 board damaged beyond repair.


A well laid out board is a thing of beauty. Anyone else here getting memories of Tron?

[Koreth]

Onto the front of the scope. Probably should have done that first, before disassembly. Live and Learn. Before you ask, no I don't have pictures of the rear of the scope. I'll take some later.

While all the main controls are intact, I'm especially thankful that the volts/div and time/div knobs were undamaged by prior owners or during shipping. Unfortuantely, there is some damage to the front panel. Note the punctured hole on the trigger indicator lights. Additionally, the knobs under the display are missing their outer part.

I can borrow the smaller knobs from B0120xx. I also have a couple face plates that I ordered, without damage to their clear portions for the indicator LEDs. That will go back on during reassembly. I will tear apart the front panel further at a later date to make sure that there isn't any other damage that needs addressing. As a bonus, the face plates ordered also have the markings for the TV trigger option board controls, and I managed to snag a couple of those. So, when this scope is brought back online, it will be damn near fully optioned. What person wouldn't want a fully-optioned Tek scope for their home lab or hackerspace?

[Koreth]

With the top plate off, now we can take a look at the options boards. One of these controls the GPIB interface, the other controls the Counter/Trigger/Timer option. I apologize for the light levels in these photos, the flash on my phone is fairly powerful and can do a pretty good job of overwhelming the sensor.

In the 1st pic, you'll see a blank label on one of the chips. When I opened this up originally, the window of that PROM was exposed, and a no-longer-sticky label came fluttering out of the chassis. I cut a piece of plain label and stuck it atop the chip to protect it. The label is tucked in a baggie with the screws and the crystal from the A5 board, so it doesn't get lost.

[Koreth]

I'm not sure what this board is, or what it does. I do believe it is a standard board, as when I opened up B0120xx for comparison, it also had this board in it. Whatever it is/does, it fits in the central cavity of the scope, between the option boards and metal wall, and has a ribbon cable that goes through the floor of the central cavity and connects to the A1 board. I was going to leave it in place, initially, but I found it easier to get at the power supply assembly with both hands with it out of the cavity.

[Koreth]

And finally! Onto the power supply!

I was especially interested in this, given that there are known failure modes with the power supply, especially the capacitors hanging off the rectifier cracking their cases from heat and age, and eventually blowing.

Of course, as soon as I got the assembly free and out into the light, I had to wince. Yeah, something caught fire and went boom there. I knew the box caps in the rectifier were problematic, but damn. They took out at least one resistor and possibly other components with them, as well. But fortunately nothing else in the rectifier section *looks* immediately damaged. Looks be damned, I'll be verifying the spec of every other component in this section and immediately downstream of the rectifier section. If only a minimum of components in the rest of the power supply assembly are damaged, then I'm tempted to turn that fried resistor into memorial, honoring its self-sacrifice to protect the rest of the PSU.


It may amuse y'all to know that the fried resistor actually has not released *all* of its magic smoke. I measured it after taking these photos. Per the schematic, it's supposed to be 30 Ohm. Well, now it's a very crispy 2.68 kOmh resistor (or 26.8 kOhm, I'll check again when I get home).

Moving onto the rest of the CPU, we can see the usual components of a switching power supply. Fortunately, none of the regular electrolytic capacitors look swollen or otherwise damaged, but they're going to be checked anyway, given their age, and especially given the current surge the PSU was subject to. I do see more of those problematic box caps. Those are getting replaced, period, given their history. I'm not waiting for them to fail.

[Koreth]

That's all I got for now. I'll post more pictures as I do more work.

I know there's a parts list (or a few of them) for the various capacitors in the PSU floating around both in this thread and the Tek Scopes group on group.io. However, I want to do a bit more research first, test and verify the components on the A2 and A3 boards, and understand what it is I am replacing and why. For that was one of the points of buying a pile of broken scopes to repair; get a better understanding of my tools by learning how to fix my tools.

A member of my local hackerspace found the service manual for these scopes. Additionally, I have found reprints for sale, and I'm tempted to pick up one of those as well. digital copies can be convenient, but paper can be pretty convenient as well, especially when you need the whole diagram in front of you in a readable size.

[tggzzz]

Of course, as soon as I got the assembly free and out into the light, I had to wince. Yeah, something caught fire and went boom there. I knew the box caps in the rectifier were problematic, but damn. They took out at least one resistor and possibly other components with them, as well. But fortunately nothing else in the rectifier section *looks* immediately damaged. Looks be damned, I'll be verifying the spec of every other component in this section and immediately downstream of the rectifier section. If only a minimum of components in the rest of the power supply assembly are damaged, then I'm tempted to turn that fried resistor into memorial, honoring its self-sacrifice to protect the rest of the PSU.

I've had that Rifa take out that resistor and some of the prepreg It was fine after cleaning and replacement.

I presume you've seen http://www.condoraudio.com/wp-content/uploads/Projects/Tektronix-2465B-Oscilloscope-Restoration-Repair.pdf

Don't forget two of the PSU electrolytic caps are swapped in the documentation. Remove and replace one at a time.

[Koreth]

I presume you've seen http://www.condoraudio.com/wp-content/uploads/Projects/Tektronix-2465B-Oscilloscope-Restoration-Repair.pdf

I have. That was going to be part of the basis of putting together my own list of parts to order from Mouser or Digikey. However, I wanted to take the time to understand the circuit at little better, and get a feel for what the components in that list (or elsewhere on the board, if needed) are supposed to do in that circuit. I figure if I can understand a component's role, then why Tek selected the parts they did during the original design should become apparent, and therefore inform what currently-available parts would best serve as replacements in that role. For all I know, that list is still current. But, I'd like to understand why, as opposed to just blindly throwing a bunch of parts on a board.

I've had that Rifa take out that resistor and some of the prepreg It was fine after cleaning and replacement.

In your experience, what other parts tend to be damaged when that RIFA and resistor blow?

[tggzzz]

From memory, there is a series resistor, and if the capacitor is a short circuit then the resistor has too much voltage across it.

Just follow the smoke and clean it up!