I’ve been using freeze spray to identify resistors that are cold sensitive, but I’d like to be able to test individual components.
In general, to measure resistor 1/f noise, you force a DC current through it and measure the low-frequency noise voltage across it or you force a DC voltage across it and measure the low-frequency noise current through it. (You could also use AC and measure sidebands.)
I guess the method with DC current would be most practical for in-circuit measurement. If the circuit contains semiconductors, you have to keep the current low enough not to turn on any semiconductor junctions.
Somehow you have to ensure that mains hum doesn't mess up the results. You also need a good contact to the resistor, otherwise you might be measuring the noise of the rotten contact to the resistor rather than of the resistor itself.
I think it's far from straightforward. What measuring equipment have you got?
I guess the method with DC current would be most practical for in-circuit measurement. If the circuit contains semiconductors, you have to keep the current low enough not to turn on any semiconductor junctions.
Somehow you have to ensure that mains hum doesn't mess up the results. You also need a good contact to the resistor, otherwise you might be measuring the noise of the rotten contact to the resistor rather than of the resistor itself.
I think it's far from straightforward. What measuring equipment have you got?
Last edited:
I guess the OP first wants to identify the noisy one, instead of random swapping.
I thought he may have narrowed down the noisey one.
If the schematic is available, those in series with the signal, especially if of high value can often be responsible for most noise.
If the schematic is available, those in series with the signal, especially if of high value can often be responsible for most noise.
For completeness, when I wrote post #2, I assumed it was all about crappy resistors with excessive 1/f noise. If it's about thermal noise, that can easiest be determined indirectly: measure the resistance, measure the temperature, calculate the thermal noise.
@JMFahey @batteryman You guys hit the nail on the head. They are through-hole, and so far I’ve been swapping each plausible resistor. I thought there might be better options. I wad just at my service persons shop, he showed me an high-gain speaker used to listen to distortion at a given point.
Your method is flawless to calculate the "normal" noise present there.
But there is "nothing broken" , that resistor is doing its thing, and more important, replacing it by another same value will improve nothing.
Your equation will equally apply to one or the other.
Here I guess we are talking about a broken/damaged resistor, which can be found and replaced.
I would try to find it, (difficult) or at least the gain stage where it resides, focusing the search immensely.
Worst case, you can still "shotgun" it, replacing all resistors in that stage, but now we are talking a handful instead of the whole amp.
A scope can help, also sequentially grounding/blocking stage by stage to find the culprit
When fighting spies or the Resistance in WW2, Germans trying to find a hidden transmitter had direction finding antennas but not with modern precision , so they resorted to cutting power block by block in the suspect area.
When transmitter abruptly stopped, they were within 50-60 meters or less (within a city block).
Same here.
Maybe the OP confuses noise with distortion?
As @JMFahey and @MarcelvdG say, resistor noise depends on resistance and temperature.
Replace one with the same resistance and thermal noise will be the same.
Only 1/f noise differs but that normally only rears its ugly head below 100Hz or so.
Maybe the OP can clarify what it is exctly that he has issues with?
Jan
As @JMFahey and @MarcelvdG say, resistor noise depends on resistance and temperature.
Replace one with the same resistance and thermal noise will be the same.
Only 1/f noise differs but that normally only rears its ugly head below 100Hz or so.
Maybe the OP can clarify what it is exctly that he has issues with?
Jan
I wouldn't be so sure about that '100Hz or so' when it is a carbon composition resistor in poor condition, for example. All we know about the resistor is that it's through-hole mounted.
Anyway, thermal noise only depends on resistance and value, so it can only be way above spec when the resistance has drifted by a huge amount (when it is doubled or halved, that only changes the noise by 3.01 dB) or it is incredibly hot (a temperature change from 20 degrees Celsius to 310 degrees Celsius only causes an increase of about 3 dB). The first is easy to measure and the second is very unlikely.
Hence my assumption that this thread is about 1/f noise of a resistor that's in bad shape.
Anyway, thermal noise only depends on resistance and value, so it can only be way above spec when the resistance has drifted by a huge amount (when it is doubled or halved, that only changes the noise by 3.01 dB) or it is incredibly hot (a temperature change from 20 degrees Celsius to 310 degrees Celsius only causes an increase of about 3 dB). The first is easy to measure and the second is very unlikely.
Hence my assumption that this thread is about 1/f noise of a resistor that's in bad shape.
@jan.didden: I didn't see him use it, he just pulled it off the shelf to show it to me. I'm almost certainly not describing it well. It has a +/- inputs, selectable gain up to ~60 dB, and a speaker. My understanding is he would use it the same way I use my speakers to hear noise/distortion being generated from my amp when it's running. Just more specifically.
@jan.didden > Maybe the OP confuses noise with distortion?
"Big loud sounds interfere with playback, sounds like wind over a microphone, squealing, or fuzz. Level is significant, reaching or exceeding program level."
@jan.didden > Maybe the OP confuses noise with distortion?
"Big loud sounds interfere with playback, sounds like wind over a microphone, squealing, or fuzz. Level is significant, reaching or exceeding program level."
Can you help me understand how to do this? Or, link me to something that explains it? I'm never sure where to place the probe, and more so, where to place the ground.
Ok, please post the schematic or at least other model similar type.
Then I can suggest "ground here and here and here ..." and why and you'll get what I mean.
In general we divide the whole circuit into individual gain blocks and mute them one by one.
We can safely ground most any point which carries signal but is at 0V relative to ground (for example most preamp tube grids, top of volume control, etc.)
IF said point cargues DC , such as a bipolar transistor stage, which will carry bias voltage, we can still audio ground it by using a large-ish capacitor, say 1uF to 10uF but it will pop/thump when connected
Again, better shown on a schematic
Then I can suggest "ground here and here and here ..." and why and you'll get what I mean.
In general we divide the whole circuit into individual gain blocks and mute them one by one.
We can safely ground most any point which carries signal but is at 0V relative to ground (for example most preamp tube grids, top of volume control, etc.)
IF said point cargues DC , such as a bipolar transistor stage, which will carry bias voltage, we can still audio ground it by using a large-ish capacitor, say 1uF to 10uF but it will pop/thump when connected
Again, better shown on a schematic
Oh sure, sorry, I thought I had already. See attached. This is the phono stage of an ARC SP10 preamp. I was highlighting all the resistors that fed the grid, anode, or cathode of one of the tubes, to help me identify what I might need to order. I'm getting bad noise out of the right channel (primarily). Freeze spray produced results in several places.
I also included a photograph of the V1/V2 area of the circuit board. My resistor labels may be wrong in some places... There's no parts on the bottom of the board, just traces.
Attachments
I'm not sure, this unit doesn't have a parts list. Is there a way to tell visually?
The resistors look like a mix of 1% metal film and wirewound power resistors. Sometimes a resistor can fail when the end caps loosen up. This failure mode can happen with both film and wirewound types. Try tapping each resistor with the eraser end of a pencil to smoke out possible offenders. In my brief (very) stint as a repair technician, I encountered one amplifier that had a carbon film part with loose end caps, which caused it to be intermittent. Swapping that particular resistor out fixed the amp, and yes, I found it by tapping it....
That checks out with with else I know. What should I look for when doing the tap trick? If tap and makes noise, bad?
Decades ago a noisy AC-30 was on my bench. It contained loads of the then infamous carbon powder resistors made by vitrohm. These produce spot noise proportional to the DC.voltage drop across them. Resistors with max voltage drop are the 100k~220k anode resistors of all these ECC83 triodes. I replaced just these with good metal film, and the job was done. The difference was clearly audible by my wife working in the kitchen!