5000 ohm resistors. 5k
I’m a beginner in the amp/ electronics diy world. I understand speaker crossovers so coming from that side of things I know what 1 ohm R can do to a signal ...
I’m putting together a balanced passive preamp and looking at the sizes it calls for 5k R on each leg before the pot to maintain load when turned fully down. I get that but what strikes me is the thousands of ohms the signal has to go through even when fully opened up. Wouldn’t this crush the signal to death?
Here is a link to what I’m going to follow
https://www.diyaudio.com/forums/pas...nuator-dual-channel-balanced.html#post3060481
I’m a beginner in the amp/ electronics diy world. I understand speaker crossovers so coming from that side of things I know what 1 ohm R can do to a signal ...
I’m putting together a balanced passive preamp and looking at the sizes it calls for 5k R on each leg before the pot to maintain load when turned fully down. I get that but what strikes me is the thousands of ohms the signal has to go through even when fully opened up. Wouldn’t this crush the signal to death?
Here is a link to what I’m going to follow
https://www.diyaudio.com/forums/pas...nuator-dual-channel-balanced.html#post3060481
OMG! I've lost my longest message in all the times! DIY audio server showed me its condoleances but it didn't make it to me to feel better
So, what's up ?
In the quoted message: line level signal carries voltage, not power.
In the link you proposed: to my unskilled view, it seems to me that the one pictured is a "common mode" attenuator, since there's no reference to ground; the potentiometer is wired like a variable resistor ( pins 2 &3 tied together).
Using a 4-ganged potentiometer wired normally is the way to go.
About the title: 1.21 GHz is RF. I admit I have no frequentation with RF. Also "balanced" is something that I've never encountered. But 1MΩ potentiometer, yess !! Nothing wrong in putting 1 MΩ in series to the signal, the problems arise when there's a cable ( with its own RLC parameters) between two stages and there might be a treble loss when the cursor of the pot introduces significant Rseries, so the low Ω potentiometers ( 10-20 kΩ) are favoured
Everything is relative.
With an 8 Ohm load, a 1 Ohm resistor makes a difference.
Especially since loudspeakers are 99% loss, and loudspeaker amplifier Watts are expensive, so we don't want to waste any.
Down at line level, loads are 10,000 to 100,000 Ohms. And milliWatts are relatively cheap. And here is where we *may* want to waste-off (or "crush") over-loud sources to get appropriate level in power amp and speaker.
Those 5k are not to "maintain load". The pot and 5ks form a Voltage Divider. Without the 5ks the pot would have to short-out the *source*. Which could be a very powerful phono preamp (in old days, a strong telephone line).
In other words: sometimes the signal is 2 or 10 times bigger than we need. This network can "crush" signal 1/2 or 1/10 "to death". It can even crush it nearly to death, as when you need to turn full-down to hear something else in the house.
With an 8 Ohm load, a 1 Ohm resistor makes a difference.
Especially since loudspeakers are 99% loss, and loudspeaker amplifier Watts are expensive, so we don't want to waste any.
Down at line level, loads are 10,000 to 100,000 Ohms. And milliWatts are relatively cheap. And here is where we *may* want to waste-off (or "crush") over-loud sources to get appropriate level in power amp and speaker.
Those 5k are not to "maintain load". The pot and 5ks form a Voltage Divider. Without the 5ks the pot would have to short-out the *source*. Which could be a very powerful phono preamp (in old days, a strong telephone line).
In other words: sometimes the signal is 2 or 10 times bigger than we need. This network can "crush" signal 1/2 or 1/10 "to death". It can even crush it nearly to death, as when you need to turn full-down to hear something else in the house.
My condolences on your lost message.
Good points.
> About the title: 1.21 GHz is RF.
Title is "1.21 Gigaohms", not Hertz. 1Gig resistors are a thing, and are used in condensor microphones. They don't pass audio, they pass teeny DC while not-loading the weak audio off the diaphragm. Again, everything is relative.
Good points.
> About the title: 1.21 GHz is RF.
Title is "1.21 Gigaohms", not Hertz. 1Gig resistors are a thing, and are used in condensor microphones. They don't pass audio, they pass teeny DC while not-loading the weak audio off the diaphragm. Again, everything is relative.
Signal is not a thing, its a concept, we define what aspect of the circuit represents the signal, it might be a voltage here, a current there, or sometimes something less direct.
Currents go through, voltages push currents through, resistors relate current to voltage. A resistor can convert a voltage signal to a current signal for instance. Nothing is crushed.
However there is always noise to worry about - thermal noise in resistors is a very small thing, but when you are amplifiing small things can become bigger and thus ultimately audible.
A resistor in series will impose its thermal noise as a voltage on the signal (and a resistor to ground imposes its thermal noise as a current). So we sometimes worry about that - for large signals its usually irrelevant though, interference is a bigger issue and things like shielding are what you worry about first.
The thermal voltage noise of a 5k resistor is 1.3µV rms in the audio frequency range. For signals of several volts that's not really an issue, for microphone signals it is an issue.
Thanks for the replies. I have 1% resistors on the way. 5k, 4.3k, 3.9k and 3.3k to experiment with. I'm not planning on ever going below 2/3 full open with the pot. (balancing a multi amp system) From what I'm getting by the responses is that fully open the 5k will not do much to the level of the signal. I'll soon find out...
As a joke I titled it 1.21 gigaohms after back to the future, funny that that is a real resistive load that is used in anything. That's an anti-bolt of lightning
As a joke I titled it 1.21 gigaohms after back to the future, funny that that is a real resistive load that is used in anything. That's an anti-bolt of lightning
The difference is that in a pre-amp you are dealing with very small currents... usually no more than a couple of milliamps but in speakers you are dealing with amperes.
If you're looking for some kind of standard, I don't know if one exists. But the engineers I've worked with usually designed outputs to have about 600 ohm impedance and inputs that were 10k ohms or higher.
One detail I would like to point out is that the attenuator topology the OP is looking at
a) has no issues with common mode to differential mode conversion
but
b) provides no common-mode attenuation either and may thus put great demand on input CMRR.
If you were to make a balanced L attenuator with two pot sections per channel going to ground, it would be exactly the opposite. Now knowing the reputation of pots for channel tracking, that approach may well be more problematic than just shooting for high input CMRR straight away. In fact, doing some simulation, a 0.6 dB mistracking is enough to drop effective CMRR by around 30 dB vs the original approach.
So I'm guessing the "single resistor in the middle" approach is in fact the way to go with pots, while for a fixed attenuator, you can take advantage of the availability of precision resistors and go for a "double L".
If you always need a lot of attenuation for whatever reason, you could in fact combine both approaches (first "double L", then "single pot"). Not like dropping your levels that low is particularly recommended or anything.
The question is - what kind of power amp has a balanced input but no input trim? Commercial studio/PA jobs usually do, and it's quite a good idea for several reasons.
a) has no issues with common mode to differential mode conversion
but
b) provides no common-mode attenuation either and may thus put great demand on input CMRR.
If you were to make a balanced L attenuator with two pot sections per channel going to ground, it would be exactly the opposite. Now knowing the reputation of pots for channel tracking, that approach may well be more problematic than just shooting for high input CMRR straight away. In fact, doing some simulation, a 0.6 dB mistracking is enough to drop effective CMRR by around 30 dB vs the original approach.
So I'm guessing the "single resistor in the middle" approach is in fact the way to go with pots, while for a fixed attenuator, you can take advantage of the availability of precision resistors and go for a "double L".
If you always need a lot of attenuation for whatever reason, you could in fact combine both approaches (first "double L", then "single pot"). Not like dropping your levels that low is particularly recommended or anything.
The question is - what kind of power amp has a balanced input but no input trim? Commercial studio/PA jobs usually do, and it's quite a good idea for several reasons.
Condenser microphone FET input bias. 10G resistors are quite a typical sight there. This minimizes input noise... bit of a longer story.
I hope not. I’ll post sweeps of the results. That was my main concern, dropping the level too low. I’m mainly using it to adjust the difference between gain of different amps so I’ll be around the top end.
Studiophone, I suggest you look at some of the fundamentals of electronics. This will help the DIY thing to be enjoyable rather than a frustrating mystery.
You are describing a voltage divider here - the math is simple fractions - so calculating the attenuation is something you really ought to be comfortable with. (type "voltage divider tutorial" into your favorite search engine).
You are describing a voltage divider here - the math is simple fractions - so calculating the attenuation is something you really ought to be comfortable with. (type "voltage divider tutorial" into your favorite search engine).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.