Topics

Update on 2465


Joseph Tatum
 

I found what was making the display appear so strange. It was like a wide ribbon almost filling the whole screen. I found the wires from the vertical termination assembly to the deflection plates had somehow gotten bent before I got the unit and were touching. Now have alpha numeric readout an the message to press A/B button. Pressing the button brings a multiple trace. Also the start screen has message: "TEST O5 FAIL 40" Have not yet found what this indicates in the manual.

Also I am not familiar with the format of this Yahoo Groups. Do I need to respond here to comments or
respond to the emails that come to my email? I do not want to clutter the group.


Roger Evans
 

Search on this forum for 'Test 05 Fail 40', you will find this is a common problem on the 2465B which I am guessing is what you have. There are four SMD electrolytic capacitors on the A5 board which leak and damage other components. If you are lucky you only need to replace the electrolytics and a few SMD resistors, more extensive corrosion will damage the mounting pads, the tracks and the vias. I was lucky I only had one mounting pad corroded away.

There is a lot more information and photographs here and elsewhere on the web.

Roger


Joseph Tatum
 

Thanks Roger. I see no evidence of corrosion on the A5 control board.and I only see 3 electrolytic caps. They are all aluminum units. They may have been replaced. I may need to check the values and if one was not replaced. The three check good with ESR meter. I will do search on here and check he photos. I am still learning to navigate these Yahoo groups.


Siggi
 

Hey Joe,

Do you have a 2465, or a 2465B?

The two are VERY different, and you will get the wrong advice if you don't
tell us which, so please do.

The 2465B is 400MHz and typically phase surface mount components on the A5
board.

Siggi
On Sat, Dec 31, 2016 at 06:58 tatumj1@... [TekScopes] <
TekScopes@...> wrote:



Thanks Roger. I see no evidence of corrosion on the A5 control board.and I
only see 3 electrolytic caps. They are all aluminum units. They may have
been replaced. I may need to check the values and if one was not replaced.
The three check good with ESR meter. I will do search on here and check he
photos. I am still learning to navigate these Yahoo groups.



[Non-text portions of this message have been removed]



[Non-text portions of this message have been removed]


Joseph Tatum
 

It is a 2465. The 300 Mhz model. Passes all test except 05. Test 05 says fail 40. Trace has constant ragged disply. Read 57 hz in top right. Make me think I need to look for 60 hz ripple on the supply.


Siggi
 

On Sat, 31 Dec 2016 at 11:32 tatumj1@... [TekScopes] <
TekScopes@...> wrote:



It is a 2465. The 300 Mhz model. Passes all test except 05. Test 05 says
fail 40. Trace has constant ragged disply. Read 57 hz in top right. Make me
think I need to look for 60 hz ripple on the supply.
Hey Joe,

sounds like a plan to me. Until the power supplies are blameless, it
doesn't make sense to check much else. Note that you should be able to turn
the scope on its own supplies, if need be.

Does your 2465 have the counter-timer option? That's a very nice option to
have, IMHO.

The 05 tests run through a whole gamut of hardware. They use the DAC to set
trigger levels, which are then tested against the line trigger. It's not
described in the service manual what the precise tests do, but presumably
the trigger level values are limit tested against sensible values. This
validates the DAC current reference, as well as the analog multiplexers &
sample-and-hold op-amps, plus of course the trigger hybrid and the line
transformer.

This is typically the first self test that fails on the 2465B when the DAC
current reference is out of whack, which commonly happens when the SMD
capacitors leak all over the reference resistor network.

If the power supplies turn out to be OK for you, then the next thing up
would be to see whether the DACs current reference is OK, which is best
done by measuring the +1.36V and -1.24V references that are derived on the
A5 board. On yours they will be brought out to TP503/TP504, so should be
easy to check.

Siggi


Joseph Tatum
 

Hello Siggi,

I spent a little time on it today. I found all power supply voltages in spec and, my old 454A, see no ac ripple on either supply so the next check I will make will be what you suggest. The scope seems now to function correctly except that the trace is wildly distorted. I pulled the vertical amp hybrid and re-seated it but nothing was changed by that.

This scope was a gift from the widow of a ham friend. I found it packed away and, knowing my friend, it is unlikely he bought it not working. I am thinking he may have used it until it quit and just boxed it up. It would be so much better for my Hammond organ repair work than my old Tek 454A if I can fix it.


Siggi
 

On Sat, 31 Dec 2016 at 17:59 tatumj1@... [TekScopes] <
TekScopes@...> wrote:



Hello Siggi,

I spent a little time on it today. I found all power supply voltages in
spec and, my old 454A, see no ac ripple on either supply so the next check
I will make will be what you suggest. The scope seems now to function
correctly except that the trace is wildly distorted. I pulled the vertical
amp hybrid and re-seated it but nothing was changed by that.

This scope was a gift from the widow of a ham friend. I found it packed
away and, knowing my friend, it is unlikely he bought it not working. I am
thinking he may have used it until it quit and just boxed it up. It would
be so much better for my Hammond organ repair work than my old Tek 454A if
I can fix it.
Hey Joe,

the 2465 is all "drive-by-wire" in that most of the analog circuitry is
controlled by the microprocessor (MPU), by way of the control voltages
coming out of the DAC and sample-and-holds (S&H). Even the nice clunky
horizontal/vertical knobs go to gray code encoders, which the MPU reads to
set up the appropriate attenuation and preamp gain levels. The horizontal
timing is similarly controlled by the MPU, by selecting and adjusting the
current sources and timing capacitors.

So, in summary, when the DAC or S&H are wonked out, pretty much anything
can happen.

That being said, it might be helpful if you describe the mode in which your
trace is distorted. When you hook up the calibrator to the scope itself, at
say 1ms/DIV sweep speed, do you get anything like 5 cycles of the signal?
Do you get anything like 0.4V? Is the front edge nice and sharp? In DC
coupling, is the top edge nice and level?

Note that at high sweep speeds, the "square wave" from the calibrator isn't
very square at all, hence best to look at it at low speeds. Since the
calibrator frequency changes with the sweep speed, you can get an idea of
how the horizontal calibration of the scope stands by hooking it up and
running up and down the sweep speeds.

Also, while you're coming across as an old hand at this, I wonder if you've
seen this document <
http://www.sphere.bc.ca/test/tek-parts/troubleshooting-scopes.pdf>? This
pre-dates your 2465 by a decade or three, but the same principles still
hold for milking the front panel :).

Siggi


Joseph Tatum
 

Hey Siggi,

Sorry to be slow responding but my Mediacom Cable internet was down yesterday all day.

Thanks for the link, I download it. I measured the voltages at TP 503/504. Here are the results:

TP503 = +1.228 VDC, TP504 = -1.113 VDC. Both with the last digit changing rapidly. Both are low by about the same amount. Is this off enough to cause my problem? I have posted pics of the screen showing the trace.

It is like this with no Channel on. Same trace with Channel 1 and 2 or both selected. Connecting the calibrate signal to either Channel 1 or 2 with 1x probe does not give a distinguishable square wave. However, switching ether channel on and off causes a momentary single straight edge of the square wave to appear and instantly go away.

Connecting the same signal to Channels 3 and 4 gives a normal appearing square wave. Using this I have been trying to isolate the problem studying the 2465 manual which is gradually beginning to make more sense to me. Been a long time since I have worked on fixing a scope and never one this complex. Recently most of my electronic work has been on older tube type Hammod Organs and modern electronic keyboards.

I measured the ripple with a Tek 454A at J119, and found no ripple even close to exceeding specs. There is more noise than ripple. The 87 VDC supply does have a radio frequency component that measures 150mv one the scope.


Siggi
 

On Mon, 2 Jan 2017 at 06:46 tatumj1@... [TekScopes] <
TekScopes@...> wrote:



Hey Siggi,

Sorry to be slow responding but my Mediacom Cable internet was down
yesterday all day.

Thanks for the link, I download it. I measured the voltages at TP 503/504.
Here are the results:

TP503 = +1.228 VDC, TP504 = -1.113 VDC. Both with the last digit changing
rapidly. Both are low by about the same amount. Is this off enough to cause
my problem? I have posted pics of the screen showing the trace.

Wow, that's one messed-up trace :).
This is off enough to cause problems to the extent that the scope will not
be anywhere near calibration, but by itself it wouldn't cause the messed-up
trace.
Note, however, that these test points should be very clean DC. If not, then
it could be that either your S&H capacitors or op-amp are bad, or the 4051
MUX could be the problem.
So, put your second scope on TP503 and TP504, and see whether there's any
kind of ripple on them. I'm betting you'll find an RC exponential "droop"
on there.

The way this works is that there's only one DAC in the scope. This is used
to generate a variety of different control voltages, which are diverted to
their respective sample-and-hold OP amps through an analog MUX. I seem to
remember the service manual mentioning the MPU refreshes the control
voltages on a period of 16ms, and it wouldn't surprise me if this one were
refreshed 4 times in the cycle, leading to the 4ms ripple you're seeing.

The MUX in question here is U2335, and the sample-and-hold op-amp is U2427A.
If you probe pins 3 and 13 of U2335, you should be able to see on pin 3 how
the DAC's output varies in a repeating pattern, and how at one point it
matches the voltage on pin 13. If the voltage on pin 13 follows pin 3, then
you conclusively have a bad U2335.

Problems we have seen on these are leaky or plain bad 4051s, leaky hold
capacitors, and I suppose it's possible for the TL074 input FETs to develop
leakage. If there's "droop" there, you should be able to see on which side
of R2330 it is at least, which'll tell you in which direction to look.

It's also possible that there'd be problems with the control signals to the
MUX - a stuck bit on the SEL inputs might route multiple control voltages
to the wrong output. You'd be able to see this by probing across the MUX, I
figure.


lamontcranston17
 

I don't want to derail this thread, but I have a 2465. No suffix.
Should I replace the caps mentioned before they cause a problem, or is the
problem not that common?
Mikek


Joseph Tatum
 

Thanks Siggi, I will print out you message and do the checks later today. My wife had a birthday yesterday and we are going with some family to lunch in a little while. I will let you know the outcome. Did you see the pics?


Siggi
 

Hey Mike,

if you're referring to the electrolytic capacitor leakage on the A5 board,
then that's a problem that seems to only afflict the A5 boards that have
surface mount capacitors. I haven't heard of this happening to a 2445/2465
as they're all through-hole. As far as I know only the 2445B/2465B had SMD
from the factory, and it seems that was only the later serial numbers, as
we've had many reports of 2465Bs with through-hold A5.

Siggi

On Mon, 2 Jan 2017 at 09:03 nojunk@... [TekScopes] <
TekScopes@...> wrote:



I don't want to derail this thread, but I have a 2465. No suffix.
Should I replace the caps mentioned before they cause a problem, or is the
problem not that common?
Mikek


Siggi
 

Hey Joe,

Yes, I saw your pics in the album you created at <
https://groups.yahoo.com/neo/groups/TekScopes/photos/albums/180181048>.

Happy new year to you, birthday to your wife, and all the best to your
family. There's no rush on my behalf - your scope, your schedule, your fix
;). I'm just glad to see one of those aging beauties restored to duty again.

As an aside, my bet is that your horizontal is going to be messed up as
well as the vertical. This would go with AC on the DACs reference voltages.
If you have a moment when you dive into this again, try and hook up the
calibrator signal and see whether the trace isn't fuzzy in the horizontal
as well as messed up in the vertical.

Siggi

On Mon, 2 Jan 2017 at 09:23 tatumj1@... [TekScopes] <
TekScopes@...> wrote:



Thanks Siggi, I will print out you message and do the checks later today.
My wife had a birthday yesterday and we are going with some family to lunch
in a little while. I will let you know the outcome. Did you see the pics?


Tom Gardner
 

On 02/01/17 14:03, nojunk@... [TekScopes] wrote:

I don't want to derail this thread, but I have a 2465. No suffix.
Should I replace the caps mentioned before they cause a problem, or is the
problem not that common?
A common problem is the X2-spec mains filtering capacitors C1016/1018, and some other X/Y caps on the A2 and A3 boards. The early warning is the outer skin becoming crazed, which lets moisture in. My 2445B was OK, but after 30s they shorted in my 2465, vapourising the series resistor in the process.

See http://www.condoraudio.com/wp-content/uploads/Projects/Tektronix-2465B-Oscilloscope-Restoration-Repair.pdf for a description and prescription.


Joseph Tatum
 

Thanks Siggi, and a happy new year to you as well.

I have to leave in a few minutes. I did take time to probe the TPs and found about 40 mv of 60 Hz AC on each. I also looked at the MUX pins. Pin 13 has a steady DC level. Pin 3, strangely enough, kind of looks like the messy trace on the pics I posted.

By "droop" do you mean a difference in DC level at each end of R2330?

I am now thinking, even though the ESR meter checks it as OK; C2224 may have some problems.

Sometimes there seem to be changes. I mean that now only the trace now only ragged (as pictured) with no channel selected and when Channel 1 is selected. Channel 2 shows a straight trace

You are right. The horizontal seems affected as well. At slow sweep in all modes the trace looks like it is intensity modulated.


Siggi
 

On Mon, 2 Jan 2017 at 11:07 tatumj1@... [TekScopes] <
TekScopes@...> wrote:



Thanks Siggi, and a happy new year to you as well.

I have to leave in a few minutes. I did take time to probe the TPs and
found about 40 mv of 60 Hz AC on each.
I also looked at the MUX pins. Pin 13 has a steady DC level.
This is strange - it doesn't rhyme if all the supplies are good, no AC at
pin 13, and 60Hz AC out of the TPs.
However, AC at the TPs is sufficient to explain the wonky traces and timing
you see, though.
So, now to find where the AC is coming from.

Are you sure (for sure) it's 60Hz? The best way to be sure is to trigger
the measuring scope on LINE (my apologies if this is preaching to the
choir).
If the ripple is stationary against your LINE trigger, then it's AC ripple
for sure, if not, it's something else.

My bet is that the trouble will have a period of 3.3ms or roughly 300Hz
(see below).

Pin 3, strangely enough, kind of looks like the messy trace on the pics I
posted.
Pin 3 should be wobbling around with a period of 3.3ms - there's a picture
of what the various portions of the wobble refer to, and the ranges they
should be in, on page 9-12 in the service manual (my copy at least). When
you turn the various pots and knobs (VAR/DIV, Vertical POSITION, TRIGGER
LEVEL, HOLDOFF, TRACE SEP, etc.) you'll see portions of the waveform move
up and down.


By "droop" do you mean a difference in DC level at each end of R2330?
Yes, one end or the other of R2330 would be drooping (first), presumably
with a period of 3.3ms.

Note that pin 13 of the MUX "pokes" -1.24V at C2550 every once in a while,
by opening the relevant analog switch as the DAC is outputting -1.24V.
Looks like it does this every 3.3ms, from the description at page 3-9 of
the service manual (I'd remembered 16ms, which is wrong). The 3.3ms may fit
your weird 2ms/DIV traces, allowing for a fair amount of fudge on the
horizontal calibration.
Now, C2550, R2330 and C2530 make a filter network, whose purpose in life is
to buffer this occasional poke into a nice DC value to the very high
impedance JFET input of U2427A.
At the positive end of C2530, which is the positive input of U2427A, there
should be DC of -1.25V, and U2427A is then a buffer for the output of the
buffer network. If there's AC at the input (pin 3) of U2427A, then U2427A
is leaky.
If the input is DC, but 60Hz AC ripple on the output, then the ripple has
to be getting in through U2427A's supply. I've a hard time believing this
could be the case, as the PSSR of this class of op-amp should be 80dB+, so
you'd need monstrous supply ripple to produce 40mV output ripple. I'm wrong
about half the time, though, so please do check ;).

... time passes ...

The 5Vp rail is instrumental in providing the +1.36V rail. Look on the
collector of Q2322, to see whether the ripple is present there?
Maybe A3C1110/A3C1111, if that rail is wonky, though I wouldn't expect 60Hz
there.


I am now thinking, even though the ESR meter checks it as OK; C2224 may
have some problems.

Sometimes there seem to be changes. I mean that now only the trace now
only ragged (as pictured) with no channel selected and when Channel 1 is
selected. Channel 2 shows a straight trace
Interesting - this might also relate to e.g. the position of the traces.
IIRC, there are three analog voltages input to each preamp. One sets the
gain factor, one the DC offset, and the third one is related to the DC
offset of the preamp.

It wouldn't surprise me greatly if you could vary the position control so
that the DC offset is set to a value that "leveled" each of the CH1/2
traces where the trouble vanishes.


You are right. The horizontal seems affected as well. At slow sweep in all
modes the trace looks like it is intensity modulated.
OK, good info.
From a DC or GND input, it's generally hard to say whether this is AC
ripple, or some other timing effect.

Two things you can easily try:

1. Hook a probe to the calibrator signal, DC coupling, NORM trigger and
a suitable trigger level. See whether you now get "fuzzy" vertical edges.
Play with the holdoff to see whether you can make them non-fuzzy.
If you get fuzzy edges, and you can get them non-fuzzy with hold-off,
then the timing interference is a multiple of 60Hz. I'm betting on 3.3ms,
or ~303Hz, so I'm betting you can't entirely eliminate the fuzz with the
holdoff.
2. Set the scope to GND coupling and LINE trigger. See whether the weird
trace is now stationary. If not, it's not 60Hz ripple causing the problem,
but something else.
I'm still betting the same way - 303Hz interference :).



So, in summary.

If you have 40mV ripple on TP 503, 504, then let's find it's period and
then origin.


Joseph Tatum
 

Thanks Siggi.

I will recheck and verify focusing on the items you think are suspicious and I may have misread. I did trigger on the line and had the 20Mhz switch on and used a 10x probe. I will take a look t see if I may have had the sweep rate higher. I did it in a hurry as the family was about ready to go to the lunch for my wife's birthday.

It looks like 60Hz sine wave but I will check again plus follow your other suggestions. I feel sure I can narrow it down.

Actually, this is the first time I have tried to do such measurements in some time. I used to do it all the time when I worked at Scientific Atlanta 40 plus years ago. I am having to remember or relearn some stuff I have not used much for years. This old 75 year old brain does not always function like it used to. I will let you know how it goes.


 

Just dropping into this thread quickly:
- Re ESR measurement (C2224): Are you sure your ESR meter doesn't show a shorted cap as having "Low ESR" (which in fact would be "correct").
- Did anyone note that with no channel selected (photo 1), there shouldn't be a trace at all?

Raymond


Joseph Tatum
 

Hey Raymond, All help is appreciated, so you are welcome to the conversation.

Tthere are three 47 uf caps and theESR meter measures them all the same. As I recall, it reads a.45 ohm.

I did not expect to see a trace with no channel selected. I cannot find it now, but I thought I read a note somewhere that said that the Channel 1 trace would appear even if the Chanel 1 button was not depressed.