Login / Register
 
Resistors: Metal film and Carbon
New Reply
Subscribe
rocksure
Thread Starter
#1
19th September 2012
Old 19th September 2012
  #1
Lives for gear
 
rocksure's Avatar
 
Joined: Jul 2007
Location: New Zealand
Posts: 1,401

Thread Starter
rocksure is offline
Resistors: Metal film and Carbon

I have a question here regarding metal film or carbon resistors. I have a vintage mic preamp module from the 70's that used carbon resitors. I also have a line module that uses basically the same circuit and components as this. I want to convert the line module to a preamp module to match the other one, which means replacing a few resistors. My question is, should I swap the necessary resistors with carbon ones, or is it ok to use metal film ones :ie will it change the sound?
Examples:
220k original was a 1/4 W 5% carbon. I don't have a 1/4 w carbon, so Should I use a 1w carbon 5%, or is a 1/2 W 1% metal film ok? Will it change the sound?
Original 15k 1/4 w 5%, should I replace with carbon 15k 1w 5% carbon, or 15k 1/2 w 1% metal film?
Does it really matter?
__________________
Tony Koretz
http://rocksuresoundz.com/
Production Music and Sound Effects
#2
19th September 2012
Old 19th September 2012
  #2
Gear addict
 
Joined: Aug 2010
Posts: 302

Temple of Light is offline
Going from a carbon film to a metal film or metal oxide type of resistor, I think would be a negligible audio difference, it may be that they both react to heat differently, but not enough to produce audible difference in your design. But you want to go with the same wattage and tolerance on the resistor. For instance, your 220K 1/4 watt 5% carbon should be replaced with a 220K 1/4 watt 5% {or 1%, better} Metal Film resistor.
If you used a 220K 1/4 watt 1% metal film, {likely more available}
that would be an improvement in the accuracy of the performance of the part.
For your 15K 1/4 watt 5%, replace it with a 15K 1/4 watt 5% metal film.
If you can, make sure that both resistors are the same tolerance either both 1% or 5%: this will minimize audio differences in the circuit. Trying to use a 1/2 watt or 1 watt resistor in that circuit wont work because the lead diameters are significantly larger in diameter. If the parts are in a "through hole" circuit board, the same wattage of the originals must be used, not to mention the electrical requirements of the circuit
being thrown off. If anything using metal film resistors as replacements will improve not only the integrity of the electrical design, but may improve the performance of the audio result, IE: either being a slight increase in frequency range, {like a little more high frequency getting through} or perhaps a slight decrease in noise floor, but I doubt you will be able to notice audible difference in the repaired product. YMMV...
rocksure
Thread Starter
#3
19th September 2012
Old 19th September 2012
  #3
Lives for gear
 
rocksure's Avatar
 
Joined: Jul 2007
Location: New Zealand
Posts: 1,401

Thread Starter
rocksure is offline
Quote:
Originally Posted by Temple of Light View Post
Going from a carbon film to a metal film or metal oxide type of resistor, I think would be a negligible audio difference, it may be that they both react to heat differently, but not enough to produce audible difference in your design. But you want to go with the same wattage and tolerance on the resistor. For instance, your 220K 1/4 watt 5% carbon should be replaced with a 220K 1/4 watt 5% {or 1%, better} Metal Film resistor.
If you used a 220K 1/4 watt 1% metal film, {likely more available}
that would be an improvement in the accuracy of the performance of the part.
For your 15K 1/4 watt 5%, replace it with a 15K 1/4 watt 5% metal film.
If you can, make sure that both resistors are the same tolerance either both 1% or 5%: this will minimize audio differences in the circuit. Trying to use a 1/2 watt or 1 watt resistor in that circuit wont work because the lead diameters are significantly larger in diameter. If the parts are in a "through hole" circuit board, the same wattage of the originals must be used, not to mention the electrical requirements of the circuit
being thrown off. If anything using metal film resistors as replacements will improve not only the integrity of the electrical design, but may improve the performance of the audio result, IE: either being a slight increase in frequency range, {like a little more high frequency getting through} or perhaps a slight decrease in noise floor, but I doubt you will be able to notice audible difference in the repaired product. YMMV...
Thank you ......that's a very helpful and informative reply. Much appreciated.
#4
19th September 2012
Old 19th September 2012
  #4
Gear addict
 
Joined: Feb 2012
Posts: 444

Speedskater is offline
One additional specification to be aware of when used in vacuum tube circuits is the resistors maximum voltage rating.
rocksure
Thread Starter
#5
23rd September 2012
Old 23rd September 2012
  #5
Lives for gear
 
rocksure's Avatar
 
Joined: Jul 2007
Location: New Zealand
Posts: 1,401

Thread Starter
rocksure is offline
Another question here:
On a couple of modules there are one or two carbon resistors that have pieces broken out of the casing of the resistor. The resistors don't appear burnt, just broken. Is this likely to have been caused by physical shock to the resistor, or is it possible for this to occur by incorrect voltage or some other electrical reason?
#6
25th September 2012
Old 25th September 2012
  #6
Gear addict
 
Joined: Aug 2010
Posts: 302

Temple of Light is offline
Discoloration is a typical sign that the part was overheated/burnt so if they are just chipped, it's more likely from mechanical shock or just plain old age.
rocksure
Thread Starter
#7
25th September 2012
Old 25th September 2012
  #7
Lives for gear
 
rocksure's Avatar
 
Joined: Jul 2007
Location: New Zealand
Posts: 1,401

Thread Starter
rocksure is offline
Quote:
Originally Posted by Temple of Light View Post
Discoloration is a typical sign that the part was overheated/burnt so if they are just chipped, it's more likely from mechanical shock or just plain old age.
Thanks again fro your help. The resistors are over 30 years old, so age could be a factor.
I hope you will bear with me for one more geeky question on this topic.
Most of the resitors in the circuit are 1/4W carbon, but there are 2 resistors that are 150 ohm 1/2 W 5% carbon. One of them is in the +V rail and the other is in the -V rail at the beginning of the circuit into the preamp module. It happens to be that one of them is the cracked/broken resistor I mentioned previously. So if I were to replace it with a 150 ohm metal film 1% resistor, should I also replace the other rail's carbon resistor with one that is the same type so they are matched rather than leave the carbon one in there, or is it not important?
I notice carbon resistors are often found in power supply circuits in some places rather than metal film ones, even though metal film resistors may be used in other parts of the circuit.........I don't know enough to understand why that is.
#8
25th September 2012
Old 25th September 2012
  #8
Lives for gear
 
Joined: Sep 2004
Location: UK
Posts: 6,410

Matt Syson is offline
Hi
Some of the points above are a trifle 'suspect' in the claims.
Generally speaking Metal film at 1 percent will be 'better' than the original 5 percent carbon for several reasons. The circuit tolerances are improved as for a pair of resistors setting bias for example COULD be 5 percent high and 5 percent low (before the ageing of 30 years) which is pretty wild tolerance but in many respects will be fine.
The 1 percent types will obviously give much closer setting to the designers initial 'theories'.
Second, metal film types tend to be 'quieter' than carbon and from the days when they were expensive, these would be specified when either low noise or specific accuracy was needed.
Carbons can usually withstand 'pulse' loads better than some others although this capability is unlikely to be required in most circuits, certainly not a preamp.
The original designer planned to use certain resistance values, but could not justify tighter than 5 percent tolerance types (bear in mind that 10 percent was probably available!) so using 1 percent types of the design 'nominal' would generally be an improvement although they may not be the 'decider' when you consider transistor hFe range from say 150 - 900 for some 'high gain' varieties and so on, which represent a very wide tolerance.
Changes to the sound comparing one unit to another may be due to your existing unit with 5 percent parts having individual components that are at the limit or even now beyond tolerance. Any actual sonic changes will also depend on exactly what each component is doing in a circuit. Anode or collector load resistors will have some effect, but a cathode or emitter resistor that is bypassed with a capacitor will not.
Matt S
__________________
Matt S
www.mseaudio.co.uk
rocksure
Thread Starter
#9
25th September 2012
Old 25th September 2012
  #9
Lives for gear
 
rocksure's Avatar
 
Joined: Jul 2007
Location: New Zealand
Posts: 1,401

Thread Starter
rocksure is offline
Quote:
Originally Posted by Matt Syson View Post
Hi
Some of the points above are a trifle 'suspect' in the claims.
Generally speaking Metal film at 1 percent will be 'better' than the original 5 percent carbon for several reasons. The circuit tolerances are improved as for a pair of resistors setting bias for example COULD be 5 percent high and 5 percent low (before the ageing of 30 years) which is pretty wild tolerance but in many respects will be fine.
The 1 percent types will obviously give much closer setting to the designers initial 'theories'.
Second, metal film types tend to be 'quieter' than carbon and from the days when they were expensive, these would be specified when either low noise or specific accuracy was needed.
Carbons can usually withstand 'pulse' loads better than some others although this capability is unlikely to be required in most circuits, certainly not a preamp.
The original designer planned to use certain resistance values, but could not justify tighter than 5 percent tolerance types (bear in mind that 10 percent was probably available!) so using 1 percent types of the design 'nominal' would generally be an improvement although they may not be the 'decider' when you consider transistor hFe range from say 150 - 900 for some 'high gain' varieties and so on, which represent a very wide tolerance.
Changes to the sound comparing one unit to another may be due to your existing unit with 5 percent parts having individual components that are at the limit or even now beyond tolerance. Any actual sonic changes will also depend on exactly what each component is doing in a circuit. Anode or collector load resistors will have some effect, but a cathode or emitter resistor that is bypassed with a capacitor will not.
Matt S
Thank you Matt, your input in this thread here (as it has been in other past threads of mine) is much appreciated.
#10
25th September 2012
Old 25th September 2012
  #10
Lives for gear
 
bogosort's Avatar
 
Joined: Jan 2010
Location: Miami, FL
Posts: 799

bogosort is online now
Matt's on point as usual, and I think he was being kind in saying that some of the previous claims are a trifle suspect.
#11
26th September 2012
Old 26th September 2012
  #11
Gear addict
 
Joined: Aug 2010
Posts: 302

Temple of Light is offline
Quote:
Originally Posted by rocksure View Post
Thanks again fro your help. The resistors are over 30 years old, so age could be a factor.
I hope you will bear with me for one more geeky question on this topic.
Most of the resitors in the circuit are 1/4W carbon, but there are 2 resistors that are 150 ohm 1/2 W 5% carbon. One of them is in the +V rail and the other is in the -V rail at the beginning of the circuit into the preamp module. It happens to be that one of them is the cracked/broken resistor I mentioned previously. So if I were to replace it with a 150 ohm metal film 1% resistor, should I also replace the other rail's carbon resistor with one that is the same type so they are matched rather than leave the carbon one in there, or is it not important?
I notice carbon resistors are often found in power supply circuits in some places rather than metal film ones, even though metal film resistors may be used in other parts of the circuit.........I don't know enough to understand why that is.
I would replace both with 150 Ohm 1/2W 1% Metal Film in both the positive and negative rails, doing so negates the effect of having to change parts.
In other words, if changing 1 part in the circuit will have an effect, changing both parts to identical parts is going to have less of an effect.
Which is I think, what you want.
#12
26th September 2012
Old 26th September 2012
  #12
Lives for gear
 
Joined: Sep 2004
Location: UK
Posts: 6,410

Matt Syson is offline
Hi
As long as both 150 Ohm resistors are within 5 percent tolerance the circuit is just fine as it is.
The designers would have used their slide rules to work out whether the circuit will work as required, when all of the components are at both extremes of their tolerance so in this case one being 5 percent up and the other 5 percent down, and vice versa. With the relative low cost of 1 percent parts now, it will be well within the design criteria.
Matt S
Deleted User #208869
#13
27th September 2012
Old 27th September 2012
  #13
Deleted User #208869
Guest
 
Posts: n/a

To add to Matt's sound advice I would recommend replacing any damaged resistors you find - even damaging a resistors outer casing can change its characteristics and tolerances.... Be careful when using pliers to move them about. The 150R mentioned sound like common mode filtering, definitely worth making sure these are OK! You can even up the power rating to be sure - 2W wirewound is fine here.

Personally for signal circuitry I would go for 1% metal film all day long...

Seeing as it is 30 years old, please consider replacing the electrolytic capacitors. You may be alright but that would be my first move with old gear.

Always make sure your ratings and tolerances are within spec. Other than that, happy modding!

Regards

Charlie
#14
27th September 2012
Old 27th September 2012
  #14
Lives for gear
 
Joined: May 2006
Location: Ann Arbor, MI
Posts: 583

digitrax is offline
A few observations on the "sound" of resistors:

Metal films are quieter and cleaner; they pick up less noise and therefore distort less. They are also (typically) designed to work at lower voltages than their carbon equivalents; this is not a consideration unless dealing with the B+ supply of valve gear.

Carbons will be more susceptible to AC noise, which gives them (to some ears) a more "organic" sound. It is also claimed that they are "slower" (i.e. smoother sounding), but I suspect that's the work of the audiophule press trying to justify the greater expense of using them.

As to whether there'll be an audible difference, it will depend on where they are in the circuit. My guess is unless it's in the signal path prior to a bunch of gain, you (and others) are not gonna hear a difference. For reliability's sake, you'd choose the MF. If you want to preserve the "vintage" aspect (i.e. resale value) of the piece, you'd use carbon to remain as close to the originals as possible.
#15
27th September 2012
Old 27th September 2012
  #15
Lives for gear
 
Joined: May 2006
Location: Ann Arbor, MI
Posts: 583

digitrax is offline
Quote:
Originally Posted by Charlie TC View Post
Seeing as it is 30 years old, please consider replacing the electrolytic capacitors.
+1

Although, you may be able to reform the electro's - see article here:
Rap on Replacing Electrolytic Capacitors
rocksure
Thread Starter
#16
28th September 2012
Old 28th September 2012
  #16
Lives for gear
 
rocksure's Avatar
 
Joined: Jul 2007
Location: New Zealand
Posts: 1,401

Thread Starter
rocksure is offline
Quote:
Originally Posted by Charlie TC View Post
To add to Matt's sound advice I would recommend replacing any damaged resistors you find - even damaging a resistors outer casing can change its characteristics and tolerances.... Be careful when using pliers to move them about. The 150R mentioned sound like common mode filtering, definitely worth making sure these are OK! You can even up the power rating to be sure - 2W wirewound is fine here.

Personally for signal circuitry I would go for 1% metal film all day long...

Seeing as it is 30 years old, please consider replacing the electrolytic capacitors. You may be alright but that would be my first move with old gear.

Always make sure your ratings and tolerances are within spec. Other than that, happy modding!

Regards

Charlie
Most of the capacitors are Tantalum, and a few ceramics so those should be fine. There are a small number of electrolytics to replace though as you suggest.
I will replace any resistors that have damaged casings. There are only a couple.
#17
28th September 2012
Old 28th September 2012
  #17
Lives for gear
 
Psychlist1972's Avatar
 
Joined: Sep 2012
Location: Maryland, USA

Psychlist1972 is offline
Quote:
Originally Posted by Temple of Light View Post
Going from a carbon film to a metal film or metal oxide type of resistor, I think would be a negligible audio difference, it may be that they both react to heat differently, but not enough to produce audible difference in your design. But you want to go with the same wattage and tolerance on the resistor. For instance, your 220K 1/4 watt 5% carbon should be replaced with a 220K 1/4 watt 5% {or 1%, better} Metal Film resistor.
Why would replacing with a 1/2 watt (assuming it fits) be an issue? It's overkill for sure, but in electronics, it's perfectly acceptable.

Quote:
For your 15K 1/4 watt 5%, replace it with a 15K 1/4 watt 5% metal film.
If you can, make sure that both resistors are the same tolerance either both 1% or 5%: this will minimize audio differences in the circuit.
This is incorrect. 5% means it varies anywhere within +/- 5% of the stated value. Two 5% resistors could easily be on opposite sides of the spectrum, especially if they're cheap no-name ones. The difference between the two could be as much as 10% of the stated value.

If you replace a 5% resistor with a 1% resistor in one of the circuits, the most variance you'll see between the two is 6% of the stated value.

That said, if they used 5% resistors to begin with, it's not that critical a value. 1%s are pretty cheap, though, so unless you're counting fractions of a penny, I usually go with 1%. It's cheaper for me to keep a stock of 1% resistors rather than keep identical values in 5% and 1%.

Quote:
Trying to use a 1/2 watt or 1 watt resistor in that circuit wont work because the lead diameters are significantly larger in diameter. If the parts are in a "through hole" circuit board, the same wattage of the originals must be used, not to mention the electrical requirements of the circuit
being thrown off.
Ahh. I see why you were concerned about the 1/2 watt ones. 1/2 watt TH resistors will fit in most through-hole boards. Even in 1/4 watt resistors there are two primary diameters with a pretty significant difference between them. If hte 1/2 watt fits (I think the OP said they already had one on hand), there's no reason not to use it.

Pete
__________________
Pete Brown
http://10rem.net | @pete_brown | SoundCloud | YouTube | Flickr | Amazon author profile
I am a developer evangelist/speaker/author at Microsoft, but participate here in a personal capacity. My opinions are my own.
rocksure
Thread Starter
#18
30th September 2012
Old 30th September 2012
  #18
Lives for gear
 
rocksure's Avatar
 
Joined: Jul 2007
Location: New Zealand
Posts: 1,401

Thread Starter
rocksure is offline
Quote:
Originally Posted by Psychlist1972 View Post
Why would replacing with a 1/2 watt (assuming it fits) be an issue? It's overkill for sure, but in electronics, it's perfectly acceptable.




Ahh. I see why you were concerned about the 1/2 watt ones. 1/2 watt TH resistors will fit in most through-hole boards. Even in 1/4 watt resistors there are two primary diameters with a pretty significant difference between them. If hte 1/2 watt fits (I think the OP said they already had one on hand), there's no reason not to use it.

Pete
It's interesting the number of replies and different opinions my original question has gotten.

I have decided to use 5% carbon resistors in all of the 1/4 W places that I need them; just to keep the look of the original circuit as much as possible. However, I don't have any 1/2 W carbon resistors to replace the 150R one I mentioned previously. The casing of this resistor is much larger than a 1/2 watt metal film ( mini-size)one of the same rating. Putting a metal film one in there is going to alter the way it looks, but it's harder to find 1/2 w carbon at this rating. Our electrical supply chain only has metal film in 1/2 W rating....so if I wanted carbon ones I would have to order them online and pay shipping etc, so it's probably not worth the hassle of doing that.
#19
30th September 2012
Old 30th September 2012
  #19
Lives for gear
 
bogosort's Avatar
 
Joined: Jan 2010
Location: Miami, FL
Posts: 799

bogosort is online now
Quote:
Originally Posted by digitrax View Post
A few observations on the "sound" of resistors:

Metal films are quieter and cleaner; they pick up less noise and therefore distort less.
Noise and distortion are two different things. Resistors don't pick up noise, they create it -- Johnson, or thermal, noise is present in the different resistor types equally; it's magnitude is dependent on resistance and temperature. Metal film resistors do have less so-called current noise than equivalent carbon types, though you can reduce the latter's by increasing the resistor's size -- a 1W carbon comp will have less current noise than a 1/2W of the same resistance.

As for resistor distortion in small signal circuits, this would be due to changes in resistance from extreme temperatures or voltages and normally not an issue.

Quote:
Carbons will be more susceptible to AC noise, which gives them (to some ears) a more "organic" sound. It is also claimed that they are "slower" (i.e. smoother sounding), but I suspect that's the work of the audiophule press trying to justify the greater expense of using them.
Again, external noise has nothing to do with a resistor's noise, which is generated internally. As for the idea that carbon-type resistors are slower than metal film types, I agree with you that such a notion is audiofoolery. Resistance is defined as voltage divided by current, and current is the rate of charge per unit of time. Thus if one resistor is "faster" than the other, all it means is that it has less resistance. By definition, a 100R resistor is exactly as fast as every other 100R resistor.

Quote:
As to whether there'll be an audible difference, it will depend on where they are in the circuit. My guess is unless it's in the signal path prior to a bunch of gain, you (and others) are not gonna hear a difference. For reliability's sake, you'd choose the MF. If you want to preserve the "vintage" aspect (i.e. resale value) of the piece, you'd use carbon to remain as close to the originals as possible.
Agreed, though, except for rare esoteric resistors, I doubt the resale value will change one cent by replacing originals with metal film.
#20
4th October 2012
Old 4th October 2012
  #20
Gear addict
 
Joined: Aug 2010
Posts: 302

Temple of Light is offline
Engineer's
#21
7th October 2012
Old 7th October 2012
  #21
Lives for gear
 
Joined: Dec 2011
Location: Cloud Peak on The Karakoram
Posts: 1,359

Frank_Case is offline
Quote:
Originally Posted by rocksure View Post
Resistors: Metal film and Carbon
....
Does it really matter?
Here's a few things you should understand when using specific resistor types in mic preamps.

1) All resistors have slight second order distortion, especially carbon resistors, which is why carbon comp resistors sound warmer than metal film resistors. This distortion figure can be obtained from the resistor manufacturer and is called the Voltage Coefficient (VC) which is expressed in PPM. A typical VC for Metal Film is 1 PPM for each resistor in the signal path which is equivalent to approx .0001% distortion when your preamp is running at a +15dBu output level.

A typical Carbon Comp resistor has a VC of 320 PPM which equates to .034% for each resistor distortion at +15dBu output. So, for every three carbon resistors you place in the circuit, you will introduce approximately .1% even harmonic distortion which is audible. FWIW, higher resistor values have greater VCs but the relationship is not linear.

2) Carbon resistors are substantially noisier than metal film resistors. All resistors have Johnson noise but what I'm talking about here is what is referred to as Excess Resistor Noise (i.e., Current Noise) beyond normal Johnson noise, and this noise is directly proportional to the voltage across the resistor as well as proportional to the resistance value. Without getting into a long drawn out discussion, suffice it to say that current noise of carbon resistors is approximately 100X that of metal film resistors. In other words, don't use carbon resistors unless you want your preamp to hiss.

To give you an idea how much noise can come from a carbon resistor: a 1/4 watt 100K carbon resistor with 10v across it will perhaps have 2uv of noise generated within it which is added to your signal. This is approximately 20 PPM which is -95 dBu which is about .002% hiss. Clearly the hiss from a single carbon resistor is audible. Multiply that by 5 similar resistors and you get approximately .01% total hiss which represents -60 dBu which is ungodly noisy.

3) Metal film resistors are about 10X more stable than carbon comp resistors in terms of temperature stability. More importantly, they are vastly more stable over the lifetime of the resistor, meaning they won't drift much with age.

4) There can be a difference in max voltage between carbon and metal film resistors, but that depends on their physical construction so I can't say much further on that other then to suggest that you check the voltage ratings of whatever resistors you use to make sure they work.

In sum........The obvious choice for your preamp is metal film. I've had good luck with cheap metal films from Xicon as well as more expensive MFs from Rodersteins among others. Both sound good.
rocksure
Thread Starter
#22
7th October 2012
Old 7th October 2012
  #22
Lives for gear
 
rocksure's Avatar
 
Joined: Jul 2007
Location: New Zealand
Posts: 1,401

Thread Starter
rocksure is offline
Aah....more interesting information coming to light. So many variations of information and opinions on resistors. Certainly plenty of food for thought. Whoddathunkit!
#23
7th October 2012
Old 7th October 2012
  #23
Lives for gear
 
dcollins's Avatar
 
Joined: Aug 2003
Location: Hollywood CA
Posts: 3,170

dcollins is offline
Quote:
Originally Posted by bogosort View Post
Again, external noise has nothing to do with a resistor's noise, which is generated internally. As for the idea that carbon-type resistors are slower than metal film types, I agree with you that such a notion is audiofoolery. Resistance is defined as voltage divided by current, and current is the rate of charge per unit of time. Thus if one resistor is "faster" than the other, all it means is that it has less resistance. By definition, a 100R resistor is exactly as fast as every other 100R resistor.
Well, there is one type of resistor that can "pick up" external fields: the wire wound.

As for the other claims of Carbon "speed" etc., I think they are just that. The actual speed factor only comes into play when considering the stray capacitance of the part, and at audio frequencies can be dismissed as it's a fraction of a pF.

A few years ago I was working on a project that used Carbon resistors and was curious about the phenomenon of value change with applied voltage (this is commonly alleged to be an issue with this construction, and I thought it was true) but at least with the Allen-Bradleys I tested I didn't see it happening. I was just proving Ohms law up to the 600V limit of my bench supply.

The other claims about Carbon resistor excess-noise is a question, as tube circuits have hundreds of volts across hundreds of kOhms in cascaded gain stages yet noise in the -80dB range, so I'm not sure if it's as bad as all that.

Yes, metal-film is better in every way, but the differences aren't that dramatic.

If you are working on a classic piece of gear, I would always try to keep it as stock as possible anyway.........

Fwiw, Linear Audio magazine has had a couple good articles of resistors - highly recommended.


DC
__________________
Dave Collins Mastering
www.collinsaudio.com
+1 323 467 5570
#24
7th October 2012
Old 7th October 2012
  #24
Lives for gear
 
bogosort's Avatar
 
Joined: Jan 2010
Location: Miami, FL
Posts: 799

bogosort is online now
Quote:
Originally Posted by Frank_Case View Post
1) All resistors have slight second order distortion, especially carbon resistors, which is why carbon comp resistors sound warmer than metal film resistors.
I agree with most of what you said, but have a hard time believing this. Where would harmonic distortion -- and specifically, asymmetrical 2nd-harmonic distortion -- come from? Resistor VCs apply equally to positive and negative voltages. I could imagine seeing it perhaps in a tube power amp, with a large DC bias across a resistor and large signal voltage swings, but that's a far cry from "all resistors". How would a resistive pad, or an RC filter network, introduce 2nd-harmonic distortion?
#25
7th October 2012
Old 7th October 2012
  #25
Lives for gear
 
bogosort's Avatar
 
Joined: Jan 2010
Location: Miami, FL
Posts: 799

bogosort is online now
Quote:
Originally Posted by dcollins View Post
A few years ago I was working on a project that used Carbon resistors and was curious about the phenomenon of value change with applied voltage (this is commonly alleged to be an issue with this construction, and I thought it was true) but at least with the Allen-Bradleys I tested I didn't see it happening. I was just proving Ohms law up to the 600V limit of my bench supply.
I suspected as much for small signals, but good to know that linearity can be seen even at 600V.
#26
8th October 2012
Old 8th October 2012
  #26
Lives for gear
 
dcollins's Avatar
 
Joined: Aug 2003
Location: Hollywood CA
Posts: 3,170

dcollins is offline
Quote:
Originally Posted by bogosort View Post
I agree with most of what you said, but have a hard time believing this. Where would harmonic distortion -- and specifically, asymmetrical 2nd-harmonic distortion -- come from? Resistor VCs apply equally to positive and negative voltages. I could imagine seeing it perhaps in a tube power amp, with a large DC bias across a resistor and large signal voltage swings, but that's a far cry from "all resistors". How would a resistive pad, or an RC filter network, introduce 2nd-harmonic distortion?
Asymmetry would imply a diode-like action, which obviously doesn't exist in resistors. Or it's so small as to be ignored.

A better question: How would you measure Frank's claim? With your distortion meter that uses resistors.............


DC
#27
3rd April 2013
Old 3rd April 2013
  #27
Gear interested
 
Joined: Apr 2013
Posts: 2

beep11 is offline
Quote:
Originally Posted by rocksure View Post
I have a question here regarding metal film or carbon resistors. I have a vintage mic preamp module from the 70's that used carbon resitors. I also have a line module that uses basically the same circuit and components as this. I want to convert the line module to a preamp module to match the other one, which means replacing a few resistors. My question is, should I swap the necessary resistors with carbon ones, or is it ok to use metal film ones :ie will it change the sound?
Examples:
220k original was a 1/4 W 5% carbon. I don't have a 1/4 w carbon, so Should I use a 1w carbon 5%, or is a 1/2 W 1% metal film ok? Will it change the sound?
Original 15k 1/4 w 5%, should I replace with carbon 15k 1w 5% carbon, or 15k 1/2 w 1% metal film?
Does it really matter?
The metal film resistors feature low noise properties and a high linearity because of a low voltage coefficient. Therefore, if you replace metal film resistors with carbon film resistors there might have noise problems. I think that it doesn't cause problems in your case.
#28
3rd April 2013
Old 3rd April 2013
  #28
Lives for gear
 
JohnRoberts's Avatar
 
Joined: Nov 2006
Location: Hickory, MS
Posts: 2,948

JohnRoberts is offline
Quote:
Originally Posted by dcollins View Post
Asymmetry would imply a diode-like action, which obviously doesn't exist in resistors. Or it's so small as to be ignored.

A better question: How would you measure Frank's claim? With your distortion meter that uses resistors.............


DC
Not to pheed the phoolery, but I have heard about THD in cheap carbon film resistors used at high signal voltage (like in a power amp negative feedback network).

In SMD resistors there seems to be some correlation with higher distortion in extremely small SMD packages.

I wouldn't make broad sweeping generalities about all resistors, but this should be well characterized in individual resistor brand/series specifications.

JR
__________________
John Roberts
www.CircularScience.com
#29
3rd April 2013
Old 3rd April 2013
  #29
Gear nut
 
Joined: Apr 2012
Location: Halifax, NS
Posts: 75

cbc6403 is offline
Quote:
Originally Posted by dcollins View Post

Yes, metal-film is better in every way, but the differences aren't that dramatic.


DC
Except in one very dramatic way. I think the OP mentioned replacing the power supply rail decoupling resistors at one point. If something goes wrong with that circuit (like a bypass tantalum shorting) a carbon film resistor will open up like a fuse, while a metal film will sit there and burn (if the PSU can supply enough current), actually going lower in value as it cooks.

This isn't just theory. I've seen it happen in several Amek 9098 preamps. When the caps shorted, the current draw wasn't enough to blow the line fuse, but it eventually burned out the custom toroidal power transformer, making what would have been a fairly simple repair much more expensive and time consuming. If the designer had used 1/4W carbon film resistors instead of the same value in metal film, the fault would still have occurred, but the resistor would have opened up like a fuse, and it would not have had such dire consequences. The originally 10? resistors had dropped down to about 5?, and stayed that way after they cooled. Whatever special properties are claimed for metal film resistors, I doubt they apply to this simple DC circuit. If you need precision, measure a bunch and select a pair that are within 1% of each other.

John Roberts has already weighed in on this subject in another forum to say that not all metal film resistors behave this way, but the ones that Amek and Crest used for their production certainly do.

Geoff
#30
3rd April 2013
Old 3rd April 2013
  #30
Gear addict
 
Joined: Mar 2013
Posts: 318

Kesh is offline
I don't know if OP is aware of it, but carbon comes in two flavours: carbon comp and carbon film. Never heard much good said about carbon film, but some people seem to like comp. Maybe they like a bit of hiss (I do), or the non linearity when driven near their watt rating. Also supposedly good for grid stoppers because of very low inductance.
New Reply Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+ 
 
Topic:
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
georgia / Post Production forum!
2
audiothings / Remote Possibilities in Acoustic Music & Location Recording
9
nukmusic / Remote Possibilities in Acoustic Music & Location Recording
5
stupid8track / Remote Possibilities in Acoustic Music & Location Recording
7
e-cue / Remote Possibilities in Acoustic Music & Location Recording
14

Forum Jump

SEO by vBSEO ©2011, Crawlability, Inc.