开云体育

Mark & Tom

Thank you for the navigation on this scope. I bought a 465
about ten years ago and it has been a gem despite its age.
It actually does everything I need, but a scope with 300 MHz
bandwidth is just to much to not repair.

I will post my findings and as closely as possible what solves
the problem. My big concern was the proprietary IC's might
be an issue but it sounds like the tantalum vexations struck
Tek also. Never imagined such an expensive cap would be
such a problem child.

Again thanks

Chuck

Hi John,

Did you find any difference in the polarization? Amazing repair though!

Aaron

Hi Miguel,
I just did the test you suggested.
Result here: /g/TekScopes/photo/281947/3538491
All other windings left open.
From calculation I obtain L = 8.3 mH. Definitely different from yours.

Max

Matt,
In terms of reading the values of the SRAM. It is critical that voltage stay always applied to the chip. Even a momentary loss of voltage for example shorting the power pins the data will be lost to the either and a recal will need to happen. This is what the battery is for to hold over the data when the scope is off.
For replacing the sram for a flash chip. This will be a larger challenge as the different types of memory are accessed written too and red from differently. There might not be the correct drive voltages present for the flash chip. Lots of data sheet diving will need to happen for a suitable replacement if any. One other issue here is timings. During boot up the chip will need to respond within the current time as well as with the correct speeds. So this will need to be taken in to account as well if looking for a replacement memory chip. Given all the hurtles the "normal" path walked is refresh the power source (battery) data can be preserved with a bench supply when switching out the battery. If caution is to the wind wiping the data and running the scope through a recal would not be a bad idea given its age. I find the calibration routine will bring any areas that need attention to light. But before going in to the deep end on the calibration make sure ALL sources are available to you the 2xxx scopes are picky when it comes to alignment/calibration tolerances especially in the high frequency compensation section. The edge speed needed here can get in to an "exotic" source depending on the speed of the scope.

Zen

Those that messaged me privately, I appreciate it. I am learning a lot as I go. I can't tell you how many times I've asked a question on an electronics forum and been eaten alive by folks because I didn't "google enough" or I'm doing something that they personally view as frivolous or unnecessary. Suffice it to say that I did not expect to get read the riot act when I've decided to undertake replacing worn-out parts that are now almost 40 years old, especially with some of the stories I've read about some parts in these units failing in rather spectacular fashion due to age, stress, or both.

I am a public school teacher, and I often get frustrated at how my students fear asking questions. This experience clarifies why that happens.

In case my point isn't clear, when someone asks a question about something, if your response is to immediately jump down their throat and tell them that what they're doing is stupid, then that's probably not going to have the effect you wanted. What it's going to do is make the person do one of a few things:
-- They are going to proceed with their plan (ill-advised or not) without hearing any advice that may have been embedded in the message
-- They are going to stop everything and abandon the project
-- They might take your advice, but they will stop asking further questions out of fear of being attacked again

I hope this clarifies my hesitance to post a further response.

Additionally, it would seem that the spreadsheet that's hosted here may either be out-of-date, incomplete, or (oddly enough) over-complete. I (now) know Menahem does recap sets for these, and have opted to use one of his kits to redo my scope. Most of the caps from DigiKey were returned (I did keep a couple of the smaller values for my parts bin).

I've prepared the system for its recapping and am just waiting on Menahem's cap kit to arrive.

The thing I'm now concerned about is retaining the calibration values, and potentially replacing the SRAM chip and battery combination with a FRAM chip instead.

Is it possible to get the calibration values out of the SRAM chip without powering the system on? I have a DIP chip test clip that would allow me to attach header pins to the SRAM in-circuit and then run those to my EPROM programmer, but I'm not sure that would do the job if the system isn't powered.

It's not a huge deal (if I have to do a re-cal on the thing after replacing the chip, oh well) but if it's possible to extract the cal values during servicing, that would be ideal.

Tom,

C861 is the one that will give trouble, 13V on a 15V tantalum. I keep 47mfd 25V ULD types in stock to replace the tantalums that are bad. Higher capacitance for others such as 100mfd can be higher in value if wanted. The decoupling off the 50V supply, I put in 15mfd in place of the ,68mfd. The original tantalum, ,68mfd, was already high in ESR. I prefer to use a 25V electrolytic to replace a 20V tantalum on a 15V supply.

Mark

Hello David,,,
I Tried the B-Trig reset with no effect unfortunately.

While probing the main board with a scope, I determined that the power on reset circuitry is working (at least the 'MAX708R' generate the reset correctly), but there are *no* signs of life from the processor.

Checking the A0 address line on various pins on the other chips, shows no activity whatsoever. I am now starting to suspect that the MPC860 processor is either dead or needs re-balling. Hmm, not the easiest of tasks to do.

Paul.

Hi Paul,

Not sure how much will help but take a look at (and download attachment):

And keep us informed about your findings, maybe a solution will show up.

TT

You can get plastic "razor blades" in either single or double sided types & several hardness grades quite cheaply on eBay if you need to scrape & are concerned about using a metal razor blade.

Ibought some & they fit straight into the holders that normal razor blades fit into.
The car detailers use them to remove decals from vehicles.
I also bought a few different types of razor blade holders off eBay at the time.

I got them to remove gummy adhesive off some engraved signs.

Regards,
Brian.

Sounds like I have the same issue, thanks for the info!
-Taylor Cotham

ISTR seeing a problem with a resistor somewhere in the U730/38/40 chain.
Check voltages and waveforms in the vicinity.

On Fri, 9 Dec 2022 at 01:44, Mark Vincent <orangeglowaudio@...> wrote:

Chuck,

Put some 91% IPA on the pins of the IC and tunnel diodes near the IC.
Then move the IC and diodes up and down to clean the pins and sockets. This
is a common problem, dirt.

Consider putting a tiny drop of Deoxit (possibly D100 but not necessarily)
on the end of a jeweller's screwdriver, and applying that directly to a
socket.

Also IPA any of the switches in the vicinity, talking the usual care with
the gold fingers.

There is a tantalum rated at 15V for a 13V supply on that board that will
suddenly short.

There are more than one :(

My notes indicate c861 c911 c965 c2101 (possibly c844 c902). Good luck
spotting them :(

Congrats and a good reason for beeing happy around the christmas tree the coming weeks. I have also problems with my 7904 and I think it is also the SMPS, so i have read this string with great interesting.
搁别苍é

Paul -

Have you tried holding the "B TRIG" button while powering on the scope?

Dave Casey

Sure enough the oscillators are both working.

So, I have a TDS3054, with good power, working oscillators but not booting......

Anyone have any suggestions of the next line of investigation?
Thanks

Chuck,

Put some 91% IPA on the pins of the IC and tunnel diodes near the IC. Then move the IC and diodes up and down to clean the pins and sockets. This is a common problem, dirt.

Change all the 0367 transistors! There are 11 in this scope. If you have more items with these in them, order more to replace the leaky ones. These are leaky, check as a diode from C to E, then replace them with KSP10BU. The pinout for the replacement transistor is BEC. This leakage of the transistors will cause odd problems. I have seen this enough times. There is a tantalum rated at 15V for a 13V supply on that board that will suddenly short. I pulled mine out and out in Nichicon ULD 47mfd 25V type. This tantalum is a known problem. If other switches have not been cleaned, clean them. Cleaning also goes for the other transistors and ICs.

Let us know what you find as the problem(s).

Mark

Gianni,

Glad to hear you found the problems and got it going. I like the military radio gear to the left of the scope.

Mark

This unit also has the disadvantage of being a non removeable probe head.

Zen

BACK TO LIFE!!! LONG LIFE THE 7844! Other two very intense day for the battle with the LV board. It was already OK, but I found the most complex way to understand it. Attaching the load to the LV, the click started again. I thought that some regulator had a fault, and so I got myself a current probe for the scope, not very precise (100mV/A) and I built an adapter cable to be able to check every wire. The probe told me nothing.
So, I had a different idea: use the differential amplifier to read the voltage across the protection resistors of the various regulators (2W resistors with very low values, like 0.1ohm). That also told me nothing: everything seemed OK.
Then, I decided to fully understand the -50V regulator and started studying it. It was OK but I had a strange phenomenon: on startup the voltage went to -30 and then in 20-30 seconds it arrived at the right value. Here started a long and complex battle. At first, I thought that the sense inputs were closed too near to the source and rewired the dummy load. No change. After, I discovered many things on the inner working of the regulator and more times I hoped I found the bug, but with no change. Eventually the solution: the problem disappears if I put a 10uF capacitor in the point of connection between the current output and the sense input (some self-oscillation, probably). Then I checked separately each voltage and all of them were correct. But all together… click again. Then, I discovered that that donkey I told you had made another one: my load has been built with two resistors in parallel, simply because the shop had not the right values. But the donkey connected only one of them, so the load was not charging enough the PSU. With the full load, the power supply was supposed to be OK, but instead I had a low +50 value and something producing heat. After incredible reflections, I noted that the donkey mounted one of the 10uF on the dummy load reversed. It was lightly swollen. Couldn't it be kind enough to burst, like all good capacitors do?
This evening I reconnected the PSU to the mainframe and tested it: it seemed OK. But it still lacked the HV, that I disabled to be freer during the work. So, I replaced the resistors and temporary remounted everything.
I was uncertain: test now and spend the night trying to fix more bugs, or wait tomorrow morning? Resisting temptations has never been my forte, so…. I pressed the switch: no clicks and after few seconds… THE TRACES! Unspeakable joy and go to bad happy. But before sleeping I wanted to update and thank you.
Note: at the end of the day, besides my silly investigations and the donkey, the real problems were two Zener diodes broken, as forecasted by Mark Vincent.
Photo:

Zentronic, if the probe was bought used eg off epay, you have no way to be sure it was never dropped.

We buy currently probes only in original Tektronix case, in like new condition

Any signs of wear, or abuse, eg frayed cables, blemished probe head, raised suspicion of cracked jaw ferrite.

Lots of them on epay.

The probes will typically have bad frequency response or inconsistent readings just by openi g and reclosing the jaws.

As the ferrite is extremely thin cross section and fragile like any ceramic, it's impossible to repair.

Send detailed photos of jaw open, showing both sides of the core, may be able to spot the hairline cracks.

Very sorry forgot this


Jon

It is not that It has "worked" I was able to zero the probe at 1 amp but it has never displayed a current amplitude of any deflection. Even 3 amps at DC. I was digging around in the unit yesterday and buzzed out the cable all was ok with the world there. Also the "balance" step in the calibration document can not be completed successfully. I did inspect the core and there is some areas that are questionable as well as some cracking of the plastics in the upper core. So unless I find a bad transistor in the differential amplifier. I am rapidly running out of options as to faults other than damage to the probe head.

Zen