[Jbucla2005]

I made a passive pre amp (pot in a box) using an Alps Blue Velvet 100k ohm potentiometer. It works but should I have used a 50k ohm pot instead? Thanks.

 

[Tjj226 Angel]

It all depends on what your source and amp are.

If your source is capacitively coupled with really low output impedance, then you want your volume pot to be pretty low in value to avoid high frequency roll off. Some people go as low as 10K.

If you don't feel like your high frequency response has suffered, then I wouldn't worry about it.

 

[tomb]

In our headphone culture, it's fairly easy to tell the difference between a 100K pot and 50K one. The 50K pot will sound cleaner. Going to 10K results in a very real chance of bass roll off, given the loading and typical output capacitors used on sources. That's why - especially seen in the DIY history of headphone amplifiers here and elsewhere - 50K volume pots seem to be a happy medium. 10K pots are used in portable applications, where attenuation of bass frequencies are often less noticeable.

 

[tomb]

Just to be clear, the phenomenon of filtering frequencies with connecting audio equipment occurs via a high-pass filter (AFAIK). IOW, the low frequencies have the potential of being cut, not the high frequencies. This is due to the RC circuit that is formed by the connected load (R) and the often-used capacitor to either filter out DC on the input of the amplifier, or filter out DC on the output of the source. Additionally, depending on the amplifier design, you can have a bass frequency cutoff occur at the headphone, where the R of the connected headphone combines with the C of the capacitor in the amplifier used to block DC. This is perfectly represented in the Bottlehead Crack, where two (one for each channel) electrolytic capacitors are used to block the DC from the tube. In the case of the Bottlehead Crack, 300 ohm Sennheisers are most often used. The 300 ohm value is high enough that the cutoff frequency of the RC circuit does not appreciably block audible bass frequencies from the Sennheiser headphone. You might notice a loss of bass with a 32-ohm headphone, however. This is not just with a Bottlehead Crack, the condition exists with any OTL tube headphone amplifier and many other amplifiers as well. Here's an example of several capacitor/headphone impedance calculations I did for bass drop:

As best I can remember, this represents a Bottlehead Crack, where two 220uf electrolytic capacitors are used (one for each channel) to block DC output from the vacuum tube. The purple is a 16 ohm IEM, whereas the red is a 300 ohm Sennheiser. You can see that the Sennheiser loses almost no bass - down about 0.25dB below 20Hz. However, the 16 ohm IEM has lost a full 3dB at about 45Hz. Physical issues come into play here as well. 220uf is about 10 times more uf than even some of the largest film capacitors available. Film capacitors sound about 10 times better, though.

Thankfully, at the input side of the amplifier, the volume pots that provide the "R" in the RC equation are much, much higher - varying, as noted before, from 10,000R to 100,000R. This is in comparison to 16 to 300 R for headphones. So, much smaller uf film capacitors can be used, resulting in much higher sound quality at the input of an amp - or the output of a DAC.


The RC circuit that results in a high-pass filter (bass frequencies are cut) is shown below from Wikipedia:

For the OP of this thread, R is the input impedance (resistance) of the amplifier. In 99% of all designs, that R is the impedance of the volume pot. (Remember that there are two channels, though.) C is usually a capacitor blocking DC output on the source. If you research many designs for DACs, you'll find capacitors that block DC on the output. (Amplified DC will blow up speakers and headphones.) With most commercial equipment, that capacitor is also in the amplifier, usually directly after the RCA jacks. The famous CMoy headphone amp uses DC-blocking capacitors on the input, for instance.

What does this mean? It means that the most common method of connecting audio equipment - either source to amp, or amp to speakers/headphones, results in a high-pass RC circuit. The danger is loss of bass frequencies if the C or R is not high enough. That is why 10K pots are not usually desired, but are OK for small portable equipment, where loss of bass frequencies is not as noticeable. 50K is a high enough resistance (impedance) to prevent losing any bass with most reasonably-sized capacitors. 100K is even better, but as noted in my earlier post, the extra noise resulting from the higher resistance is noticeable, at least with headphones - maybe not so much with speakers.

A low-pass circuit, where high frequencies are in danger of being cut, is an RC circuit as shown below (again, from Wikipedia):

In this circuit, the C is actually connected as the "load" and R is in series with the current flow, but is divided between the output and the capacitor. As far as I know, it is not really analogous to anything we normally do in connecting audio equipment. I think the only time this is seen is when there is actual need to filter out very high-frequencies (MHz) that can cause instabilities in DAC chips and opamps, or when defining the upper bounds of digital recording processes.

 

[Tjj226 Angel]

Just to be clear, the phenomenon of filtering frequencies with connecting audio equipment occurs via a high-pass filter (AFAIK). IOW, the low frequencies have the potential of being cut, not the high frequencies. This is due to the RC circuit that is formed by the connected load (R) and the often-used capacitor to either filter out DC on the input of the amplifier, or filter out DC on the output of the source. Additionally, depending on the amplifier design, you can have a bass frequency cutoff occur at the headphone, where the R of the connected headphone combines with the C of the capacitor in the amplifier used to block DC. This is perfectly represented in the Bottlehead Crack, where two (one for each channel) electrolytic capacitors are used to block the DC from the tube. In the case of the Bottlehead Crack, 300 ohm Sennheisers are most often used. The 300 ohm value is high enough that the cutoff frequency of the RC circuit does not appreciably block audible bass frequencies from the Sennheiser headphone. You might notice a loss of bass with a 32-ohm headphone, however. This is not just with a Bottlehead Crack, the condition exists with any OTL tube headphone amplifier and many other amplifiers as well. Here's an example of several capacitor/headphone impedance calculations I did for bass drop:

As best I can remember, this represents a Bottlehead Crack, where two 220uf electrolytic capacitors are used (one for each channel) to block DC output from the vacuum tube. The purple is a 16 ohm IEM, whereas the red is a 300 ohm Sennheiser. You can see that the Sennheiser loses almost no bass - down about 0.25dB below 20Hz. However, the 16 ohm IEM has lost a full 3dB at about 45Hz. Physical issues come into play here as well. 220uf is about 10 times more uf than even some of the largest film capacitors available. Film capacitors sound about 10 times better, though.

Thankfully, at the input side of the amplifier, the volume pots that provide the "R" in the RC equation are much, much higher - varying, as noted before, from 10,000R to 100,000R. This is in comparison to 16 to 300 R for headphones. So, much smaller uf film capacitors can be used, resulting in much higher sound quality at the input of an amp - or the output of a DAC.


The RC circuit that results in a high-pass filter (bass frequencies are cut) is shown below from Wikipedia:


For the OP of this thread, R is the input impedance (resistance) of the amplifier. In 99% of all designs, that R is the impedance of the volume pot. (Remember that there are two channels, though.) C is usually a capacitor blocking DC output on the source. If you research many designs for DACs, you'll find capacitors that block DC on the output. (Amplified DC will blow up speakers and headphones.) With most commercial equipment, that capacitor is also in the amplifier, usually directly after the RCA jacks. The famous CMoy headphone amp uses DC-blocking capacitors on the input, for instance.

What does this mean? It means that the most common method of connecting audio equipment - either source to amp, or amp to speakers/headphones, results in a high-pass RC circuit. The danger is loss of bass frequencies if the C or R is not high enough. That is why 10K pots are not usually desired, but are OK for small portable equipment, where loss of bass frequencies is not as noticeable. 50K is a high enough resistance (impedance) to prevent losing any bass with most reasonably-sized capacitors. 100K is even better, but as noted in my earlier post, the extra noise resulting from the higher resistance is noticeable, at least with headphones - maybe not so much with speakers.

A low-pass circuit, where high frequencies are in danger of being cut, is an RC circuit as shown below (again, from Wikipedia):


In this circuit, the C is actually connected as the "load" and R is in series with the current flow, but is divided between the output and the capacitor. As far as I know, it is not really analogous to anything we normally do in connecting audio equipment. I think the only time this is seen is when there is actual need to filter out very high-frequencies (MHz) that can cause instabilities in DAC chips and opamps, or when defining the upper bounds of digital recording processes.

I think the idea is that a given power amp will have X amount of input capacitance. The input capacitance in combination with the volume pot is what forms the low pass filter.

This is why you always want to design an active preamp with low output impedance.

There is something about a volume control back loading your source and causing low frequency to suffer, but I think that is mainly an issue with things like CD players and other low output sources. Hopefully I am remembering everything correctly.

 

[Jbucla2005]

I changed the 100k for a 50k and it sounds MUCH better. Not distant/faded/bassy like it was before.