Skorpio
Are you intending on putting this attenuator before or after the buffer (ie at the input or the output of the buffer)?
If you are putting it at the output then I agree with Andrew's choice of 10k.
If you are putting it at the input then it is worth experimenting with higher values.
Are you intending on putting this attenuator before or after the buffer (ie at the input or the output of the buffer)?
If you are putting it at the output then I agree with Andrew's choice of 10k.
If you are putting it at the input then it is worth experimenting with higher values.
Last edited:
There is no point in locating a buffer at the input of anything.
The buffer is the low source impedance that is necessary to drive cables and capacitance at the receiver's input.
The buffers can be located at the output of a substandard source (eg CD player) and/or at the output of an attenuator.
A 100k pot with 15pF to 20pF after it will have an RC timeconstant of around 0.4us to 0.5us.
That is the standard treble filter that some decerning listeners demand.
There is not much leeway for cable, parasitic and receiver capacitance in that 15pF to 20pF allowance using a 100k pot in front of the b1.
However a 10k pot allows 150pF to 200pF of total load capacitance for that 0.4us to 0.5uS roll off.
The buffer is the low source impedance that is necessary to drive cables and capacitance at the receiver's input.
The buffers can be located at the output of a substandard source (eg CD player) and/or at the output of an attenuator.
A 100k pot with 15pF to 20pF after it will have an RC timeconstant of around 0.4us to 0.5us.
That is the standard treble filter that some decerning listeners demand.
There is not much leeway for cable, parasitic and receiver capacitance in that 15pF to 20pF allowance using a 100k pot in front of the b1.
However a 10k pot allows 150pF to 200pF of total load capacitance for that 0.4us to 0.5uS roll off.
Last edited:
Thanks for further clarifying your point of view.
What about the capacitance in the rca cable which precedes the buffer.
Will this become an rcr filter eg output impedance of cd player then capacitance in cable, then input impedance of buffer.
How do you calculate the -3dB point for rcr?
What about the capacitance in the rca cable which precedes the buffer.
Will this become an rcr filter eg output impedance of cd player then capacitance in cable, then input impedance of buffer.
How do you calculate the -3dB point for rcr?
Last edited:
My system vil be:
CD-player -> 80 cm cable -> Input selector -> Relay volume (10/50/100k input resistance) -> B1 buffer -> 80 cm cable -> power amp 10k input resistance....
Between relay att. output and B1 input will be few inches of wire (inside same cabinet)
Perhaps 50k will be good compromise? What is benefit from 50k to 100k?
CD-player -> 80 cm cable -> Input selector -> Relay volume (10/50/100k input resistance) -> B1 buffer -> 80 cm cable -> power amp 10k input resistance....
Between relay att. output and B1 input will be few inches of wire (inside same cabinet)
Perhaps 50k will be good compromise? What is benefit from 50k to 100k?
Yes, each cable and any other capacitance in that section will create a Low Pass filter with the preceeding source resistance/impedance.
If you want a 0.3us rolloff for the RF filter (Low Pass) then with Rs = 200r you can have or add 1n5F. if the parasitic and input and cable capacitance came to 100pF you would need to add a further 1n4F to get the RF filter into the right frequency range.
If the CD player or other source had Rs = 1000ohm then the RF filter capacitance would need to be ~300pF. For this combination the added capacitance would be ~ 200pF to give the 0.3us RF filter.
F = 1 / 2 / Pi / R / C
rearrange to find C
C = 1 / 2 / Pi / R / F
Where:
C is in Farads
R is in ohms
F is in Hertz
Pi = 3.14159
If you want a 0.3us rolloff for the RF filter (Low Pass) then with Rs = 200r you can have or add 1n5F. if the parasitic and input and cable capacitance came to 100pF you would need to add a further 1n4F to get the RF filter into the right frequency range.
If the CD player or other source had Rs = 1000ohm then the RF filter capacitance would need to be ~300pF. For this combination the added capacitance would be ~ 200pF to give the 0.3us RF filter.
F = 1 / 2 / Pi / R / C
rearrange to find C
C = 1 / 2 / Pi / R / F
Where:
C is in Farads
R is in ohms
F is in Hertz
Pi = 3.14159
Last edited:
Buy your self a couple of very cheap carbon track pots (ie 100k and 50k) try both of them. Don't twist wires at the pot (I don't do this anyway) to minimise any capacitance and then see if you can hear any difference.
If the 100k pot is ok then make up your 100k relay controlled attenuator.
By the way, I have already done this with both 100k and 25k pots and measured bandwidth using signal generator and cro (ie real measurements in the lab) and I was able to run a 500kHz square wave on both without any reduction in amplitude at the output. From memory bandwidth went up to around 1MHz on both.
So I don't see you having any issues with this, and as you say it is very easy to change a few resistors if it doesn't work out.
If the 100k pot is ok then make up your 100k relay controlled attenuator.
By the way, I have already done this with both 100k and 25k pots and measured bandwidth using signal generator and cro (ie real measurements in the lab) and I was able to run a 500kHz square wave on both without any reduction in amplitude at the output. From memory bandwidth went up to around 1MHz on both.
So I don't see you having any issues with this, and as you say it is very easy to change a few resistors if it doesn't work out.
Last edited:
if bandwidth of 1MHz means F-3dB then this unmeasurable difference seems unlikely.
25k pot and 1MHz requires ~25pF after the pot. 100k and 1MHz requires ~6pF after the pot. I don't believe your measurements/methods.
that bandwidth does not seem unreasonable.
That is more than two octaves away from 1MHz using 100k pot.
In post 1858 you said "Most require the Low Pass filter to have F-3dB >=50kHz and some ask for >=300kHz."
Well I am pretty certain I got well over those requirements with both pots. So where is the problem.
Anyway for everyones benefit I will try and do the measurements again today if I get a spare moment.
Well I am pretty certain I got well over those requirements with both pots. So where is the problem.
Anyway for everyones benefit I will try and do the measurements again today if I get a spare moment.
The problem is you said the bandwidth difference between using a 100k pot and a 25k pot was unmeasurable and probably around 1MHz.
1MHz and 100k requires the capacitance after the pot ~6pF. This seems near impossible.
Your statement gives the impression that treble roll off can be ignored even when using what by normal practice are quite unsuitable component values.
Treble roll of due to low set RF filters can be severe and when slight is still detectable by some.
To the extent, some builders even refuse to fit RF filters, that too is not advised.
By the way, I normally advise that the RF filter be between 0.5us and 1us and I use 0.68us (1k0 + 680pF) in a power amp and slightly higher frequency, RC<0.5us in any preceeding source/pre-amp input.
1MHz and 100k requires the capacitance after the pot ~6pF. This seems near impossible.
Your statement gives the impression that treble roll off can be ignored even when using what by normal practice are quite unsuitable component values.
Treble roll of due to low set RF filters can be severe and when slight is still detectable by some.
To the extent, some builders even refuse to fit RF filters, that too is not advised.
By the way, I normally advise that the RF filter be between 0.5us and 1us and I use 0.68us (1k0 + 680pF) in a power amp and slightly higher frequency, RC<0.5us in any preceeding source/pre-amp input.
Last edited:
I repeated the measurements today.
I used 2V rms in. Then set the pots (linear pots) to get 1 V rms out (ie set at 12 O'Clock).
For the 25k pot I got exactly 1MHz (+/- 2%), for the other pot I got 470kHz.
The pots were on left and right channel. I did not test on the same channel, as the pots are already mounted (waiting to do listening tests), and one of them is an absolute pain in the **** to remove because it sits right against one of the 10uF capacitors.
I used 2V rms in. Then set the pots (linear pots) to get 1 V rms out (ie set at 12 O'Clock).
For the 25k pot I got exactly 1MHz (+/- 2%), for the other pot I got 470kHz.
The pots were on left and right channel. I did not test on the same channel, as the pots are already mounted (waiting to do listening tests), and one of them is an absolute pain in the **** to remove because it sits right against one of the 10uF capacitors.
Last edited:
While I am confident in the measurements. I am not confident in the values for the pots.
I installed these things quite a few months ago and my memory is like a sieve.
My intention was to do listening tests between different pot values, but since I haven't completed any amps yet, this preamp has been sitting around.
I purchased 3 cheap pots, they were either 25k, 50k, and 100k, or were 50k, 100k, and 200k. I know the 50k pot isn't installed because it is in my parts bin, so I have either a 25k pot and 100k pot installed or 100k pot and 200k pot installed.
I was pretty sure I had installed a 25k and a 100k but these measurements don't make sense to me. I would have expected a lower value for the 100k pot if indeed the other pot is a 25k pot.
So it is possible I have a 200k pot and 100k pot installed. I can't confirm this till next monday when I go back to work.
If the resuls make sense to you Andrew could you please exlain.
I installed these things quite a few months ago and my memory is like a sieve.
My intention was to do listening tests between different pot values, but since I haven't completed any amps yet, this preamp has been sitting around.
I purchased 3 cheap pots, they were either 25k, 50k, and 100k, or were 50k, 100k, and 200k. I know the 50k pot isn't installed because it is in my parts bin, so I have either a 25k pot and 100k pot installed or 100k pot and 200k pot installed.
I was pretty sure I had installed a 25k and a 100k but these measurements don't make sense to me. I would have expected a lower value for the 100k pot if indeed the other pot is a 25k pot.
So it is possible I have a 200k pot and 100k pot installed. I can't confirm this till next monday when I go back to work.
If the resuls make sense to you Andrew could you please exlain.