Rank Inexpensive Great Quality Metal Film Resistors

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
madisonears said:
You probably can't hear one resistor unless it's really awful or incredibly good, but you can most likely hear 40 or 140 resistors in a single circuit. You might like it, or you might not. For those of us who build one or two at a time, we have the chance to combine various flavors to obtain a more savory, and probably closer to neutral, blend.
Audio tends to fall into two camps: high open-loop gain, high feedback; simple circuits with little or no feedback (apart from followers and triodes). In the first case the performance is determined by the feedback resistors, so the other 38 or 138 will have little effect. In the second case there might be only 4 or 5 resistors which have a direct effect on the sound, and even then they will be swamped by active device behaviour.

If resistors have random non-linearities, then mixing them will lead to cancellation at low orders (due to the Central Limit Theorem) but some higher order stuff will be unavoidable (although at a very low level). If, as is more likely, resistors have family characteristics then it is more likely that they all pull in the same direction so you could end up with pleasant distortion, albeit at a fairly low level (except for carbon comp - which has significant distortion which is why some people prefer it for guitar amps).
 
Perhaps I could challenge PauloPT to do a double blind test to identify the "sonic signature" of each resistor manufacturer, including whether the steel used in the leads has an audible component.
Incidentally I have never seen reference to such by a manufacturer.
The outcome may help to moderate some of the pretentious drivel written about passive components used in audio.
 
Different resistors do sound different. And they sound different because they measure different. Of course, it depends where they are located. And if you use only one brand/type of resistor all across the amp the differences will be bigger and you might not like the result. I could do a controlled ABX test and spot the differences without a problem. I prefer to use industrial type resistors - like the Daless RN/CFM series - to use some takmans or shinkoh because the industrial ones are not designed with some sound signature in mind. They're more neutral.

Measure different? So a 1k resistor in a feedback network will sound different than a 10k resistor? What a surprise. :D. A few years later and still no proof with ABX testing? Not a surprise.
 
If people would take time and read some of the vintage articles about tube amp building they would clearly understand that the designers/builders always stressed using the newest/freshest components available during their period. Just so happened that Carbon Composition was the newest technology in electronics, so that is what they used. Even then they stressed using tighter tolerances in certain parts of the amp (1%-2%) as opposed to looser spec'd tolerances in other parts of the amp build (5%-10%-20%).

Can you make a signal behave differently by placed different spec'd components in its path? I would think so, but like DF96 elegantly explained far better than a non-technical person such as myself... there are laws/science that apply to tube topology that stand hard and fast no matter what brand/type of resistor is used.

Many, many blind tests have proven that when you take component mixing out of the hands of a listener and just tell them you changed something... it is very hard for them to discern any change in the music. If you have a new type resistor(s) in your hand and you start desoldering/resoldering them into your circuit with the plans to see if they make a difference in your 'sound'... guess what? You have predisposed your psyche to expect a result... either better/same/worse.

Laws of electronics and audiology have not changed much since they were discovered. Neither has snake oil or their peddlers.

Cheers
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.