Topics

New member with a currently-dead 2465


Siggi
 

On Sun, Jul 12, 2020 at 6:21 PM Vincent Mallet <vmallet@...> wrote:

I removed U2556 and replaced it with a new one from TI that I had around
and ta-da! It lives! Super exciting!

Congrats!

I had a couple of questions on the oscillating circuitry:

1/ R2553 is a 10K tied to -5V. What does that do? Why do we need to bring
-5V into the picture?
For the same reason that there's a 680Ohm pullup to the +5 I suspect, e.g.
to bias the inverter just so, to make sure it oscillates. That pullup is
particularly interesting, as it's on the driven output of a gate. As you
found, the circuit seems to be on a knife edge.


2/ Figure 3831-50 “Test waveform setup information” for A5 shows 3
waveforms, including “1” taken at U2556-4 and “2” taken at U2468-5. Both
waveforms seem to be centered around 0V with a +/- 2V swing around 0. How
do I interpret this? Why negative? If I probe U2556-4 (now that it lives) I
get a waveform that looks like what’s in the manual but it swings from 0V
to ~3.6V. Is it because I am supposed to be AC-coupled and ignore the
offset?

That seems weird. There's no way you see a TTL signal symmetrical about
ground so I guess those oscillograms were recorded by the intern?


Now I need to find some probes for this scope so I can try to quantify its
performance. The only probes I have are the ones bundled with my Siglent
DSO ("PP215 200MHz") and I don't know how good or bad they are.
If you read through the "performance verification" portion of the service
manual you'll see no mention of using probes for the verification nor for
the calibration. To verify or calibrate this scope you input the set of
signals specified in the "performance verification" portion of the service
manual and then interpret the results on the CRT. These signals are
generally hooked up with good quality coax cables.

See e.g. this video (https://www.youtube.com/watch?v=oxJQr4dKnys) and this
one (https://www.youtube.com/watch?v=OiAmER1OJh4) for some of the whats and
hows of probes.
If you can compensate your Siglent probes with the 2465, they'll probably
work OK to the limits of their frequency spec. If you can't compensate
them, you'll need probes with a suitable compensation range, but depending
on what you expect to look at, they don't need to match the scope's
bandwidth.


Vincent Mallet
 

It lives!!

I'm definitely in here for the journey but it's nonetheless interesting to
see how someone with experience would zero in on the potential culprit and
how they'd move forward with the investigation.

Here I put things back together, removed all the probes except TP505 (far
enough from the circuit not to influence it) and tried a few power on/off
cycles to try to catch the oscillator napping... It was mostly hyper (i.e.
30MHz at pin-1, 3.75MHz at TP505). Caught it napping once, went to measure
voltage drop over R2559 with a DMM and that was enough to wake it up!

Gave up on that plan; I lifted a leg of each of the five resistors in there
to measure them out of circuit; they look impeccable:
R2549: 685R (specs: 680R 5%)
R2564: 330R (specs: 330R 5%)
R2571: 986R (specs: 1K 5%)
R2573: 997R (specs: 1K 5%)
R2553: 10K01 (specs: 10K 5%)

I removed U2556 and replaced it with a new one from TI that I had around
and ta-da! It lives! Super exciting!

I managed to remove U2556 without turning it into a sacrifice so I will try
to exercise it offline to see how it behaves.

I had a couple of questions on the oscillating circuitry:

1/ R2553 is a 10K tied to -5V. What does that do? Why do we need to bring
-5V into the picture?

2/ Figure 3831-50 “Test waveform setup information” for A5 shows 3
waveforms, including “1” taken at U2556-4 and “2” taken at U2468-5. Both
waveforms seem to be centered around 0V with a +/- 2V swing around 0. How
do I interpret this? Why negative? If I probe U2556-4 (now that it lives) I
get a waveform that looks like what’s in the manual but it swings from 0V
to ~3.6V. Is it because I am supposed to be AC-coupled and ignore the
offset?

Now I need to find some probes for this scope so I can try to quantify its
performance. The only probes I have are the ones bundled with my Siglent
DSO ("PP215 200MHz") and I don't know how good or bad they are.

Thanks!

Vince.

On Sat, Jul 11, 2020 at 8:21 AM Siggi <siggi@...> wrote:

On Fri, Jul 10, 2020 at 8:02 PM Tom Miller <@tmiller>
wrote:

Just replace U2556. It is a cheap part. Cut the leads close to the chip
package, remove the package, then unsolder and remove each pin one at a
time. Clear the holes and put in a new 74LS04. I wager a cheeseburger
that is the problem.
In an over-the-counter repair shop, this would be sage advice, and I'm not
sure I'd take you up on that bet.
I believe Vincent is, however, more interested in the journey than the
destination, and by measuring and thinking through all the things, you
learn more :).




Siggi
 

On Fri, Jul 10, 2020 at 8:02 PM Tom Miller <@tmiller> wrote:

Just replace U2556. It is a cheap part. Cut the leads close to the chip
package, remove the package, then unsolder and remove each pin one at a
time. Clear the holes and put in a new 74LS04. I wager a cheeseburger
that is the problem.
In an over-the-counter repair shop, this would be sage advice, and I'm not
sure I'd take you up on that bet.
I believe Vincent is, however, more interested in the journey than the
destination, and by measuring and thinking through all the things, you
learn more :).


Siggi
 

On Fri, Jul 10, 2020 at 7:33 PM Vincent Mallet <vmallet@...> wrote:

Tested +5V and GND pins of U2556 and they are as expected.
I tested the 3-4 inverter by bringing pin 3 to +5V or GND (like you
mentioned, although using a 22R shunt briefly instead as this is what I had
laying around) and hmm I didn't see much reaction on its output 4.
That's not too surprising. The only inverter input where it is reasonable
to do is the 1-2 inverter. The input there is weakly biased, rather than
being positively driven from another logic output.


Surprising things happened next when I tested the 1-2 inverter. I started
by putting my scope probe on pin 1 to make sure my 22R shunt was having the
desired effect on the input. The input was sitting at about 1.05V. I
briefly applied +5V to pin 1 (via 22R) and bam! I started seeing a sine on
pin 1. The sine had a 840mV pk-pk amplitude sitting right above a 1V
offset, 10Mhz. I touched pin 1 again and a 2.24V pk-pk sine appeared (again
sitting right above 1V), and I saw something appear on the 2465 display!
The scope had booted up! Things vanished shortly after.

I redid the experiment after putting another probe on TP505 U2092-37 and I
saw different results. Once I got a nice 2.24V pk-pk sine on pin 1 which
gave me a 1.25Mhz (almost) square on TP505 which is matching the expected
value from the service manual. It lasted 5s, then the pin1 sine shrunk to
~840mV pk-pk again. Touched it again and I got a 400mV pk-pk (offset 1V)
but this one was ~30Mhz! The resulting TP505 more-or-less square wave was
at 3.75Mhz, three times the expected frequency. Touched it again and pin 1
sine became a not-so-sine wave, 10Mhz, 560mV pk-pk.

Sounds like there's something just marginal in the oscillator. If you catch
the oscillator napping, you can infer the values of the biasing resistors
(R2571/R2573/R2553) by the voltage drop across them. You also want to
validate that the output of the 1-2 inverter is making the right output
voltages. Low is below 0.8V and high is above 2.0V, if the output is in
between there, you're nowhere. Same for the inputs. You measured 1.05V,
which is in no-man's land, so now you need to figure out why. Either the
biasing resistors have drifted (this happens with carbon comp resistors in
particular) or your 1-2 inverter is marginal or bad, or there's leakage
somewhere.


I pulled the legs of C2565, C2566, C2572 to measure them with an LCR
meter.
C2565: 84.5pF (specs: 82pF, 5%)
C2566: 33.3pF (specs: 33pF 5%)
C2572: 100.1nF (specs: 100nF 20%)

I do not know how to measure leakage current yet.
If you catch the circuit in a non-oscillating state, you can measure the
voltage drops across the respective resistors. From Ohm's law and the
resistor nominal values, you should then be able to guestimate the current
going through them. If things don't add up, you either have resistors that
are out of spec, or you have leakage through the capacitors - measuring the
resistors will tell you which. You can probably get quite near the truth
by measuring the resistors in-circuit.


I should have access to a better desoldering tool soon and I'll pull the
resistors and U2556 so I can measure / test them out of circuit.

Good luck!
Siggi


 

Just replace U2556. It is a cheap part. Cut the leads close to the chip package, remove the package, then unsolder and remove each pin one at a time. Clear the holes and put in a new 74LS04. I wager a cheeseburger that is the problem.

Regards

On 7/10/2020 7:32 PM, Vincent Mallet wrote:
Tested +5V and GND pins of U2556 and they are as expected.
I tested the 3-4 inverter by bringing pin 3 to +5V or GND (like you
mentioned, although using a 22R shunt briefly instead as this is what I had
laying around) and hmm I didn't see much reaction on its output 4.
For kicks I tested the 9-8 inverter (not used according to the schematics)
and this one responded as expected.

Surprising things happened next when I tested the 1-2 inverter. I started
by putting my scope probe on pin 1 to make sure my 22R shunt was having the
desired effect on the input. The input was sitting at about 1.05V. I
briefly applied +5V to pin 1 (via 22R) and bam! I started seeing a sine on
pin 1. The sine had a 840mV pk-pk amplitude sitting right above a 1V
offset, 10Mhz. I touched pin 1 again and a 2.24V pk-pk sine appeared (again
sitting right above 1V), and I saw something appear on the 2465 display!
The scope had booted up! Things vanished shortly after.

I redid the experiment after putting another probe on TP505 U2092-37 and I
saw different results. Once I got a nice 2.24V pk-pk sine on pin 1 which
gave me a 1.25Mhz (almost) square on TP505 which is matching the expected
value from the service manual. It lasted 5s, then the pin1 sine shrunk to
~840mV pk-pk again. Touched it again and I got a 400mV pk-pk (offset 1V)
but this one was ~30Mhz! The resulting TP505 more-or-less square wave was
at 3.75Mhz, three times the expected frequency. Touched it again and pin 1
sine became a not-so-sine wave, 10Mhz, 560mV pk-pk.

I pulled the legs of C2565, C2566, C2572 to measure them with an LCR
meter.
C2565: 84.5pF (specs: 82pF, 5%)
C2566: 33.3pF (specs: 33pF 5%)
C2572: 100.1nF (specs: 100nF 20%)

I do not know how to measure leakage current yet.

I should have access to a better desoldering tool soon and I'll pull the
resistors and U2556 so I can measure / test them out of circuit.

Thanks,

Vince.





On Fri, Jul 10, 2020 at 12:29 PM Siggi <siggi@...> wrote:

On Fri, Jul 10, 2020 at 12:31 PM Vincent Mallet <vmallet@...> wrote:

Thanks Siggi. The scope is a 2465 300Mhz, S/N B027189. I don't see any
smd
components on the A5 board.
Yeah, only (some of) the -B version has SMD on the A5 board, as I
understand.


I don't think the scope has any options but I'm
not too sure how to tell yet.

I looked at the joints of the oscillating circuit components under a
microscope and they looked fine (to my untrained eyes); I tested
conductivity between leads around this area and it matched expectations.
I
wiggled things a bit and flexed the board very gently and powered things
back on, still a flat line there.
Well then, I guess you need to dig in, see what's up. In your shoes I'd
start by verifying that U2556 is getting 5V and ground.
You should then be able to check whether the U2556A inverter is driving
its output by measuring the voltages on pin 1 & 2. It should be safe to
test the inverter by shunting its input to +5V and GND through - say - a
100Ohm resistor to see whether it can drive both up and down.
Maybe you want to test R2571 & R2573 as well as C2572. There have been
cases where these 0.1uF capacitors have gone leaky, and that might be
enough to bring down this oscillator.

Good luck,
Siggi



--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


Vincent Mallet
 

Tested +5V and GND pins of U2556 and they are as expected.
I tested the 3-4 inverter by bringing pin 3 to +5V or GND (like you
mentioned, although using a 22R shunt briefly instead as this is what I had
laying around) and hmm I didn't see much reaction on its output 4.
For kicks I tested the 9-8 inverter (not used according to the schematics)
and this one responded as expected.

Surprising things happened next when I tested the 1-2 inverter. I started
by putting my scope probe on pin 1 to make sure my 22R shunt was having the
desired effect on the input. The input was sitting at about 1.05V. I
briefly applied +5V to pin 1 (via 22R) and bam! I started seeing a sine on
pin 1. The sine had a 840mV pk-pk amplitude sitting right above a 1V
offset, 10Mhz. I touched pin 1 again and a 2.24V pk-pk sine appeared (again
sitting right above 1V), and I saw something appear on the 2465 display!
The scope had booted up! Things vanished shortly after.

I redid the experiment after putting another probe on TP505 U2092-37 and I
saw different results. Once I got a nice 2.24V pk-pk sine on pin 1 which
gave me a 1.25Mhz (almost) square on TP505 which is matching the expected
value from the service manual. It lasted 5s, then the pin1 sine shrunk to
~840mV pk-pk again. Touched it again and I got a 400mV pk-pk (offset 1V)
but this one was ~30Mhz! The resulting TP505 more-or-less square wave was
at 3.75Mhz, three times the expected frequency. Touched it again and pin 1
sine became a not-so-sine wave, 10Mhz, 560mV pk-pk.

I pulled the legs of C2565, C2566, C2572 to measure them with an LCR
meter.
C2565: 84.5pF (specs: 82pF, 5%)
C2566: 33.3pF (specs: 33pF 5%)
C2572: 100.1nF (specs: 100nF 20%)

I do not know how to measure leakage current yet.

I should have access to a better desoldering tool soon and I'll pull the
resistors and U2556 so I can measure / test them out of circuit.

Thanks,

Vince.

On Fri, Jul 10, 2020 at 12:29 PM Siggi <siggi@...> wrote:

On Fri, Jul 10, 2020 at 12:31 PM Vincent Mallet <vmallet@...> wrote:

Thanks Siggi. The scope is a 2465 300Mhz, S/N B027189. I don't see any
smd
components on the A5 board.

Yeah, only (some of) the -B version has SMD on the A5 board, as I
understand.


I don't think the scope has any options but I'm
not too sure how to tell yet.

I looked at the joints of the oscillating circuit components under a
microscope and they looked fine (to my untrained eyes); I tested
conductivity between leads around this area and it matched expectations.
I
wiggled things a bit and flexed the board very gently and powered things
back on, still a flat line there.
Well then, I guess you need to dig in, see what's up. In your shoes I'd
start by verifying that U2556 is getting 5V and ground.
You should then be able to check whether the U2556A inverter is driving
its output by measuring the voltages on pin 1 & 2. It should be safe to
test the inverter by shunting its input to +5V and GND through - say - a
100Ohm resistor to see whether it can drive both up and down.
Maybe you want to test R2571 & R2573 as well as C2572. There have been
cases where these 0.1uF capacitors have gone leaky, and that might be
enough to bring down this oscillator.

Good luck,
Siggi




Siggi
 

On Fri, Jul 10, 2020 at 12:31 PM Vincent Mallet <vmallet@...> wrote:

Thanks Siggi. The scope is a 2465 300Mhz, S/N B027189. I don't see any smd
components on the A5 board.

Yeah, only (some of) the -B version has SMD on the A5 board, as I
understand.


I don't think the scope has any options but I'm
not too sure how to tell yet.

I looked at the joints of the oscillating circuit components under a
microscope and they looked fine (to my untrained eyes); I tested
conductivity between leads around this area and it matched expectations. I
wiggled things a bit and flexed the board very gently and powered things
back on, still a flat line there.
Well then, I guess you need to dig in, see what's up. In your shoes I'd
start by verifying that U2556 is getting 5V and ground.
You should then be able to check whether the U2556A inverter is driving
its output by measuring the voltages on pin 1 & 2. It should be safe to
test the inverter by shunting its input to +5V and GND through - say - a
100Ohm resistor to see whether it can drive both up and down.
Maybe you want to test R2571 & R2573 as well as C2572. There have been
cases where these 0.1uF capacitors have gone leaky, and that might be
enough to bring down this oscillator.

Good luck,
Siggi


Vincent Mallet
 

Thanks Siggi. The scope is a 2465 300Mhz, S/N B027189. I don't see any smd
components on the A5 board. I don't think the scope has any options but I'm
not too sure how to tell yet.

I looked at the joints of the oscillating circuit components under a
microscope and they looked fine (to my untrained eyes); I tested
conductivity between leads around this area and it matched expectations. I
wiggled things a bit and flexed the board very gently and powered things
back on, still a flat line there.

Thanks,

Vince.

On Fri, Jul 10, 2020 at 8:01 AM Siggi <siggi@...> wrote:

Hey Vincent,

congrats on your 2465, I'm sure you'll get'er going again in no time.
Just to make sure though, is this a 2465 or a 2465A/B? There are important
differences between the three - notably the A5 board in the 2465B suffers
from surface mount capacitor leakage.

Inline.

Siggi

On Thu, Jul 9, 2020 at 9:05 PM Vincent Mallet <vmallet@...> wrote:

Question: how likely is it that the oscillator circuit has failed?

It happens.


Or could
there be an external factor preventing it from oscillating that I don't
see?
This can probably happen also :).
You may want to try and gently flex the board and maybe wiggle the
components involved. IIRC the crystal stands fairly proud, and so may
suffer vibration or shock damage. Best case you have a poor solder joint
somewhere.



My next step would be to pull the few components from the oscillating
circuit out so I can test them but I figured maybe I'd ask around if this
rang a bell first.
I don't remember oscillator death being a common issue.




Siggi
 

Hey Vincent,

congrats on your 2465, I'm sure you'll get'er going again in no time.
Just to make sure though, is this a 2465 or a 2465A/B? There are important
differences between the three - notably the A5 board in the 2465B suffers
from surface mount capacitor leakage.

Inline.

Siggi

On Thu, Jul 9, 2020 at 9:05 PM Vincent Mallet <vmallet@...> wrote:

Question: how likely is it that the oscillator circuit has failed?

It happens.


Or could
there be an external factor preventing it from oscillating that I don't
see?
This can probably happen also :).
You may want to try and gently flex the board and maybe wiggle the
components involved. IIRC the crystal stands fairly proud, and so may
suffer vibration or shock damage. Best case you have a poor solder joint
somewhere.



My next step would be to pull the few components from the oscillating
circuit out so I can test them but I figured maybe I'd ask around if this
rang a bell first.
I don't remember oscillator death being a common issue.


Vincent Mallet
 

Hi group,

I just joined TekScopes after hunting for info on how to go about
resurrecting a dead 2465 I acquired recently. I picked up electronics as a
hobby a little over a year ago and it has been a learning experience for
sure, and my new challenge is to see if I can do something about this
'scope :)

The 2465 powers on and a lot of lights light up and stay lit on the front
panel; the display stays absolutely blank (no hint of a beam anywhere no
matter the position of the intensity settings). No blinking TRIG'D light,
pressing A/B does nothing (actually, pressing any button doesn't change
anything).

I started getting familiar with the service manual and going through the
basic troubleshooting steps. The power supply measurements appear to be
within specs as per table 5-1; ripples _seem_ to be within spec but I'm not
sure I'm measuring them right with my DSO.

I moved on the troubleshooting steps until I reached the clock signal
coming to U2092-37 (TP505) where I get a flatline rather than a 5MHz
signal. I looked at the oscillator circuit on A5 and the test points at the
output of U2556B and U2468A are both flat lines too (instead of a 10MHz
square and 5MHz square). I do get a -5V on one side of R2553 and +5V on
R2549.

Question: how likely is it that the oscillator circuit has failed? Or could
there be an external factor preventing it from oscillating that I don't see?

My next step would be to pull the few components from the oscillating
circuit out so I can test them but I figured maybe I'd ask around if this
rang a bell first.

Thanks!

Vince.