Topics

TEK 7704A pre-death behaviour a clue to the fault?


Simon Owen
 

Hi all
I am embarking on the likely arduous journey to resurrect my 7704A from
tick mode.

It was actually in tick mode when I bought it so I got a good deal (I
hope!). I got it home and pulled some plugins out and gave them a good
shove home and presto, it hummed to life.

I got a couple of months of tinkering with it but alas back in tick mode.

I have put together a test load and I'll be measuring voltages and ripple
shortly.

It just occurred to me that before its most recent death there was
potentially a warning sign: the scope would tick a few times before coming
to life. Then I remember it would tick for longer and longer, but I found
that if I pushed the beam finder button it would 'snap out of it' and get
down to business. This clearly didn't last.

Does anyone have any wisdom on this symptom so perhaps I will have an idea
of the likely problem? When it was running during its 'pre-death' period I
didn't notice anything else peculiar.

Assuming I can rule out the power supply as the cause of the trouble, I
wouldn't know where to begin!

Thanks a million
Simon Owen


Eric
 

First thing I would suspect is a leaky electrolytic or a tant that is
shorted. I am leaning more to the electrolytic though. Seems like a "warm
up" with a cap charging in time. Sounds like one of the rails is coming up
slowly

Eric

On Wed, Oct 7, 2020, 6:50 PM Simon Owen <Sowenbd@...> wrote:

Hi all
I am embarking on the likely arduous journey to resurrect my 7704A from
tick mode.

It was actually in tick mode when I bought it so I got a good deal (I
hope!). I got it home and pulled some plugins out and gave them a good
shove home and presto, it hummed to life.

I got a couple of months of tinkering with it but alas back in tick mode.

I have put together a test load and I'll be measuring voltages and ripple
shortly.

It just occurred to me that before its most recent death there was
potentially a warning sign: the scope would tick a few times before coming
to life. Then I remember it would tick for longer and longer, but I found
that if I pushed the beam finder button it would 'snap out of it' and get
down to business. This clearly didn't last.

Does anyone have any wisdom on this symptom so perhaps I will have an idea
of the likely problem? When it was running during its 'pre-death' period I
didn't notice anything else peculiar.

Assuming I can rule out the power supply as the cause of the trouble, I
wouldn't know where to begin!

Thanks a million
Simon Owen






Simon Owen
 

Thanks Eric
I forgot to mention I've replaced all the tantalum and electrolytic caps on
the rect/filter board.

Now I recall a tantalum cap on the board on top of the CRT burned out
shortly after its first resurrection so I did a wholesale replacement of
tantalum caps throughout.

What you're saying makes sense to me though. Are you thinking it could be
one or both of the huge beercan caps?

Thanks,
Simon

On Thu., 8 Oct. 2020, 09:56 Eric, <ericsp@...> wrote:

First thing I would suspect is a leaky electrolytic or a tant that is
shorted. I am leaning more to the electrolytic though. Seems like a "warm
up" with a cap charging in time. Sounds like one of the rails is coming up
slowly

Eric

On Wed, Oct 7, 2020, 6:50 PM Simon Owen <Sowenbd@...> wrote:

Hi all
I am embarking on the likely arduous journey to resurrect my 7704A from
tick mode.

It was actually in tick mode when I bought it so I got a good deal (I
hope!). I got it home and pulled some plugins out and gave them a good
shove home and presto, it hummed to life.

I got a couple of months of tinkering with it but alas back in tick mode.

I have put together a test load and I'll be measuring voltages and ripple
shortly.

It just occurred to me that before its most recent death there was
potentially a warning sign: the scope would tick a few times before
coming
to life. Then I remember it would tick for longer and longer, but I found
that if I pushed the beam finder button it would 'snap out of it' and get
down to business. This clearly didn't last.

Does anyone have any wisdom on this symptom so perhaps I will have an
idea
of the likely problem? When it was running during its 'pre-death' period
I
didn't notice anything else peculiar.

Assuming I can rule out the power supply as the cause of the trouble, I
wouldn't know where to begin!

Thanks a million
Simon Owen










 

I concur with Eric that most likely, it's a matter of one or more bad caps.
With the 7704A, the display unit may simply be separated from the base unit, at the "Main Interconnect". You just have to loosen the screws holding down the rails linking the top and bottom sections, to the left and right of the mainframe.
After a quick look at the schematics, I can't find a reason why running the base unit powered on with the display unit disconnected, could do any harm.
Checking the base unit under power on its own not only may allow to narrow down the problem area but measuring resistances across the power connections of the display unit may indicate the problem area(s).

Raymond


 

On Thu, Oct 8, 2020 at 01:13 AM, Simon Owen wrote:


What you're saying makes sense to me though. Are you thinking it could be
one or both of the huge beercan caps?
No, the beercan caps sit straight across the mains input, before the oscillator causing the ticking, so no tick mode. They fail not often but if they do, they usually cause the main fuse to blow and, with some bad luck, damage the input rectifier,

Raymond


Simon Owen
 

Thanks a million, great advice both of you. I shall proceed as you suggest
and report back!

On Thu., 8 Oct. 2020, 10:21 Raymond Domp Frank, <@Raymond> wrote:

I concur with Eric that most likely, it's a matter of one or more bad caps.
With the 7704A, the display unit may simply be separated from the base
unit, at the "Main Interconnect". You just have to loosen the screws
holding down the rails linking the top and bottom sections, to the left and
right of the mainframe.
After a quick look at the schematics, I can't find a reason why running
the base unit powered on with the display unit disconnected, could do any
harm.
Checking the base unit under power on its own not only may allow to narrow
down the problem area but measuring resistances across the power
connections of the display unit may indicate the problem area(s).

Raymond






Simon Owen
 

Update: I just measured the +50V supply as being within tolerance but with
1.8 full volts of ripple.

Further evidence for the case against an electrolytic capacitor?

On Thu., 8 Oct. 2020, 10:26 Simon Owen via groups.io, <Sowenbd=
gmail.com@groups.io> wrote:

Thanks a million, great advice both of you. I shall proceed as you suggest
and report back!

On Thu., 8 Oct. 2020, 10:21 Raymond Domp Frank, <@Raymond>
wrote:

I concur with Eric that most likely, it's a matter of one or more bad
caps.
With the 7704A, the display unit may simply be separated from the base
unit, at the "Main Interconnect". You just have to loosen the screws
holding down the rails linking the top and bottom sections, to the left
and
right of the mainframe.
After a quick look at the schematics, I can't find a reason why running
the base unit powered on with the display unit disconnected, could do any
harm.
Checking the base unit under power on its own not only may allow to
narrow
down the problem area but measuring resistances across the power
connections of the display unit may indicate the problem area(s).

Raymond










Simon Owen
 

Meant to say the AVERAGE voltage is within tolerance at 50.1V but with the
measured ripple the peaks fall outside both bounds of tolerance (49.2V to
51.0V)

On Thu., 8 Oct. 2020, 10:37 Simon Owen, <sowenbd@...> wrote:

Update: I just measured the +50V supply as being within tolerance but with
1.8 full volts of ripple.

Further evidence for the case against an electrolytic capacitor?

On Thu., 8 Oct. 2020, 10:26 Simon Owen via groups.io, <Sowenbd=
gmail.com@groups.io> wrote:

Thanks a million, great advice both of you. I shall proceed as you suggest
and report back!

On Thu., 8 Oct. 2020, 10:21 Raymond Domp Frank, <@Raymond>
wrote:

I concur with Eric that most likely, it's a matter of one or more bad
caps.
With the 7704A, the display unit may simply be separated from the base
unit, at the "Main Interconnect". You just have to loosen the screws
holding down the rails linking the top and bottom sections, to the left
and
right of the mainframe.
After a quick look at the schematics, I can't find a reason why running
the base unit powered on with the display unit disconnected, could do
any
harm.
Checking the base unit under power on its own not only may allow to
narrow
down the problem area but measuring resistances across the power
connections of the display unit may indicate the problem area(s).

Raymond










 

On Thu, Oct 8, 2020 at 01:21 AM, Raymond Domp Frank wrote:


No, the beercan caps sit straight across the mains input, before the
oscillator causing the ticking, so no tick mode. They fail not often but if
they do, they usually cause the main fuse to blow and, with some bad luck,
damage the input rectifier,
There's a failure mode of the big caps that may cause tick mode. I didn't realize earlier because I'm in a 230 VAC country and you're probably in a 115 VAC area.
The big caps usually fail shorted where I live but in your area, before becoming shorted they may become (almost) open because of loss of capacity or detached internal links.
It should be very easy for you to check that the Vpeak of your mains supply should be across the caps; something like 160 VDC if you're in a 115 VAC country. With low-capacity cans, a pulsating DC voltage may drive the inverter. That may possibly cause ticking. Far from sure, just keeping it in mind.

*** Be aware that the big cans are directly connected to the mains, so use an isolation transformer if you can! ***

Raymond


 

On Thu, Oct 8, 2020 at 01:37 AM, Simon Owen wrote:


Update: I just measured the +50V supply as being within tolerance but with
1.8 full volts of ripple.

Further evidence for the case against an electrolytic capacitor?
You're not saying 60 or 120 Hz ripple, are you? If not, what approximate frequency?

Raymond


Simon Owen
 

Thanks Raymond,
240V 50Hz give or take (Sydney AU)

On Thu., 8 Oct. 2020, 10:42 Raymond Domp Frank, <@Raymond> wrote:

On Thu, Oct 8, 2020 at 01:21 AM, Raymond Domp Frank wrote:


No, the beercan caps sit straight across the mains input, before the
oscillator causing the ticking, so no tick mode. They fail not often but
if
they do, they usually cause the main fuse to blow and, with some bad
luck,
damage the input rectifier,
There's a failure mode of the big caps that may cause tick mode. I didn't
realize earlier because I'm in a 230 VAC country and you're probably in a
115 VAC area.
The big caps usually fail shorted where I live but in your area, before
becoming shorted they may become (almost) open because of loss of capacity
or detached internal links.
It should be very easy for you to check that the Vpeak of your mains
supply should be across the caps; something like 160 VDC if you're in a 115
VAC country. With low-capacity cans, a pulsating DC voltage may drive the
inverter. That may possibly cause ticking. Far from sure, just keeping it
in mind.

*** Be aware that the big cans are directly connected to the mains, so use
an isolation transformer if you can! ***

Raymond






Simon Owen
 

I measure 1.8V peak-to-peak at approximately 7MHz. Yes that's Megahertz!

On Thu., 8 Oct. 2020, 10:44 Raymond Domp Frank, <@Raymond> wrote:

On Thu, Oct 8, 2020 at 01:37 AM, Simon Owen wrote:


Update: I just measured the +50V supply as being within tolerance but
with
1.8 full volts of ripple.

Further evidence for the case against an electrolytic capacitor?
You're not saying 60 or 120 Hz ripple, are you? If not, what approximate
frequency?

Raymond






 

On Thu, Oct 8, 2020 at 01:45 AM, Simon Owen wrote:


Thanks Raymond,
240V 50Hz give or take (Sydney AU)
Aah, 240 VAC even!
Do you have a Service Manual (SM)?
If you do, it should be clear that no significant 50 Hz (i.e. mains) ripple should (can) come out of the low voltage power supply (LVPS), if the full-wave bridge rectifier works ok, because it produces 100 Hz!
Do you also see 50 Hz ripple on the other outputs?
If 50 Hz is indeed coming out of the LVPS, my idea of bad capacity value of one or both big cans becomes closer or, even much more likely, the bridge rectifier could be partly defective, because otherwise, you'd see the full-wave-rectified 100 Hz that the bridge produces.
The idea would be that during the half-wave that the bridge functions ok, not enough voltage can be generated to produce a DC voltage that's flat enough. With a linear regulator, I'd bet 90% on that with a 50 Hz ripple!

Raymond


 

On Thu, Oct 8, 2020 at 01:58 AM, Raymond Domp Frank wrote:


With a linear regulator, I'd bet 90% on that with a 50 Hz ripple!
To be clear: Obviously, it isn't a linear regulator.
If it isn't the bridge, which can be excluded with a good 340 VDC across the big caps and you do have 50 Hz across the LV outputs, the circuit around U3105 may be the culprit. That's on the lower part of DWG 8, "Inverter/rectifiers".

OTOH: You say the power supply is in tick mode but you do have at least one voltage that is close (50 VDC)?

Raymond


Simon Owen
 

I have put together a dummy load that demands about half of the total power
the scope can consume, built from info I found in the forum.

The PSU ticks a few times and then fires up with the dummy load connected.

I thought: Great! The power supply is OK! Then: Oh no, if it's not the
power supply... but now the measurements I am getting for the different
supplies are baffling my novice brain.

The +5V supply is putting out a waveform which is sinusoidal 1.2Vppk and
centred on zero volts, the frequency of the wave is 7MHz.

The -15V supply has 1.2V of ripple at the same 7MHz.

+15V has 50Hz ripple of about 10mV.

Probing continues.

On Thu., 8 Oct. 2020, 11:16 Raymond Domp Frank, <@Raymond> wrote:

On Thu, Oct 8, 2020 at 01:58 AM, Raymond Domp Frank wrote:


With a linear regulator, I'd bet 90% on that with a 50 Hz ripple!
To be clear: Obviously, it isn't a linear regulator.
If it isn't the bridge, which can be excluded with a good 340 VDC across
the big caps and you do have 50 Hz across the LV outputs, the circuit
around U3105 may be the culprit. That's on the lower part of DWG 8,
"Inverter/rectifiers".

OTOH: You say the power supply is in tick mode but you do have at least
one voltage that is close (50 VDC)?

Raymond






 

On Thu, Oct 8, 2020 at 01:55 AM, Simon Owen wrote:


I measure 1.8V peak-to-peak at approximately 7MHz. Yes that's Megahertz!
OK, timing in our messages probably caused my misunderstanding that you saw 50 Hz ripple on the 50 VDC. I now understand it's 7 MHz on the 50 VDC output.
Tick mode applies to all low-voltage outputs together, so if 50 VDC is present, so should the other low voltages.
Currently, I'm not sure I have a clear picture.

Raymond


Simon Owen
 

Nor I. I will continue and put together my findings. Thanks for your help.
S

On Thu., 8 Oct. 2020, 11:29 Raymond Domp Frank, <@Raymond> wrote:

On Thu, Oct 8, 2020 at 01:55 AM, Simon Owen wrote:


I measure 1.8V peak-to-peak at approximately 7MHz. Yes that's Megahertz!
OK, timing in our messages probably caused my misunderstanding that you
saw 50 Hz ripple on the 50 VDC. I now understand it's 7 MHz on the 50 VDC
output.
Tick mode applies to all low-voltage outputs together, so if 50 VDC is
present, so should the other low voltages.
Currently, I'm not sure I have a clear picture.

Raymond






 

On Thu, Oct 8, 2020 at 02:25 AM, Simon Owen wrote:


I have put together a dummy load that demands about half of the total power
the scope can consume, built from info I found in the forum.

The PSU ticks a few times and then fires up with the dummy load connected.

I thought: Great! The power supply is OK! Then: Oh no, if it's not the
power supply... but now the measurements I am getting for the different
supplies are baffling my novice brain.

The +5V supply is putting out a waveform which is sinusoidal 1.2Vppk and
centred on zero volts, the frequency of the wave is 7MHz.

The -15V supply has 1.2V of ripple at the same 7MHz.

+15V has 50Hz ripple of about 10mV.

Probing continues.
Are you sure that your loads are all connected correctly between the output and ground each and that you're measuring between the outputs and GND?
I'm not sure what to think of the 7 MHz. It's far too high for the inverter and looks more like a (parasitic) oscillation. That could happen around U3105 because of a missing or bad cap. Could it be there from when you recapped it? Did you forget a small cap or connected it wrongly? Can you try and look at the inputs of the linear regulators (dwg 9 and figure 6-16) and report on ripple, esp. if you're seeing the 7 MHz as well?

I'm going to bed now (I'm in NL, GMT+2). Have fun!

Raymond


 

On Thu, Oct 8, 2020 at 02:48 AM, Raymond Domp Frank wrote:


I'm going to bed now (I'm in NL, GMT+2). Have fun!
Quick afterburner: I understand now that the PSU is out of the 'scope and you're running with just a load connected? So it's still just the PSU after all?

Raymond


Simon Owen
 

Thank you Raymond, good night, by the way I think I have found several
power diodes not behaving properly with reverse voltages of 2 to 3 volts
only. I think we might be zeroing in.
Cheers,
Simon

On Thu., 8 Oct. 2020, 11:51 Raymond Domp Frank, <@Raymond> wrote:

On Thu, Oct 8, 2020 at 02:25 AM, Simon Owen wrote:


I have put together a dummy load that demands about half of the total
power
the scope can consume, built from info I found in the forum.

The PSU ticks a few times and then fires up with the dummy load
connected.

I thought: Great! The power supply is OK! Then: Oh no, if it's not the
power supply... but now the measurements I am getting for the different
supplies are baffling my novice brain.

The +5V supply is putting out a waveform which is sinusoidal 1.2Vppk and
centred on zero volts, the frequency of the wave is 7MHz.

The -15V supply has 1.2V of ripple at the same 7MHz.

+15V has 50Hz ripple of about 10mV.

Probing continues.
Are you sure that your loads are all connected correctly between the
output and ground each and that you're measuring between the outputs and
GND?
I'm not sure what to think of the 7 MHz. It's far too high for the
inverter and looks more like a (parasitic) oscillation. That could happen
around U3105 because of a missing or bad cap. Could it be there from when
you recapped it? Did you forget a small cap or connected it wrongly? Can
you try and look at the inputs of the linear regulators (dwg 9 and figure
6-16) and report on ripple, esp. if you're seeing the 7 MHz as well?

I'm going to bed now (I'm in NL, GMT+2). Have fun!

Raymond