Resistors and caps with exaggerated power rating

[LightSoundGeometry]

I thought they gained capacitance over time and started turning into one big resistor eventually if not taken care of ?

I have a bunch of old phillis and they all read twice as high as normal ..same with the allen bradly resistors ..for example all my 470K AB read 500-550K and my 10uf electos all read 17-30 uF

unless I am getting sent sorted junk from some sellers stash

How a capacitors value drifts over time depends on the dielectric type, chemistry, and how it's been used (or abused). In general, an electrolytic will drift high with age, but it may also drift low - this is mostly dependent on the composition of the electrolyte. Paper capacitors also tend to drift high as the oil/wax reacts with the paper/oxidizes and becomes acidic. Ceramic/film/styrene/mica capacitors don't generally drift to any significant degree (in relation to audio circuits, RF and oscillators are another topic entirely) unless they are physically damaged.
As electrolytics drift, their ESR also increases.
For both paper and electrolytic capacitors, as they drift their leakage current will increase. This can be equated to their EPR (equivalent parallel resistance) decreasing in value, so in a sense, they do become smaller and smaller resistors, which allows more and more DC to get through. Since we usually want capacitors to block DC, this is a problem.

Carbon composition resistors (like your Allen Bradley's) are notorious for drifting, both with temperature and applied voltage. This leads some to claim audible effects like compression, but these effects are minuscule compared to even minor drifts from nominal values which can cause circuits to behave differently from their design intentions (which may or may not be subjectively "good"). What is certain is that the circuit will no longer sound like it was designed to.
Carbon composition resistors are already very noisy compared to other technologies, but once they begin drifting, they become even noisier. There's little reason to use them other than aesthetics.

yeop, I quit buying them and tropical fish caps ..pure cosmetics like you said

while we are talking esr ..I picked up one of these for 14 dollars shipped to me in under 7 days

http://www.ebay.com/itm/322011201779?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

worth every penny so far ..LCR-T4 ESR - has stand offs so it can be enclosed .its really nice for the price ..I am measuring fool right now lol ..going to build a holder/case for it soon

anyone want to donate an old O scope to a poor guy who would give it lots of love and use? lol

[greaser_au]

During my days in the service industry, low value-high voltage electros were always a bugbear. Often just replacing all the 0.47/1.0/2.2u63V electros would restore correct operation.

david

[GGBB]

Carbon composition resistors (like your Allen Bradley's) are notorious for drifting, both with temperature and applied voltage. This leads some to claim audible effects like compression, but these effects are minuscule compared to even minor drifts from nominal values which can cause circuits to behave differently from their design intentions (which may or may not be subjectively "good"). What is certain is that the circuit will no longer sound like it was designed to.
Carbon composition resistors are already very noisy compared to other technologies, but once they begin drifting, they become even noisier. There's little reason to use them other than aesthetics.

Interesting. What changes in a carbon comp resistor as it drifts that makes it noisier? Are we merely talking noisier because of higher resistance or is there more to it? For example is a 100k resistor that has drifted to 120k any noisier than a 120k resistor that hasn't drifted (other things being equal)?

[amptramp]

For a resistor, there is a phenomenon known as 1/f noise where you have more noise than would be predicted from the noise based on Boltzmann's constant.  Op amps also show this; where you have an op amp with a noise spec, there is a midband noise and a rising noise level at low frequencies, usually within the audio spectrum.  This excess noise is very small for metal film or wirewound where the attachment of the wire is done properly and larger and starting at a higher frequency for carbon comp.

[GGBB]

I was aware of that stuff in general, but I am more interested in knowing about how carbon comp resistors change as they age, and how they become or what causes them to become noisier. I know that resistance drifts higher, but is the increase in noise merely a factor of the higher resistance or of something else?

[Phoenix]

Interesting. What changes in a carbon comp resistor as it drifts that makes it noisier? Are we merely talking noisier because of higher resistance or is there more to it? For example is a 100k resistor that has drifted to 120k any noisier than a 120k resistor that hasn't drifted (other things being equal)?

Actually, that's a good question, and I've never directly compared it, but I have a feeling that a 100k carbon composition resistor that's drifted to 120k will be noisier than a nominal 120k that measures as such - at least that is what my experience has led me to think. I'm not certain of the reason why this would be the case, but I wonder how much of the drift is due to changes in the carbon/ceramic slug as opposed to degradation of the end-cap bonds. I'd speculate that perhaps the increased noise is flicker/excess noise - which is already caused by fluctuations in resistance - and once a carbon composition resistor begins drifting on top of that, the amplitude of the excess noise increases as a result. We already know that excess noise is different with different manufacturing techniques (even of the same technology), carbon composition resistors can range anywhere from a few hundred nanovolts/volt per decade to a few microvolts/volt per decade, depending on manufacturer and specification, so I think it's safe to assume that if something is happening with the chemistry or physical construction of the resistor that's causing it to drift in resistance outside of specified tolerance, all bets are off as to the original 1/f noise specs.
Remember though, this is conjecture, I don't have any sources for this and I haven't conducted my own experiments on the subject, so excess noise may not get worse with resistance drift as I speculate, but I can tell you with certainty that it most definitely won't get better.

[R.G.]

The actions inside an electrolytic cap are determined by the electrochemistry of the oxide layer of aluminum grown on the outside of one (normal) or both (bipolar) foils.
Thin foils of high purity aluminum are bathed in aqueous chemical baths while being charged by power supplies. The nature of the chemicals and current flow causes aluminum on the surface to combine with oxygen in the bath to form aluminum oxide on the surface of the foil. The thickness of the foil  determines the withstanding voltage. The chemistry of the bath and the voltage/current nature of the power supply determine how fast it forms, and to some extent it's other characteristics. Each maker of electro caps no doubt has their own "secret sauce" that optimizes the profit of sales of these caps, driven in part by warranty costs and reputation as well as by actual electrical specs. The more voltage that is applied at forming and the longer forming takes, the higher the voltage the oxide layer will resist. That's another way of saying its voltage rating increases. Current must be carefully controlled, as too much current makes hot spots that impede the oxide forming a continuous layer.

The current that forms the insulating oxide layer is formed by a DC current. It resists current flow from that direction as its thickness builds up. Reverse the polarity by even a volt or so, and it starts UN-forming back into the conducting electrolyte. As it does this, the oxide gets thinner, perhaps unevenly, some spots conducting more than others. The conduction destroys that spot faster, so holes eventually get eaten in the oxide, hot spots develop - well, you see where that goes. Electro caps can actually be a form of sloppy, short-lived rectifier if you feed them AC, just as copper-oxide plates can.

The capacitance of a cap depends on the physical setup: plate surface area and insulator thickness. Higher voltages need thicker oxide, so you have to add more surface area (that is, more wound up aluminum foil) to get the same capacitance at higher voltage. Higher voltage caps are bigger for the same capacitance.

Even if the cap is not abused by reverse voltage, it slowly UN-forms into the semi-liquid electrolyte chemicals in the can, and over years gets thin spots. The thinner the oxide, the higher the capacitance and the lower the voltage, so old caps get higher capacitance and lower voltage, but also "poison" the electrolyte goo in the cap, which changes the apparent capacitance from the outside.

Re-forming "weak" old caps can be done, but it's lengthy, tricky, and may not form high quality oxides, as the insides of the cap are not the same as the secret forming sauce that was set up to make the oxide first.

I have seen an explanation about carbon comp resistors being noisier because the insides are a slurry of carbon particles suspended in a binder. The percentage of carbon to binder sets the resistance per unity volume. They resist because there is a multitude of contacts of particle to particle through the body. It was opined that the excess noise was caused by the current crowding at the microscopic, even molecular, contact spots between carbon particles. Made a good read.

[LightSoundGeometry]

The actions inside an electrolytic cap are determined by the electrochemistry of the oxide layer of aluminum grown on the outside of one (normal) or both (bipolar) foils.
Thin foils of high purity aluminum are bathed in aqueous chemical baths while being charged by power supplies. The nature of the chemicals and current flow causes aluminum on the surface to combine with oxygen in the bath to form aluminum oxide on the surface of the foil. The thickness of the foil  determines the withstanding voltage. The chemistry of the bath and the voltage/current nature of the power supply determine how fast it forms, and to some extent it's other characteristics. Each maker of electro caps no doubt has their own "secret sauce" that optimizes the profit of sales of these caps, driven in part by warranty costs and reputation as well as by actual electrical specs. The more voltage that is applied at forming and the longer forming takes, the higher the voltage the oxide layer will resist. That's another way of saying its voltage rating increases. Current must be carefully controlled, as too much current makes hot spots that impede the oxide forming a continuous layer.

The current that forms the insulating oxide layer is formed by a DC current. It resists current flow from that direction as its thickness builds up. Reverse the polarity by even a volt or so, and it starts UN-forming back into the conducting electrolyte. As it does this, the oxide gets thinner, perhaps unevenly, some spots conducting more than others. The conduction destroys that spot faster, so holes eventually get eaten in the oxide, hot spots develop - well, you see where that goes. Electro caps can actually be a form of sloppy, short-lived rectifier if you feed them AC, just as copper-oxide plates can.

The capacitance of a cap depends on the physical setup: plate surface area and insulator thickness. Higher voltages need thicker oxide, so you have to add more surface area (that is, more wound up aluminum foil) to get the same capacitance at higher voltage. Higher voltage caps are bigger for the same capacitance.

Even if the cap is not abused by reverse voltage, it slowly UN-forms into the semi-liquid electrolyte chemicals in the can, and over years gets thin spots. The thinner the oxide, the higher the capacitance and the lower the voltage, so old caps get higher capacitance and lower voltage, but also "poison" the electrolyte goo in the cap, which changes the apparent capacitance from the outside.

Re-forming "weak" old caps can be done, but it's lengthy, tricky, and may not form high quality oxides, as the insides of the cap are not the same as the secret forming sauce that was set up to make the oxide first.

I have seen an explanation about carbon comp resistors being noisier because the insides are a slurry of carbon particles suspended in a binder. The percentage of carbon to binder sets the resistance per unity volume. They resist because there is a multitude of contacts of particle to particle through the body. It was opined that the excess noise was caused by the current crowding at the microscopic, even molecular, contact spots between carbon particles. Made a good read.

some of the top tube amp restoration guys, and radio guys, pull the old electros outer layer/shell/exterior off and carefully cover new electros up to make them look time period correct

[greaser_au]

some of the top tube amp restoration guys, and radio guys, pull the old electros outer layer/shell/exterior off and carefully cover new electros up to make them look time period correct

usually they just cut the wires off the can and tack a new electro under the chassis where you can't see it!   (but for some reason stuffing a 276-P box with AA's will get you kicked out of the vintage radio club...)

david

[amptramp]

Here is the definitive article on restuffing old capacitors with new ones:

http://antiqueradio.org/recap.htm

Fortunately, new capacitors tend to be smaller, so they fit quite well.  I just got through an exercise of this sort with a Philco 37-116X radio and I used electrolytic capacitors in series with equalizing resistors in parallel with each of them because it appeared the maximum voltage across the 450 volt capacitors could be as high as 523 volts.  And I still had room to spare.

[AndreasOlausson]

Very useful information!

Thank you very much everyone