Dear all, can you please enlighten me, what is the truth behind (external) preamps in context with number of amp. stages and gain cell like volume control ?
To my understanding there are basically passive and active preamp types.
- Passive is used to attenuate input signal and let the signal pass towards the amplifier(s), it also might have an input selector but more or less that's it. For a monoblock setup it's mostly an external preamp, for integrated amplifiers, the preamp is built into the same amplifier chassis, alongside with the gain and output stages.
- Active can have active component too and raise signal level over standard (CD player-like) line-out voltage level (these outputs are called Pre-Out, right ? )
Volume control:
- for passive preamps, it's attenuation actually and lower pot settings (for normal or less room listening levels) have lower SNR than higher pot settings (although when implemented well, unnoticable)
- for active preamps, volume control can still be passive (attenuation) OR an active solution, like a gain cell, where an inverting opamp circuitry is taking care of attenuation/amplification and the volume pot itself is only changing the opamp's gain, this way achieving a similar volume-control effect.
Now I'd like to have some questions based on all these:
Conventional Class A/AB amplifiers (monoblocks too) usually have 3 amplification stages: input stage, driver stage, output stage.
- Does an external active preamp mean a 4th amplification stage in the whole gain chain ? (Or better said, it's actually the 1st stage when we count from source towards the output stage).
- Does an additional active circuitry, like the gain-cell, really achieve better sound (whatever this means) while introducing possible additional errors to the signal, than a single good quality attenuator and staying with 3 amplification stages ?
- Conventional Line-In-s on monoblocks or integrated amps have limits regarding input signal voltage. This is normal. When an active preamp's Pre-Out signal is somewhat stronger, than a standard Line-Out, what is it used for ? Are there amps where they have a Line-In and a Pre-In connector too ? I assume Pre-In is fed by an active preamp's Pre-Out and the final amp's Pre-In is bypassing its normal input stage (to avoid overdrive) so its Pre-In is connected to the driver stage ? Kind of an integrated-preamp bypass.
- When I have a HT Processor with Pre-Outs (alongside with Line-Outs), I assume I can use these Pre-Outs to feed some external amps via their Pre-Ins, again bypassing their own input stages, right ? If I used Line-Outs, I would have to use the final amps' Line-Ins, right ?
- My father used to build quite a lot of tube and transistor amps couple of years ago. He told me, due to SNR considerations, the best to use volume control (attenuation) is before the output stage. I haven't asked him further about this but I don't see such implementation as a common practice at all.. why ? Comfort, complexity, other reasons ? For separates I can understand, but for integrated amps ?
To my understanding there are basically passive and active preamp types.
- Passive is used to attenuate input signal and let the signal pass towards the amplifier(s), it also might have an input selector but more or less that's it. For a monoblock setup it's mostly an external preamp, for integrated amplifiers, the preamp is built into the same amplifier chassis, alongside with the gain and output stages.
- Active can have active component too and raise signal level over standard (CD player-like) line-out voltage level (these outputs are called Pre-Out, right ? )
Volume control:
- for passive preamps, it's attenuation actually and lower pot settings (for normal or less room listening levels) have lower SNR than higher pot settings (although when implemented well, unnoticable)
- for active preamps, volume control can still be passive (attenuation) OR an active solution, like a gain cell, where an inverting opamp circuitry is taking care of attenuation/amplification and the volume pot itself is only changing the opamp's gain, this way achieving a similar volume-control effect.
Now I'd like to have some questions based on all these:
Conventional Class A/AB amplifiers (monoblocks too) usually have 3 amplification stages: input stage, driver stage, output stage.
- Does an external active preamp mean a 4th amplification stage in the whole gain chain ? (Or better said, it's actually the 1st stage when we count from source towards the output stage).
- Does an additional active circuitry, like the gain-cell, really achieve better sound (whatever this means) while introducing possible additional errors to the signal, than a single good quality attenuator and staying with 3 amplification stages ?
- Conventional Line-In-s on monoblocks or integrated amps have limits regarding input signal voltage. This is normal. When an active preamp's Pre-Out signal is somewhat stronger, than a standard Line-Out, what is it used for ? Are there amps where they have a Line-In and a Pre-In connector too ? I assume Pre-In is fed by an active preamp's Pre-Out and the final amp's Pre-In is bypassing its normal input stage (to avoid overdrive) so its Pre-In is connected to the driver stage ? Kind of an integrated-preamp bypass.
- When I have a HT Processor with Pre-Outs (alongside with Line-Outs), I assume I can use these Pre-Outs to feed some external amps via their Pre-Ins, again bypassing their own input stages, right ? If I used Line-Outs, I would have to use the final amps' Line-Ins, right ?
- My father used to build quite a lot of tube and transistor amps couple of years ago. He told me, due to SNR considerations, the best to use volume control (attenuation) is before the output stage. I haven't asked him further about this but I don't see such implementation as a common practice at all.. why ? Comfort, complexity, other reasons ? For separates I can understand, but for integrated amps ?
For best SNR the signal must be as strong as possible in all stages. In other words, the signal must be strong in as many stages as possible. That is achieved when the volume control is downstream as far as possible, which usually means right before the power amplifier. It is common practice in integrated amplifiers.
Last edited:
Some power amps are 0dB gain and will rely on the preamp to basically serve as the 3-stages of a typical Class AB amp. This kind of a preamplifier will have some challenging requirements of needing to swing as high as the rails of the power amp. For a 25w amp that is 40Vpp and 50w amp, it is 57Vpp for 8ohm loads.
The phrase "passive pre-amp" makes no sense. A pre-amp amplifies, which needs active parts. A passive device does not amplify.
To understand how we made this so confusing, you have to go back in history.
We have phono pickups, microphones, radio tuners. The signal from a needle or a microphone is VERY weak. The *first* ("pre") thing we have to do is boost it up (with active amplifiers). But not so it distorts. Disks and singers vary a lot. Typically we aim for a part-Volt. (Tuners are a lot of gain also.)
If we have multiple sources we have a way to select them.
Sources and tracks vary, and we may want loud or soft. Now that the signal has been pre-amplifed to a part-volt, it is OK to cut it down a little as needed. Volume control.
To get the next step you have to know that for most of audio history there was big money in putting audio on telephone lines. The earliest PA systems often relayed a speech in Washington DC to a crowd in Philadelphia. Later radio networks sent programs all around the USA (and this was most of Bell Telephone's profit in the 1930s). Radio studios moved away from their transmitters and had to carry the signal on a telephone line.
"Line Level" can't be weak or it would drown in stray interference. It can't be too strong, for cost, and so it won't cross-talk into other telephone conversations. Line Level is "about 1 Volt".
It does not make sense to build the several pre-amps to make such a strong signal (risks overload). Instead we follow the volume control with a Line Amp. This is also the place to put tone controls.
Well, all this is too complicated for the average buyer. Makers bundled all these stages together in one box called "Preamplifier".
Power amplifiers typically take Line Level and boost it to the desired output. Loudspeaker power amps often have a gain of "about 20". Large transmitters have much higher gain as needed.
If you look at all the classic hi-fis of the 1950s-1970s they have gain-structure and functional blocks about like this.
But then came the CD. The signal from the laser-eye is very tiny gibberish, so a chip bumps it bigger. And for whatever reason, Philips made the maximum level 2 Volts.
If you look at the classic level diagram, 2V is so big that most of these amplifying stages are not needed. All you need is a passive volume potentiometer to knock the signal *down* to get the desired power amplifier level. So now the "pre-amplifier" is not an amplifier at all.
To understand how we made this so confusing, you have to go back in history.
We have phono pickups, microphones, radio tuners. The signal from a needle or a microphone is VERY weak. The *first* ("pre") thing we have to do is boost it up (with active amplifiers). But not so it distorts. Disks and singers vary a lot. Typically we aim for a part-Volt. (Tuners are a lot of gain also.)
If we have multiple sources we have a way to select them.
Sources and tracks vary, and we may want loud or soft. Now that the signal has been pre-amplifed to a part-volt, it is OK to cut it down a little as needed. Volume control.
To get the next step you have to know that for most of audio history there was big money in putting audio on telephone lines. The earliest PA systems often relayed a speech in Washington DC to a crowd in Philadelphia. Later radio networks sent programs all around the USA (and this was most of Bell Telephone's profit in the 1930s). Radio studios moved away from their transmitters and had to carry the signal on a telephone line.
"Line Level" can't be weak or it would drown in stray interference. It can't be too strong, for cost, and so it won't cross-talk into other telephone conversations. Line Level is "about 1 Volt".
It does not make sense to build the several pre-amps to make such a strong signal (risks overload). Instead we follow the volume control with a Line Amp. This is also the place to put tone controls.
Well, all this is too complicated for the average buyer. Makers bundled all these stages together in one box called "Preamplifier".
Power amplifiers typically take Line Level and boost it to the desired output. Loudspeaker power amps often have a gain of "about 20". Large transmitters have much higher gain as needed.
If you look at all the classic hi-fis of the 1950s-1970s they have gain-structure and functional blocks about like this.
But then came the CD. The signal from the laser-eye is very tiny gibberish, so a chip bumps it bigger. And for whatever reason, Philips made the maximum level 2 Volts.
If you look at the classic level diagram, 2V is so big that most of these amplifying stages are not needed. All you need is a passive volume potentiometer to knock the signal *down* to get the desired power amplifier level. So now the "pre-amplifier" is not an amplifier at all.
Attachments
PRR, many thanks for your time and great explanation. I've never thought there's so much history behind this seemingly simple topic. 
Would you please have a look at this short vid ? (5 mins). YouTube
It's about the same "no pot" philosophy I wanted to know, where to implement.
So if I understand you right, a REAL (active) preamp nowadays is actually equivalent to a monoblock's input stage, right ? Technically speaking.
A CD player's RCA Line-Out (or let's talk about a DAC) maxes at around 2V. It can be (and usually is) connected directly to the integrated amplifier's Line-In which is followed by the input stage, driver stage and output stage.
If we have a different approach, like working with separates, we have:
- passive preamp (which isn't a pre-amp actually, you said), with an input selector switch and a main volume control (attenuator)
- DSP or active 3-way crossover behind that
- 6 monoblocks (in case of 3-way stereo)
Now, individual gains representing the speaker sensitivities can be corrected on the monoblocks themselves (again a pot in the signal path, either letting the whole signal through, or attenuating a bit).
But the main volume control to control system-wide overall volume is still in the "preamp" box, again an attenuator pot.
I'm still curious, if this classic solution is better OR to make an active preamp with precise opamp and modifying volume by modifying its gain. A max pot setting would be then 1:1 amplification (neither amplifies nor attenuates the input signal), min pot setting is then a less-than-one multiplier to the signal so it gets smaller and smaller..
And I'm talking here about the system-wide volume control.
Same question goes for the 6 individual amps:
- for class D the gain control remains an attenuator, due to circuit integration and complexity
- for class A/AB (either tube or solid state), the same kind of opamp circuitry could replace the pots before the input stage, right ? OR even replace the complete input stage with it and have it connected directly to the driver stage .. ?
Would you please have a look at this short vid ? (5 mins). YouTube
It's about the same "no pot" philosophy I wanted to know, where to implement.
So if I understand you right, a REAL (active) preamp nowadays is actually equivalent to a monoblock's input stage, right ? Technically speaking.
A CD player's RCA Line-Out (or let's talk about a DAC) maxes at around 2V. It can be (and usually is) connected directly to the integrated amplifier's Line-In which is followed by the input stage, driver stage and output stage.
If we have a different approach, like working with separates, we have:
- passive preamp (which isn't a pre-amp actually, you said), with an input selector switch and a main volume control (attenuator)
- DSP or active 3-way crossover behind that
- 6 monoblocks (in case of 3-way stereo)
Now, individual gains representing the speaker sensitivities can be corrected on the monoblocks themselves (again a pot in the signal path, either letting the whole signal through, or attenuating a bit).
But the main volume control to control system-wide overall volume is still in the "preamp" box, again an attenuator pot.
I'm still curious, if this classic solution is better OR to make an active preamp with precise opamp and modifying volume by modifying its gain. A max pot setting would be then 1:1 amplification (neither amplifies nor attenuates the input signal), min pot setting is then a less-than-one multiplier to the signal so it gets smaller and smaller..
And I'm talking here about the system-wide volume control.
Same question goes for the 6 individual amps:
- for class D the gain control remains an attenuator, due to circuit integration and complexity
- for class A/AB (either tube or solid state), the same kind of opamp circuitry could replace the pots before the input stage, right ? OR even replace the complete input stage with it and have it connected directly to the driver stage .. ?
Oh. I never connected "line level" with telephone lines, but this makes perfect sense.
The signal from the laser pickup is not only small but digital to boot, and it has to be sync'd up, decoded (including CIRC error correction) and D/A converted.
2 Vrms is 17 dB above the 300 mV nominal consumer line level, which was deemed adequate headroom for an audiophile medium. Early CDs also had corresponding average levels much like existing sources. That content got louder and louder and louder over the years is not Philips' (or Sony's) fault. It is not unusual to see average levels of -10 dBFS or even -7 dBFS these days.
Last edited:
Oh, it's this guy...
I think you should start thinking in engineering terms. Dynamic range, levels, noise, distortion, gain. What you have is a classic gain staging problem, governed by Friis' formula and its relatives.
At the power amp, you generally want a noise floor about 110 dB below the voltage corresponding to about 110 dB SPL, so there isn't any audible hiss when nothing is playing. It gets more tricky than you might think if you consider that speakers may vary in sensitivity from below 85 dB SPL / 2.83 V / m (small bookshelves) to over 100 dB (PA / horns). The latter would place correspondingly more stringent requirements on output noise level while potentially only requiring a fraction of maximum output power. (It's even more extreme in headphones.)
Sources in front of a volume control may get away with as little as 70 dB instantaneous, since you may not even want to listen that loud, or if so, a bit of background noise won't matter (the instantaneous dynamic range of our hearing actually isn't all that great).
There are multiple ways of achieving this goal. Which one is chosen will generally depend on what makes the most sense with the technology available.
If you get the levels too low, you may run into trouble with noise (/hum/interference) and may need to jump through some hoops in order to get that down.
If you select the levels too high, you may need inconveniently large power supply voltages, may have to go discrete rather than being able to pick from (inexpensive) high-performance ICs, or may have trouble keeping distortion related to slewing and general nonlinearity down.
There basically is a good bit of leeway there.
Your "inverting amplifier with variable gain" idea is not exactly new - see Musical Fidelity A1 (late '70s I think). It has one big downside - where less than perfect wiper contact will generate rather benign dropouts in a conventional circuit, in this one it creates big gain spikes. The human ear is quite insensitive to short dropouts, but short spikes are plainly audible.
When a large variability in transducer sensitivity has to be accomodated, usually a combination of variable attenuation plus variable (stepped) gain is being used. PGAs tend to incorporate both on-chip.
Many problems in the real world stem from not being able to choose the parameters of all the components at will but rather buying them off the shelf. Your typical hi-fi power amp will have a gain of 26-29 dB, sometimes more, while output noise level with input shorted can be as low as 30-60 µVrms. Your typical hi-fi preamp (with volume upfront) will have a gain of 16.5 dB and and output noise level of maybe 4 µVrms. That's about 110 µVrms of output noise from the preamp alone, or about 12 dB SPL @ 1 m for 100 dB / 2.83 V / m speakers. So if horns are your thing, you may be forced to install a passive attenuator in between to get noise in check, or hunt down a fancy preamp with a 2-stage volume control.
Basically, if you want to end up with a system that works well, you have to consider
* output noise level
* maximum output amplitude
* input level
* input sensitivity
at each junction in the signal chain. As-is, SNR will be limited by the ratio of input sensitivity to the RMS sum of output and equivelent input noise.
It's all a bit complex but you were basically asking for it. You may want to sketch out what it is that you actually want to do and what components you already have, so that we could apply the general rules to a specific example. There are many way to skin the proverbial cat.
Uh, yeah.
Preamps were necessary when LP's mikes & tape heads were the sources.
Now we have Cd's, cellphones, D/A converters putting out about 2 v at 600 to 1000 ohms impedance. Televisions and radios can put out 7 to 20 vac on their speaker lines, but have *****y internal speakers. (And "sound bars" for TV's cost $180 up, which is rediculous IMHO for a device with a 3" woofer).
The term "passive preamp" is used around here to denote a box with source selector switch, volume pot, maybe pre-switch volume pots to even devices up. You can't actually buy such a device in a physical store.
A "passive preamp" has one more function: all those cables connecting the devices to it are out there making a RF antenna and conducting radio modulations, lightning pops and loop hum from the AC wiring in the room to the power amplifier. The grounds in the cables make an AC loop pickup with the safety grounds on the various sources. So some sort of RF & hum filtering is in order on the inputs. The connectors cable to box can make diodes that reduce the radio waves to audible frequencies.
Then the cable from "passive preamp" to the power amplifier can also pick up radio, lightning, hum. The higher power amps have fans, and putting them in a remote location where these can't be heard is desirable. That lengthens the cables.
In many cases, putting an active element in the preamp is desirable to overcome the reduction by the filtering and swamp out the current generated in the cable to the amp by RF pickup and AC loop hum pickup.
I certainly need active drive on my 4 meter cables from control hub to the power amp. Else they pick up KSUG radio. It helps to justify the mixer since I do play LP's, which requires 50x gain and a RIAA filter.
Multiple input amps are made, to eliminate the "preamp", but they are called "mixer amps" and tend to be mono for the building paging market. Stereo mixer amps are sold but tend to come in 12 to 20 input boards, which is a lot of control area to put in your living room. Also they have a lot of devices like effect loops, monitor plus main amplifiers, frequency contouring which run the costs up and don't pertain to home use.
"Preamps" with all the inputs on at the same time are called "mixers". I use one now instead of a preamp, as switching the various inputs around when I change media is a nuisance. Requires walking around the table from the LP to the switch, or the CD to the switch, or the radio to the switch. I leave them all on, and don't pay a huge hiss hum penalty.
Happy shopping, or even better, building your entertainment hub.
Preamps were necessary when LP's mikes & tape heads were the sources.
Now we have Cd's, cellphones, D/A converters putting out about 2 v at 600 to 1000 ohms impedance. Televisions and radios can put out 7 to 20 vac on their speaker lines, but have *****y internal speakers. (And "sound bars" for TV's cost $180 up, which is rediculous IMHO for a device with a 3" woofer).
The term "passive preamp" is used around here to denote a box with source selector switch, volume pot, maybe pre-switch volume pots to even devices up. You can't actually buy such a device in a physical store.
A "passive preamp" has one more function: all those cables connecting the devices to it are out there making a RF antenna and conducting radio modulations, lightning pops and loop hum from the AC wiring in the room to the power amplifier. The grounds in the cables make an AC loop pickup with the safety grounds on the various sources. So some sort of RF & hum filtering is in order on the inputs. The connectors cable to box can make diodes that reduce the radio waves to audible frequencies.
Then the cable from "passive preamp" to the power amplifier can also pick up radio, lightning, hum. The higher power amps have fans, and putting them in a remote location where these can't be heard is desirable. That lengthens the cables.
In many cases, putting an active element in the preamp is desirable to overcome the reduction by the filtering and swamp out the current generated in the cable to the amp by RF pickup and AC loop hum pickup.
I certainly need active drive on my 4 meter cables from control hub to the power amp. Else they pick up KSUG radio. It helps to justify the mixer since I do play LP's, which requires 50x gain and a RIAA filter.
Multiple input amps are made, to eliminate the "preamp", but they are called "mixer amps" and tend to be mono for the building paging market. Stereo mixer amps are sold but tend to come in 12 to 20 input boards, which is a lot of control area to put in your living room. Also they have a lot of devices like effect loops, monitor plus main amplifiers, frequency contouring which run the costs up and don't pertain to home use.
"Preamps" with all the inputs on at the same time are called "mixers". I use one now instead of a preamp, as switching the various inputs around when I change media is a nuisance. Requires walking around the table from the LP to the switch, or the CD to the switch, or the radio to the switch. I leave them all on, and don't pay a huge hiss hum penalty.
Happy shopping, or even better, building your entertainment hub.
Last edited:
Uh Gentlemen, heavy inputs. Respect, what a great forum.
@sgrossklass:
3-way active setup:
- 95dB PA woofer (driven by Class D Hypex UcD2k based monoblock, gain at around 33, attenuated to match the others)
- 96dB PA midranges (driven by Class AB KT150 tube monoblock)
- 98/100dB AMT tweeter (depending on which I'm choosing of the 2 types I need, driven by Class AB Mosfet based monoblock)
- 4-way Linkwitz 4th order.. (1st way is a highpass at ~22Hz to filter out subsonics - either analog-electronic OR via DSP)
6 monoblocks alltogether, all DIY.
Sources:
HTPC with Pink Faun's PCIe I2S Card (PS Audio-like connection via HDMI to a DAC)
DAC: DSP or a Matrix X-Sabre Pro
LP: maybe later
Bluetooth board from Aliexpress.
All in DIY case, all with true balanced (XLR) outputs. The monoblocks will also have XLR inputs only. All AB monoblock based on "fully balanced" principle (pushpull) so I expect low noise levels if designed and built carefully.
That's it so far.
Oh, and maybe also a remote control for volume changing, muting, input selection etc. Some kind of IC based design or even a 256-step relay attenuator, doesn't really matter. Also from aliexpress..
@sgrossklass:
3-way active setup:
- 95dB PA woofer (driven by Class D Hypex UcD2k based monoblock, gain at around 33, attenuated to match the others)
- 96dB PA midranges (driven by Class AB KT150 tube monoblock)
- 98/100dB AMT tweeter (depending on which I'm choosing of the 2 types I need, driven by Class AB Mosfet based monoblock)
- 4-way Linkwitz 4th order.. (1st way is a highpass at ~22Hz to filter out subsonics - either analog-electronic OR via DSP)
6 monoblocks alltogether, all DIY.
Sources:
HTPC with Pink Faun's PCIe I2S Card (PS Audio-like connection via HDMI to a DAC)
DAC: DSP or a Matrix X-Sabre Pro
LP: maybe later
Bluetooth board from Aliexpress.
All in DIY case, all with true balanced (XLR) outputs. The monoblocks will also have XLR inputs only. All AB monoblock based on "fully balanced" principle (pushpull) so I expect low noise levels if designed and built carefully.
That's it so far.
Oh, and maybe also a remote control for volume changing, muting, input selection etc. Some kind of IC based design or even a 256-step relay attenuator, doesn't really matter. Also from aliexpress..
If you need to drive 6 amps with different frequency ranges, you have frequency filtering in your hub. 6 different cables is an invitation to rf/hum/pop interference. Each twisted pair cable introduces about 20 pf per foot which filters highs off high impedance sources. I would say you need 6 different op amps (maybe 2 to 4 per IC) to drive the 6 cables. Each has separate gain pot on the panel, and several of them are not full frequency and need a pre-filter which is most economically done with op amp stages these days. This would be custom, even full feature mixers have typically 2 stereo output channels, main & monitor. You could use the headphone out as a third ajustable level amp drive, but the frequency control of headphone in a commercial mixer is limited.
Your three high level sources have XLR cables, which can pick up RF/hum/pops, so some filtering at the input is indicated. I use 33 pf ceramic to case ground. The balanced input lends itself to op amp in the differential mode to discriminate common mode noise, although you might be able to get away with passive pot attenuation. However, mixing them to the power amp bus requires 1000 ohm minimum mixer resistors to keep from blowing up the drivers in the sources, and that attenuation is usually followed by an op amp to boost the signals back up to 2 vac. 10000 ohm resistors doing the mixing are more usual, there are reasons covered in westsound's long discussions of mixers. No point IMHO mixing both top and bottom of the xlr IMHO. You could use transformers to input the differential signals from the XLR and make them ground referenced, but these are more expensive than op amps, running about $90 each for the good ones (jensen, sennhauser). Instead of mixing the various inputs you could put a rotary stereo switch which is expensive these days, or various LDR attenuators controlled by toggle or slide switches. Farnell/digikey/mouser don't carry stereo rotary switches anymore, but I think dynaco upgrade hubs like dynacodoctor do. the dynaco rotary switches shorted the unused inputs though to prevent pops, which is not compatible with active drive sources. But if you don't follow the LDR attenuators by 1000 ohm resistors & a mixer op amp, if you turn two on by accident, you can blow up both source output (op)amps. Better in that case just to switch cables on a single input xlr jack.
BTW a steel box is required to keep out the RF lightning pops etc. If you put the power transformer in the device, a second steel box or bulkhead is required to keep the hum out of the internal amps. Particularly if you have a 50x gain amp for mm LP. Moving coil LP is more like 80 x gain. I proved this point when I turned a hummy hissy $15 disco mixer into my entertainment hub Improving a disco mixer to mid fi is the name of the thread where I detailed that exploration. & I am at about 75 db s/n not 110 as sgrossklass and PRR talk about.
I safety ground the hub box, and not the power amp to avoid the ground loops case ground to hub case through both the xlr shield and the building wiring. Three of my power amps are single ended and have RCA jacks grounded to case with a 2 prong AC cord. The ungrounded power amps live behind the organ between the speakers and are not a safety hazard because no babies crawl around my floors. Modern power amps with an op amp on the input to differentially reject the common mode hum can have the case grounded. Op amp ICs became useable for low hiss low hum hifi about 1985. Even the ~1980 Peavey cs800 rev A, B, & C that had differential input op amps were too hissy. CS800x & CS800s are dead quiet.
I recommend low hiss op amps like MC33078, njm2068, NE5532. Maybe LM4562 on output, although it is a bit high powered and fast for audio. Peavey gets away with NJM4580 for 100' XLR cables from mixer to stage amps.
Have fun building your system.
Your three high level sources have XLR cables, which can pick up RF/hum/pops, so some filtering at the input is indicated. I use 33 pf ceramic to case ground. The balanced input lends itself to op amp in the differential mode to discriminate common mode noise, although you might be able to get away with passive pot attenuation. However, mixing them to the power amp bus requires 1000 ohm minimum mixer resistors to keep from blowing up the drivers in the sources, and that attenuation is usually followed by an op amp to boost the signals back up to 2 vac. 10000 ohm resistors doing the mixing are more usual, there are reasons covered in westsound's long discussions of mixers. No point IMHO mixing both top and bottom of the xlr IMHO. You could use transformers to input the differential signals from the XLR and make them ground referenced, but these are more expensive than op amps, running about $90 each for the good ones (jensen, sennhauser). Instead of mixing the various inputs you could put a rotary stereo switch which is expensive these days, or various LDR attenuators controlled by toggle or slide switches. Farnell/digikey/mouser don't carry stereo rotary switches anymore, but I think dynaco upgrade hubs like dynacodoctor do. the dynaco rotary switches shorted the unused inputs though to prevent pops, which is not compatible with active drive sources. But if you don't follow the LDR attenuators by 1000 ohm resistors & a mixer op amp, if you turn two on by accident, you can blow up both source output (op)amps. Better in that case just to switch cables on a single input xlr jack.
BTW a steel box is required to keep out the RF lightning pops etc. If you put the power transformer in the device, a second steel box or bulkhead is required to keep the hum out of the internal amps. Particularly if you have a 50x gain amp for mm LP. Moving coil LP is more like 80 x gain. I proved this point when I turned a hummy hissy $15 disco mixer into my entertainment hub Improving a disco mixer to mid fi is the name of the thread where I detailed that exploration. & I am at about 75 db s/n not 110 as sgrossklass and PRR talk about.
I safety ground the hub box, and not the power amp to avoid the ground loops case ground to hub case through both the xlr shield and the building wiring. Three of my power amps are single ended and have RCA jacks grounded to case with a 2 prong AC cord. The ungrounded power amps live behind the organ between the speakers and are not a safety hazard because no babies crawl around my floors. Modern power amps with an op amp on the input to differentially reject the common mode hum can have the case grounded. Op amp ICs became useable for low hiss low hum hifi about 1985. Even the ~1980 Peavey cs800 rev A, B, & C that had differential input op amps were too hissy. CS800x & CS800s are dead quiet.
I recommend low hiss op amps like MC33078, njm2068, NE5532. Maybe LM4562 on output, although it is a bit high powered and fast for audio. Peavey gets away with NJM4580 for 100' XLR cables from mixer to stage amps.
Have fun building your system.
Last edited:
"20 years worrying about pots" I bet he'd have a heart attack if he ever saw a mixing desk
Uh that's heavy stuff, still digesting. Many-many thanks for these tips.
For OpAmp selection: one of my friends, an op-amp specialist gave me these hints:
- Opamps have low distortion in Class A mode
- Recommended types are OPA2209 or OPA2140
- PSU: strictly use K-multiplier type, nothing else
Last edited:
I use speakers, and mine have 2nd harmonic distortion down 20 db at all frequencies 54 hz-14khz at 1 W. That is the volume I listen at, and I think guys that claim they can hear .01% HD are full of it. On headphones, maybe, if a guy has never fired a gun or a firecracker without earplugs,never been too close to an airplane or a poorly muffled motor, never used a power saw without earplugs, never been near the stage at a rock/techno/house/ concert without protection. Most men at age 25 have no hearing above 8 khz: mine was tested to 14 khz 11 years ago.
I can hear 1 % hd, my ST70 amp is about that. Fuzzy sound. My "world's worst amp" the ST-120 with djoffe bias mod, and my no-capacitors in the sound path .02% HD CS800x sound exactly the same on these speakers at 1/4 to 1 watt my normal soft passage level.
There is nothing wrong with the OPA2140 except the $6.50 each price for the TSOP package that takes a $2 adapter to be used by ordinary mortals without a wave solder machine. I think my Peavey mixer with 4560 & 4580 op amps sounds just fine, as does my RA-88a after I changed the 4558 to 33078 and purified the power supply hum and straightened out the grounding. 33078 cost me a whole $.38 each and the legs went nicely into the phosphor bronze sockets I installed so I can try some op amp more exotic if I find more purity is in order. Not likely this decade. My next projects are a 30W power amp for my summer camp so I can hear the FM radio while outside & working, and a $38 PV8 mixer for the church so I can mix my wimpy voice from a face mike into the Allen organ speaker system, while I play piano and sing.
My power supply hum problems were solved with a $1 surplus race car 18 v wall transformer, a toroid choke at the inlet to the mixer, 3 470 to 1000 uf caps with intermediate resistors as a pi filter, and two 1n5344 zeners to regulate the +-7.8 voltage into 2 caps for the op amp rails. The biggest fault I see in the RA-88a is not those, it is the approximate RIAA curve that could use much more expensive caps to be more accurate. If they would fit.
I can hear 1 % hd, my ST70 amp is about that. Fuzzy sound. My "world's worst amp" the ST-120 with djoffe bias mod, and my no-capacitors in the sound path .02% HD CS800x sound exactly the same on these speakers at 1/4 to 1 watt my normal soft passage level.
There is nothing wrong with the OPA2140 except the $6.50 each price for the TSOP package that takes a $2 adapter to be used by ordinary mortals without a wave solder machine. I think my Peavey mixer with 4560 & 4580 op amps sounds just fine, as does my RA-88a after I changed the 4558 to 33078 and purified the power supply hum and straightened out the grounding. 33078 cost me a whole $.38 each and the legs went nicely into the phosphor bronze sockets I installed so I can try some op amp more exotic if I find more purity is in order. Not likely this decade. My next projects are a 30W power amp for my summer camp so I can hear the FM radio while outside & working, and a $38 PV8 mixer for the church so I can mix my wimpy voice from a face mike into the Allen organ speaker system, while I play piano and sing.
My power supply hum problems were solved with a $1 surplus race car 18 v wall transformer, a toroid choke at the inlet to the mixer, 3 470 to 1000 uf caps with intermediate resistors as a pi filter, and two 1n5344 zeners to regulate the +-7.8 voltage into 2 caps for the op amp rails. The biggest fault I see in the RA-88a is not those, it is the approximate RIAA curve that could use much more expensive caps to be more accurate. If they would fit.
Last edited:
To #11: not sure if I could place all this in context for myself.. let's try it: I don't want to mix my sources. 
Just switch between them with an input selector (analog).
Mostly the computer would be used, almost 99% of the time.
DAC's XLR Output :
4.5Vrms Fixed at 0 dBFS DAC MODE
0-4.5Vrms Adjustable at 0 dBFS PRE MODE (using it as a digital preamp)
If I connect an analog electronic crossover like this one after the DAC (with values adjusted), that should be enough I think and I can connect the XO's outputs directly with the power amps.
I'm just thinking about where to adjust gains for the channels due to somewhat different speaker sensitivities. Either on the amps themselves (Class D: attenuation, mid/tweeter amps: either attenuation or gain cell as input stage).
A master volume "pot" is only needed for the non-DAC source, else I can adjust my master volume from the DAC itself, using it in PRE mode.
Just switch between them with an input selector (analog). Mostly the computer would be used, almost 99% of the time.
DAC's XLR Output :
4.5Vrms Fixed at 0 dBFS DAC MODE
0-4.5Vrms Adjustable at 0 dBFS PRE MODE (using it as a digital preamp)
If I connect an analog electronic crossover like this one after the DAC (with values adjusted), that should be enough I think and I can connect the XO's outputs directly with the power amps.
I'm just thinking about where to adjust gains for the channels due to somewhat different speaker sensitivities. Either on the amps themselves (Class D: attenuation, mid/tweeter amps: either attenuation or gain cell as input stage).
A master volume "pot" is only needed for the non-DAC source, else I can adjust my master volume from the DAC itself, using it in PRE mode.
That electronic crossover from sound.au.com seems a marvel of op amp technology. Not only splits your sound 3 ways by frequency, it contains buffer amps to adjust the levels to the 3 individual amps. I can't see on the picture or schematic what IC they are using.
If you have two sources you can use a dpdt toggle switch, but if you have 3 you need a rotary or a logic selector for 1 of 3 plus relays or LDRs. I did find a 3 position rotary that would work but no telling if the contacts are gold: https://www.newark.com/electroswitch/d4g0603n/switch-rotary-6p3t-1-5a-115v/dp/06M4673
There is also a 5 position 4 pole version to allow for future expansion for LP + CD sources: https://www.newark.com/electroswitch/d4g0405n/switch-rotary-4p5t-1-5a-115v/dp/06M4668
If you have two sources you can use a dpdt toggle switch, but if you have 3 you need a rotary or a logic selector for 1 of 3 plus relays or LDRs. I did find a 3 position rotary that would work but no telling if the contacts are gold: https://www.newark.com/electroswitch/d4g0603n/switch-rotary-6p3t-1-5a-115v/dp/06M4673
There is also a 5 position 4 pole version to allow for future expansion for LP + CD sources: https://www.newark.com/electroswitch/d4g0405n/switch-rotary-4p5t-1-5a-115v/dp/06M4668
Last edited:
if you have all the high-resolution DSP, DAC following each other, you can control volume from there in the digital domain and skip the volume control.
The easiest route however is to have a pre-amp with selector connected to a power amp. You feed it with DAC or whatever source, like phono, ipod, and send the pre to the power amp.
For the woofer stage you take that power amp output, and feed it to an integrated amplifier with tone controls to filter the high frequency.
This enables you to have smaller cabinets , by using dedicated woofers.
Mid range and tweeter are better used with a XO and driven by separate amps with same pre-signal.
So in the ideal world:
Sources:
AUX 1(iphone)
2(computer)
3 (TV)
PHONO
DAC server
sources -----) pre-amp (volume control + gain +buffer)
Pre-amp has an output splitter.
Pre-signals to a)----- power amp 1 (same gain as power amp 2)---- Tweeter XO
b)------ power amp 2 ----- mid/woof XO
Signal from either power amp 1 and 2 -------- attenuator ---- filter ---- integrated high power subwoofer amplifiers (can be either stereo with 2 subs or /mono/2 channels to mono)
The easiest route however is to have a pre-amp with selector connected to a power amp. You feed it with DAC or whatever source, like phono, ipod, and send the pre to the power amp.
For the woofer stage you take that power amp output, and feed it to an integrated amplifier with tone controls to filter the high frequency.
This enables you to have smaller cabinets , by using dedicated woofers.
Mid range and tweeter are better used with a XO and driven by separate amps with same pre-signal.
So in the ideal world:
Sources:
AUX 1(iphone)
2(computer)
3 (TV)
PHONO
DAC server
sources -----) pre-amp (volume control + gain +buffer)
Pre-amp has an output splitter.
Pre-signals to a)----- power amp 1 (same gain as power amp 2)---- Tweeter XO
b)------ power amp 2 ----- mid/woof XO
Signal from either power amp 1 and 2 -------- attenuator ---- filter ---- integrated high power subwoofer amplifiers (can be either stereo with 2 subs or /mono/2 channels to mono)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.