Re: Hello from newcomer Fabio Trevisan - My first Tek Scope 464 + DM44

Fabio Trevisan
 

Hi Dave,
Thanks again for your feedback... I am suppressing my original post and
leave only your comments and my replies.
I hope this won't mangle the whole thing this time (Ì'm doing it from my
mail client).
Rgrds,
Fabio

2016-09-22 2:15 GMT-03:00 David @DWH [TekScopes] <
TekScopes@...>:

On 16 Sep 2016 17:07:00 -0700, you wrote:
Parsing a problem so it can be described to someone else is a powerful
technique.
Thanks for having the patience to go through it.

This supply is unregulated so the variation from nominal is expected.
A change from 1Vpp to 2Vpp indicates that the current draw is
doubling.
Since our last exchange, I did further progress in measurements, as well as
in troubleshooting resources..
I measured the current drawn by the inverter (while it's still operating
normally), and it's around 200mA.
I didn't have chances to measure it after HV drops, after the inverter runs
hot.

Yep, doubling the capacitance will halve the ripple. The rule of
thumb for the 120 Hz output from a full wave 60 Hz rectifier is that
8200uF will yield 1 Vpp of ripple at 1 amp.
From your rule of thumb, and from my previous conclusion that C1487 is
still at its nominal value,
the current being drawn from the +15V UNREG must be around 0.67A (making up
for a 1Vpp ripple on 5500uF).
Since the Inverter alone, draws just 0.2A, (while in normal operation) and
since the ripple doubles to 2Vpp when the inverter is in "fault"...
the net current out of the +15V UNREG is doubling to ~ 2 x 0.67A = 1.33A.
This means that the current drawn by the inverter alone is raising from
0.2A to 0.87A (almost 4.5 times).

My notes say it is 8.5 kilovolts. That is enough difference that I
would suspect failure of the high voltage multiplier is causing the
problem.
I concur with you that the 464 should be 8.5kV, But I have another theory!
I think the correct anode voltage is indeed in the 7kV ballpark for 2
reasons:
1. because this is exactly what is marked in the 464's HV's Cage Warning.
It says: CAUTION 7000V inside !!!
2. 7kV of PDA plus 1500V of cathode voltage makes up exactly for 8500V
acceleration !!!

Moreover, I`m considering that even the 7.22kV I`m measuring may be already
HIGH by 220V
and this can be exactly an indication of the inverter being driven harder
to compensate for some
deficiency on the cathode circuitry (either the winding or something else).
This rational is consistent with the other winding (600V) which also
measures around 620V
So, it really seems the problem is at the Cathode circuitry...
I will try to confirm that by measuring the HV doubler voltage or the 600V
while the scope
heats up towards to failure.

The resistors could not cause the problem anyway.
Agree, but I replaced them anyway, by glass insulated (3kV class) metal
film resistors. (but no appreciable change in status)

On my suggestion of possible replacements of the HV diodes... you wrote...
The PDA voltage will decrease a little bit but the effect on
deflection calibration will be minor.
Agreed! So much that I decided to take the step and replace all 3 HV diodes
(the 2 at the HV doubler and at the cathode circuit).

Leakage through either of the diodes will cause problems. The
capacitors could also be causing the problem.
I have not ordered from them yet, but these guys look like a good
source for high voltage diodes and capacitors:
<http://hvstuff.com/>
About hvstuff.com... I have seen their website before, I think that I will
ultimately need to source from
them if I need to replace the capacitors, as the sources in Brazil for HV
capacitors are either
unreliable or only affordable for the industries (no hobby market).
We used to have Newark/Element14 in Brasil but they closed their operations
here.

As I mentioned above, I decided to take my chances with the ESJA53-12
diodes I found on a loca store...
After I replaced the HV diodes, I noticed a great improvement... but
unfortunately it wasn't the definitive fix...
Now the HV doesn't seem to fail anymore if I run the scope without the
outer casing (cooler).
While this is good news in some sense, it made things more difficult to
troubleshoot.
When I install the outer casing, the time to fail increased from 20~30
minutes to about 1 hour.
I measured the leakage of the original cathode diode (CR1503) and under the
same voltage of about 3600V,
it's 25 times bigger than the 1.1nA that I measured in the new diodes,
but 28nA is still very small leakage and I think that it may just have
given the inverter a little bit more breath
so now it takes longer to fail.

I would change the high voltage capacitors before messing with the
transformer. Make sure the board around the high voltage multiplier
is clean when you finish working on it.
That's my next step... It will be difficult to find 6.8nF x 5kV around
here...
But at least for testing, I may resort to an association (parallel)
It's not HV friendly to have the uninsulated connections that would be
required
but there's plenty of space where the original capacitors are (they're
huge!)

The oscilloscope and CRT will still work with the PDA missing although
deflection will be lower and brightness and sharpness will be
affected. You could temporarily remove C1582 to verify if the high
voltage multiplier is the problem.
That tip is one that I wouldn't expect!
It would never occurred to me to remove the PDA altogether.
I didn't think it would ever work.
For the reasons I stated before, I don't think my problem in on the HV
or the HV doubler, but I will keep this advice if I need to troubleshoot
this area further.
For now, I will focus on the Cathode Supply circuitry and see if I get to
any conclusion.

Then I would take a close look at the cathode supply rectifiers and
capacitors around CR1503. And then the +600 volt bias supply built
around CR1512.
I've already checked the +600V supply and it seems alright.
This one is simpler to troubleshoot as the components are not that hard to
source.

Thanks again...
I will answer shortly the other message you answered on another topic...
The "double-peak"... That's another story.

Rgrds,

Fabio




Join TekScopes@groups.io to automatically receive all group messages.