EEVblog #1203 - REPAIR: Tektronix 2465B Oscilloscope

[EEVblog]

Dave repairs a classic Tektronix 400MHz 2465B oscilloscope that's been sitting on the mailbag shelf for years.
Arguably the greatest analog oscilloscope ever made.
BONUS magic smoke escape!
Although it was repaired to a functional level, this is kinda Part 1, as not all functions were exercised and tested, and not all parts were refurbished.

[cengland0]

Regarding your question about what to do with the case, simply get some sheet metal and bend it to shape.  Probably 2 cuts and two bends is all you need.  If you want to curl the edges so the sharp side isn't exposed, that would be a nice touch.

[BravoV]

Dave, from the giant 2465B discussion thread (HERE), few points I gathered regarding your 2465B.

1. The leaking SMD electrolytic caps at the logic A5 board are known to be really nasty, desolder all of it now even you don't have the replacement yet. As the leaked electrolyte known to destroy the surrounding PCB traces. As yours is still working, you simply very lucky. Once desoldered, you will have to clean those nasty leaked electrolyte as its very-very corrosive.

Example of corroded thin traces surrounding the leaked cap at two different A5 boards at forum member's 2465B, also the corrosive liquid crawled underneath the components and destroyed traces, click to enlarge.




2. Those hybrid ICs at the big bottom board (called A1), are well known to heat up/thermal overloaded if the scope is powered on without the blue cover installed, put a fan blowing the whole bottom A1 board to cool them. The case is designed with those vent holes that suck air and blow directly at the hybrid ICs. Just don't power on the scope too long without that cover attached.

The special designed vent holes formations at the bottom of the case, where air get suck in and cool straight at those unobtainium hybrid ICs.


3. The Dallas DS1225Y, stores the scope's calibration, your TL866 writer supports it and can read the old one, and write the fresh one, and its still produced by MaximIC and named DS1225AD as the "Y" version is obsolete. This is another high priority work to be done asap, as the battery inside will die anytime.

4. The soft faint blink at all LEDs at the panel when any button was pushed, is perfectly normal, if I'm not mistaken, read it somewhere, up to certain serial numbers of Tek 2465B produced have this flaw, mine too has this problem. 

5. Subject to arguments/debates, the vulnerable U800 chip, my opinion only, unscrew both nuts as they serve no purpose at all and probably may do more harm than good. I posted my arguments -> HERE



... not sure I've covered all for your scope, but there you go, hope this help, for the 2nd part of this video.

Actually, all of these above issues already covered and discussed in the giant 2465B thread, but I guess you won't have time to read it all as they're buried somewhere in there, and that thread grows almost 1.5K posts now.

[tautech]

I'm quite sure among the membership there will be a significant # of 'parts mule' 2465B's and so potential spare cases. A plea to members with parts mules should provide a case for this scope Dave.

Get 'er all fixed and finished (using a bench fan) and surely a case will materialize.

[wilfred]

It's maroon because it's a QLD'er. State-of-Origin scope repairs. Yay!

[David Hess]

4. The soft faint blink at all LEDs at the panel when any button was pushed, is perfectly normal, if I'm not mistaken, read it somewhere, up to certain serial numbers of Tek 2465B produced have this flaw, mine too has this problem. 

It is a side effect of the design and not a flaw.  The front panel indicators are driven by a chain of serial to parallel shift registers which lack registered outputs, 74x164s as I recall, so when any of the outputs need to be changed, the entire serial bit stream gets shifted through the indicators.  They could have used 74x595s which have a registered output but they were quite a bit more expensive and would have required an extra bit to control the parallel load function.

The front panel indicators on the 2247A series work the same way.  Honestly it does not bother me because it provides feedback that the processor is responding to user input.

[EEVblog]

3. The Dallas DS1225Y, stores the scope's calibration, your TL866 writer supports it and can read the old one, and write the fresh one, and its still produced by MaximIC and named DS1225AD as the "Y" version is obsolete. This is another high priority work to be done asap, as the battery inside will die anytime.

Yep, next step, along with those processor caps.

[calin]

Dave, the pretty blue screen is noting more than a piece of blue film/pespex. The 2465B has a transparent screen "sheet" while the other one is a blue one. I am almost 100% sure you can transplant the blue screen from the 2465 to the B.

[Cnoob]

@ calin

You are correct, I replace the blue filter screen off my tek2645  with the one below I brought off ebay.

Tektronix 378-0199-03 - CRT Filter Blue - 2465B, 2465A, 2445 - Loc: BB1-02-04

[Howardlong]

Ooooo my 15 seconds of fame at 22:30!

https://www.eevblog.com/forum/testgear/tektronix-2465b-oscilloscope-teardown/msg644951/#msg644951
https://www.eevblog.com/forum/testgear/tektronix-2465b-oscilloscope-teardown/msg645469/#msg645469

If memory serves me right, I made the mistake of pulling off all the electrolytic caps first, without manually noting the values as I did it, incorrectly assuming that the schematics and board layouts would be correct. That was an expensive mistake in terms of time, and I was a bit surprised it hadn't been noted publicly before :-(

Regarding the assumed incorrect functioning when pressing the A/B button, ISTR that is standard operation when you haven't set up the B trigger first. I'll have to try it when I'm back in the lab later today. Edit: I just checked on my 2467B, that is how it works. See https://nscainc.com/wp-content/uploads/pdf/T_C30BP.pdf page 2-18. I'll warn you that unless you use this scope frequently, working with the dual trigger function is not always very intuitive.

If you don't get any errors at boot, then the calibration parameters in the Dallas chip are kosher. Having said that, while you're at it, you might want to consider lifting the Dallas chip anyway, taking an image, and putting a low profile socket in for a replacement when the time comes. Unless you have the right test equipment and enough time, re-calibrating is quite time consuming. I replaced the Dallas chip with an FRAM. Unbelievably the the source code for the DS1225Y/FRAM programmer I made from bits I had in the lab has had 540 downloads.

https://www.eevblog.com/forum/testgear/tektronix-2465b-oscilloscope-teardown/msg583511/#msg583511

[aqarwaen]

you where lucky that you didnt measure anything while that exploded..if you had you may eneded with some nasty injur,very lucky..

[jnissen]

No matter how experienced you are it's funny how those exploding caps will scare the he** out of you! Just be glad it wasn't a tantalum. There are burn marks on the lab wall at one of my earlier employers after the tech installed several tantalum's backwards. Looked like a firework fountain.

[Wim_L]

Not sure where the idea comes from that the 2465b is the greatest oscilloscope ever made.

The mainframes from that period, with a selection of plugins, would obviously be more versatile (because you can choose input types) while giving you the same performance.

If you insist on portable scopes (which at the time merely meant it had a handle and a single person who was not a professional weightlifter could lift it) the 2467b got you the same performance with the benefit of a microchannel plate CRT. At the cost of containing even more unobtainium.

Having worked with those about ten years ago, they were excellent to deal with some of the issues in temperamental measurement equipment, and provided better image quality than the digital scopes (even early DPO ones) available.

Though one could argue the 2457b is only superior to the 2465b if you needed its high performance CRT. For many normal debugging tasks, they would be interchangeable. You'd only notice a difference when reaching the intensity limit of the 2465b. You were never limited by intensity on a 2467b. Set to the fastest time base, even a single trigger event would flash a clearly visible trace on the screen.

And the 7104 mainframe would get you the benefits of a microchannel plate CRT, plugins, and 1 GHz bandwidth. Never got my hands on one of those though.

[mnementh]

https://youtu.be/v9KaTgChudw

Of course the 2467BrightEye was even better... it was DESIGNED to build on the 2465 and be exactly that.   

And yes, by the way you couched your statement it's obvious you know full-well that the arguments for the 246xx scopes being the pinnacle of Tek analog design have as much to do with the fact of how much they packed into a "portable package" as it did with the fact that they invented dozens of new manufacturing processes in its development. This was literally space-age tech from the days when that actually meant something.

Okay... you've played devil's advocate. We all still know that between the documentation and the design, nobody has come close to it, and never will unless analog computing somehow has a renaissance.

That last part of your statement is the most telling... "I haven't got my hands on one of them..." and that is exactly what the 2465x family is... the finest analog scope most mere mortals will ever lay their hands on.

mnem
 

[BravoV]

As pointed at other thread on fake 2465B -> Fake 2465B in EEVblog #1203.

Dave, either its fake or not, I think you should raise this issue in your upcoming 2nd part of this 2465B repair video.


Attached below comparison of the badges between Dave's vs mine.

[mnementh]

Who cares? These 'scopes were sent into the mailbag years ago as DOA units reclaimed from the recycle bin; nobody bought them. 

The phony labels may very well have been applied as replacement parts for actual 2465s with damaged faceplates. These "reprinted" labels and front covers have been available as "repair parts" on fleaBay literally for decades. The fact that one of these front covers showed up randomly in one of Dave's videos is no proof of any skullduggery of any kind. We need to actually poke around in its innards to know that. 

Jeebus folks... get some proportion here. Sometimes a cigar is just a effing cigar. 

mnem
 

[helius]

The fakes have been circulating for years, and the seller who originated it has done it to thousands of scopes.
At this point, the probability that a non-orginal nameplate is because of this fraud is more likely than a simple cosmetic part replacement.

[rf-loop]

Who cares? These 'scopes were sent into the mailbag years ago as DOA units reclaimed from the recycle bin; nobody bought them. 

The phony labels may very well have been applied as replacement parts for actual 2465s with damaged faceplates. These "reprinted" labels and front covers have been available as "repair parts" on fleaBay literally for decades. The fact that one of these front covers showed up randomly in one of Dave's videos is no proof of any skullduggery of any kind. We need to actually poke around in its innards to know that. 

Jeebus folks... get some proportion here. Sometimes a cigar is just a effing cigar. 

mnem
 

If this seller is honest, he do not change original nameplate. He keep original 2445B nameplate and then tell "modified for 400MHz BW, functionally like 2465B" if he is honest. Period.

After modification it is NOT 2465B. It is 2445B modified for imitate 2465B as long as you do not open it.  Even if it meet same BW and functionally is like 2465B but still it is not 2465B, it is modified 2445B. If seler tell it is 2465B it means he lie.
Even if you do not understand this.

You can continue your "popcorns" and talk "who cares". 
Example I care. So do not come and tell "who cares". Now you know who cares. Least I care.
If I collect to my museum or collection one  Tek 2465B  this kind of counterfeit is the worst class of deception.

Of course there are sellers who have moral blackout and ethics broken - we can't do anything for it.
But we can spread information about known scams.

If one is happy it works like 2465B then why he need wrong model number in front panel. No, only scammer need it for that he can cheat the buyer to pay more.

[med6753]

The fakes have been circulating for years, and the seller who originated it has done it to thousands of scopes.
At this point, the probability that a non-orginal nameplate is because of this fraud is more likely than a simple cosmetic part replacement.

Thousands? Really? Where do you come up with this number? I have no doubt that this guy is a douche bag but I'd be hard pressed to think he's even done hundreds.

It's too bad the master S/N build list isn't readily available which would tell you immediately how the scope left Tektronix. I doubt Tek would release that data anyway because it would reveal their production stats which most likely is privileged information. And given that the last 24XX series was built in 1990's that data is probably archived or possibly even destroyed.         

[mnementh]

Yes, I know about this guy. I even bought parts from him to resurrect my own 2465 scopes, more than a decade ago. I think the rebadging thing he does is outright fraud, and it's a shame, because when I needed parts the guy was helpful and gave me fair prices, and even sent me pics to help.

Fraud is fraud, and I loathe it just like you do.

But you guys are adding apples and imaginary a-holes and coming up with "OMG WTFBBQ!!!" and running around with your hair on fire. It is mob mentality, pure & simple.

Needless sensationalism does no good to this venue. It makes people discount the value of articles by people like Dave. Sure... there's a slim chance that this one unit made it from an unscrupulous seller in the US all the way to OZ and wound up in a skip with a bad power supply... then got sent to Dave and sat around on the mailbag shelf for years. 

But REALLY... that is highly unlikely, and using this video as a thin pretense to AGAIN trash this guy in ANOTHER public forum is just... trashy. Folks have brought the possibility to Dave's attention. That could have been done in PM, or with a single post.

We have no evidence at all that this scope is a fake anything, or that 2465b ever even came within missile range of the thing.

We have NOTHING that ties this guy to this scope, except VERY SUPERFICIAL APPEARANCE and the fact that a bunch of people are PISSED OFF AT HIS EXISTENCE.

mnem
 

[David Hess]

After modification it is NOT 2465B. It is 2445B modified for imitate 2465B as long as you do not open it.  Even if it meet same BW and functionally is like 2465B but still it is not 2465B, it is modified 2445B.

Meeting the bandwidth specification is not difficult.  The part removed is the bandwidth limiter.  The part which is missing is the phase compensation network for the delay line so these modified instruments can never meet their transient response specifications. (1)

I think I saw the phase compensation network in Dave's video when he was perusing the insides so my guess is he has a genuine one.

(1) The phase compensation network is not required when the bandwidth is limited because no signal content will be present to require compensation.

[aqarwaen]

dave then you going to make video,then you fix this scope completely?i would love to see this bad boy completely repaired and functional.....