Hello,
I’m looking for advice on “passive preamp” potentiometer value, in order to avoid damage and avoid filtering audible frequencies.
I’m using an Arcam irDAC, which outputs 2.2V, probably around 47 ohms (I could only find this figure for the second version).
I’d like to experiment with a cheap Nobsound passive to begin with, which can be found in 20k and 50k varieties. Please can you let me know a) do the impedance/voltage values of my equipment indicate a passive will not cause damage, and b) do you think a 20k or 50k (or something else) would sound best?
I’m looking for advice on “passive preamp” potentiometer value, in order to avoid damage and avoid filtering audible frequencies.
I’m using an Arcam irDAC, which outputs 2.2V, probably around 47 ohms (I could only find this figure for the second version).
The amp is a Naim NAP 100, which has a published voltage gain of +29db, probably about 1V input sensitivity (for the NAP 200). Input impedance is 18k ohms.
I’d like to experiment with a cheap Nobsound passive to begin with, which can be found in 20k and 50k varieties. Please can you let me know a) do the impedance/voltage values of my equipment indicate a passive will not cause damage, and b) do you think a 20k or 50k (or something else) would sound best?
Post #21 makes me think 20k may be better: Power amps which are happy with a passive pre ... | Page 2 | pink fish media
Thanks very much,
Matt
You want a linear pot with the input impedance that low, else at low rotation you will have significantly higher attenuation than desirable if you were to use a logarithmic attenuator. I would use 50k for this kind of application, in order to have an effective impedance higher than 10k, as seen by the source. Not that 10k is a problem, but some tube stages will have issues.
But if you don't have a buffer in the middle you will get some degradation of high frequencies anyway, as the capacitance of the cable is effectively forming a RC filter with the 18k input impedance.
But if you don't have a buffer in the middle you will get some degradation of high frequencies anyway, as the capacitance of the cable is effectively forming a RC filter with the 18k input impedance.
The control taper (linear vs log) is present regardless of total pot value, and works the same way until loaded with an input impedance that approaches the pots value, at which point the taper is significantly modified. You always want a log/audio taper pot for a volume control otherwise the first half of the rotation range only modifies volume by 6dB.
That's not now it works. The impedance seen by the input will vary with pot position.
No input will have issues with a source impedance that is lower. The problem occurs when the input impedance is not effectively bridging the source impedance. In other words, a pot that is of too high value, not too low.
With a 10K pot at the position of highest source impedance, and a typical 1M audio cable, the effective -3dB point will be around 200kHz.
This is actually not how it works. The input impedance of the amplifier is usually set by a resistor to ground (except for some topologies where a part of the output also factors into the input impedance). This resistor appears in parallel to the effective resistance between the wiper and ground. This modifies the taper ratio and the use of a linear pot is well-established, it's called law faking.
This is an example:
ESP - A Better Volume Control
Any input capacitor is effectively a short circuit for AC, and can be treated as a straight wire.
You are correct in saying that the impedance seen by the input will vary as seen by the input. However, I was talking about the impedance seen by the source. The other end of the horse, so to speak.
I would like to know the definition of this term 'bridging' wrt impedance. I've not come across it before, perhaps you have some references that can be studied? And again, I wasn't talking about source impedance, but load impedance.
This is an example:
ESP - A Better Volume Control
Any input capacitor is effectively a short circuit for AC, and can be treated as a straight wire.
You are correct in saying that the impedance seen by the input will vary as seen by the input. However, I was talking about the impedance seen by the source. The other end of the horse, so to speak.
I would like to know the definition of this term 'bridging' wrt impedance. I've not come across it before, perhaps you have some references that can be studied? And again, I wasn't talking about source impedance, but load impedance.
I said pretty much that in my previous post, but faking a log pot that way requires a nearly 10:1 ratio of pot value to load value. Generally just not as good as a real log pot.
[/quote]
Any input capacitor is effectively a short circuit for AC, and can be treated as a straight wire.
[/quote]Right...in general.
Both need to be considered>
A bridging load is at least 5 to 10x the source impedance. Ideally 10X or greater. Pretty common term. I’m sure you can find your own references.
There you go, that's the idea.
The whole thought behind a passive volume control is that it's clean, pure, etc., blah, blah. But when you get down to doing it there are compromises at every corner that actually do negatively impact SQ.
Buffering gets you around every bit of that, except the psychology.
JFET Buffer – Bartola(R) Valves
this or B1 if executed correctly will be more musical than passive
ps: you can get original board and toshiba jfets from passlabs
PassDiy
i still have many original transistors for JLH buffer, old but good
A Paul Kemble web page - John Linsley Hood preamp designs.
this or B1 if executed correctly will be more musical than passive
ps: you can get original board and toshiba jfets from passlabs
PassDiy
i still have many original transistors for JLH buffer, old but good
A Paul Kemble web page - John Linsley Hood preamp designs.
Nice value.
I would even use 5k, since it´s still easy to find.
Both are perfectly fine, of course, no audible difference.
Just don´t put the passive volume control here and the power amp there, 100 yards away
Did you look at the "Sticky" thread at the TOP of this section?
Attachments
Thanks for the thoughts everyone.
The sticky says avoid using low impedance pots in order to avoid source damage. According to the DACT calculator, even using pots up to 250k with my 18k input amp shouldn't cause audible low-pass-filtering (although I'm not sure what my capacitance is; cables are short at least). What's the reasoning behind 10k pots?
Also, I'm UK-based and have struggled to find passive with a buffer although will look into it further.
Certainly did, thanks, several times. I found the part about avoiding damage to your source a little unclear, so I thought I would check with specific equipment/figures. Apologies if it would have been better to reply on there instead of creating a new thread.
The sticky says avoid using low impedance pots in order to avoid source damage. According to the DACT calculator, even using pots up to 250k with my 18k input amp shouldn't cause audible low-pass-filtering (although I'm not sure what my capacitance is; cables are short at least). What's the reasoning behind 10k pots?
Also, I'm UK-based and have struggled to find passive with a buffer although will look into it further.
That's not the definition of "preamplifier". Look it up for yourself.
A pre-amplifier implies amplification. If no amplification is required, then it's not a pre-amplifier, its an attenuator.
Transformers do not provide gain, they transform the relationship between voltage and current, with impedance being the end result, and with unity (or slightly less) power gain. If you need increased voltage and can tolerate reduced current and increased source impedance, they work fine, but that's not what "amplifiers" do. They are not interchangeable functions.
There is no such at thing as a passive "amplifier". A transformer is not an amplifier, and always functions at unity power gain minus insertion loss.
"Passive Preamplifier" is a term made up by people who don't understand electronics, and are pursuing the goal "purity" that is neither achievable nor real with the passive topology. The are making a passive attenuator, possibly with an impedance transformer. Neither has zero impact on the signal, neither results in anything "pure" other than the psychological (non-real) and subjective perception of what's going on.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.