开云体育

The Delta Ref and Delta pots on my 2465 are very erratic (both cursors jump around when the pots are turned whether in independent or tracking mode), but all of the other pots on the front panel seem to work fine. Any troubleshooting suggestions would be appreciated. Thanks. Dave

On Mar 3, 2019, at 4:55 PM, Reed Dickinson <reed714@...> wrote:

I just re-read your trouble note and I suspect you have identified the problem when you said the trace is only 2/3 of the normal width. This points immediately to a defective U1321 voltage tripler. Unsolder the lead from T1320, pin 12, tripler and power on, look for normal waveforms around Q1318. If your tripler is defective a replacement 3X can be had in an NTE539 or ECG539 (an NTE538 or ECG538 would work too).
By momentarily connecting the base of Q1318 to +5 through a 1K resistor you have bypassed all the regulation that you get from Q1306 through Q1316 and you have proved that the transformer and associated circuit are working. The shortened trace you saw was probably due to the unregulated HV giving a shorter trace or a defective voltage tripler U1321. I would still look for problems around C1304. When you power up the time constant of R1304 and C1304 is about 680ms. This means that you have under a second for the HV to stabilize since the base of Q1306 is being pulled toward +110V until it is nailed at +0.6 by the conduction of CR1304. After about 1 second the base of Q1306 should go to about +0.6V and oscillations should occur via Q1318.
To troubleshoot the regulator circuit pull F1318 out and disconnect P1313, the transistor Q1318 is now out of the circuit. Connect a 25K pot between +15 and - 5 and connect the wiper of the pot through a 10K resistor to the Q1306 base node. Connect a sensitive voltmeter to the - side of C1316 and power on your scope. Connect another voltmeter to the Q1306 base node and vary the pot. When the base of Q1306 is around +0,6V you should see some variation at the - side of C1316 (a 4.7K temporary resistor to ground will be necessary to give a little current through Q1316), If no variation is seen when rotating the pot then you have a problem in the regulator circuit, troubleshoot it and you will find the problem. If a variation is seen then the regulator is working and you need to fix problems on the secondaries of the transformer. Before using the following shotgun replacement procedure unsolder the lead from T1320 pin 12 going to the 3X voltage multiplier. If the multiplier is bad it would put a load on the oscillator circuit and kill the oscillations. Next, use the shotgun approach and replace C1320, CR1320, CR1321, C1322, C1323, C1328 and CR1329. I recommend you use HV caps rated at least 5KV as the surge of HV during the first second is high.
Good luck:
Reed Dickinson
On Sunday, March 3, 2019, 12:01:12 PM PST, SuddenLink <bob.ripley@...> wrote:

I missed identified C1304. Instead it was C1468 with a broken lead that I pulled from the circuit board. A perfect example of a mistake made when one works on a problem to long without stepping away for a while.

C1436 on the -8 volt rail was replaced with a 33uf capacitor.

Reed,

I checked C1304 and it is fine, measures 2.7uf on the dot. CR1304 was replaced with a IN4007 per your recommendation and I tried your test to the base of Q1318.

The first time I taped the base of Q1318 with a 1K resistor in series with the +5 volt rail the waveform at TP1318 more than doubled in amplitude from 760mv p-p to 2.2 volts p-p and has stayed there. The frequency however, is now down to 14.4Khz

Now when I touch the lead of the 1K resistor from the +5 volt rail to the base of Q1318 the Oscillator waveform at TP1318 jumps up to 7.2 volts p-p still at 14.4 Khz. When I take the 1K resistor away the waveform drops back to 2.2 volts.

BREAKING NEWS:

While typing this report up I turned back to the scope, taped the 1K resistor from the +5 volt rail to the base of Q1318 again. This time the waveform at TP1318 jumped up to 30.8 volts p-p and now I have a trace on the CRT. Take the 1K resistor away and the waveform at TP1318 keeps running at 30.8 volts p-p and there is still a trace on the CRT. Note: The trace is only covering about a two-thirds of the screen and the horizontal position is behaving strange. Instead of sliding the whole trace left or right. It erases part of trace moving towards the middle of the screen if the Horizontal position is turn left or right. I suspect that is because the HV oscillator is only running at 13.8Khz.

I shut the scope off then turn it back on. The waveform at TP1318 is down to 2.2 volts p-p. Tap the base of Q1318 with the lead from the 1k resistor going to +5 volt rail and the oscillator waveform is goosed to 30.8 volts p-p again. The trace on the CRT is present as before and stays there until I shut the scope off and turn it back on.

How weird is that!

So now what? Is Q1318 flaky? I kind of doubt that. C1316 was replaced with a new 1uf tantalum capacitor. C1317 was pulled and checked. The LCR meter says it’s a 10uf capacitor. I don’t remember what the ESR was but I don’t remember it raising an eyebrow with me. CR1317 was check with the diode function on my HP 3466 DMM but it could be leaky…maybe?

What the Heck?

Ripley

Sent from Mail for Windows 10

From: Steph L
Sent: Saturday, March 2, 2019 5:43 PM
To: [email protected]
Subject: Re: [TekScopes] 475 with no Display and no HV

Page 3-35 of manual has a good functional description including point that ambient current supplied to base (of Q1318) determines frequency (ie not LC arrangement) and as such there isn’t a specific cap component for resonate frequency (a function oscillator still relies upon however to function).

You said “The rest of the voltage measurement compared to the manual are no where near correct” so perhaps a closer look at the whole setup noting Q1308/10 push-pull, gain and phase reversal Q1306 and bias arrangement behind Q1310 (ie the Q1312 and associated components).

I have an old Tek544 where the HT falls dramatically when ambient room temp too hot and was amazed in the frequency shift involved.

It is a different circuit but 475 still has resonance involved and the output voltage is regulated by driving current harder into the base.

Just a thought. Please excuse me/any error as nubie here. Interesting fault.

Hi Bill,

I have the same amplifiers and when I first got them, had the same or
at least similar issues.

I tore them down, one at a time, and soaked 24lb bond with IPA, to
clean those switches. Yes, they are a bear to get to! IIRC, I had to
disassemble them down a ways to get into them. There were shields on
the inner sides and on the outer sides, too. Once I had them broken
down, it was a simple thing to get the soaked paper into the switch
fingers and carefully pull it through, in the same way that a
turntable moves a record under the stylus... never backward!

I recommend doing this somewhere where you can leave it laying,
disassembled, so that you can see how you got it apart, making it
easier to reassemble it. I took pictures on my phone at each step of
the way. It was a "trick" of a deal getting it broken down to the
point where it could be cleaned, however. Pictures helped me to stay
on track and get it all back together! Tektronix did it's homework...
they were made to be repaired... gotta love 'em!

I hope I've inspired you and not the opposite. I did have to work up
my nerve to tear into them, but the result was two operable
amplifiers, I am glad that I took my time and worked carefully.

Phil

On 3/2/2019 2:26 PM, Bill Carns wrote:

I'll have to go put the unit back in the mainframe and investigate
but my memory says that most were bad.

I have investigated more and remove the upper capacitor cover and,
boy are those things hard to get at. Gonna be real hard to get
cleaning paper in there on all the switches. Some are not bad,
others really bad to get to. One set has an additional little cover
over them for some reason.

Working on this thing is not for sissies.


B

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of
Albert Otten
Sent: Saturday, March 02, 2019 3:53 PM
To: [email protected]
Subject: Re: [TekScopes] Help with 7A18 problem

On Sat, Mar 2, 2019 at 07:58 PM, Colin Herbert wrote:

---
On the subject of the input attenuators, they are the type which are
operated by cam-switches. They have springy gold-plated contacts
that get moved up-and-down by cams on the controls shaft. The
circuit board that they are on has gold-plating, too, and the board
itself is made of an easily-damaged material. You need to use thin
strips of paper moistened in IPA and trapped between the moving and
stationary contacts, they *gently* pulling the paper strip out. You
may have to do this a few times. Don't use any other solvents or
contact-cleaner as you will probably ruin the board. Of course it
may only be one or two of these contacts that is dirty, as has
already been suggested; there are only four attenuators there - they
get switched in-and-out to get the different VOLTS/DIV settings. In
the most sensitive VOLTS/DIV setting, they are all switched out, so
if your 5mV/div setting shows lousy bandwidth, then it is likely to
be caused by some other problem, not just the attenuators and their
switching . Of course, there might be a problem there, too, but it
won't be the only one.

Hi Colin,

You probably were too quick in your statement about the 5 mV/div
setting. In that setting all attenuators are bypassed by means of a
series of closed contacts, so any dirty contacts there destroy the
signal path.
I'm not sure about the 7A18 construction. My impression is that
including AC/DC there are 10 contact pairs above the drum and 10
below the drum. (Each contact pair forming one switch).

It would still be nice to hear from Bill which V/div settings are
good and which are bad.

Albert

Hi Bill,

I have the same amplifiers and when I first got them, had the same or at least similar issues.

I tore them down, one at a time, and soaked 24lb bond with IPA, to clean those switches.? Yes, they are a bear to get to!? IIRC, I had to disassemble them down a ways to get into them.? There were shields on the inner sides and on the outer sides, too.? Once I had them broken down, it was a simple thing to get the soaked paper into the switch fingers and carefully pull it through, in the same way that a turntable moves a record under the stylus... never backward!

I recommend doing this somewhere where you can leave it laying, disassembled, so that you can see how you got it apart, making it easier to reassemble it.? I took pictures on my phone at each step of the way.? It was a "trick" of a deal getting it broken down to the point where it could be cleaned, however.? Pictures helped me to stay on track and get it all back together!? Tektronix did it's homework... they were made to be repaired... gotta love 'em!

I hope I've inspired you and not the opposite.? I did have to work up my nerve to tear into them, but the result was two operable amplifiers, I am glad that I took my time and worked carefully.

Phil

On 3/2/2019 2:26 PM, Bill Carns wrote:

I'll have to go put the unit back in the mainframe and investigate but my memory says that most were bad.

I have investigated more and remove the upper capacitor cover and, boy are those things hard to get at.? Gonna be real hard to get cleaning paper in there on all the switches.? Some are not bad, others really bad to get to.? One set has an additional little cover over them for some reason.

Working on this thing is not for sissies.


B

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Albert Otten
Sent: Saturday, March 02, 2019 3:53 PM
To: [email protected]
Subject: Re: [TekScopes] Help with 7A18 problem

On Sat, Mar? 2, 2019 at 07:58 PM, Colin Herbert wrote:

---
On the subject of the input attenuators, they are the type which are
operated by cam-switches. They have springy gold-plated contacts that
get moved up-and-down by cams on the controls shaft. The circuit board
that they are on has gold-plating, too, and the board itself is made
of an easily-damaged material. You need to use thin strips of paper
moistened in IPA and trapped between the moving and stationary
contacts, they *gently* pulling the paper strip out. You may have to
do this a few times. Don't use any other solvents or contact-cleaner
as you will probably ruin the board. Of course it may only be one or
two of these contacts that is dirty, as has already been suggested;
there are only four attenuators there - they get switched in-and-out
to get the different VOLTS/DIV settings. In the most sensitive
VOLTS/DIV setting, they are all switched out, so if your 5mV/div
setting shows lousy bandwidth, then it is likely to be caused by some
other problem, not just the attenuators and their switching . Of course, there might be a problem there, too, but it won't be the only one.

Hi Colin,

You probably were too quick in your statement about the 5 mV/div setting. In that setting all attenuators are bypassed by means of a series of closed contacts, so any dirty contacts there destroy the signal path.
I'm not sure about the 7A18 construction. My impression is that including AC/DC there are 10 contact pairs above the drum and 10 below the drum. (Each contact pair forming one switch).

It would still be nice to hear from Bill which V/div settings are good and which are bad.

Albert

I'll have to go put the unit back in the mainframe and investigate but my memory says that most were bad.

I have investigated more and remove the upper capacitor cover and, boy are those things hard to get at. Gonna be real hard to get cleaning paper in there on all the switches. Some are not bad, others really bad to get to. One set has an additional little cover over them for some reason.

Working on this thing is not for sissies.


B

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Albert Otten
Sent: Saturday, March 02, 2019 3:53 PM
To: [email protected]
Subject: Re: [TekScopes] Help with 7A18 problem

On Sat, Mar 2, 2019 at 07:58 PM, Colin Herbert wrote:

---
On the subject of the input attenuators, they are the type which are
operated by cam-switches. They have springy gold-plated contacts that
get moved up-and-down by cams on the controls shaft. The circuit board
that they are on has gold-plating, too, and the board itself is made
of an easily-damaged material. You need to use thin strips of paper
moistened in IPA and trapped between the moving and stationary
contacts, they *gently* pulling the paper strip out. You may have to
do this a few times. Don't use any other solvents or contact-cleaner
as you will probably ruin the board. Of course it may only be one or
two of these contacts that is dirty, as has already been suggested;
there are only four attenuators there - they get switched in-and-out
to get the different VOLTS/DIV settings. In the most sensitive
VOLTS/DIV setting, they are all switched out, so if your 5mV/div
setting shows lousy bandwidth, then it is likely to be caused by some
other problem, not just the attenuators and their switching . Of course, there might be a problem there, too, but it won't be the only one.

Hi Colin,

You probably were too quick in your statement about the 5 mV/div setting. In that setting all attenuators are bypassed by means of a series of closed contacts, so any dirty contacts there destroy the signal path.
I'm not sure about the 7A18 construction. My impression is that including AC/DC there are 10 contact pairs above the drum and 10 below the drum. (Each contact pair forming one switch).

It would still be nice to hear from Bill which V/div settings are good and which are bad.

Albert

---
Something that occurred to me today, after what Albert pointed out
about the 5mV/DIV setting being possibly compromised by dirty switch
contacts, was that it is possible to see if the attenuator assembly
has any problems by putting an appropriate signal into the input and a
scope probe on to the output, which is the junction of R130, C210 and
R210 or the equivalent components on the other channel. This of course
requires either an 7000-series extender or some other way to access
the 7A18 plug-in with a probe, plus another fully-functioning scope.
Turning the VOLTS/DIV switch will allow examination of the
signal-path, to some extent. This could be achieved without an
extender, with one scope of the 7000-series having two vertical
amplifiers, by removing both the side-panel of the DUT amplifier and
the left side-panel of the mainframe. If the DUT amplifier plug-in is
in the left bay, then I think it would be possible to connect the
scope probe from the working vertical amplifier plug-in to the DUT
one? You wouldn't need to be accepting the signal from the DUT amplifier, though this may be of use. Is there any sense in my suggestion, or have I got it completely wrong?

Colin.

---
On the subject of the input attenuators, they are the type which are
operated by cam-switches. They have springy gold-plated contacts that
get moved up-and-down by cams on the controls shaft. The circuit board
that they are on has gold-plating, too, and the board itself is made
of an easily-damaged material. You need to use thin strips of paper
moistened in IPA and trapped between the moving and stationary
contacts, they *gently* pulling the paper strip out. You may have to
do this a few times. Don't use any other solvents or contact-cleaner
as you will probably ruin the board. Of course it may only be one or
two of these contacts that is dirty, as has already been suggested;
there are only four attenuators there - they get switched in-and-out
to get the different VOLTS/DIV settings. In the most sensitive
VOLTS/DIV setting, they are all switched out, so if your 5mV/div
setting shows lousy bandwidth, then it is likely to be caused by some
other problem, not just the attenuators and their switching . Of course, there might be a problem there, too, but it won't be the only one.

The reseating fixed the channel I worked on.

---
On the subject of the input attenuators, they are the type which are
operated by cam-switches. They have springy gold-plated contacts that
get moved up-and-down by cams on the controls shaft. The circuit board
that they are on has gold-plating, too, and the board itself is made
of an easily-damaged material. You need to use thin strips of paper
moistened in IPA and trapped between the moving and stationary
contacts, they *gently* pulling the paper strip out. You may have to
do this a few times. Don't use any other solvents or contact-cleaner
as you will probably ruin the board. Of course it may only be one or
two of these contacts that is dirty, as has already been suggested;
there are only four attenuators there - they get switched in-and-out
to get the different VOLTS/DIV settings. In the most sensitive
VOLTS/DIV setting, they are all switched out, so if your 5mV/div
setting shows lousy bandwidth, then it is likely to be caused by some
other problem, not just the attenuators and their switching . Of course, there might be a problem there, too, but it won't be the only one.

---
Something that occurred to me today, after what Albert pointed out about the
5mV/DIV setting being possibly compromised by dirty switch contacts, was that
it is possible to see if the attenuator assembly has any problems by putting
an appropriate signal into the input and a scope probe on to the output, which
is the junction of R130, C210 and R210 or the equivalent components on the
other channel. This of course requires either an 7000-series extender or some
other way to access the 7A18 plug-in with a probe, plus another
fully-functioning scope. Turning the VOLTS/DIV switch will allow examination
of the signal-path, to some extent. This could be achieved without an
extender, with one scope of the 7000-series having two vertical amplifiers,
by removing both the side-panel of the DUT amplifier and the left side-panel
of the mainframe. If the DUT amplifier plug-in is in the left bay, then I
think it would be possible to connect the scope probe from the working
vertical amplifier plug-in to the DUT one? You wouldn't need to be accepting
the signal from the DUT amplifier, though this may be of use. Is there any
sense in my suggestion, or have I got it completely wrong?

Colin.