Tektronix 2465B oscilloscope teardown

[AMR Labs]

Correct me if I am wrong as I have never bothered with any additional U800 heat sinking solutions and probably never will as they are not really required unless your internal air flow should go south and the situation is not noticed for some time, then perhaps additional heat sinking might be useful and potentially save your chip under certain heavy duty complex high frequency waveform display usage situations, but....

As I understand one of the U800 threaded studs is at +5V potential and the other is at ground, so joining both with a piece of aluminum will in essence create a short circuit. Again, I am up for being corrected if this is not correct, but I think I am not wrong.

Edit: And in my general opinion messing at all with the U800 setup as it came from the factory might create serious problems were there where none, specially if undoing and reinstalling those two nuts onto the studs. Be very careful about the torque and evenness of both nuts or the delicate IC substrate will be damaged during the process, of possibly even further down the line due to thermal stresses added to improperly torqued down nuts.

[WastelandTek]

Hmmmm,

First I have heard of this being in any way controversial.  I was under the impression that it was a standard modification.

Certainly going to look into the +5V stud and isolate if necessary, thank you.

[BravoV]

As I understand one of the U800 threaded studs is at +5V potential and the other is at ground, so joining both with a piece of aluminum will in essence create a short circuit. Again, I am up for being corrected if this is not correct, but I think I am not wrong.

Nope, those two studs are connected to -5 Volt (negative 5 volt). Or connected to U800 pin 19.

Edit: And in my general opinion messing at all with the U800 setup as it came from the factory might create serious problems were there where none, specially if undoing and reinstalling those two nuts onto the studs. Be very careful about the torque and evenness of both nuts or the delicate IC substrate will be damaged during the process, of possibly even further down the line due to thermal stresses added to improperly torqued down nuts.

Even they're installed with "right and perfect" torque, think again, these nuts serve no purpose at all. And may do more harm than good, its the thermal cycle that the chip endures, while its secured firmly, the expansion and contraction will stress out the die attached to the metal tab.

- Are they holding the U800 chip from falling out ? Nope, as the whole chip is soldered firmly on the board.

- Are they serve to improve the thermal conduction to the studs ? Nope, as both nuts use sandwiched 2 star washers, which is clearly not a good way to let the heat generated transferred from the tab to the studs. Beside the studs are not a good heat sink anyway.

Pic speaks thousands words.


Pay attention or maybe zoom in really close, watch for the tiny2 circled indentations around the hole, made by the star washer at the metal tab, definitely not good for thermal transfer from the metal tab to the stud.

[med6753]

Yep, BravoV makes a lot of good points concerning the funky construction and mounting of U800. And we've had a lot of debate about whether to heat sink it or leave it alone.

I'm in the "leave it alone" camp. The funky construction and mounting of U800 is just asking for trouble. And to my knowledge Tek never published the torque specs. It would be all too easy to over torque it and damage the die.

Look at it this way....these 24XX scopes are over 30 years old and MOST of the U800's out there are functioning just fine. And no one to my knowledge has ever posted in this thread that they had one crap out. I have two 2465's and I have no intention of touching U800. But I DO insure the fan is working properly and there's adequate air flow at all times.

[GerryR]

It is my understanding that Tek subbed out the making of the U800 IC to anther vendor (I'd have to look up to whom) and they had production problems, where the die would come loose from the heat sink.  These overheated and failed.  That problem was addressed and corrected, but the "legend" of the overheating U800 lives on.

When I had my unit opened for calibration, I monitored the U800 temperature (mainly because of the legend) and found that even with no fan blowing air across the board, the temperature of the IC never went above 135 deg F.  With a little air running across it, it settled to about 120 deg F.  I didn't think there was any reason for concern and left everything as stock.

[David Hess]

Tektronix sold their fab to Maxim.  I have some Tektronix parts marked with the Maxim logo.

[WastelandTek]

Well, it certainly was not my intention to stir up controversy.

I can verify that the 2 studs in mine are electrically connected.

Cautions about torque are well taken, but I believe I will go ahead with my heatsink with moderate torque applied.

[med6753]

And allegedly those U800 Maxim dies are more failure prone than the Tek dies. If you have a late 2465A and most 2465B's they have Maxim U800's.   

[siggi]

Cautions about torque are well taken, but I believe I will go ahead with my heatsink with moderate torque applied.

The problem with this plan - allegedly - is that when you torque down on the tab, you risk pulling the lead frame away from the die, which will kill your U800.
Good luck!

[BravoV]

Subject to debate of course, my opinion only, the ideal case for U800 chip is de-solder the whole chip, then put a SIP socket (not DIP socket).

With dual SIP IC sockets, the chip body once installed will be lifted and has more space underneath it.

This opened up opportunity to have some custom heatsink with 2 pieces of metal + thermal pad for sticking underneath the chip, and using both studs + washers and nuts to sort of clamp the chip body from the top and bottom without stressing the chip body as its now socketed, this will make an effective thermal path thru the chip's die from the bottom.

Again, imo this is not worth it, unless the owner happened to have a case that the A1 board must be de-soldered and lifted from the scope for a serious repair, with this condition it will be a good opportunity to fix this funky U800 chip installation.

[LarryS]

Hi,

I thank you for your post.

Black body radiation is how you begin your studies of radiation in Physics 10? ( I forget 1,2 or 3 - it was more than 50 years ago).  i.e. it is the simplest and best radiator.  I know not the black substrate on which you would be putting the heat sink, but I would think it adhearent to whatever lay below.
-------
It seams that my comment is posted as the last comment even though I was commenting about the 1st or 2nd page of this subject header. i.e. "Reply #5 on: November 04, 2011, 04:32:39 am".

[syau]

Scored a 2465B, while try to replace the NVRAM with STK16C68-W45 found the board with 2 feed through capacitor from the factory instead of the SMD leaking one 

[BravoV]

Scored a 2465B, while try to replace the NVRAM with STK16C68-W45 found the board with 2 feed through capacitor from the factory instead of the SMD leaking one 

Congratulation  , yeah, your A5 board revision is exactly like mine (see 1st post), its using the thru hole caps, instead of those evil SMD electrolytic ones.

I've replaced mine, and just fyi, if my memory serves me well, that both thru hole electrolytic caps were basically still in spec & healthy when I de-soldered them, but I replaced them anyway as I already bought their replacements.

Is it made in Guernsey UK ?


Just curious, what are you planning on your U800 chip ? Leave it as is ?

[syau]

The unit S/N is B06xxxx and made in USA.

For the U800, I just whipped the dust off and leave it as is.

[syau]

For the power supply, I just replaced 2 68nF RIFA cap.

[BravoV]

For the power supply, I just replaced 2 68nF RIFA cap.

Good call  , from the photo, those 2 RIFAs already cracked, aren't they ?

Any chance you could share the close up photo for those de-soldered cracked 2 RIFAs, another good example and will serve as a good reference.

[syau]

At least no visible crack on the casing

[BravoV]

Thanks, still better safe than sorry. 



According to TEA dictionary ... RIFA Madness - May refer to the state of the equipment or its owner when old RIFA capacitors suddenly let out copious amounts of pungent magic smoke. With some devices, this can occur even when the power switch is turned off. Beware! 

[Satbeginner]

At least no visible crack on the casing

(Attachment Link)

(Attachment Link)

(Attachment Link)

Not yet you mean... 

I have seen the other capacitors go dry as well, but never the two big ones.
Beware, there is a mix-up of a 30V and a 150V capacitor in the silkscreen printed partnumbers, C1132 and C1115 are swapped by mistake.
If you put them according to the silkscreen you will have magic smoke and more damage done.....

Good luck with your 2465, they are the best!!

Un saludo,

Leo

[BravoV]

At least no visible crack on the casing

Did you ever power it when the RIFAs were installed ? Consider your self lucky.

I saw your photos for just a quick glance using my phone at the 1st time, and didn't catch the details, and when I visited this thread again and used desktop big monitor, was curious and tried to zoom in  , well, picture speaks thousand words, attached of the zoomed & tweaked photo of your RIFA below. 

[WastelandTek]

she's fixin' ta BLOW

[BravoV]

Yep, big chance it will go kablooey once the ingressed water thru those cracks is accumulated enough.

Also those visible weird egg shaped, darken patterns, suspect those are actually the trapped water vapor ? 

[syau]

At least no visible crack on the casing

Did you ever power it when the RIFAs were installed ? Consider your self lucky.

I saw your photos for just a quick glance using my phone at the 1st time, and didn't catch the details, and when I visited this thread again and used desktop big monitor, was curious and tried to zoom in  , well, picture speaks thousand words, attached of the zoomed & tweaked photo of your RIFA below. 

I did power it up for an hour or so to video record  the NVRAM using exec 02

[syau]

Scored a 2465B, while try to replace the NVRAM with STK16C68-W45 found the board with 2 feed through capacitor from the factory instead of the SMD leaking one 

(Attachment Link)

(Attachment Link)

Just open it up again to replace the 5V light bulb.

Below showing STK16C68-W45 fitted in the place of NVRAM.

[AMR Labs]

If anyone with a plain 2465 or the A version is in need of the original Siemens fan motor, over at the Tekscopes-IO group someone found a seller on ebay that seems to have a big inventory of what seems to be brand new exact replacement motors. As far as I know the B series scopes never used this motor but rather a more common muffin fan type. But I might be wrong on that, and even some early B SN also use this fan. Since this motor is not easy to source, let alone in new condition, and a pain to repair, I thought I pass along this information. Even if yours is still working fine, perhaps a good idea to keep a spare on hand just in case for future needs.

Edit: this fan will also fit the plain 2445 scopes, and as pointed out by the seller, it will also fit some 46x and 47x series as well. Do your own homework, and probably a good idea to check the Tekscopes thread to be sure as I have no experience with those older scopes.

I am in no way connected to the sale.
https://www.ebay.com/itm/274013390896

Here is the original thread at Tekscopes group:
https://groups.io/g/TekScopes/topic/71491839