Component selection

[eitanwaks]

Hi everybody,
I finally find a schematic that I would like to build. The schematic comes with a part list. The schematic is a very old one (1950s/60s I presume). The part list recommends different types of resistors in different places. Sometimes they recommend Silver Mika at other places orange drops and one wire wound. Same goes for capacitors.

Should I follow these recommendations or can I use common carbon resistors if they have the same specifications?

For the wire wound resistor I presume it would be best to adhere to that specification.

What about capacitors?

Thanks for your help,
Eitan Waks

[alex278]

Everybody knows that you should use silver everywhere. The more esoteric, the better.

[eitanwaks]

Maybe gold, diamond or platinum
but on a more serious note, does it really matter? If so, what should I use?

[kevinkr]

Actually this is a serious question. You didn't elaborate on what you are actually going to build, however resistors in the old days ranged from noisey to well noisier.. We are talking about noise well in excess of the Johnson noise inherent to a given resistance value.

I wouldn't recommend carbon resistors in a phono stage for example, due to excess noise. In the old days there were special low noise sputtered carbons which are no longer available.

In a power amplifier carbons would be fine.

Incidentally most cheap modern carbon film resistors have serious linearity issues and shouldn't be used IMHO.

I would generally select a metal film if stability and low noise are criteria. Good ones come from Roederstein (some types discontinued) draloric, dale/vishay and others.

Note that some MF and many CF types can exhibit high levels of THD if their end terminations are poorly swaged to the ceramic, or if they are oxidized. I had an old lot of 15K RN60 MEPCO Electra resistors that generated 0.15% thd at 1kHz with a couple of volts rms across them in a simple attenuator circuit - they sounded pretty bad actually, far worse than you would expect for this %. Incidentally a "perfect" resistor will introduce no meaningfully measurable distortion.

Metal Oxide resistors should be restricted to power supply circuits, they can produce significant distortion with large voltage swings across them.

Mills makes excellent wirewound resistors in a large range of values, packages and tolerances, both non- inductive and conventional. They sound fine at high currents, but inductive types should be avoided in circuits with high source impedances to avoid frequency response aberations and possible resonances at high (RF) frequencies.

For real carbon resistors consider Allen Bradley as the gold standard, just be aware that initial tolerance and long term drift are not good in carbons. Issues with drift and noise can be reduced by using 1W or 2W resistors in place of 1/2W if there is any significant dissipation through them. I believe these are no longer made, but there are usually plenty of NOS ones out there on eBay and elsewhere.


You can get a lot of these parts online from various sources, but take a look at Michael Percy Audio Consultant http://www.percyaudio.com/ download his catalog. He does sell internationally.

There are a lot of choices in capacitors as well. In general imho film & foil types sound better than their metalized counter parts, but this is not always the case. Jensen PIO sound good, but in low current applications sound "sluggish" to me, probably due to significant DA.
I prefer teflon films for long term stability and low DA. Polystyrene films are good as well. I don't like mylar except in power supply bypass and filter capacitor applications.

In low feedback, no feedback applications the non ideal characteristics of passive components can have an audible and often measurable effect on the performance.

Remember chosing components can be a highly subjective experience if you let it be, if you don't believe in these sorts of things just choose based on basic engineering requirements and cost. Everything above is my opinion - your mileage may and probably will vary..

Good Luck!

[anatech]

Hi kevinkr,
Just a note. I haven't had any linearity problems with metal oxide resistors. They are my preferred component in tube gear. Watch the flash over rating on metal film and metal oxide resistors.

-Chris

[alex278]

The advantage of carbon film is that they're available at higher voltage ratings. Most metal film resistors I've seen stop at 350V.

[kevinkr]

I run all my resistors well inside of their voltage ratings, and I have never heard a metal oxide or carbon film resistor I liked, including some exotic and expensive ones that I could have distributed here in ancient times..

Kevin

[Looneytunes]

Hi Kevin and all.

I remember reading that plain ol' carbon resistors were best in certain locations, but I can't remember why. From reading these posts, I'm getting that the reason is because they are less prone to failure due to excessive voltage. Is this right? Are there certain locations where carbon resistors are usually a better choice than the more "exotic" ones?

I hope that this is not too obvious to everyone but me. I'm a certified NewB.

Thanks

[Miles Prower]

Originally posted by Looneytunes
Hi Kevin and all.

I remember reading that plain ol' carbon resistors were best in certain locations, but I can't remember why. From reading these posts, I'm getting that the reason is because they are less prone to failure due to excessive voltage. Is this right? Are there certain locations where carbon resistors are usually a better choice than the more "exotic" ones?

I hope that this is not too obvious to everyone but me. I'm a certified NewB.

Thanks

"I remember reading that plain ol' carbon resistors were best in certain locations, but I can't remember why."

Carbon comp resistors are pretty much required in RF applications. Metal film resistors, especially the larger values, are internally constructed with the film deposited in a spiral pattern. That forms an inductor. Even if it doesn't exceed a few uH, that's bad news at the top of the ham bands and up. For audio frequencies, it is of no consequence, and metal film is much quieter. (Again, thermal noise falls off with increasing frequency.)

As for "exotic" components, watch out! There are a whole bunch of overblown claims being made for stuff like this. There are plenty of places that'll gladly rip you off for $20.00 capacitors and resistors that you could get from Mouser for a few tens of cents that won't make any noticeable difference.

[kevinkr]

Carbon resistors are preferred where rf is concerned because of their very low inductance.

Hi Miles,
Not to take particular issue with what you are saying as I used to feel and believe exactly the same thing, however I have found that passives and particularly capacitors do produce noticeable colorations in designs that don't use feedback, and that frequently although not always this can be attributed to measurable deficits in the component's characteristics. (Dielectric absorption is quite significant.)

I must be a sucker for preferring the "sonics" of certain rather expensive teflon coupling capacitors after receiving samples of same to evaluate for a review a couple of years ago.

I started off with CDE WMF types I bought cheaply at surplus, noticed after a while that the background seemed to filled with spurious grundge, tried what I thought was the ridiculously expensive rel rtx (now multicap) and thought that an improvement in resolution, and a huge reduction in the grundge, but exhibiting a bright top end I didn't much care for, moved on to the TFT and found that better still (ouch and expensive) - more resolution still and fewer artifacts. The VH caps I currently use I think are the best I have tried.

Being sceptical I did a blind a/b with a group of fellow audiophiles some years ago, built a simple box where I could switch 3 pairs of closely matched caps in and out of the circuit seemlessly and whoa, most of the people could hear differences and after some time were able to reliably identify that I had switched in a different capacitor, to about 90% overall certainty, and all had preferences - not necessarily the same ones either.

Again this is my belief, opinion only, yours and your mileage may vary..

Kevin