Using 1% resistors in an amp as apposed to various tolerance resistors

[Luxman1050]

Cheers Craig

[Radio Wrangler]

Originally Posted by Luxman1050

I think carbon film are less noisy these days but the trend seems to be metal film. So which is best?

They're a bit better, mostly due to better process control and choice of materials in the termination areas, but this is small compared to the difference between carbon film and metal film. The difference between carbon composition and carbon film is also a big step.

There is a theoretical noise floor of any resistor, set by quantum mechanics and the Maxwell-Boltzmann constant. Metal film resistors are very close to this limit.

Other resistors produce what's called "excess noise" from any DC flowing in them.

Watch out for thick-film surface mount resistors. Some are Tantalum Nitride, others are Ruthenium Oxide. These things can be 20dB noisier than metal film equivalents. This can really show up in resistors in the tuning lines og PLLs and in low noise amplifiers at audio or RF.

In a previous life I was behind the Agilent/HP Noise figure analysers and designed the smart noise sources to go with them. There was a lot of work went into choosing national standards labs for traceability to their thermal noise standards for calibration.

Hefty carbon resistors are good for RF loads where you want minimum stray reactance and brute power handling, so they do have a use. The only other benefit they have is in the minds of some audio peple who eschew scientific method.

David

[Luxman1050]

Thanks for info David. I think I'm going to struggle and find all the values I need on this one the only option would be to go down the route of series or parallel to get the rightist value. Seems to be defeating the objective. I did come across one brand but at £15.00 each yes each my mouth dropped at that one. Who on earth would payout that much per resistor? Anyhow I'll keep trawling the Web.

[factory]

If replacing all carbon composite resistors in test gear, which some folk do (not me) you can find certain circuits where a spiral cut film resistor, causes a fault that didn't exist before.

David

[Radio Wrangler]

For those circumstances we bought 'non inductive' metal film resistors.

But any resistor with length must have some inductance, it just meant there was no more than the minimum.

For ultra-critical circuits, such as making attenuators to microwave frequencies we arranged grounded shrouding shaped around the resistor so that it added distributed capacitance to suit the distributed inductance and form a transmission line. Shaped to match the resistive impedance down the length of the resistor. Looked wild but we got DC to microwave performance. Carbon comp has ageing problems so it's no solution for serious work You might get things reasonable to start with, but it doesn't last.

David

[kalee20]

The thread is moving off the 1%-ness and onto general resistor technology!

If it's an AF amplifier then inductance of a spiral-cut path won't matter - this starts to kick in at tens of MHz. But David has much more experience here.

Voltage rating of resistors does start to matter, as has been mentioned - here I'm going to display ignorance and ask why a 1 megohm 1/4W resistor might have a voltage rating of 250V (say) when this voltage subjects the resistor to only 25% of its power capability. Yet, over-volting a metal-film resistor can result in it going open-circuit. Hopefully someone can tell us what the failure mechanism is!

Originally Posted by Radio Wrangler

Hefty carbon resistors are good for RF loads where you want minimum stray reactance and brute power handling, so they do have a use. The only other benefit they have is in the minds of some audio peple who eschew scientific method.

They do have one other virtue - because power is dissipated uniformly in the bulk of the material, there are no hot-spots. They have the best surge withstand capability of any resistor technology. (That's if you can accept the mediocre tempco, the aging, the non-linearity, the excess noise, the wide tolerance...)

[GrimJosef]

Originally Posted by kalee20

... I'm going to display ignorance and ask why a 1 megohm 1/4W resistor might have a voltage rating of 250V (say) when this voltage subjects the resistor to only 25% of its power capability. Yet, over-volting a metal-film resistor can result in it going open-circuit ...)

Each turn of the spiral resistive path will be separated from its neighbours by a thin insulating gap. I'd always assumed that the failure mechanism was electrical breakdown, probably at the interface between the resistor substrate and any insulating coating, of that gap. If one gap were to go short-circuit then the voltage across the whole resistor would be redistributed between the remaining gaps, raising the voltage across each one and increasing the chances of another gap failing.

If the source impedance of the applied voltage is relatively low then the falling impedance will lead to a rising current and it may be that the power dissipated becomes large enough to blow the resistive material away. That would give us an open circuit and the resistor would then be completely ruined.

This explanation works for me, but I'm not a resistor expert, so if anyone knows better then I'd be interested to hear.

Cheers,

GJ

[G.Castle]

I've just seen this thread:

The question that the OP put can be answered as, yes 1% Metal film are perfectly acceptable for all the learned reasons that have been posted subsequently. 5, 10, or, 20% components would have been used originally because of availability and cost. No quality manufacturer would select components for specific values from an inferior range of components to use in a precision duty, the only example of this I've seen was easy to identify is where a padding resistor is needed to supplement another where the greater tolerance will have little effect on the combination.

Manufacturing of components is now an exact science and one doesn't find strange tolerance spreads in modern components as they once had due to manufacturer selection.

Good quality resistors will normally be bang on, capacitor manufacture is different though, and if the range you are using has a wide tolerance, it is most often found that the actual value of the component will be on the low tolerance boundary. They can be made accurately, but the manufacturer can also save money on materials by making them accurately, but only just within tolerance.

I do have one question for the OP, why would he be considering replacing all the resistors in an amplifier? Are they all out of tolerance?

Regards,
Greg.

[Radio Wrangler]

I seem to remember that in our bulk quantities, we were paying ONE US cent per resistor for 1/4 watt 1% metal film parts in the 1980s.

Carbon film resistors are also often spiral cut.

Melf resistors have useful surge ratings.

RF work at high frequencies is usually low impedance and this militates for lower inductance resistors. When you get to really high frequencies, resistors become exotic fan shapes on substrates. The idea is to treat them as broadband transmission lines.

David

[Craig Sawyers]

Microwave resistors are more common now with 5G and other exotic high frequency applications. Even Vishay make flip chip microwave resistors that are good to 60GHz.

Of course you are into exotic circuit board materials and strip lines when you are are at those frequencies.

Craig

[Luxman1050]

Well in the end I went with ebay only place I could find the values required plus they're all the same make.
Some interesting info there to digest for sure.
I'll take voltage/current measurements before and after as I go be interesting.
So there could also be a slight difference in musicality using these resistors or would it not be noticeable?

Chris

[GrimJosef]

You might have to define 'musicality'. (Good luck with that - the best I could find was 'having a pleasant sound'. But then you have to say what you mean by pleasant. My wife thinks the taste of Brussels sprouts is pleasant.)

Cheers,

GJ

[Luxman1050]

Ha ha 😂
More clarity in the highs.
I just read also that metal film has polarity? And to watch inductance.

Chris

[GrimJosef]

More clarity than the original music had ?

I can see why that might be attractive if, like me, you're old enough that your ears have lost much of the high-frequency response they had when they first heard Minnie Riperton.

But I think I'd sooner have a power amp that reproduced the audio signal exactly the way the studio people laid it down - no more of anything, no less of anything. Then if I feel I need to alter the music I can do it using the controls on my pre-amp. Just my two penn'orth of course.

Getting back on topic, I wouldn't expect metal film resistors, in and of themselves, to change the amp's frequency behaviour at all. They're as near perfect resistors as we can reasonably get. I'd expect them to be beautifully neutral.

Cheers,

GJ

[Station X]

Originally Posted by Luxman1050

I just read also that metal film has polarity?

Chris

Where did you read that?

[Sideband]

You should read the 'audiophoolery' thread. When I see complete nonsense about polarised resistors and cables.......!!!

[Luxman1050]

Well I did wonder! Just reading about resistors on the Web I'll see what site it was.

Chris

[Luxman1050]

Called electronics shack and I quote
"Do metal film resistors have polarity?
Metal film resistors have polarity. This means that there is a (+) or (-) sign near the terminals on the resistor. When connecting metal film resistors in an electronic circuit, it is important to make sure that the correct polarity is maintained."

[Wellertcp]

Did you call them on the 1st April

[Radio Wrangler]

There is every opinion imaginable on the internet in one place or another. Plus, there are also opinions you probably couldn't imagine.

If you use the internet as a source of information, you have to be selective in trerms of who you believe.

Ideally, you need to favour sources where what they say can produce repeatable results and that other people can repeat these experiments getting similar results. Having an ability to detect the same effects with unliving, unthinking test gear is another plus in that it avoisd psychological effects whether intended or not.

The audiophoolery thread on here is a bit of a laugh (or it makes you cry at some of the things) and it's a safety valve to let people let off steam.

There are people trying to sell you twaddle to make money off of you. There are people trying to convince you of twaddle and get you to buy in to make theselves feel better about their purchases.

There are also quietly sensible people with good views of what is going on. The loonies make more fuss and noise, though.

It does get quite difficult to avoid getting sucked into all sorts of strange beliefs, and sometimes thereby get suckered.

David