On Fri, 03 Feb 2017 16:38:26 -0600, wrote:

On Thu, 02 Feb 2017 18:36:08 -0800, Jeff Liebermann
wrote:

And what are these newer ones made from?

For electrolytics, try polymer caps:
https://en.wikipedia.org/wiki/Polymer_capacitor#Lifetime.2C_service_life
http://www.mouser.com/pdfdocs/Panasonic_Capacitors_WP_final.PDF

So, what are the BEST ones? For example, I am getting a Hallicrafters
sx-99 radio, to recap it, what should I use for the small caps (not
electrolytics)? Should I use the "orange drops", or is there something
better? I'd rather spend a few bucks more and get the best.

I'll defer to the those from the antique radio forum, who have more
experience with this than me.

I have helped various friends rebuild old HF radios. I tend to
replace parts involving RF with parts that have the same temperature
coefficient. So, mica caps get replaced with silver mica caps.
Ceramics get replaced by ceramics of approximately the same value,
voltage, and tempco. Film caps are potentially a problem, but I've
seen few of those in old tube radios. Bumble bee, black beauty, and
orange drop caps are junk. There's no temperature coefficient
involved with those, so just whatever I can find that has similar
values in polycarbonate or polypropylene. This should help:
http://www.audioasylum.com/forums/set/messages/7/75075.html

By the way, why are all the caps now rated at some oddball figure.
For example, instead of .05, it's ,047? or instead of .003 it's .0033.

Get out your calculator and estimate the acceptable range of values
based on the tolerance. For example, a 0.047uF +/- 10% cap can vary
between 0.042 and 0.052uF. Therefore, a 0.050uF is just fine.
Actually, if you measure the caps involved with an LRC meter, you'll
find that they vary well outside of the original specifications. Just
because a capacitor is marked +/-10% doesn't mean that the circuit
requires the same tolerance range.

Same for the 'lytics instead of 30mf, they are 33mf and so on.....

Most cheapo electrolytics are +100% -20%. So for 30uF, anything
between 60uF and 24uF should work. Again, electrolytics are not that
critical (or very accurate). Occasionally, there will be an
electrolytic in some kind of audio filter that has to be fairly
accurate. You can use the schematic to find those. The rest (power
supply, decoupling, interstage audio coupling, cathode bypass, etc),
can be fairly loose with the tolerances.

Not all ceramics are that reliable. MLCC (multi-layer ceramic caps)
are rather fragile and microphonic.

What do those look like? Are they the ones with colored dots that look
like dominos? (But I think those are mica caps, if I'm not mistaken).

You won't find any MLCC caps in tube radios. You'll find them in
computah equipment in the form of large SMT chip caps, sometimes with
leads and dipped in epoxy:
https://www.google.com/search?q=mlcc+capacitor&tbm=isch
The caps are evil and fragile. They vary in capacitance with
mechanical pressure and make a tolerable capacitance microphone. Touch
one end, but not the other, with a soldering iron, and the temperature
differential will cause internal cracking and eventually a short.

I remember those squarish brown ones with the leads on the bottom, those
were supposed to be superior. (Silver mica, maybe?)

Silver mica. They're very good parts and rarely die unless you cram
too much power through them as in a transmitter.
https://www.google.com/search?tbm=isch&q=silver+mica+capacitors
Watch out for the tempco on those. It's printed on the case as NPO,
N750, N1500, Y5V, Y5P, etc.

And the round ceramics were said to be good too.

Those are called "disc ceramic" capacitors:
https://www.google.com/search?tbm=isch&q=disc+ceramic+capacitor
Same warnings are silver mica.

A for electrolytic caps, it seems that the newer ones have a much
shorter life than the old ones did.

Nope. The old ones filtered at 120 Hz. The new caps filter at 100 to
300 KHz. Internal dissipation follows frequency.

Can you explain that. I dont understand...
(I would think that a 'lytic in a power supply would only need to filter
at 120hz, or would some filter at 60hz also, depending on the
configuration?

Well, ok. I don't know of any tube radios that use switching power
supplies, so yes, the highest frequency a tube radio power supply will
see is 120 Hz. However, you comment was "for electrolytic caps, it
seems that the newer ones have a much shorter life than the old ones
did". By "newer one's", I assumed newer capacitors in newer circuits,
namely switching power supplies. My guess(tm) is that newer
capacitors will have the same long life as the originals (20+ years)
when used as a replacement in a 120 Hz power supply.

I guess going back to the tube stuff makes me feel young again!

Sigh. My collection of test equipment looks more like a museum than a
working test bench. I also find myself fixing 40+ years old test
equipment as I do fixing radios etc. Looking at the old stuff, all it
does is make me feel old and tired.

Good luck.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558