Tektronix 2465B oscilloscope teardown

[FireDragon]

I have discovered that both RT1010 and RT1016 were sourced from Ketema's Rodan division. That division was sold to Selas and renamed RTI Electronics. RT1016 has the manufacturer part number SG200-S. RTI still sells the SG200. The "S" was for "straight" leads. However, the current tolerance is 15% and not 10%. Otherwise, the part appears to be a dead match. That is Mouser part number 995-SG200 and is $1.33.

I have not found a direct match for RT1010. Its manufacturer part number is 75DJ7R5K-RO-220 which appears to be an older product. It is 7.5 ohm, 10% and 3.9%/C. The value of IMax is not given, but is 7A for RT1016 and so should be fairly close. These are used in opposite input legs of the power supply. However, the circuits are slightly different so the current flows will also be slightly different. The RT1016 has 3.3%/C and is curve "A". There is a curve "B" which is 3.9%, but it is only available for a 40 ohm part with much lower maximum current rating. The closest that I have found is SG64 (Mouser part 995-SG64, $3.62) which is 7 ohms, 15%, curve J (3.5%/C) and 10A for IMax. It may be acceptable. It may even be almost the same because RT1010 is larger than RT1016 which may be an indication of a higher IMax.

I have a query out to RTI to find the best modern match for RT1010 - they have several product lines and one of those may be better.

[electrongeek]

I have no idea how NTC thermistors age or typically fail. Since at room temp they are close to their specified resistance, I'm kind of doubting that is your problem. I would keep looking for something else, but wouldn't hurt to replace them anyway I suppose.

I wonder if using a variac to bring voltage up more slowly while monitoring current with a clamp meter would be helpful? That might depend on how the switching power supply starts up, but might get you past the surge and you could at least tell if current draw was normal or not in the steady operating state.

Chip

[FireDragon]

I have no idea how NTC thermistors age or typically fail. Since at room temp they are close to their specified resistance, I'm kind of doubting that is your problem. I would keep looking for something else, but wouldn't hurt to replace them anyway I suppose.

I wonder if using a variac to bring voltage up more slowly while monitoring current with a clamp meter would be helpful? That might depend on how the switching power supply starts up, but might get you past the surge and you could at least tell if current draw was normal or not in the steady operating state.

Chip

The bridge has shorted and is almost certainly my problem. The NTC thermistors are used as surge limiters and do fail and do age. Mine are significantly out of specification. I am also preemptively replacing the surge suppressors.

[FireDragon]

For those who might want (now or in the future) a replacement for the RT1010 thermal resistor, 75DJ7R5K-RO-220. This is 7.5 ohms, 10% 3.9%/Cpart. Contact

   Mehdi Samii
   mehdis@ametherm.com

at Ametherm. He is making a couple of free samples for me and stated that he is the contact for any future requests.

The RT1016 is SG200-S 5 ohm, 10% 7A and is still currently available. It is part SG200 (the -S means straight leads, kinked leads are now SG311). The Mouser part number is 995-SG200 for $1.33. It is also available at DigiKey.

These parts are used in many Tektronix products as surge limiters. They may not fail often, but do age. Mine haven't filed, but appear to be significantly out of specification.

[MSO]

I've completed recapping the A3 and A2A1 boards on my 1995 era 2467B. I found only the C1101 on the A3 board actually leaking electrolytic and the RIFA capacitors (C1016 & C1018 on A2A1; C1020, C1052, C1051 on A3) with the gold tinged translucent cases showed mild crazing, but no actual breakage. All in all, everything was pretty clean.

Now I'm moving on to the A5 digital control board I intend to replace all the aluminum electrolytic capacitors (C2011, C2113, C2331, C2965) with new aluminum electrolytic caps.  Is it recommended to replace the tantalum caps as well?  I've ordered the replacements for the tantalums (C2010, C2350, C2420) as they were cheap enough, but I don't know if these are at all problematic.

The real challenge on the A5 board is replacing U2460, the DS1225Y-200 NVRAM with the built-in battery.  I've ordered a new chip and a socket and will perform a direct replacement.  My hope is that I'll be able to make a copy of the data in the old NVRAM and copy the data to the new NVRAM.  It seems that the chances of success doing this is pretty limited.

I've hand copied all of the calibration data (EXER 02) from the NVRAM so I have that information saved, but I don't know if it's necessary to save any/all other data from the NVRAM.  Perhaps the ROM information (EXER 04) should be saved as well?

If I'm unable to read the current NVRAM, I thought I'd read the new chip just to get a clean binary file that I would then edit to insert the calibration data or anything else I might need. The calibration data, as I read somewhere is this thread is located starting at address 1E00.  Now EXER 02 returns the calibration in the form of 16bit words instead of bytes. The 2467 uses the Motorola MC68B02P microprocessor so I assume it uses the Motorola convention for byte ordering.  The most significant byte is written at the lowest address followed by the least significant byte at the next address. Does anybody know if this is correct?

Any help here would certainly be appreciated.

[FireDragon]

The 2467 uses the Motorola MC68B02P microprocessor so I assume it uses the Motorola convention for byte ordering.  The most significant byte is written at the lowest address followed by the least significant byte at the next address. Does anybody know if this is correct?

The Motorola 68xx processors use the big endian byte ordering, so you are correct.

The tantalums are sealed, so they don't have the same failure modes as aluminum electrolytic capacitors. They do have their own failure modes, but generally they don't need to be replaced. Some people suggest measuring the ESR, but that is difficult to do without lifting one pin. Lifting a lot of pins unnecessarily just increases the risk of damage due to recapping. If you are going to lift one pin to measure, you might as well just replace it and be done with it. I would recommend leaving the tantalums alone unless you can trace a failure to one because the likelihood of failure is comparatively low.

The only things I have replaced (during a full recap) was all of the electrolytic capacitors, and some resistors on A1A2 which tend to fail due to overheating. I had two that were seriously burned (and two of my RIFA capacitors had exploded). I replaced the 0.5W resistors (and two 0.25) with 1W versions. Some people replace those with 3W or 5W versions, but 1W should be enough to prevent the problem and are the same size as the originals.

If your scope has failed, you might want to consider replacing more, but sometimes just the capacitors is enough. In my case, my scope had failed and the capacitors / resistors weren't enough. I found that the bridge was shorted, so I am replacing it, the surge suppressors and the two thermal resistors. Only the bridge is really necessary.

As far as the battery is concerned, I have a non-SMD version of the 2465BCT so I have an explicit battery. Even so, I plan on replacing the ROM with an FRAM. You can do something similar, but either way you need to save the calibration data. EXER 02 allows you to read the data, but doesn't do squat for putting it back. So you need to read out the data and then write the data to the replacement IC. That is a complicated process and I can't advise on that. There are people who have replaced their Dallas chips with an FRAM and a piggyback battery. You need to do some searching to make sure you fully understand what you are doing for that. If you can't read / write the data, then you need to do a full calibration. That takes specialty equipment.

[Old-E]

My first posting!   Been following this thread for some months with interest, because it's one of the best.  I purchased a 2465B months ago with problems.   It worked great - except when it intermittently didn't.   Decided to keep and fix it because it was so clean inside and out with no physical damage.

Through this process I've learned that this scope is an electronic engineers dream.   That is; a manager puts together a team of engineers with free rein to design the best scope possible.   The result is; short of a signal trace memory, it includes just about every detail known to man at that time.   Consequently, it can be a dream for users - but a nightmare to fix when something goes wrong.   But at least it is designed to be repairable.

While searching for the cause of the signal trace going crazy, the first unrelated thing I replaced was the fan.   It was a bit noisy and I suspected it to be the original fan (Nidac, Mod. D08G-12th).   I replaced it with an NMB Technologies Corp. #3108NL-04W-B30-P00 which seemed to be a perfect match.

While confirming the fan polarities and air flow direction, I found the voltage at the fan feed through to be 6.7 volts.   This did not sound right as both the old and new fans were rated for 12 volts.   Further investigation revealed that they didn't simply just drive the fan with 12 volts when the scope was powered on - they put in a temperature sensing voltage controller, powered from the +15 volt power supply.   So when the shop and scope are cold, the fan will see a floor of 6.7 volts.   And as proven with hot air from a heat gun on the temperature sensor, the fan voltage increases - then decreases when blowing room temperature air.

Taking it a step further, with the scope back in its case on a clear bench with the front elevated a few inches by the carry handle, I turned it on.   Over time, I monitored the voltage through added leads extending out through the rear grille, and the exhaust temperature with a thermocouple in the exhaust port.   Over the course of an hour the exhaust T rose from 62 degrees F (shop T), to 83 deg. F where it stabilized.   That's a delta T of 21 deg. F.   The fan voltage increased from 6.7 to 9.7.   Therefore, I'm assuming that in a hotter environment the fan will see 12 or more volts.  A closer look at the new fan spec's found it's rated at 6-13.8 volts.   So from all this, I assumed the scopes cooling fan is working correctly.   Ok - - this is an elegant way to control the scopes temperature, but I question the value to cost trade off.

The other thing I found with the fan was that the sheet aluminum assembly that clamps the fan in place was a bit too tight.   It was deforming the fan spokes enough to cause a noise increase in the fan bearings.   looking at the original fan revealed that it's frame spokes were permanently deformed by the excess pressure on the center of the fan, rather than the outboard ears that were designed for mounting.   They were EE's not ME's    The old original fan now runs quieter on the bench with the mounting pressure removed.   To fix this, I tweaked one of the bends in the aluminum clamp, ever so slightly with long nose pliers, to remove most of the pressure on the fan.   it runs quieter now while still being firmly held in place.

More later
Old-E 

[MSO]

Thanks for the tantalum info FireFragon, I'm from the firmware side of things, so my EE skills cause me more problems then they solve.

My scope was working, so my goal was to replace all the electrolytic caps before they ruined the boards.  I did replace all the electrolytic caps with new electrolytic caps rather than attempting to spec tantalum replacements. From what I could gather, the tantalums require a bit more knowledge of their usage than I possess, so I stuck with the 'lytics.

I looked at replacing the DS1225Y-200 chip with the FM1608 family, but decided to just put in a new socketed DS1225Y-200 instead. The device currently in use is over 20 years old and I figured that I could easily program another one ten years from now and be on my merry way.

You did set a bit of a high standard when you replaced all the 'lytics on the main board, though, and it seems foolish of me not to do the same thing.  Did you replace the main board caps with tantalums or did you use new electrolytics?

OLD-E, thanks for the fan info.  Mine fan is a bit noisy too and I was going to order  a new one as well.  With only size, voltage and current specs to draw upon, I didn't feel comfortable specifying a replacement. I'll follow your lead and then blame you if it doesn't work out. (I'm kidding!).

I did order a new MiniPro TL866A programmer from amazon, but I didn't pay enough attention; seems mine is coming via slow boat from China and I won't see it until the end of March. So I've put the scope back together and it's still running fine.  I'll look into the replacement of the capacitors on the main board and double check all the info I can find about loading new data into the DS1225Y while waiting for my programmer to arrive. I have my fingers crossed and am hoping I'll be able to read an image from the old chip.  I expect that I'll have to parallel the battery in the old chip, but that doesn't appear to be as difficult as removing the chip from the board.

Thanks for all the contributions on this thread.

[Bryan]

For those replacing the Dallas you may find some good information here.

http://worldphaco.com/uploads/TEKTRONIX_2465b_OSCILLOSCOPE_CALIBRATION___REPOWERING_THE_DS1225.pdf

[Old-E]

Thanks Bryan for the article.   I had found that one before starting on the scope and it has a lot of good info.

Back to cooling - Actually MSO, if that fan does turn out to be a problem, you would do me a favor by letting me know.   I bought it 3 months ago from Digi-Key (their #P15646-nd) for $13.61.   I think it has 15-20 hours on it so far and running good.

Another parallel effort was cooling U800, the IC that has been reported by others as running too hot.   As I recall from reading further back in this thread, Tektronix turned their production of this chip over to Maxim to produce.   As it turns out, my chip is an original Tektronix one which is said to run cooler.   Running uncovered on the bench, it runs fairly warm to touch, but not hot.   Looking closer, my finger indicated that the heat sink end of the chip was hotter than the opposite end.   So it would appear that their was a reason for putting the heat sink on that end.   Conclusion was that this Tek chip probably did not need additional cooling.   But, since it's all apart, a little more frosting on the cake can only help.   So the decision was to add a heat sink to the existing heat sink on the hot end of the chip.   Laying a heat sink across the top of the chip will certainly help, but the concern was the questionable thermal resistance through the plastic case.

This heat sink was aluminum machined to 5/8 " OD x 3/8" high with 8 radial fins on sides and top.   It screws onto the existing stud on the hot end of the chip.   I drilled the tap hole a few thousands smaller than normal to get slightly more thread contact for better thermal conduction.   Then I lubed the threads and the seating surface on top of the existing nut with a thermal conducting paste to further aid in conducting the heat out.   The heat sink bottoms out on the existing nut which was just high enough so that the bottom of the heat sink cleared the top of the IC.   The top of the heat sink is < the height of the surrounding parts.   This way, none of the existing hardware had to be disturbed, but it does rely on the thermal energy conducting up through the face of the nut and the threaded stud.   Could have made the heat sink larger in diameter, but didn't want to shade to much of the IC from the cooling air coming in through the bottom of the case.   It was an intuitive tradeoff.   The new heat sink feels to be the same elevated temp as the chip, so apparently it is conducting as planned.   But, the heat sink fins need air moving past them to do any good, as through the vent holes in the bottom of the case. 

I have pictures to insert, but can not figure out how to add them.   Tried Copy & Paste, and I found the Insert Image button, but nothing works.   "Help" only says that it can be done.   Can someone please advise?

More later with photos.
Old-E

[tautech]

@Old-E
Using Atttachments and other options, select a forum compatible file and upload with your post. Pics do not have to be large to show good detail and normally ~100Kb is plenty.
When your post is made you'll have a thumbnail at the foot of the post and if you copy that URL you can edit your post with the URL inserted between the IMG quotes.

Use Quote on somebody's post with imagery to see the syntax used. 

[FireDragon]

Thanks for the tantalum info FireFragon, I'm from the firmware side of things, so my EE skills cause me more problems then they solve.

. . .

You did set a bit of a high standard when you replaced all the 'lytics on the main board, though, and it seems foolish of me not to do the same thing.  Did you replace the main board caps with tantalums or did you use new electrolytics?

I only replaced the aluminum electrolytics. I did buy many of the tantalums, but decided to leave them alone. There really aren't that many tantalums, the problem is that the parts list does not distinguish between aluminum and tantalum and bipolar. The only way to find out is to have someone tell you or to actually look at the original parts. The latter can be difficult before removing the boards.

Some tips on removing the main board. The removal instructions give a complex process to remove the power switch assembly. Completely unnecessary. Just unscrew the screw at the front that holds in it, then you can just twist it out of the piece of connecting plastic and (holding it carefully!) remove the entire assembly in one piece. Putting it back is just as easy. Just be really, really careful with the CRT leads. They are very delicate. If you don't have a steady hand, lots of patience and plenty of small tools then I wouldn't recommend removing the main board - you really don't want to fark up the CRT!

Make sure to get lots of pictures on how things go back together - especially for the CRT leads which are in a dark hole. Also be careful replacing the capacitors under the delay line. I propped up the delay line with something to give me just enough room to work. It was still tight. There is also a wire that is held down by a retainer in the board. Mine broke! Pretty trivial, I will use a piece of Kapton tape to hold the wire down because I don't have a replacement, but hopefully forewarned is forearmed!

Here is my parts list for the main board. My scope is pre-B050000, but the parts lists does not list any changes that affect the electrolytics so my replacement list should be adequate. I purchased everything at Mouser. For a small shipment Digi-Key has slightly less shipping, but includes tax. In this case, Digi-Key either doesn't have the parts that I used or are as much as 50% higher in cost. To be fair, there are a lot of parts that Digi-Key has and Mouser does not (but not for this project).

   C0102  290-0973-00  100uF    20%     25VDC                 UPW1V101MPD  100uF  35V 20% 105c 0.24
   C0107  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0114  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0121  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0130  290-0776-01   22uF    20%     10V                   UPW1E220MDD   22uF  25V 20% 105c 0.16
   C0152  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0185  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0218  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0221  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0307  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0325  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0335  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0512  290-0246-00    3.3uF  10%     15V  Axial Tantalum
   C0536  290-0246-00    3.3uF  10%     15V  Axial Tantalum
   C0723  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0731  290-0944-01  220uF    20%     10V                   UPW1C221MPD  220uF  16V 20% 105c 0.25
   C0732  290-0944-01  220uF    20%     10V                   UPW1C221MPD  220uF  16V 20% 105c 0.25
   C0733  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0738  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0740  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0957  290-0804-00   10uF   +50-20%  25V                   UPW1V100MDD   10uF  35V 20% 105c 0.16
   C0977  290-0246-00    3.3uF  10%     15V  Axial Tantalum

[FireDragon]

Thanks for the tantalum info FireFragon, I'm from the firmware side of things, so my EE skills cause me more problems then they solve.

My scope was working, so my goal was to replace all the electrolytic caps before they ruined the boards.  I did replace all the electrolytic caps with new electrolytic caps rather than attempting to spec tantalum replacements. From what I could gather, the tantalums require a bit more knowledge of their usage than I possess, so I stuck with the 'lytics.

I am also from the software side - including firmware. My EE skills definitely need improvement, but I am getting there.

I have seen a lot of statements that replacing a tantalum with an electrolytic requires around 10 times the capacitance. I think that is a hang over from the state of affairs ages ago. The reasoning seems to be that tantalums have a low ESR compared to electrolytics.

Things have changed! Tantalums are more reliable, but also most tantalums appear to have an ESR of 1-3 ohms. There are lower ESRs, but they are hard to find and apparently not common. So presumably not the subject of the older advice. Most of the electrolytics that I used for replacement have ESRs which are well below 1 ohm. So I don't see any need to replace electrolytics with tantalums. And if you REALLY need low ESR use ceramics which can be 0.01 ohms or even lower.

[FireDragon]

With this set of replacements, the scope passes the smoke test and all diagnostics. 

Replaced bridge, surge suppressors, one more resistor and one of the thermal resistors. Only the bridge was necessary.

A major milestone, but the saga isn't over yet. I am still waiting on the replacement for the other thermal resistor. That is a custom part. Also I am waiting for the LEMO style connectors (knock off, from China) so that I can add two probe power sockets. And calibration along with replacing the RAM with FRAM. At this point none are critical, so I will take those in the order they become possible. I still have to LEARN how to use the scope! But it looks like all of the features are present and working, but without more testing I could still find problems.

Replaced parts.

Diagnostic test result.

[Muxr]

Nice!

[Old-E]

Thanks tautech for the instructions on posting photos.   But I need more information!   I have the forum compatible photos ready to go.   Then - if I understand correctly -

1.  Place the curser where the photo is to be located on the text page and left click.   This pegs the location for the photo.
2.  Click on the "Insert Image" icon (located above on the far left side).   This displays the "IMG's" where the photo is to be located.
3.  Click on "Attachments and other options" (below).   This displays the "Attach:" path.
4.  Click on "Browse."   This displays the "Choose file to up load" page.
5.  Locate intended photo on the up load page, then click on the "Open" button (lower right).   This displays the photo file info in the "Attach:" path.

If this is correct so far, I got that.   But now what?   Clicking on "Preview" displays the message text, but no indication of a photo.   It appears that something is missing.   And if I click on "(Clear Attachments)," nothing happens.   I would like to get it right before posting on the blog, but there does not seem to be a way to do that!

Frustrated. 

[Old-E]

Well, some how one of the photos attached to the last message, but I don't know how.   I'll experiment more to see if I can make it happen again.

Frustrated.

[FireDragon]

I have been playing with my 2465BCT. So far everything that I know to try has worked. My only signal source is the calibrator signal, but it works on all four channels. The scope can do at least some measurements - but I need to learn more to really test that. I recreated everything I was doing when the scope failed and it all worked as expected.

I do appear to be missing a bolt. One of the two on the middle back that is about 3/4" long. I think that I probably used it somewhere internally because I have an extra short one. I got a replacement (Philips head, though) from my junk screw bowl and used that. If the original doesn't turn up, I will probably buy a replacement so that they all match.

There is one possible glitch - I am not sure the fan is working. After I had the scope on for some time (not sure, perhaps an hour or so), I noticed that the top was mildly warm to the touch. However, I was unable to feel any fan exhaust. Perhaps it wasn't hot enough to kick on or it was running very slowly.

But - perhaps the fan isn't working. I need to investigate that. It could be the fan, the soldered connections, the pass through capacitor or the fan control circuit. I'll look at that tomorrow, its about my bedtime now. I don't like the pass through capacitor. I shouldn't need to solder / unsolder the fan every time I want to do some maintenance. Also, it is much too close to the chassis. Just for safety, I put some Kapton tape on the chassis wall to prevent accidental contact with the fan's positive supply.

[MSO]

FireDragon:

Thanks for the part list; it'll make the ordering process a lot easier.  I'll order those caps for the main board and the two caps on the A9 board.  Once they're in hand, I'll be pushed into actually replacing them.

Congrats on bringing your scope back to life; it's nice to see it performing again. I've made note of the surge protectors for future reference.  I didn't test mine when I had the scope apart.  BTW, I also came up with one extra screw; it's a short one with no washer attached. I'll keep you posted if I ever find where it came from.

OLD-E, that looks like a nice heat sink. Mine has one of those grounding fingers mounted on the heat sink end, but it is formed such that it won't touch the bottom of the case.  I don't know who did it, but it's there.

Here's (I hope) a picture of my scope with a 500MHz signal @ 13dBm fed into channel 1 with the x10 mag enabled. The old scope is pretty quick for its age.

[MSO]

FireDragon,

 When I put my scope back together, I had mounted it backward so the support bracket was pushing on the fan axel preventing it from moving.  My 'lab' is kept cool, 60f or so, but the fan turns on immediately; at least it does after mounting it correctly.

I was suckered by the indents on the back side of the fan housing that matched up perfectly with the two little pegs incorporated in the fan mounting plate.  They fit together perfectly, they aligned the fan perfectly and there were no such indents on the front side of the fan. Hey, that's the way its got to go right? 

[FireDragon]

FireDragon,

 When I put my scope back together, I had mounted it backward so the support bracket was pushing on the fan axel preventing it from moving.  My 'lab' is kept cool, 60f or so, but the fan turns on immediately; at least it does after mounting it correctly.

I was suckered by the indents on the back side of the fan housing that matched up perfectly with the two little pegs incorporated in the fan mounting plate.  They fit together perfectly, they aligned the fan perfectly and there were no such indents on the front side of the fan. Hey, that's the way its got to go right?

I may have done that. I will have to check after breakfast and a grocery store trip.

Is the fan supposed to blow into the case or out of it? Normally, a fan will blow into the case.

[BravoV]

   C0102  290-0973-00  100uF    20%     25VDC                 UPW1V101MPD  100uF  35V 20% 105c 0.24
   C0107  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0114  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0121  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0130  290-0776-01   22uF    20%     10V                   UPW1E220MDD   22uF  25V 20% 105c 0.16
   C0152  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0185  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0218  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0221  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0307  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0325  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0335  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0512  290-0246-00    3.3uF  10%     15V  Axial Tantalum
   C0536  290-0246-00    3.3uF  10%     15V  Axial Tantalum
   C0723  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0731  290-0944-01  220uF    20%     10V                   UPW1C221MPD  220uF  16V 20% 105c 0.25
   C0732  290-0944-01  220uF    20%     10V                   UPW1C221MPD  220uF  16V 20% 105c 0.25
   C0733  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0738  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0740  290-0943-02   47uF    20%     25V                   UHV1V470MDD   47uF  35V 20% 105c 0.212
   C0957  290-0804-00   10uF   +50-20%  25V                   UPW1V100MDD   10uF  35V 20% 105c 0.16
   C0977  290-0246-00    3.3uF  10%     15V  Axial Tantalum

FireDragon  , thanks a lot for listing down the caps for the A1 board.

To recap this big A1 board is already in my to do list, its just I'm too scared, really appreciate the tips on removing it, noted in my to do list for my 2465B.

Just curious if there is any bad effects if I replace them with quality polymer cap like Oscon caps ? Apart from the price of course.

[BravoV]

....<snip>.... Hey, that's the way its got to go right?

Is the fan supposed to blow into the case or out of it? Normally, a fan will blow into the case.

The fan is blowing out of the case at the rear as I dissembled it for the 1st time.

I'm not the expert, but after looking at the vent holes like these, I believe their purpose are to have fresh cooler air get suck in and to blow straight at the hot parts, like the U800 and others.

Reposting my tear down photos to show what I mean.

Scope's bottom view, the U800 is located at the left top one.


Close up view on the vent holes, you can see the U800 is located straight below.

[MSO]

Is the fan supposed to blow into the case or out of it? Normally, a fan will blow into the case.

Straight from page 2-1 of the service manual:

    "INSTRUMENT COOLING" 
To prevent instrument damage from internally generated
heat, adequate air flow must be maintained. Before
turning on the power, verify that the spaces around the
air-intake holes on the bottom of the cabinet and the fan exhaust
holes in the rear panel are free of any obstruction
to airflow."

So the fan sucks in from the bottom and blows out the rear.  I wanted to make a joke here, but decided not to.

[Old-E]

Ok - - we'll see if I can make the picture attachment work in order to conclude the cooling issues from a few days ago.

This is a picture of the U800 IC that runs warm - hot.



This shows the new aluminum heat sink added.



The heat sink is designed to shed heat best when cooling air is being pulled in through the vent holes in the bottom of the case.   It's not as efficient when operating in the open like this in still air.   But, while the case was off the temperature was measured with a digital thermocouple (TC) meter anyway.   This was done by holding the TC in-between the fins of the heat sink with a tooth pick and a dab of thermal conducting paste.   Then it was moved to the stud at the opposite end of the IC.   The stabilized temps were -
120 deg. F on the heat sink.
114.4 deg. F on the threaded stud at the opposite end of the IC.
60.6 deg. F shop air temp.

A 55.4 deg. rise in temp seems like a lot, but this is with the case off.

Then at a later date the temp was measured again, in still air and with the old removed fan blowing, with the scope laying on its side as in the picture below.   The temps were -
122 deg. F on the heat sink in still air.   This is probably better for convection cooling than with the scope upside down.
84 deg. F on the heat sink in moving air.
65.1 deg. F shop air temp as seen on the digital meter.   In this photo the thermocouple can be seen suspended in mid air.

This illustrates how moving air helps in cooling the heat sink.



Unfortunately, I didn't measure the temp before installing the heat sink, or while in the case.   So all I can say is that the heat sink should help and I'm not too concerned with the temp of this Tek chip ether way.   But, if it was running inside a tent on the Sahara Desert in the summer, I might be concerned.