Discussion:
Frequency response of a guitar amplifier
(too old to reply)
Robert Manktelow
2003-08-07 13:49:24 UTC
Permalink
Hello All
Does anybody know, or have measured, the frequency response of any good
quality guitar amplifiers. I am not interested in bass guitar amplifiers
here.

By frequency response I mean the low & high frequencies where the output
power is -3db down on its centre frequency, measured using a dummy load and
a signal generator.
For example with HI-FI amps this range is typically 20Hz to 20KHz.

Having just measured my home made amp - which is still under construction
(my excuse) - I found out why it has very low hum - it cuts off in the low
hundreds of Hz.

Any tips or pointers welcomed.
--
Robert Manktelow
Phil Crooks
2003-08-07 15:29:58 UTC
Permalink
Post by Robert Manktelow
Hello All
Does anybody know, or have measured, the frequency response of any good
quality guitar amplifiers. I am not interested in bass guitar amplifiers
here.
By frequency response I mean the low & high frequencies where the output
power is -3db down on its centre frequency, measured using a dummy load and
a signal generator.
For example with HI-FI amps this range is typically 20Hz to 20KHz.
Having just measured my home made amp - which is still under construction
(my excuse) - I found out why it has very low hum - it cuts off in the low
hundreds of Hz.
Any tips or pointers welcomed.
The 'speakers are the obvious limiting factor. Freq. response will be
someting like 50Hz to 5KHz (perhaps less). Amplifier frequency response
isn't normally quoted - different amps tailor frequency response to create a
diferent "sound".

Phil
steve at fivetrees
2003-08-07 15:43:44 UTC
Permalink
Post by Robert Manktelow
Having just measured my home made amp - which is still under construction
(my excuse) - I found out why it has very low hum - it cuts off in the low
hundreds of Hz.
Are you talking about the amp itself or the speakers?

From recollection:
- Guitar amp response should be 50Hz or so up to about 12kHz or so. Can
often be higher (essentially HiFi).
- Guitar speaker response should be around 150Hz or so up to about 5-7kHz
or so. (Unless you use horns on your guitar amp, which is generally not a
great idea.)

(All above are very variable and are subject to correction by the
better-informed ;).)

Summary: speakers are usually the main limiting factor.

Steve
http://www.fivetrees.com
http://www.sfdesign.co.uk
Robert Manktelow
2003-08-08 07:46:14 UTC
Permalink
Post by steve at fivetrees
Post by Robert Manktelow
Having just measured my home made amp - which is still under
construction
Post by steve at fivetrees
Post by Robert Manktelow
(my excuse) - I found out why it has very low hum - it cuts off in the low
hundreds of Hz.
Are you talking about the amp itself or the speakers?
- Guitar amp response should be 50Hz or so up to about 12kHz or so. Can
often be higher (essentially HiFi).
- Guitar speaker response should be around 150Hz or so up to about 5-7kHz
or so. (Unless you use horns on your guitar amp, which is generally not a
great idea.)
(All above are very variable and are subject to correction by the
better-informed ;).)
Summary: speakers are usually the main limiting factor.
Steve
http://www.fivetrees.com
http://www.sfdesign.co.uk
Thanks Steve

It was the amp on its own I was asking about. Will get the scope & soldering
out again.
I was most surprised what a dramatic effect the speaker has on the sound
when I changed mine from an Eminience to a Celestion Vintage 30 recently.

Robert
Phil Crooks
2003-08-08 14:02:00 UTC
Permalink
Post by Robert Manktelow
Thanks Steve
It was the amp on its own I was asking about. Will get the scope & soldering
out again.
Who cares?
Post by Robert Manktelow
I was most surprised what a dramatic effect the speaker has on the sound
when I changed mine from an Eminience to a Celestion Vintage 30 recently.
Which explains why amps are designed to work with particular speakers. The
frequency response of the combination (amp and speaker) is far more
important.

Phil
Chris Berry
2003-08-08 21:54:22 UTC
Permalink
Steve, sure the OT is one part of the limitation with respect to "response"
Here's where a lot of the real limitation occurs: Tone stacks.
http://www.duncanamps.com/tsc/index.html
It's really worth the download especially if you want to see what you're
making your amp do.
Who would have thought that our good old friends at Marshall and Fender
seriously limited the tone circuitry to trick us into what sounds good or
not - eh?
As for the speakers here's a good example of response:
http://editweb.iglou.com/eminence/eminence/pages/products02/legend/legm12.htm
You can see that the speaker's response in hifi terms is about 100-1kHz but
it's the artifacts and harmonics above that that contribute greatly to tone.
You can also see that guitar amp manufacturers could save some cash by
putting in a weenie transformer that's efficient from 80Hz up. Would they do
something like that? Hmmm Businessmen.
When talking solid state, the response is entirely different. These guys
would have to pay money to give you a worse response.
cb
Robert Manktelow
2003-08-11 13:00:38 UTC
Permalink
Post by Robert Manktelow
Post by Robert Manktelow
Thanks Steve
It was the amp on its own I was asking about. Will get the scope &
soldering
Post by Robert Manktelow
out again.
Who cares?
Post by Robert Manktelow
I was most surprised what a dramatic effect the speaker has on the sound
when I changed mine from an Eminience to a Celestion Vintage 30 recently.
Which explains why amps are designed to work with particular speakers. The
frequency response of the combination (amp and speaker) is far more
important.
Phil
I was unaware that guitar amps are designed to work with particular
speakers. Presumably the amplifier design also takes account of the speakers
enclosure type (vented/open backed/tuned/..) as well?
Is the aim of the amplifier design to compensate for the non - linearity of
the speaker in question or what?
How would I obtain this information - from the speaker manufacturers?

Robert
Chris Berry
2003-08-11 13:46:07 UTC
Permalink
Post by Robert Manktelow
Post by Robert Manktelow
Post by Robert Manktelow
Thanks Steve
It was the amp on its own I was asking about. Will get the scope &
soldering
Post by Robert Manktelow
out again.
Who cares?
Post by Robert Manktelow
I was most surprised what a dramatic effect the speaker has on the sound
when I changed mine from an Eminience to a Celestion Vintage 30
recently.
Post by Robert Manktelow
Which explains why amps are designed to work with particular speakers. The
frequency response of the combination (amp and speaker) is far more
important.
Phil
I was unaware that guitar amps are designed to work with particular
speakers. Presumably the amplifier design also takes account of the speakers
enclosure type (vented/open backed/tuned/..) as well?
Well the amp part of things is pretty straightforward. It's designed for
cost (you need smaller output transformers for 50Hz-20kHz than you do for
20Hz-20kHz but the rest of the power section is plain vanilla. Then you've
got speaker/amp interactions that should be designed into a system. A good
example is the use of 4x celestion Blues (25W) for use in an overdriven
marshall (120-140W) The saturation is a big part of the sound as is cone
break-up and how harmonics are produced/handled. Here you've got a big
difference between different amplifier types in terms of negative feedback
(often called presence) which basically controls how the amplifier tries to
compensate for differences between its input and output signals. Different
valves also behave differently a difference that you can hear pretty well by
comparing overdriven el34's and 6550's in the same circuit. The key question
is how the amp is designed to operate. Take a classic Marshall - It's not
designed to produce a Fender "clean" sound but can be modified to do so.
Output transformers play a huge role here as well where you have frequency
response as well as saturation properties and resonances. Complex.
The type of tone stack (talking preamps here) - every head has one - is a
lot more complex in its relationship to the choice of speakers. In a tone
stack, you've got a few things happening. The presence control (isn't really
part of the tone stack but is thought of as such) is already covered. Then
based on the different designs, you can have a frequency cut or boost (or
both) at the predefined frequencies. Also there's the general predisposition
of the circuitry to colour the sound e.g. bass and treble rolloff, where
humps are in the frequency response (setting the amp as flat as possible for
instance) and how the controls interact (e.g. a fender mid control does a
lot more than just control the mids and Marshall tone stacks are almost
always V shaped regardless of the knob settings).
Then go into how hot the amp is biased, cabinet design and construction,
cabinet materials (Solid pine is great for baffles for instance), speaker
breakup character, power handling, efficiency... and you've got such a huge
number of variables that you're going to need a huge budget to get your head
around all of it and understand it in practice that it's worth paying for
the experience of someone else to be applied.
Interestingly enough it's the "fudges" that amp makers integrate (try a
sound city 120 or similar) that make the tone have more character
I bet you never knew that "basic good tone" can mean so much being a
statement so small.
cb
Post by Robert Manktelow
Is the aim of the amplifier design to compensate for the non - linearity of
the speaker in question or what?
How would I obtain this information - from the speaker manufacturers?
By reading about all of the above in a lot of detail, looking at the
circuits of a lot of classic amps, understanding them, understanding how
speakers work, understanding materials, tone stacks and of course components
and fudges.

cb
Phil Crooks
2003-08-11 15:45:43 UTC
Permalink
Post by Robert Manktelow
Post by Robert Manktelow
It was the amp on its own I was asking about. Will get the scope &
soldering
Post by Robert Manktelow
out again.
Who cares?
Well, yeah, except it sounds like the OP really does have an amp problem
(cutting off below a few hundred Hz - it should go lower than that,
otherwise the bottom-end response will be, er, absent).
Perhaps I was a little short in my response. Most amps hum (AC heaters) and
many amps have a low frequency roll off (to keep the distortion sounding
"nice"). What, specifically, is the problem?

Phil
Robert Manktelow
2003-08-12 07:49:38 UTC
Permalink
Post by Phil Crooks
Post by Robert Manktelow
Post by Robert Manktelow
It was the amp on its own I was asking about. Will get the scope &
soldering
Post by Robert Manktelow
out again.
Who cares?
Well, yeah, except it sounds like the OP really does have an amp problem
(cutting off below a few hundred Hz - it should go lower than that,
otherwise the bottom-end response will be, er, absent).
Perhaps I was a little short in my response. Most amps hum (AC heaters) and
many amps have a low frequency roll off (to keep the distortion sounding
"nice"). What, specifically, is the problem?
Phil
I have an old (1960 ish) 25 watt valve PA amplifier made by Grampian which I
have converted to a guitar amplifier by replacing the pre-amplifier/tone
controls circuits with the "Verberator" by Harmonic Appliances front end.
The phase invertor / output valves / OP transformer & PSU remain unchanged.
Upon testing the modified amplifier with a signal generator / dummy load /
scope I found the lower
-3db point was approx 500Hz with the controls flat!
There may be a wiring error causing this early cut-off however I realised I
did not know what frequency response I should be aiming for - hence the
question.
The speaker to be used is a Celestion Vintage 30 in a open back cabinet made
of 15 mm plywood.

Robert
Phil Crooks
2003-08-12 11:32:59 UTC
Permalink
Post by Robert Manktelow
Post by Phil Crooks
Post by Robert Manktelow
Post by Robert Manktelow
It was the amp on its own I was asking about. Will get the scope &
soldering
Post by Robert Manktelow
out again.
Who cares?
Well, yeah, except it sounds like the OP really does have an amp problem
(cutting off below a few hundred Hz - it should go lower than that,
otherwise the bottom-end response will be, er, absent).
Perhaps I was a little short in my response. Most amps hum (AC heaters)
and
Post by Phil Crooks
many amps have a low frequency roll off (to keep the distortion sounding
"nice"). What, specifically, is the problem?
Phil
I have an old (1960 ish) 25 watt valve PA amplifier made by Grampian which I
have converted to a guitar amplifier by replacing the pre-amplifier/tone
controls circuits with the "Verberator" by Harmonic Appliances front end.
The phase invertor / output valves / OP transformer & PSU remain
unchanged.
Post by Robert Manktelow
Upon testing the modified amplifier with a signal generator / dummy load /
scope I found the lower
-3db point was approx 500Hz with the controls flat!
There may be a wiring error causing this early cut-off however I realised I
did not know what frequency response I should be aiming for - hence the
question.
The speaker to be used is a Celestion Vintage 30 in a open back cabinet made
of 15 mm plywood.
Robert
I tried to find a schematic for the Verbarator but the Harmonic Appliances
web site seems to be unavailable at the moment. I think someone mentioned
this already but you can get some idea of the effect of your tone stack by
using this marvellous piece of software called "Tone Stack Calculator" at
http://www.duncanamps.com/tsc/index.html. It may help.

What do you mean when you say that the tone controls are flat? Guitar tone
stacks - usually - have a very pronounced mid notch caused by the mid
frequencies being cut more than any other frequencies. There is a possiblity
that this is what you are seeing (500Hz could easily fall into the notch).
The TSC program will help you see that.

Other than that, low end cut can be caused by cathode bypass caps that are
too small (22uF is the norm in the preamp) or small interstage coupling caps
47nF is a fairly normal value here (could be 22 or 100). The output
transformer can also cut bass (a physically small transformer probably won't
manage a lot of bass). Having said that 500Hz is very high to be seeing
significant bass roll off caused by these factors.

Guitar amps (and speakers) are very non-linear in their sound reproduction
so there is no right answer to frequency response. I understand that Denis
Cornell tunes his amps to your speaker of choice AND your guitar. It's how
it sounds not what the specs say. How does your amp sound?

Phil
Robert Manktelow
2003-08-12 14:43:53 UTC
Permalink
Post by Robert Manktelow
Upon testing the modified amplifier with a signal generator / dummy load /
scope I found the lower
-3db point was approx 500Hz with the controls flat!
500Hz is definitely too high.
Have you tried measuring the bandwidth at the junction of the preamp and
power amp? This may tell you which of the two sections is the problem.
Steve
http://www.fivetrees.com
http://www.sfdesign.co.uk
The tone controls used in the Verberator were, I understand, designed using
Mr Duncans program and are of the Baxendall type. I.e. a bass and treble
centred on the guitar frequency range, not hi-fi. These provide both cut and
boost with central settings giving (but not in my case) a "flat" response
all other things being equal.The web site, when up again is excellent.

Anyway i have just carried out some more measurements along the pre-amp
section and the attenuation is in the tone controls. The power stage is
faithfully reproducing what it gets and the first valve, an ef86, is
providing a linear response down to aprox 30Hz.
So it's likely to be my wiring that is at fault, oh dear.

With the tone controls in the flat position the amplifier + guitar sound
good at low levels but degrades at increasing levels - goes "tiney". I need
a better location (not my house) to really quantify this.
Will figure out what is going wrong first, fix it, and then re-do the
listening tests.

Thank you all for your help on the amplifier frequency response question.
I am intrigued with tuning an amplifier to match its speaker cabinet. Is the
power stage's frequency response modified, by the use of frequency dependant
feedback, to achieve this?
Any schematics available to look at?

Robert
Phil Crooks
2003-08-12 14:59:52 UTC
Permalink
Post by Robert Manktelow
Post by Robert Manktelow
Upon testing the modified amplifier with a signal generator / dummy
load
Post by Robert Manktelow
/
Thank you all for your help on the amplifier frequency response question.
I am intrigued with tuning an amplifier to match its speaker cabinet. Is the
power stage's frequency response modified, by the use of frequency dependant
feedback, to achieve this?
Any schematics available to look at?
Very few guitar amps use feedback (Vox AC30 being a very notable exception).
Where obvious tonal shaping occurs (using RC filters) it is usually to keep
distortion under control and this trend is continued by Celestion who are
famous for producing rawk speakers that neatly filter out some of the
nastier artifacts of heavy clipping within the amplifier. The Vintage 30
that you are using is excellent at the job. So good, in fact, that I had to
sell mine - it was starting to sound "closed in" and middly to my ears (the
frequency response of my ears may have something to do with this). Vintage
amplifiers (particularly Fenders) are much more straight forward in design;
if they produced distortion at all it was as a by-product rather than as a
deliberate "effect". They tended to use "hi-fi" speakers. Things have
changed since then.

The three channel Bogner amp has an interesting schematic showing three very
different preamps. Other, simpler designs, that may be of interest are
Soldano and old Fender.

Hope the amp rework goes well (I do like the sound of an input pentode),

Phil
Chris Berry
2003-08-14 17:01:46 UTC
Permalink
Post by Phil Crooks
Very few guitar amps use feedback (Vox AC30 being a very notable exception).
Taking Fender:
Bassman: NFB
Tweed deluxe: NFB
Deluxe reverb: NFB
Marshall:
1959: NFB
...
The only non NFB amp that comes to mind is a gibson GA6
Negative feedback is most often used as a presence control.

Kind of want you to define very few...

cb
Robert Manktelow
2003-08-18 08:08:08 UTC
Permalink
To close this thread off, problem found.

There is a missing decoupling capacitor in the schematic of the Verberator
pre-amp I used.
This results in the poor low frequency response, moving the bass control pot
changes the DC bias on the valve after the tone controls.

Add capacitor and the floor moves, for me anyway :)

Robert
(snip)
Post by Robert Manktelow
-3db point was approx 500Hz with the controls flat!
Robert, depending on the tone circuit, flat may be at 0, 12%, 30% 50%
100% -
almost anything.
The job of the tone circuit is to allow you to "dial in" a certain
response.
A marshall has a pretty flat response (within 3 db) for all of the scale
(25-10kHz+) with the controls at (Bass, Mid, treble) 38%, 100% 1% while a
fender type is flat at 3%, 100%, 0% so you've basically got different
types
of controls...
cb
Phil Crooks
2003-08-19 11:23:50 UTC
Permalink
Post by Robert Manktelow
To close this thread off, problem found.
There is a missing decoupling capacitor in the schematic of the Verberator
pre-amp I used.
This results in the poor low frequency response, moving the bass control pot
changes the DC bias on the valve after the tone controls.
Add capacitor and the floor moves, for me anyway :)
That's great. Any chance of a circuit diagram or sound samples one day?

Phil
Robert Manktelow
2003-08-19 16:12:04 UTC
Permalink
Post by Phil Crooks
Post by Robert Manktelow
To close this thread off, problem found.
There is a missing decoupling capacitor in the schematic of the Verberator
pre-amp I used.
This results in the poor low frequency response, moving the bass control
pot
Post by Robert Manktelow
changes the DC bias on the valve after the tone controls.
Add capacitor and the floor moves, for me anyway :)
That's great. Any chance of a circuit diagram or sound samples one day?
Phil
Happy to share the schematic, but at present its a series of pencil drawings
in a note book.

Does anybody have recommendations of Windows programs (preferably free) for
drawing valve amplifier schematics that can produce, say, .pdf format
documents or other "universal" formats.

The amp is still very much 'work in progress' - learning as I go.
Am aiming towards a Dr Z Route 66 type sound.
John Mayall's guitarist used one of these locally, with a Strat and no
pedals, and the range of sounds he dragged from it was astounding - but then
he was *very* good.

Robert
Richard Kaulfuss
2003-08-19 23:56:15 UTC
Permalink
On Tue, 19 Aug 2003 20:08:31 +0100, "steve at fivetrees"
Post by Robert Manktelow
Does anybody have recommendations of Windows programs (preferably free)
for
Post by Robert Manktelow
drawing valve amplifier schematics that can produce, say, .pdf format
documents or other "universal" formats.
Ouch. Tough call. I wish I knew.
Schematic capture programs tend to be expensive: Orcad is my favourite, but
it's around £2000. There is an Orcad-Lite version which is available for
download; it will restrict you to quite small drawings, but if you do it in
a modular fashion (always a good idea), it might work for you.
The alternative is to use a drawing package (e.g. Visio), but this sucks.
Orcad "understands" schematic capture, and can do a lot more than just draw
lines (design rule checks, netlists, bills of material etc). A drawing is
just a bunch of lines, a picture, with no understanding of what is actually
being drawn. And it's really a pain to do things this way (I've done both).
If you find a decent (free/cheap) schematic capture for either Windows or
*nix, I'd love to know about it.
Try SIMetrix. (http://www.catena.uk.com/Pages/download.html)
The demo version has only a few restrictions and is adequate for
most small projects.
--
Dick
steve at fivetrees
2003-08-20 02:23:17 UTC
Permalink
Post by Richard Kaulfuss
Try SIMetrix. (http://www.catena.uk.com/Pages/download.html)
The demo version has only a few restrictions and is adequate for
most small projects.
Thanks - underway ;).

Steve
http://www.fivetrees.com
http://www.sfdesign.co.uk

Loading...