Topics

5116+5D10 intensity problem

Yeun-Jung Wu
 

Hello:

Today I dig out my 5116+5D10 to display signal from FG501 and FG501A. I used "channal addition" to display added signal from two oscillators tuned to nearly 7kHz sine wave. At equal amplitude the expected envelope was beautifully displayed on the screen. Switch to X-Y mode display a dotted Lissajous pattern. Read out and cursor all function as expected.

After operation about one and a half hour the intensity gradually decay. It was accompanied by the expansion of horizontal and vertical graticutes. Now one dotted horizontal and vertical division expanded to include four real inner CRT divisions. Read out information was pushed outside of screen. Pushing the "Beam finder" button display a image filled out the entire screen (save + Power on), instead of a scaled down version of display. The intensity of the pattern was barely visible. Weak wave forms were still displayed on the screen; mostly with a red LED warning: possible under-sampling.

I had tried to adjust intensity range R245 on A3 board with little effect.

Should I go further to poke the wave forms at 21 and 26A/B? Was the problem in Z-axis amp or HV (pin 2 of CRT socket)?
Q222, Q234, Q226 all located in a hard to reached region so I had not checked their status yet.

Any suggestions? Thank you in advance.

Yeun-Jung Wu

Yeun-Jung Wu
 

BTW +-30V. 200V and 5V all were within spec. Haven't checked their ripple content yet.

 

If you switch it off and allow it to cool down for (say) a few hours or overnight, does it work again for a while?

I'm thinking this might be one of the problematic epoxy potted HT transformers

David

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Yeun-Jung Wu
Sent: 14 February 2020 10:43
To: TekScopes@groups.io
Subject: [TekScopes] 5116+5D10 intensity problem

Hello:

:

After operation about one and a half hour the intensity gradually decay. It was accompanied by the expansion of horizontal and vertical graticutes. Now one dotted horizontal and vertical division expanded to include four real inner CRT divisions. Read out information was pushed outside of screen. Pushing the "Beam finder" button display a image filled out the entire screen (save + Power on), instead of a scaled down version of display. The intensity of the pattern was barely visible. Weak wave forms were still displayed on the screen; mostly with a red LED warning: possible under-sampling.

Yeun-Jung Wu
 

Hello David:

Thank you for your prompt reply. I will wait overnight and try again tomorrow. It looked like e-beam was pushed twice as much in each direction. This could happen if electrons move slower, i.e. insufficient electron acceleration. Slower electron impact creates dimmer fluorescent light. It would all fit together as a reasonable explanation.

Best Regards,

Yeun-Jung Wu

Yeun-Jung Wu
 

BTW I did not see neon, DS274, (DS271+DS272+DS273 in series) lit.

Yeun-Jung Wu
 

I just digged out the A3 HV board and checked all the transistors, diodes except CR247, CR298, which is a 10kV, 5mA silicon diode and a 800V, 25mA silicon diode, respectively. I forgot to check the thick film hybrid resistor R272, which had multiple wire juction at A, B, C, D, E, F.

I also checked two aluminum electrolytic capacitors, 10uF/100V, and 10uF/50V. Their capacitance were 12uF with serial resistor about 2Ohms at 100Hz.

Putting everything back all the symptoms remain the same. I wondered if there was leakage of CR247 or CR298?

Now the horizontal and vertical expansion was about 1.5 graticules and varing slowly. Read out was still way outside the limit of screen. "Possible under-sampling" red LED warning lit all the time, stable trigger was not possible.

Harvey White
 

Have you (and can you) safely measure the high voltage when this wandering happens?  First, though, I'd check ripple and voltage on the supplies.  That you have no stable trigger tends to make me think (in general terms) that there's ripple on the power supply somewhere.

It's possible that the supplies start good, and then something changes in the supply goes bad and the ripple goes up.

Harvey

On 2/14/2020 10:57 AM, Yeun-Jung Wu wrote:
I just digged out the A3 HV board and checked all the transistors, diodes except CR247, CR298, which is a 10kV, 5mA silicon diode and a 800V, 25mA silicon diode, respectively. I forgot to check the thick film hybrid resistor R272, which had multiple wire juction at A, B, C, D, E, F.

I also checked two aluminum electrolytic capacitors, 10uF/100V, and 10uF/50V. Their capacitance were 12uF with serial resistor about 2Ohms at 100Hz.

Putting everything back all the symptoms remain the same. I wondered if there was leakage of CR247 or CR298?

Now the horizontal and vertical expansion was about 1.5 graticules and varing slowly. Read out was still way outside the limit of screen. "Possible under-sampling" red LED warning lit all the time, stable trigger was not possible.


Yeun-Jung Wu
 

Hello Harvey:

Thank you for your suggestion.

I had checked the ripple content of power supply:

+30V: 20mV p-p; should be 3mV or less
-30V: 8mV p-p; should be 3mV or less
+200: 195, 1.5V p-p; should be 3V or less
+5: 8mV; should be 3mV or less

Plus 30V branch had significant larger ripple content, which may attributed to aging filter capacitor.

According the spec. the accelerating voltage was 4.5kV, pin 2 at CRT socket. My pocket DMM can measure voltage up to 1kV. A HV probe is needed.

Maybe I can order a new piece of 10kV/5mA diode first.

With Tek 485 probed at TP 12, I got an interesting picture of the vertical input signal when 5D10 display a standard graticule grid. It was digitally precise!

Best Regards,

Yeun-Jung Wu

Harvey White
 

OK, then what I was looking for was a significantly larger ripple.  What happens, voltage wise, at the very beginning when you turn on power, and then after the trace starts to dim?  Is there a voltage change there?

Naturally, you do need to measure the HV, since that may be a problem.  Lower HV typically means dimmer trace, and depending on how the CRT is done, can make for a larger trace.

However, and you need to check when the trace goes bad, does the waveform at TP12 change?  If not, then the fault lies beyond that in the circuit.  (could likely be in the CRT circuit).  Since this is a 5116, did you mention if the problem exists with another set of plugins?  IF not, then it's looking as if the problem is rather specifically in the 5D10.

Not familiar with the setup, so I'm just working on symptoms.

Harvey

On 2/15/2020 1:15 AM, Yeun-Jung Wu wrote:
Hello Harvey:

Thank you for your suggestion.

I had checked the ripple content of power supply:

+30V: 20mV p-p; should be 3mV or less
-30V: 8mV p-p; should be 3mV or less
+200: 195, 1.5V p-p; should be 3V or less
+5: 8mV; should be 3mV or less

Plus 30V branch had significant larger ripple content, which may attributed to aging filter capacitor.

According the spec. the accelerating voltage was 4.5kV, pin 2 at CRT socket. My pocket DMM can measure voltage up to 1kV. A HV probe is needed.

Maybe I can order a new piece of 10kV/5mA diode first.

With Tek 485 probed at TP 12, I got an interesting picture of the vertical input signal when 5D10 display a standard graticule grid. It was digitally precise!

Best Regards,

Yeun-Jung Wu


Yeun-Jung Wu
 

Hello Harvey:

Since the intensity went down in recent event it never came back. What this 5116 did after turned on was nothing on the screen, even if the intensity control knob was turned fully clockwise. I can tell it was powered on only when scale graticule illumination was fully on.

I guess turn down the vertical and horizontal gain would shrink the display to normal siize. This couldn't be a cure since most important task is to restore a visible trace in normal viewing condition. Once the intensity problem was solved the graticule scale would be normal again.

I did not have any other 5000 series plug-in's. I bought this broken 5116+5D10 with smashed LCD color shutter a few years ago and fixed 5D10 digitizer.

I up load three photos under the file name 5116. Signal to vertical amplifier was interesting: an old video game?

Regard,

Yeun-Jung Wu

Harvey White
 

You definitely want to look into one of two things, either bad/reduced high voltage (symptoms would be magnified deflection because a slower electron stream is easier to deflect), and then the cathode/grid bias, but right now, I'm betting on the high voltage.

I wouldn't change the horizontal or vertical gain, but I'd check the outputs of the amplifiers for normal signal levels.  I seem to remember that when the 5000 series was mentioned, there may have been some problems with an epoxy coated HV transformer, which may need to be replaced.  However, I'd check the HV oscillator circuit to see what's going on there.  If, for instance, there is an HV multiplier block (possible), feeding a separate accellerator, then failure of that block causes the image to bloom.

I'd suggest doing a search on the group archives for scope HV problems, which shows the steps you can take to troubleshoot the system.   You do need an HV probe.  It should be specified to work with the meter impedance you have, since it's a 1000 to 1 high voltage divider using the meter itself as the lower part.  10 meg meter needs 9990 meg resistive divider.

You might want to check the discussions on those, too, for additional information

Harvey

On 2/16/2020 2:46 AM, Yeun-Jung Wu wrote:
Hello Harvey:

Since the intensity went down in recent event it never came back. What this 5116 did after turned on was nothing on the screen, even if the intensity control knob was turned fully clockwise. I can tell it was powered on only when scale graticule illumination was fully on.

I guess turn down the vertical and horizontal gain would shrink the display to normal siize. This couldn't be a cure since most important task is to restore a visible trace in normal viewing condition. Once the intensity problem was solved the graticule scale would be normal again.

I did not have any other 5000 series plug-in's. I bought this broken 5116+5D10 with smashed LCD color shutter a few years ago and fixed 5D10 digitizer.

I up load three photos under the file name 5116. Signal to vertical amplifier was interesting: an old video game?

Regard,

Yeun-Jung Wu


Yeun-Jung Wu
 

Hello Harvey:

Thank you for your hint on future diagnosis. I will start collecting more information on epoxy coated transformer first.

5D10 seemed to work fine except read out information had been expanded outside the screen. Now I understand why red LED "possible under-sampling" would lit: I was using slow time base to watch two nearly identical 7kHz since wave superimposed together to form a nice envelope. There were lot of wave form variation during long accuqsition time but the available memory space for storing waveform was limited. When I select faster time base such that several cycles was within the range of one horizontal division that warning went away.

More than 5 year's ago I bought a Tek high voltage probe from eBay. There was an empty can supposed to contain liquid with high dielectric strength, something similar to banned CFC freon. For 4.5kV maybe I can fill some other liquid and give it a try.

Best Regards,

Yeun-Jung Wu

Chuck Harris
 

When the Anode voltage on the CRT is lower than it is supposed to
be, the electrons in the beam travel proportionately slower than
they should. Because they are slower, they spend a longer amount
of time to passing between the deflection plates. This means
longer time means that the force the electric field between the
plates applies to the electrons, has a longer time to work at
deflecting the electron beam. Which further means the electron
beam will be deflected more than it would at the proper voltage.

What you should take away from this is: If the anode voltage is
low, everything on the CRT screen will be expanded (magnified) out
much larger than it should be.

That your readout is expanded outside of the screen is nearly
perfect proof that your anode voltage is low. You can guestimate
how low by how much the expansion is. But it is most definitely
lower than it should be.

-Chuck Harris

Yeun-Jung Wu wrote:

Hello Harvey:

Thank you for your hint on future diagnosis. I will start collecting more information on epoxy coated transformer first.

5D10 seemed to work fine except read out information had been expanded outside the screen. Now I understand why red LED "possible under-sampling" would lit: I was using slow time base to watch two nearly identical 7kHz since wave superimposed together to form a nice envelope. There were lot of wave form variation during long accuqsition time but the available memory space for storing waveform was limited. When I select faster time base such that several cycles was within the range of one horizontal division that warning went away.

More than 5 year's ago I bought a Tek high voltage probe from eBay. There was an empty can supposed to contain liquid with high dielectric strength, something similar to banned CFC freon. For 4.5kV maybe I can fill some other liquid and give it a try.

Best Regards,

Yeun-Jung Wu

Harvey White
 

You're welcome.  I don't know if you have the particular HV transformer, but as Chuck (next) says, you're definitely looking at a low high voltage, so the question, of course, is why....

For the probe:  That's likely a P6015 or the equivalent.  It may be used with or without the dielectric liquid.  WIth, I think that the rating is 30KV, without, about 18KV.  You should be fine without it.   You may want to try using it with an adaptor with a meter, but it is designed to work into a 1 meg input impedance, and most meters are 10.  Paralleling the meter input to make it 1 meg is recommended if you try this.  I'd also try a test run (if you do that) on a known 100 volt or so supply to make sure you get the required 0.1 volt on the meter.  If not, then you're using it on a scope, and while you'll get a good idea, of voltage, it will only be within about 5% or so unless you have an A/D plugin.  The 5D10 might just do for that if you could see the readouts.

I recommend hunting the internet for the manuals for everything you have, if you haven't already.  I have an electronic library in several places.  Some of that information is available for only a limited time.

Harvey

On 2/17/2020 3:22 AM, Yeun-Jung Wu wrote:
Hello Harvey:

Thank you for your hint on future diagnosis. I will start collecting more information on epoxy coated transformer first.

5D10 seemed to work fine except read out information had been expanded outside the screen. Now I understand why red LED "possible under-sampling" would lit: I was using slow time base to watch two nearly identical 7kHz since wave superimposed together to form a nice envelope. There were lot of wave form variation during long accuqsition time but the available memory space for storing waveform was limited. When I select faster time base such that several cycles was within the range of one horizontal division that warning went away.

More than 5 year's ago I bought a Tek high voltage probe from eBay. There was an empty can supposed to contain liquid with high dielectric strength, something similar to banned CFC freon. For 4.5kV maybe I can fill some other liquid and give it a try.

Best Regards,

Yeun-Jung Wu


Paul Cohen
 

Or maybe SF6 (nasty) gas which is used in HV

בתאריך יום ב׳, 17 בפבר׳ 2020, 10:22, מאת Yeun-Jung Wu ‏<
B0190@...>:

Hello Harvey:

Thank you for your hint on future diagnosis. I will start collecting more
information on epoxy coated transformer first.

5D10 seemed to work fine except read out information had been expanded
outside the screen. Now I understand why red LED "possible under-sampling"
would lit: I was using slow time base to watch two nearly identical 7kHz
since wave superimposed together to form a nice envelope. There were lot of
wave form variation during long accuqsition time but the available memory
space for storing waveform was limited. When I select faster time base such
that several cycles was within the range of one horizontal division that
warning went away.

More than 5 year's ago I bought a Tek high voltage probe from eBay. There
was an empty can supposed to contain liquid with high dielectric strength,
something similar to banned CFC freon. For 4.5kV maybe I can fill some
other liquid and give it a try.

Best Regards,

Yeun-Jung Wu



Harvey White
 

Ok, several things after searching the  manuals.

1) the liquid is fluorocarbon 114.  SF6 is used in the AWACS transmitter unless they rebuilt it.

2) the P6015 has the liquid, the P6015A has a silicone filler and no liquid.

3) the P6013 (IIRC) may have neither and is rated for about 12KV or so, while the P6015 is rated from 20KV to 30KV depending on circumstances.

4) I did not find the specs on the 6015 without dielectric fluid, not in the internet material I have at the moment.


Harvey

On 2/17/2020 3:26 AM, Paul Cohen wrote:
Or maybe SF6 (nasty) gas which is used in HV

בתאריך יום ב׳, 17 בפבר׳ 2020, 10:22, מאת Yeun-Jung Wu ‏<
B0190@...>:

Hello Harvey:

Thank you for your hint on future diagnosis. I will start collecting more
information on epoxy coated transformer first.

5D10 seemed to work fine except read out information had been expanded
outside the screen. Now I understand why red LED "possible under-sampling"
would lit: I was using slow time base to watch two nearly identical 7kHz
since wave superimposed together to form a nice envelope. There were lot of
wave form variation during long accuqsition time but the available memory
space for storing waveform was limited. When I select faster time base such
that several cycles was within the range of one horizontal division that
warning went away.

More than 5 year's ago I bought a Tek high voltage probe from eBay. There
was an empty can supposed to contain liquid with high dielectric strength,
something similar to banned CFC freon. For 4.5kV maybe I can fill some
other liquid and give it a try.

Best Regards,

Yeun-Jung Wu



Chuck Harris
 

The manuals on the internet are the old 1963 versions of
the P6015 manual. If you look at the picture, you should
notice that the connector on the compensation box is an
old PL259 UHF connector, like was used on the early 500
series scopes.

The manual that came with my P6015 is APR 87, and it has
two curves, one with Freon, and one without. The curve
with Freon shows 20KVDC or RMS from DC to 1MHz. With air
is 13KVDC or RMS from DC to 1MHz.

The old probe shows the same curve as mine, only it limits
the probe to 28KV, at less than 100KHz.

I am pretty sure that if you over voltage the probe, it
has a spark gap that will arc over before the probe is
ruined. Obviously, the spark gap will arc over at a higher
voltage with freon, and a lower voltage with air.

The patent indicates that R12 is suitable if the probe is
to be used and stored below 10C.

If I had a probe that was sans fluid, and I couldn't buy
more R114, I would either live with the 13KV limit, or
try another refrigerant with a high boiling point...
probably butane. SF6 is also bound to work.

[Please don't bother mentioning how butane is flammable.
Everyone already knows that. They should also know that
it isn't if there is no air in the probe... If you
fill the probe with liquid in the probe, you won't be
able to screw on the probe's cap before the butane has
boiled enough to totally displace the air.]

With SF6, you would have to fill with vapor only, as its
liquification pressure is too high.

-Chuck Harris


Harvey White wrote:

Ok, several things after searching the  manuals.

1) the liquid is fluorocarbon 114.  SF6 is used in the AWACS transmitter unless they
rebuilt it.

2) the P6015 has the liquid, the P6015A has a silicone filler and no liquid.

3) the P6013 (IIRC) may have neither and is rated for about 12KV or so, while the
P6015 is rated from 20KV to 30KV depending on circumstances.

4) I did not find the specs on the 6015 without dielectric fluid, not in the internet
material I have at the moment.


Harvey


On 2/17/2020 3:26 AM, Paul Cohen wrote:
Or maybe SF6 (nasty) gas which is used in HV

בתאריך יום ב׳, 17 בפבר׳ 2020, 10:22, מאת Yeun-Jung Wu ‏<
B0190@...>:

Hello Harvey:

Thank you for your hint on future diagnosis. I will start collecting more
information on epoxy coated transformer first.

5D10 seemed to work fine except read out information had been expanded
outside the screen. Now I understand why red LED "possible under-sampling"
would lit: I was using slow time base to watch two nearly identical 7kHz
since wave superimposed together to form a nice envelope. There were lot of
wave form variation during long accuqsition time but the available memory
space for storing waveform was limited. When I select faster time base such
that several cycles was within the range of one horizontal division that
warning went away.

More than 5 year's ago I bought a Tek high voltage probe from eBay. There
was an empty can supposed to contain liquid with high dielectric strength,
something similar to banned CFC freon. For 4.5kV maybe I can fill some
other liquid and give it a try.

Best Regards,

Yeun-Jung Wu






magnustoelle
 

Good Day Yeun-Jung,
apart from the discussion on HV probing, my recommendation is to check the HV capacitors in the high volt section. I have seen a few which failed in other 5000 series oscilloscopes, but I have never seen a failed HV transformer in that family of devices. Faults might be spurious. If you have access to an insulation tester, that would be a relatively easy test, i.e. remove the board, unsolder the capacitors and test them. Carefully increase the test voltage while measuring the leakage current and follow the tester‘s manual.  Good luck,
Magnus

Chuck Harris
 

I have never seen a fault on a 5000 series EHT before either.

The epoxy used on the 5000 series, and the 7000 series, EHT's
is very different from the epoxy I find on 547's, etal. It is
clear amber, and very hard. The epoxy on failed 547's is a
muddy, brownish red, and is much softer.

I am coming to the conclusion that the epoxy failure is some
sort of fire retardant failure... much like computer cases that
turn from white to brown over time.

-Chuck Harris

magnustoelle via Groups.Io wrote:

Good Day Yeun-Jung,
apart from the discussion on HV probing, my recommendation is to check the HV capacitors in the high volt section. I have seen a few which failed in other 5000 series oscilloscopes, but I have never seen a failed HV transformer in that family of devices. Faults might be spurious. If you have access to an insulation tester, that would be a relatively easy test, i.e. remove the board, unsolder the capacitors and test them. Carefully increase the test voltage while measuring the leakage current and follow the tester‘s manual.  Good luck,
Magnus



Yeun-Jung Wu
 

Hello every Tek expert:

I just finished a long reply for thank you all for a long list of helpful hints. Unfortunately my wi-fi connection was lost at the time I keyin "send" command. Everything was lost.

Here I reype my reply short. First I want to thank everyone on the help of HV probe and the dielectric fluid. HV measurement was still a lethal barrier I had not cross before. Probably it was also a basic skill every devoted Tek collector should master. This will be my next great lesson to learn.

I managed to tune down X/Y gain of 5D10 and vertical/horizontal gain of 5116 by marking their original position first. On 5D10 there was little effect even the gain was set to minimum. Setting 5116's gain of vertical/horizontal indeed shrunk the scale on bother direction back a little; last portion of character of top readout was barely visible. As Harris and other experts pointed out sideway deflection should increase since slower electron spent more time across those electrostatic plates. Trace was dim because final kinetic energy was less due to insufficient acceleration. Finally I returned all settings to their original position.

I went on to probe waveforms on A3 HV board. TP267 at HV oscillator measured 0.56V, exactly as manual described. TP24 had similar waveform with the same period (oscillator worked) but the amplitude was smalle; 100Vpp v.s. 400Vpp. TP25 I measured similar waveform and amplitude while manual indicate a positive half waveform. (TP25 was at the junction of CR245, C246, R246. C246 was a 10kOhm R29 away from TP24. at least there was discrepencies on waveforms.

68nF C248/C249, 1nF C241, 47nF C242 were all cylindrical polyester film capacitors. I wonder whether LCR meter would reveal any leakage at high voltage? I also ordered a 10kV, 5mA silicon diode CR247. But expect long sipping delay.

Filter capacitors C837, C839, C815, C810 were at correct capacitance and low serial resistance. I had not trace where larger than normal ripple of +30V was from.

I wish the transformer, as many of you explained, would be O.K. Digging through HV problem of 5000 series scope I spotted Dennis's post of replacement of 577 curvier tracer by older 5110's. I also bought a broken 577 from eBay several years ago. Brought it back to work a few days then the intensity control became inoperative, rendered storage mode not function. Since it still displayed curves I leave it as it was. Hopefully there will no epoxy problem.

Dennis post also reminds me about the possibility of displaying 5D10's screen on a big screen scope, eg. R7603?

Best Regards to all of you!

Yeun-Jung Wu