Maker Pro
Maker Pro

Replace tantalum caps with aluminum?

W

Walter Harley

Jan 1, 1970
0
I'm replacing the electrolytics in a nearly-40-year-old HP 654A test
oscillator. This is a 10Hz to 10MHz sinewave oscillator. I have the
service manual and schematics for it.

Many of the caps are specified in the service manual to be tantalum. Some
are quite large values, up to 390uF. The caps are all axial (cylinders with
leads at each end), which is a relatively uncommon package these days.

Can I replace these large-value tantalum caps with aluminum caps? I'm
thinking that in 1967, the ESR, tolerance, and stability of aluminum was
pretty lousy, but maybe now it's as good as tantalum was back then?

Thanks for any input.
 
A

Al

Jan 1, 1970
0
Walter Harley said:
I'm replacing the electrolytics in a nearly-40-year-old HP 654A test
oscillator. This is a 10Hz to 10MHz sinewave oscillator. I have the
service manual and schematics for it.

Many of the caps are specified in the service manual to be tantalum. Some
are quite large values, up to 390uF. The caps are all axial (cylinders with
leads at each end), which is a relatively uncommon package these days.

Can I replace these large-value tantalum caps with aluminum caps? I'm
thinking that in 1967, the ESR, tolerance, and stability of aluminum was
pretty lousy, but maybe now it's as good as tantalum was back then?

Check the operating temperatures. The tantalums may have a higher
operating range.

Tantalums are also made from slugs. Aluminum electrolytics tend to be
wound; they may have some peculiar resonances.

Al
 
W

Walter Harley

Jan 1, 1970
0
Al said:
Check the operating temperatures. The tantalums may have a higher
operating range.

It's a low power device; basically sits pretty close to room temp.
Tantalums are also made from slugs. Aluminum electrolytics tend to be
wound; they may have some peculiar resonances.

Hmm. I'll take a look at the mfr spec sheets for the aluminums. Thanks.
 
I

Ian Stirling

Jan 1, 1970
0
In sci.electronics.equipment Walter Harley said:
I'm replacing the electrolytics in a nearly-40-year-old HP 654A test
oscillator. This is a 10Hz to 10MHz sinewave oscillator. I have the
service manual and schematics for it.

Many of the caps are specified in the service manual to be tantalum. Some
are quite large values, up to 390uF. The caps are all axial (cylinders with
leads at each end), which is a relatively uncommon package these days.

Can I replace these large-value tantalum caps with aluminum caps? I'm
thinking that in 1967, the ESR, tolerance, and stability of aluminum was
pretty lousy, but maybe now it's as good as tantalum was back then?

If they are mostly working, you might measure the ESR now, and get an
idea.
 
J

Jim Yanik

Jan 1, 1970
0
It's a low power device; basically sits pretty close to room temp.


Hmm. I'll take a look at the mfr spec sheets for the aluminums. Thanks.

I suspect good low-ESR al. caps would work fine,and be much less expensive.
 
C

Costas Vlachos

Jan 1, 1970
0
Jim Yanik said:
I suspect good low-ESR al. caps would work fine,and be much less expensive.


May also be worth checking out the "Solid Aluminium" caps. These are
different from the aluminium electrolytics. They have very low ESR
(comparable to tantalums) and very long life.

Costas
 
J

Jim Adney

Jan 1, 1970
0
I'm replacing the electrolytics in a nearly-40-year-old HP 654A test
oscillator. This is a 10Hz to 10MHz sinewave oscillator. I have the
service manual and schematics for it.

Many of the caps are specified in the service manual to be tantalum. Some
are quite large values, up to 390uF. The caps are all axial (cylinders with
leads at each end), which is a relatively uncommon package these days.

If these are in the metal cans with the glass seal around the positive
lead, then these tend to be VERY reliable and I would not replace them
with aluminums under any circumstances. They are quite pricey, but if
you should actually have a bad one, you should replace it with another
tantalum.

These are a much better cap than the teardrop tantalums.

I've actually never seen one of these good tantalum caps fail, but
I've increased the reliability of a lot of my gear by replacing failed
(and commonly failing) low voltage alum electrolytics with these. If
you keep your eyes open you can often get good deals on these at swap
meets. I've been picking them up for years and now have a good supply
whenever I need one.

-
 
C

CWatters

Jan 1, 1970
0
Electrolytics can have low lifetime specifications. I once saw a some with a
spec of only 1000 hours. Thats about 41 days! Must have been a mistake?
 
M

Michael A. Terrell

Jan 1, 1970
0
CWatters said:
Electrolytics can have low lifetime specifications. I once saw a some with a
spec of only 1000 hours. Thats about 41 days! Must have been a mistake?

No, just cheap caps. You need to select electrolytics with a higher
temperature rating to get a longer life. In other words, the closer you
run it to its rated temperature, the quicker it fails. Cheap caps are
rated to 85°C, better caps are 125°C. You have to consider the operating
temp of the equipment, plus the heat generated by any AC passing though
the cap. A low ESR cap will generate less heat and last longer. Both
higher temperature grade and low ESR cost more money, so lots of bean
counters use the cheapest piece of crap they can sweep off the floor at
the factory.
--
22 days!


Michael A. Terrell
Central Florida
 
W

Walter Harley

Jan 1, 1970
0
CWatters said:
Electrolytics can have low lifetime specifications. I once saw a some with a
spec of only 1000 hours. Thats about 41 days! Must have been a mistake?

I believe I've read that lifetime doubles for each 10C under the rated temp.
So if they were 1000 hours at 85C, then when run at 25C they'd be good for
64k hours. That's 7.5 years of continuous operation.

So they probably put them in TV motherboards at 35C, and then the TV dies
after 3 years and the owner buys a new one. The cap not only saves you
money, it *earns* you money!
 
J

Jerry G.

Jan 1, 1970
0
This has to be a mistake. Can you think of the hundreds of electro-caps
used in each of every TV set and appliance???

--

Greetings,

Jerry Greenberg GLG Technologies GLG
=========================================
WebPage http://www.zoom-one.com
Electronics http://www.zoom-one.com/electron.htm
=========================================


Electrolytics can have low lifetime specifications. I once saw a some with a
spec of only 1000 hours. Thats about 41 days! Must have been a mistake?
 
J

Jim Yanik

Jan 1, 1970
0
This has to be a mistake. Can you think of the hundreds of electro-caps
used in each of every TV set and appliance???

Look thru the Digi-Key catalog,the list the typical life of some of their
electrolytics.I think that's a worst-case spec though.
 
J

Jeff

Jan 1, 1970
0
Hi Jerry,
No mistake but the rated life is done in hours over time
at a given temp for so much change in value at a max.
If ran that way most would still work longer if the circuit
allows it but it would be a poor choice to use in that case.
Jeff
 
B

Boris Mohar

Jan 1, 1970
0
I believe I've read that lifetime doubles for each 10C under the rated temp.
So if they were 1000 hours at 85C, then when run at 25C they'd be good for
64k hours. That's 7.5 years of continuous operation.

So they probably put them in TV motherboards at 35C, and then the TV dies
after 3 years and the owner buys a new one. The cap not only saves you
money, it *earns* you money!

Only if the customer is stupid enough to buy your brand again.
 
W

Walter Harley

Jan 1, 1970
0
Boris Mohar said:
Only if the customer is stupid enough to buy your brand again.

....Unless the customer figures that all the other brands suck just as much.

Far as I can tell, practically everything out there these days is made to
the same low standards. It all seems to be a race to pinch every last $0.01
out of the retail cost, regardless of whether it halves the useful life of
the device, makes it less usable, or what. Except for "audiophile" audio
gear, which is just as absurd in the opposite direction.

That's part of why I'm fixing the old HP 654A that triggered this thread -
because it was built to be fixed, and to work and last a good long time. I
could buy some new Chinese-crap-factory test oscillator for less than the
time to fix the HP is going to cost me, but I don't wanna.

(Sorry to offend any Chinese engineers out there. I'm sure there are
companies in China making superb, high-quality, reliable, robust products.
Unfortunately, there don't seem to be any USA companies that want to import
them.)
 
B

Boris Mohar

Jan 1, 1970
0
That's part of why I'm fixing the old HP 654A that triggered this thread -
because it was built to be fixed, and to work and last a good long time. I
could buy some new Chinese-crap-factory test oscillator for less than the
time to fix the HP is going to cost me, but I don't wanna.

Are those HP tantalums bad? You would be surprised how long tantalums last if
not abused.
 
W

Walter Harley

Jan 1, 1970
0
Boris Mohar said:
Are those HP tantalums bad? You would be surprised how long tantalums last if
not abused.


This morning I measured their ESR, and they all seemed fine, though one or
two have a bunch of crystallization around the terminals.

Is that the definitive test? Is there a better test? (I suppose I should
also measure their capacitance, which I've not yet done.)

-w
 
C

ctsbillc

Jan 1, 1970
0
In the military world, we treat reliability extremely seriously. We use
Tantalum for most of the high C applications.

There is an attempt to cut the price of military electronics by using
commercial off the shelf components (COTS), but the reliability is not
there. Apart from electrolytics drying out, we have issues with plastic
packaged ICs which are hygroscopic, and even the in-built tendency of
semiconductors to die after a few year due to diffusion of the metalization
through the Silicon, and the migration of the metal across the surface of
the die.

In summary, the electrolytics are just one of the components which, for
consumer level electronics, tend to die more often than we would like, but
believe me, you do not want to incur the cost of going high reliability on
your next television.

Cheers,

Bill
 
Top