One other minor point: a typical vacuum cleaner generates about 70 dB SPL, according to a number of sources I looked up. I consider that about the loudest sound that typical neighbours in a typical apartment block will put up with.
With that benchmark, we can calculate how much power our "apartment-friendly" guitar amp should put out.
As an example, I have a speaker with a nominally 91 dB sensitivity - it puts out 91 dB at 1 metre when driven with 1 watt of audio power. Reducing the power by a factor of ten (-10 dB) will reduce SPL to 81 dB, still way louder than a vacuum cleaner - and we're now talking about one-tenth of a watt from the amp!
Reducing that 81 dB by another 10 dB gets us to 71 dB, and now we're in vacuum-cleaner noise level territory. We are also talking about driving the speaker with only one-hundredth of a watt of power!
I've approached this estimate from several different directions, and always end up with the same conclusion: particularly for heavily overdriven tones, with a typical guitar speaker, we don't need a 5-watt amp, we need an amp with output power somewhere between one tenth and one hundredth of a watt. 10 to 100 milliwatts!
-Gnobuddy
With that benchmark, we can calculate how much power our "apartment-friendly" guitar amp should put out.
As an example, I have a speaker with a nominally 91 dB sensitivity - it puts out 91 dB at 1 metre when driven with 1 watt of audio power. Reducing the power by a factor of ten (-10 dB) will reduce SPL to 81 dB, still way louder than a vacuum cleaner - and we're now talking about one-tenth of a watt from the amp!
Reducing that 81 dB by another 10 dB gets us to 71 dB, and now we're in vacuum-cleaner noise level territory. We are also talking about driving the speaker with only one-hundredth of a watt of power!
I've approached this estimate from several different directions, and always end up with the same conclusion: particularly for heavily overdriven tones, with a typical guitar speaker, we don't need a 5-watt amp, we need an amp with output power somewhere between one tenth and one hundredth of a watt. 10 to 100 milliwatts!
-Gnobuddy
two comments:
1.) the amp needs some range (overhead) to cope with transients above the average levels from your source/playing. For a guitar performance that could be +20B or maybe even as high as +30dB.
2.) the vacuum cleaner noise might be 70dBspl but at what distance? 10m or 100m? and at what filtering, C or A, or some other?
What is it at 1m to be comparable to your 91dB speaker?
You can generate some pink noise and input that to your speaker. Adjust until you have 0.283Vac, as best as your DMM can manage with that big range of frequencies. That test noise is -20dB ref 1W (i.e. 10mW) and should let you hear what 71dB @ 1m sounds like in the next room with the doors closed, most of the noise leakage will come around the gaps in the door opening.
1.) the amp needs some range (overhead) to cope with transients above the average levels from your source/playing. For a guitar performance that could be +20B or maybe even as high as +30dB.
2.) the vacuum cleaner noise might be 70dBspl but at what distance? 10m or 100m? and at what filtering, C or A, or some other?
What is it at 1m to be comparable to your 91dB speaker?
You can generate some pink noise and input that to your speaker. Adjust until you have 0.283Vac, as best as your DMM can manage with that big range of frequencies. That test noise is -20dB ref 1W (i.e. 10mW) and should let you hear what 71dB @ 1m sounds like in the next room with the doors closed, most of the noise leakage will come around the gaps in the door opening.
Last edited:
Good analysis for 99% of Hi Fi users, but NOT applicable to a Rock Guitar.
Which needs 0 (zero) dB dynamic range, the user *wants* to clip the guitar sound.
Correct that to heavily clip it.
This is what you get from a 20W Marshall guitar amp when overdriven:
An externally hosted image should be here but it was not working when we last tested it.
Guitar amps are driven by signal typically 10 to 20dB *above* the clipping point.
Surely your analysis only applies to the transients. When the playing is at moderate volume, then the clipping does not happen and the sound changes.
Adding some -10dB before the amp will have some of the effect of reducing the clipping. Again that changes the sound production.
And if he plays quietly to avoid disturbing the neighbours he ends up even further from clipping.
How does one set up the guitar preamp so that consistent sound comes from the valve poweramp & speaker?
Adding some -10dB before the amp will have some of the effect of reducing the clipping. Again that changes the sound production.
And if he plays quietly to avoid disturbing the neighbours he ends up even further from clipping.
How does one set up the guitar preamp so that consistent sound comes from the valve poweramp & speaker?
Last edited:
Interesting thread. I'm in a similar situation where I love the sound of my 2W amp when the master volume is wide open and the push-pull output stage is starting to saturate.
My speaker comes in around 98dB@1W, so 2W is 101dB. Too loud for home use. 2mW would be better.
However...
With a valve amp, you've got a fairly linear output impedance (maybe a little inductive towards the top end) driving a very nonlinear load. The frequency response will vary with the impedance of the speaker. That's fine, because we like that sound.
How do we go about designing something that absorbs 99.9% of the output power, while maintaining the same driving impedance as the amplifier?
Chris
My speaker comes in around 98dB@1W, so 2W is 101dB. Too loud for home use. 2mW would be better.
However...
With a valve amp, you've got a fairly linear output impedance (maybe a little inductive towards the top end) driving a very nonlinear load. The frequency response will vary with the impedance of the speaker. That's fine, because we like that sound.
How do we go about designing something that absorbs 99.9% of the output power, while maintaining the same driving impedance as the amplifier?
Chris
Fixed bias (grid biased) amps need the aux bias supply. Cathode-biased amps don't. For cathode-biased amps, you can use Richie Hall's VVR, which works along similar lines, or Trinity Amp's VRM. Both are very effective.
For in-ear monitored gigging, I've used an 8 ohm resistive load on the output of a 10W tube amp with a Behringer GI100 DI box, which simulates the frequency curve of 4X12 Greenback cab. Still sounds harsh with distortion. So I don't think the problem is related to Fletcher-Munsters.
I think I should now look into Randall Aiken's dummy load design. For a cheap and cheerful attenuator, I've also had fairly decent results with those auto-transformer type speaker volume controls, designed for in-wall mounting. However, at extreme attenuation, they get a little mushy.
Last edited:
With due respect, you clearly are not into Rock Guitar
The OP does not just want to "play quiet", the standard volume control is more than enough to achieve that, but to heavily overdrive his tube power amp and then attenuate *that*.
Of course, that produces unadjustable "fixed volume" output, so only way to attenuate and control room volume/SPL is to place an attenuator (passive or active) between power amp and speakers.
The second option is to apply a variable power supply feeding the power amp, so voltage is heavily reduced and the poorly fed power amp, not only reduces output but overdrives even more.
The third option is to use a speaker where the magnetic field can be reduced to almost zero.
I understand it´s a "crazy" concept for Hi Fi lovers
But the OP **WANTS** power stage clipping !!!!!
Like Pappo Napolitano, one of our guitar heroes once said: "you get a (MarkII - Plexi) Marshall, set all knobs to 10 and start hammering."
Here´s a couple classic examples: the amp does not clip "on transients" but all the time, even during the quieter passages.
https://www.youtube.com/watch?v=424Hv4dg3Yw
It is quite difficult to find a number of acceptable wattage - just because musical taste differs.
One simple example that opened my eyes many years ago. At that time I played in a commercial band and during a saturday nite gig some neighbours complained. So a Policeman with a phonemeter was called and measured about 96dB on the dance floor.
He claimed this was an acceptable level under these circumstances.
A short time later a marching band came in , with drums, horns section etc.
I called the man to measure again, and he read out 106dB.
Needless to say that nobody complained.
So thats it with the numers
One simple example that opened my eyes many years ago. At that time I played in a commercial band and during a saturday nite gig some neighbours complained. So a Policeman with a phonemeter was called and measured about 96dB on the dance floor.
He claimed this was an acceptable level under these circumstances.
A short time later a marching band came in , with drums, horns section etc.
I called the man to measure again, and he read out 106dB.
Needless to say that nobody complained.
So thats it with the numers
Guitarists tend to use both "clean tones" and "overdriven tones". As J.M. Fahey said, there is very little dynamic range when it comes to heavily overdriven sounds, because the amp is being pushed far into clipping.
But clean tones do have some dynamic range. Empirically, by ear, I found that I could use roughly 10 times as much power for clean tones, to sound about equally loud (or equally annoying, if you are the neighbour next door trying to take a nap!)
With a typical vintage guitar speaker (around 90 dB/W @ 1m), and shooting for "no louder than a vacuum cleaner", that still works out to somewhere in the ballpark of 10 mW for overdriven tones, and 100 mW for clean tones.
Keep in mind that there are plenty of modern guitar speakers that are actually much more sensitive than this. As an example, I've attached the manufacturer's published frequency response for the Eminence Rajin' Cajun - it has a sensitivity around 98 dB/W@1m over much of the guitar frequency range, and over 105 dB/W@1m in the range where the human ear is most sensitive, around 3 kHz.
( Here's a direct link to the manufacturer spec-sheet: https://www.eminence.com/pdf/Ragin_Cajun.pdf )
At 98 dB speaker sensitivity, you need a whopping 1.6 mW of audio power from the guitar amp to get 70 dB SPL at a 1-metre distance!
I think AndrewT has a good point that some sort of frequency weighting does matter, and Voltwide has an even better point about the social acceptability (or otherwise!) of certain types of sounds. There is also the question of time of day - most neighbours will tolerate the sound of a vacuum cleaner during the day, but try vacuuming at midnight, and you might find the police at your front door.
And when do many of us want to play guitar at home? Yup, late at night, after the work and the chores are finally done. At this time of day, I would suggest that the appropriate "don't annoy the neighbours!" SPL reference level is that of normal spoken conversation. That's in the ballpark of 60 dB SPL, according to various online sources I found. So now we can knock those guitar amp power levels down one more order of magnitude; a thundering 160 microwatts for the Rajin Cajun.
Because there are so many inexact variables, I don't think there is any way to calculate exactly the "right" guitar amp power. The point I'm trying to make is a simple order-of-magnitude estimate: for quiet home usage, with an actual guitar speaker, we should be looking at milliwatts of guitar amp power output, not watts.
You can use Hi-Fi speakers, which are typically 10 dB less sensitive. It's also perfectly possible to attenuate any guitar amp down to milliwatt output powers, of course. I've tried both, and my amps usually sound utterly uninspiring once attenuated that low.
And this is usually the point at which people throw up their hands, and go get a pair of headphones and an amp-modelling guitar pedal or software for their laptop.
I've tried the modelling pedal approach too, and my guitar strap, guitar cable, and headphone cable always seem to get into some sort of power struggle. I spend more time trying to disentangle them than actually playing guitar.
I tried Bluetooth headphones, too. Too much latency. You play a note on the guitar now, and hear it in your ears a fraction of a second later. It makes it impossible to play anything fast or complex.
Old-fashioned analogue FM headphones would probably solve that problem. I haven't tried those.
In the end, this might simply be a case of "You can't always get what you want", as the Rolling Stones put it.
-Gnobuddy
Attachments
This is actually not too hard, if you know (or can measure) the output impedance of your guitar amp. You can use the three-resistor L-pad schematic I drew up and attached.
For large attenuation factors (like the 99.9% you describe), you can use a simplified design method. Start by setting R3 equal to your guitar amp output impedance.
Now calculate the attenuation caused by R3 interacting with the speaker impedance. Suppose this is 10 dB, and you need a total of 20 dB.
That means R1 and R2 have to provide the remaining 10 dB of attenuation. R1 plus R2 should equal 8 ohms (or whatever your amp is expecting). R2/(R1+R2) should -10 dB, so you can now calculate R1 and R2. (For large attenuation factors, R1 will end up basically the same as Rspkr, and R2 will decide the attenuation amount.)
Keep in mind we don't need exact values here. For practical purposes, assuming your amp is happy feeding an 8 ohm speaker, I would just make R1 6.8 ohms, and fiddle with R2 to get the attenuation you need.
If you try this, I hope you'll tell us how you liked the result. (I'm pessimistic about how good the sound will be, for all the reasons discussed previously on this thread. But I may be wrong.
)-Gnobuddy
Attachments
That 99.9% attenuation is -30db! I reckon most would really want a bit more volume than that.
I was thinking about comparative tones and volumes around the volume knob dial on the amp. On my amps, I go for a tone that is fairly clean, with dirt on it. The best sounds at high volume are probably mostly cleanish with preamp crunch, with some significant power-stage distortion but probably not dominating, say 20% THD (wild guess). ie, the master volume control is still effective at changing volume.
The MV's on both my amps are post phase invertor, which is significant and allows the PI tube to be driven harder, t capture the first part of what power-stage distortion is, even at low volume.
Lets say the tone I like best is at 7 on the volume pot. The Vol pot is an A audio taper, so lets say 10% at mid turn, 5. It has a bi-linear taper. So 7 is 46% (call it 50%).
Suppose I want to attenuate down to make that tone but with the volume as if it was at 2 on the volume. At 2, the volume pot is at around 4%. The attenuation I need is about x 0.08 on voltage signal, which is 20log0.08 -22db for power.
But what ive built is much less than that, its job is not too do the whole thing, but just to lift the volume setting off the floor to let the MV do most of the work.
I was thinking about comparative tones and volumes around the volume knob dial on the amp. On my amps, I go for a tone that is fairly clean, with dirt on it. The best sounds at high volume are probably mostly cleanish with preamp crunch, with some significant power-stage distortion but probably not dominating, say 20% THD (wild guess). ie, the master volume control is still effective at changing volume.
The MV's on both my amps are post phase invertor, which is significant and allows the PI tube to be driven harder, t capture the first part of what power-stage distortion is, even at low volume.
Lets say the tone I like best is at 7 on the volume pot. The Vol pot is an A audio taper, so lets say 10% at mid turn, 5. It has a bi-linear taper. So 7 is 46% (call it 50%).
Suppose I want to attenuate down to make that tone but with the volume as if it was at 2 on the volume. At 2, the volume pot is at around 4%. The attenuation I need is about x 0.08 on voltage signal, which is 20log0.08 -22db for power.
But what ive built is much less than that, its job is not too do the whole thing, but just to lift the volume setting off the floor to let the MV do most of the work.
If it works for you, you're already way ahead of the game compared to most of us who are looking for the same thing.
At night, I cannot strum an acoustic guitar in my apartment without bothering neighbours. So the electric-guitar solution I've been hunting for has to be much quieter than that.
So far, I've not had much luck finding anything that really nails it. The two approaches that come closest for me, are my DIY guitar amp, set up for preamp distortion, master volume well down, and with the speaker set at ear-height less than a metre from my head; or the "attenuate, EQ, re-amp" approach I mentioned earlier.
-Gnobuddy
Yes...some neighbours with a cleaning fetish who love to vacuum late at night.
Our neighbour is 100m away. I only have to not annoy my wife too much or else she'll want another horse.
Didn´t know you were talking 3AM levels
If so, I published this simple one a couple yers ago:
FWIW I am deeply aware of neighbour problems.
So much so that I had to *move* from Recoleta, a posh Buenos Aires neighbourhood, to La Boca, the marginal old Port area, complete with surrounding naval yards, ship engine shops, and wool - cotton - raw leather - shipping container warehouses and deposits all around.
But now I can test 300W amplifiers in an open patio at 3 AM, literally, with no complaints.
Well, practically nobody lives 60 to 100 meters around me :O
No "families" I mean, only shops one kind or another.
If so, I published this simple one a couple yers ago:
An externally hosted image should be here but it was not working when we last tested it.
FWIW I am deeply aware of neighbour problems.
So much so that I had to *move* from Recoleta, a posh Buenos Aires neighbourhood, to La Boca, the marginal old Port area, complete with surrounding naval yards, ship engine shops, and wool - cotton - raw leather - shipping container warehouses and deposits all around.
But now I can test 300W amplifiers in an open patio at 3 AM, literally, with no complaints.
Well, practically nobody lives 60 to 100 meters around me :O
No "families" I mean, only shops one kind or another.
Last edited:
Oh, the neighbours are all really nice. They're Canadian, remember. It's the defining national characteristic.
I've never had a neighbour actually complain - and I intend to keep it that way. I'm not about to try and find out how much of a jerk I can be, before someone loses their patience and calls the cops on me. (I'm not Canadian, yet, but clearly, I have the right personality type to become one.
)Where I live, a stand-alone house on it's own land costs somewhere north of a million bucks. Shared housing is the best any working stiff can hope for - and you're lucky to be able to afford that.
So my guitar-playing situation is just the nature of living in housing with shared walls and ceilings and floors. It's an increasingly common situation for more and more people, as our cities become increasingly crowded, and more and more people are forced to live in crowded pigeon-coops that our ancestors would have despised. Yamaha has a whole line-up of "silent" (they're not really silent, they're quiet) instruments, designed for just these sorts of living situations: Yamaha Silent Instruments - Quietly Making Your Mark | Dolphin Music
-Gnobuddy
Well, even if it's not 3 AM, when it comes to shared housing, I think benchmark home SPL levels should be based on "normal" sounds that most neighbours would accept and not be upset about. Benchmarks like a home vacuum cleaner, TV audio, human speech, maybe a small dinner-party.Didn´t know you were talking 3AM levels
We had exactly the same idea. See post #30!
Now, that's a rare thing! I used to rent a small suite on a 19-acre farm, surrounded by other farms. Even there, I would not have dared to play even a 15-watt amp full blast at 3 AM!
-Gnobuddy
I've used the same Behringer DI box for home recordings, straight from my old Superchamp XD to my multi-track Zoom recorder. With an 8 ohm resistive load for the amp, of course.
It worked well for clean tones for me, though I found I preferred the sound with the (Behringer DI's) cab emulation switched off.
If this sounds harsh with distortion, why not try following it with a graphic or parametric EQ? Those things can work sonic wonders sometimes.
I bought one at a surplus outlet once, but never dared to use it on my (valve) guitar amps, because I wasn't sure how it worked, particularly whether the primary impedance would stay anywhere near 8 ohms as you twiddled the attenuation knob. I was concerned that it was just a variable-ratio stepdown transformer, and therefore that the primary impedance might go very high at larger attenuation settings.
Do you happen to know how these work, and/or if my worry was justified?
-Gnobuddy
Strap a 20ohm resistor across the primary, it'll be fine.
Thanks for the hint on the 3-resistor L-pad. I'd been doodling something similar but wasn't sure how well it'll work.
With my 2w amp, I can use small components and probably add a bypass switch!
Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.