Re: Yet another attempt to fix a 2465B


Chuck Harris
 

I prefer a capacitor that is higher voltage, and larger
diameter... more like the original's diameter... over a
smaller diameter, lower voltage capacitor for a couple
of reasons.

First, to make a higher voltage capacitor, you have to
apply a thicker layer of dielectric oxide. If you keep
the foil area the same, a higher voltage capacitor will
have lower capacitance than a lower voltage capacitor of
the same size.

To compensate for the reduced capacitance of a higher
voltage capacitor, the manufacturer can do a few things:
1) increase the number of turns of foil, which will
increase the ESL of the capacitor.
2) increase the height of the capacitor, which will lower
the resistance of the aluminum foil, and reduce the
ESR of the capacitor.
3) or do both.

Second, if the higher voltage results in a larger
capacitor, it will have more surface area from which to
radiate heat.

If you look at the tables for a given capacitor family,
you will almost always find that a higher voltage capacitor
of the same capacitance will have a lower ESR than its lower
voltage cousin.

My other reason has to do with lead spacing. If you
use a smaller capacitor, it will almost always have
narrower lead spacing than the original. If you simply
stick the leads in the holes, and press down until the
capacitor stops going down, the leads will be under
stress, and the seal will be stretched. This makes
the capacitor more likely to leak its electrolyte.

In the olden days, it was a very bad thing to use
an electrolytic capacitor that was much higher in WV
than the circuit requires. Today, it isn't. Today's
electrolytic capacitors will maintain their oxide layer
regardless of the voltage applied. This is because
the oxide layer is anodized onto the foil surface before
the capacitor is assembled. They can do this now because
the modern electrolytes are perfectly neutral to aluminum
and aluminum oxide.

-Chuck Harris

Stephen Hanselman wrote:

Chuck,

If you have low ESR, high ripple current and high temp, close to or higher capacitance right at or above working voltage (which is what we do) why is physical size important?

steve

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Chuck Harris
Sent: Monday, October 5, 2020 6:22 AM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Yet another attempt to fix a 2465B

Grayson,

To some people, a capacitor is a capacitor is a capacitor....

Not all electrolytic capacitors work well in switching power supplies. To get the best life out of your power supply capacitors You have to use the best long life (5K hours+), low ESR, low impedance, high ripple current capacitor you can afford. Further it has to be 105C rated or better.

I recommend looking for capacitors that are similarly sized to the originals. To get that size, go to higher value and higher voltage in your replacements. Higher voltage capacitors generally have lower ESR and higher ripple current ratings, rated for switching power supply use.

The guy that did the recap botched it, you should replace them again, only this time with properly selected capacitors.

I replace all of the 180uf40V and 250uf20V caps with a single type, 330uf 50V from United Chemicon. Panasonic also makes a suitable capacitor in that value. Nichicon doesn't (didn't?).

These capacitors are the most critical in the scope. I always find one or two of the original caps in these 6 positions that is bad. They handle +/-5V, +/- 15V, and the 5V digital.

The 5V digital pair are C1110 and C1111.

What I generally find with bad recap jobs is the capacitor's values are slavishly matched, with a capacitor that is about
1/4 the size of the original. Usually something that is a general purpose capacitor, not a switching rated capacitor.

-Chuck Harris

Grayson Evans wrote:
I have had a nice 2465B for about 6 years that I bought at Dayton (hamvention). It worked. About a year ago it quit when I turned it on one morning. It has the same symptoms as Dave's in his EEvlog #1203 on youtube. All the front panel LEDs light up but that's it.
Voltages are all in spec except the +5V, it is 4.5V with 400 mVpp ripple. The +15 volt supply also has the same 400 mVpp ripple, but voltage is good.
Whoever had it before me did a very nice recap of the power supply, very nice resoldering work. Looks like a factory job. Replaced the SMDs on the logic board with regular electrolytic.
THe only thing I can figure out is that at least some caps around the +5 and +15 supply need replacing again.
Looks like Tek assumed the power supply would not need servicing. Hard to get in and out and impossible to work on while it's installed. If it breaks, board level replacement. I can't see anyway to troubleshoot the board set.
Has anyone ever hooked up 120V to the boards out of the scope to try to troubleshoot it? Doesn't seem practical or safe.
BTW, I am plowing through the giant eevblog thread, but only up to page 11 (of 68!). Don't know if I can make it.
Thanks,
Grayson















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