**Hi,

what specs does an audio generator need to make it ideal for servicing guitar amps? I suspect most would say practically any budget priced bench generator will be fine - but is it ?

Most guitar amps have very high gain, wide range tone controls and built in treble boost. They also can output plenty of hum, noise and harmonic distortion all of which is unlike regular audio gear.

So, there are some special needs the generator ought to meet.

1. Ability to adjust the output level to well below 1mV and smoothly up to several volts.

2. Circuitry not linked to the AC safety ground to avoid hum loops.

3. Have a fixed level output to connect to the external trigger input on a scope - so the trace remains in synch despite noise, hum and harmonics.

3. Square wave output for checking tone circuits and amplifier stability.

4. Frequency from 10Hz to 100kHz, preferably in four overlapping ranges.

5. Amplitude remains steady when frequency setting or range is changed.

6. THD not more than 0.1% with low order harmonics.

If #6 is not met, then those amps that have large amounts of fixed treble boost (like Mesa Boogie) will amplify the harmonics coming from the generator far more than the fundamental - resulting in a dirty looking wave on the scope. Function generators are best avoided.

BTW: use an analogue scope - using DSOs with guitar amp repairs just makes life hard for yourself.


..... Phil

On Mon, 29 Dec 2014 16:32:16 -0800, Phil Allison wrote:

**Hi,

what specs does an audio generator need to make it ideal for servicing
guitar amps? I suspect most would say practically any budget priced
bench generator will be fine - but is it ?

Most guitar amps have very high gain, wide range tone controls and built
in treble boost. They also can output plenty of hum, noise and harmonic
distortion all of which is unlike regular audio gear.

So, there are some special needs the generator ought to meet.

1. Ability to adjust the output level to well below 1mV and smoothly up
to several volts.

2. Circuitry not linked to the AC safety ground to avoid hum loops.

3. Have a fixed level output to connect to the external trigger input on
a scope - so the trace remains in synch despite noise, hum and
harmonics.

3. Square wave output for checking tone circuits and amplifier
stability.

4. Frequency from 10Hz to 100kHz, preferably in four overlapping ranges.

5. Amplitude remains steady when frequency setting or range is changed.

6. THD not more than 0.1% with low order harmonics.

If #6 is not met, then those amps that have large amounts of fixed
treble boost (like Mesa Boogie) will amplify the harmonics coming from
the generator far more than the fundamental - resulting in a dirty
looking wave on the scope. Function generators are best avoided.

BTW: use an analogue scope - using DSOs with guitar amp repairs just
makes life hard for yourself.


.... Phil

Do you have a favourite?

On Monday, December 29, 2014 4:32:20 PM UTC-8, Phil Allison wrote:
**Hi,

what specs does an audio generator need to make it ideal for servicing guitar amps?

In the 'convenience' category, consider an iPod with a few recorded test tracks.

1. Ability to adjust the output level to well below 1mV and smoothly up to several volts.

Probably that means you need calibration of the output, too? A ten-turn pot and a
trueRMS meter would be helpful.

2. Circuitry not linked to the AC safety ground to avoid hum loops.

Battery powered, then; the iPod and an old HP204C (Wien bridge) generator would fit.
Or, anything with a good signal-level transformer.

3. Have a fixed level output to connect to the external trigger input on a scope - so the trace remains in synch despite noise, hum and harmonics.

Any generator, with an external attenuator, will do that.

3{a}. Square wave output for checking tone circuits and amplifier stability.

Or, maybe a white-noise output? Mainly, (3) and (3a) will require an arbitrary waveform
generator, nowadays; other sine/square combos are mainly inaccurate.

4. Frequency from 10Hz to 100kHz, preferably in four overlapping ranges.

Arb or HP204 will qualify.

5. Amplitude remains steady when frequency setting or range is changed.
Arb should do that.

6. THD not more than 0.1% with low order harmonics.

After warmup, the HP204 will do that. Or, an expensive arb. Or, a good CD player.

I can't think of any single item that does all these (but simple, easily
available equipment can do any of 'em).

whit3rd wrote:

**Hi,

what specs does an audio generator need to make it ideal for servicing guitar amps?

In the 'convenience' category, consider an iPod with a few recorded test tracks.

** Several times worse than useless.



1. Ability to adjust the output level to well below 1mV and smoothly up to several volts.

Probably that means you need calibration of the output, too?

** Nope. Markings on a dial are plenty accurate.


2. Circuitry not linked to the AC safety ground to avoid hum loops.

Battery powered, then;

** Nope. A plug-pak supply woul be fine.




3{a}. Square wave output for checking tone circuits and amplifier stability.

Or, maybe a white-noise output?


** Worse than your other mad ideas.


5. Amplitude remains steady when frequency setting or range is changed.
Arb should do that.

6. THD not more than 0.1% with low order harmonics.

After warmup, the HP204 will do that.

** You may have one of them, but they are not plentiful worldwide.


I can't think of any single item that does all these

** I had to build my own ideal bench generator.

Combination of a 2Hz to 200kHz low THD oscillator, squarer circuit, 1000:1 stepped attenuator, fine frequency adjust, fixed level transformer isolated output for scope synch, another fixed output for the frequency counter.

Amplitude stabilisation is selectable between a glass bead thermistor or a
diode limiter - the latter resulting in zero bounce but 0.2% THD.

Frequency adjustment uses a log wound, double gang 15kohms WW pot - for stability and low noise.

THD of the oscillator is 0.002% and with the aid of a -100dB notch filter I can test nearly anything.



..... Phil



.... Phil

On Mon, 29 Dec 2014 17:32:16 -0700, Phil Allison
wrote:

...snip...
1. Ability to adjust the output level to well below 1mV and smoothly up
to several volts.

2. Circuitry not linked to the AC safety ground to avoid hum loops.

3. Have a fixed level output to connect to the external trigger input on
a scope - so the trace remains in synch despite noise, hum and harmonics.

3. Square wave output for checking tone circuits and amplifier stability.

4. Frequency from 10Hz to 100kHz, preferably in four overlapping ranges.

5. Amplitude remains steady when frequency setting or range is changed.
6. THD not more than 0.1% with low order harmonics.

..snip..

You can do all this with some very inexpensive stuff. To cover the whole
range from 10Hz to 100kHz [well more like 10Hz to 89kHz, or any specific
smaller range you want] with a little C/C++ coding and the ASIO interface
to directly work with a 24 bit soundcard, like Creative Labs EMU1212 [old
and cheap], you can make a generator to pretty much accomplish everything
you want.

You can drive the generator with variable amplitude, with a set of
simultaneous tones that will show you passband, phase, and distortion. Use
a scope, or the soundcard's inputs to get a lot of information. From
memory the drive is low noise, better than 16 bits with amplitude fairly
well calibrated. The output is like a pure tone accurately/repeatably set
to selectable amplitude.

For passband, I suggest driving with an independent set of tones, not
square wave, but tones that just cover the complete range. Monitor the
amp's drive with one input, monitor the amp's output with the other, do
FFT's on both channels preserving the phases. Then comparing the FFT's
will provide a great deal of information regarding not just passband and
phase shift, but also the amount of distortion AND noise in the amp. Plus,
you can even measure the spectral noise across the band, spot excessive
[well, relatively excessive] noise peaks, like AC mains related.

I have my own set up and software for this, but more readily available
with a simple to use GUI try Bob Masta's SW and use your own soundcard
already in your system. The SW is small and non-obtrusive to a system. I
think there's a free trial time for the full blown and if purchased you
keep with rights to ALL future downloads. If you opt out after trial you
get to keep some features anyway, like the generator capability, a very
useful generator. Also this version has LCR Meter capability.

DAQARTA v7.60
Data AcQuisition And Real-Time Analysis
http://www.daqarta.com
Scope, Spectrum, Spectrogram, Sound Level Meter
Frequency Counter, Pitch Track, Pitch-to-MIDI
FREE Signal Generator, DaqMusiq generator
Science with your sound card!

Robert Macy wrote:


1. Ability to adjust the output level to well below 1mV and smoothly up
to several volts.

2. Circuitry not linked to the AC safety ground to avoid hum loops.

3. Have a fixed level output to connect to the external trigger input on
a scope - so the trace remains in synch despite noise, hum and harmonics.

3. Square wave output for checking tone circuits and amplifier stability.

4. Frequency from 10Hz to 100kHz, preferably in four overlapping ranges.

5. Amplitude remains steady when frequency setting or range is changed.
6. THD not more than 0.1% with low order harmonics.

..snip..

You can do all this with some very inexpensive stuff.


** Which does NOT qualify in any way, shape or form as a BENCH AUDIO GENERATOR suitable for day in day out, knock down drag out bench repair work on smelly, dirty, rusty old guitar amps.

Bet a software ****** like YOU has never even seen the insides of one.



.... Phil

On Tue, 30 Dec 2014 21:27:17 -0700, Phil Allison
wrote:

...snip...
** Which does NOT qualify in any way, shape or form as a BENCH AUDIO
GENERATOR suitable for day in day out, knock down drag out bench repair
work on smelly, dirty, rusty old guitar amps.

Bet a software ****** like YOU has never even seen the insides of one.

not a software guru, just had to learn once for a project.

I have two guitar amps. One is a Fender, and yes they are robust
construcion [they're in storage and I can't remember model numbers]

My main contribution is pure analog design of a unique mag pickup. Turns a
'cheap' $1100 guitar into sounding like a $10k acoustic guitar. Absolutely
the first time I've ever heard great sounding E-A string harmonics. Maybe
been around too many cheap guitars, but usually I HATE when those two
strings are strummed. Always sounds a bit sour, but NOT with my pickup.

"My main contribution is pure analog design of a unique mag pickup. Turns a
'cheap' $1100 guitar into sounding like a $10k acoustic guitar. Absolutely
the first time I've ever heard great sounding E-A string harmonics. Maybe
been around too many cheap guitars, but usually I HATE when those two
strings are strummed. Always sounds a bit sour, but NOT with my pickup. "

I was going to tell him to just use a damn guitar. That's what I do. Want 0..1 distortion in a guitar amp ? ****, the more distortion one of them things has the more people like it.

The E and A played together makes an A chord, but it is sorta inverted. You couild think of it as you want the fundamental of the chord to be the lowest note in the chord. If you play the E open and the A at the second fret you get an E chord. If you play the A open and the D at the second fret you also get an A chord but it is with the A as the lowest note in the chord, which some seem to find sounds better. At this point the chords are undefined, meaning there is no major or minor or any of that other stuff.

But everything is a chord. You can play all the strings on a guyitar open and it is some form of A or E chord. Like a seventh fifth with and augmented 13th and a diminishe 15th or some such ****. Seriously. Very rarely used but is definable. There is a site online that will tell what it is and I used it a few times but don't remember what it said. Also, I along with quite a few people can simply strum oll the strings open and immediately tell if it is in tune. Not that I have perfect pitch, the whole thing could be flat or sharp but it is at least in tune with itself. When you get that good you should be making some money somehow off of it.

Some chords just sound like **** on an electric. Take the regular G for example. The first string is played at G and the A is played at the second fret which makes it a B. That is not what makes it a chord, that is what makes it a major chord. If the B is an octave up, the fuzzy will sound good, but the way it is played conventionall that B is too close in frequency and the distortion inherent in regular magnetic guitar picks can make it sound like ****. Same with the C when you play it by the book.

I play a couple of little ditties on guitar which absolutely MUST be pl;ayed on an aacoustic because of the chords. I think I invented a bunch of the chords but of course that is not true. I DISCOVERED a bunch of unconventional chords. Everything is a chord no matter what as long as all the notes are on the standard scale. Twelve steps to an octave is what makes the math work.

I have an acoustic floating around here with a pickup added to it. Truthfully I would rather put a microphne in front of it than to use the pickup. However I have heard the new special pickups they put in acustic guitars now and have been properly inpressed. They sound real if they are going through an amp clean. Listen to Treetop Flyer by Steven Stills for an idea. They sound like that. I'm not sure if he used those type of pickup or a mic, but that ****ing guitar is reproduced faithfully. High fidelity.

And that's the difference. In fact talk to guitar players and alot of them will agree that playing an acoustic is a different ballgame than an electric, and some say the electric is not really a guitar. Les Paul disagreed a few decades ago and may have won the argument in light of all the rock and roll in the world. (my Uncle took guitar lessons from Les Paul)

But Les Paul in a way effectively invented a new instrument. You could say it is like the difference between a piano and a harpsichord. Seriously different but sdo alike.

You could almost say though that the electric guitar is limiting. There are so many chords that simply sound like **** on it, but when played on a nice big bodied acoustic sound not only good, but could be described as rich. In the last few days at work, the guitar guy has been in and just when he tunes them up and tests them, the chord prgression he goes through is IMPRESSIVE. He is also a master of appregriation. (sp?) I guess if you are going to build guitars you should be able to play them eh ?

In fact I am a witness to what happens otherwise. Friend of mine has a surrrogate Grandfather. Lives with him on occasion, lkeeps in touch and when he was young they did everyrhing. The Man things like fishing, welding and scrapping. That old Man built his own guitar. Of course he bought the neck, but the body was made out of the wood from a box for a casket. ot the casket, but the box it came in.

He was off. He did not realize that the twelfth fret has to be in the middle of the strings. I adjusted it but it put the bridge in a place that seems awkward. Doesn't sound bad though as long as that is where it needs to be.

So this pickup, is that your claim to fame ? These pickups I have heard, I consider them fantastic. Inside the ghuitar they pickup less feedback and they really do sound good. Friand of mine (hmm, about time to call him) bought a new Ibanez with those pickups in it. Not the kind mounted to the hole which work like electric pickups, this is those good ones inside that sound like it is miced. Well that electronic preamp in there went bad.

He paid $300 for it and when the electronics went bad under warranty they said it could not be fixed and gave him a full refund. I found out why because I gave him $100 for it. So he made a hundred on the deal but it is a good damn playing acoustic. I thought I could fix it but the way they made it, they are right, it can't be fixed.

Well it could be but the body of it would have to be taken apart. They must ount one of those peize things or whatever in there before the wood is glued together. On a manufacturain level you can do that.

But the fact is that on that guitar the action is almost good as an electric and the intonation is perfect. (that is from my ears and comparing to a digital piano, not by any laboratory testing or freq counter, which would actually be LESS accurate IMO)

Wish I oculd play the ****en things. Yeah, I can pick out a few chords and **** but really, that is not really playing. I had a buddy who said Jimi Hendrix was nothing but a second rate blues guitarist. When he plays you know he is damn qualified to say that. He'll sit there and play and sing Stairway To Heaven no problem. Playing a song like that alone rather than just part of it in a band is a whole different thing. You have one instrument and you must play all the parts. I've tried it. It's a bitch. I can BARELY manage to play and sing at the same time.

OK this is probably already a TLDR so, later.

"** Which does NOT qualify in any way, shape or form as a BENCH AUDIO GENERATOR suitable for day in day out, knock down drag out bench repair work on smelly, dirty, rusty old guitar amps."

Rusty ? you're lucky they got metal in them to rust. In my country it is hard to get metal, everything is plastic.

Anyway, what is so knock down drag out about it ? Are you going to do some sort of product testing and rating ? If you are talking guitar amps maybe instead of a low distortion generator you want a high distortion generator. they crave distortion in thoise things. In fact the crave the distortion by their favorite amp, at times the will build a seald box and mic the amp for a concert, or at least they used to. Really, all those big ass speakers and in the back is a little 25 watt Fender or some **** with THE ORIGINAL SPEAK, in a wood box with foam all over the sides, playing its guts out, but the microphone in the box feeds the big ass 2,000 watt amps outside.

Two basic things I do not see you needing for this applicatiion. you do not need anything over 20 kHz and you do not need low THD. The thing you probably could use is a calibrated output level to make sure the amp distorts correctly at a certain level input.

When you are lookng for bad outputs and **** like that, almost any signal will do. In fact I tesat regular audio amps sometimes just on a regualr input,like for FM or something. I just null the output and input on the scope and see what is left. Divide it up mathematically and you get a total distortion figure, which includes any phase shift induced error as well as any variations in the frequency response. It always reads worse than the THD or IM specs but if it is under 1 % you are doing good. At least you can be relatively sure that there is no malfunction in the amp.

We are not the Institute of High Fidelity, nor are we Julian Hirsch. That's something maybe to get across at Audiokarma. If you can REALLY hear the difference, there was something wrong. It is hard to tell 1 dB in frequency response and most people cannot hear 1 % distortion, but that varies. Studies done a long time ago indicated that people were much more tolerant of even order harmonic distortion than odd order. And the fact is that most speakers have so much THD that the spec is never published. It is a secret. A few do, but they are not among the cheap.

Anyway, as far as a generator just use resistors to divide it down and use the function genny. Float the DUT with an isolation transformer like the book says. That's the way you do it. However it will not get your money for nothing and your chicks for free.

wrote:

"** Which does NOT qualify in any way, shape or form as a BENCH AUDIO GENERATOR suitable for day in day out, knock down drag out bench repair work on smelly, dirty, rusty old guitar amps."

Rusty ? you're lucky they got metal in them to rust. In my country it is hard to get metal, everything is plastic.

** So you have neve seen a guitar amp - right ?


Anyway, what is so knock down drag out about it ?

** Means getting the job done in the most efficient way.


Are you going to do some sort of product testing and rating ? If you are talking guitar amps maybe instead of a low distortion generator you want a high distortion generator. they crave distortion in thoise things.

** So you have neve seen a guitar amp - right ?

Two basic things I do not see you needing for this applicatiion. you do not need anything over 20 kHz and you do not need low THD.

** So you haver neve seen a guitar amp - right ?

The thing you probably could use is a calibrated output level to make sure the amp distorts correctly at a certain level input.


** Yeah - all you need for that is a linear pot and a 10:1 step attenuator.

A DMM will do for calibration of the highest range at say 100Hz at a few spots on the dial - then it remains the same for all lower ranges and frequencies.


Anyway, as far as a generator just use resistors to divide it down and use the function genny.

** Function generators are **** awful for audio repair work, the sine wave is full of high order harmonics that get exaggerated by the treble boost in many guitar amp models. Sounds bad through the speaker and looks bad on a scope.

So you have never seen a guitar amp - right ?

..... Phil

On Tuesday, December 30, 2014 2:15:55 PM UTC-8, Phil Allison wrote:
whit3rd wrote:

3{a}. Square wave output for checking tone circuits and amplifier stability.

Or, maybe a white-noise output?


** Worse than your other mad ideas.

A/B switch, headphones, attenuator; it's not hard to check uncalibrated tone
controls that way. Or, with a math-equipped DSO, it can be quick to measure.

A real square wave has lots of bandwidth, beyond the important, audible, spectrum.

whit3rd wrote:


3{a}. Square wave output for checking tone circuits and amplifier stability.

Or, maybe a white-noise output?


** Worse than your other mad ideas.

A/B switch, headphones, attenuator; it's not hard to check uncalibrated tone
controls that way. Or, with a math-equipped DSO, it can be quick to measure.

** So you never tested a tone control or stack using a square wave and a scope ?

It is a VERY quick, easy and unambiguous way of seeing if all the controls do what they are supposed to.


A real square wave has lots of bandwidth, beyond the important, audible, spectrum.

** But a guitar amp does not - plus the amplitude of all harmonics fall in proportion to their frequency.

Square waves contain only the odd numbered harmonics, predominately 3rd, 5th and 7th with amplitudes 1/3, 1/5 and 1/7 of the fundamental.

With guitar amps, I use a 250Hz square wave and firstly try to set the controls so the output looks square on a scope. Next, I turn each control through is full range and then back to the "flat" position.

Treble boost causes overshoots to appear at the leading edges of the square wave, treble cut rounds them. Mid and bass controls cause the horizontal parts of the square wave to tilt sharply uphill and down. At the same time, you can see if the controls are working smoothly, free of noise and intermittent operation.



..... Phil

On Wed, 31 Dec 2014 15:06:00 -0800 (PST) Phil Allison
wrote in Message id:
:


** Function generators are **** awful for audio repair work, the sine wave is full of high order harmonics that get exaggerated by the treble boost in many guitar amp models. Sounds bad through the speaker and looks bad on a scope.

It doesn't meet all your specs, but what about a Keithley 2015? They can
be had for about $500 on Ebay...

On 01/05/2015 03:37 AM, JW wrote:
On Wed, 31 Dec 2014 15:06:00 -0800 (PST) Phil Allison
wrote in Message id:
:


** Function generators are **** awful for audio repair work, the sine wave is full of high order harmonics that get exaggerated by the treble boost in many guitar amp models. Sounds bad through the speaker and looks bad on a scope.

It doesn't meet all your specs, but what about a Keithley 2015? They can
be had for about $500 on Ebay...


Look on the bay for a Heath IG-58 (several variations). They are very
low distortion for the price. (Typically 0.04 to 0.02 % THD, depending
on level and frequency). You can get a guaranteed one for $125 or less.
Or wait for a bargain and grab one for $50. The Elenco with the big
pointer is around 1%. Your smartphone can be used, too. I have not
measured the distortion of such apps.

dave wrote:


Look on the bay for a Heath IG-58 (several variations). They are very
low distortion for the price. (Typically 0.04 to 0.02 % THD, depending
on level and frequency). You can get a guaranteed one for $125 or less.
Or wait for a bargain and grab one for $50.


** You meant a Heath IG-18 right?

Yeah - that looks pretty good, except for the lack of a decent frequency dial.

My generator has a 3inch aluminium disk that turns with the frequency pot, marked across 290 degrees, turned by a 6:1 reduction drive using ball bearings.

I also have a ancient, valve Wien Bridge generator that uses a large, dual tuning gang coupled to a 30:1 reduction worm drive - as you might imagine, frequency adjustment is very silky and resolution is genuinely infinite.

Lotsa bench generators use dual tuning gangs for frequency, cos they give ultra smooth, noise free adjustment that lasts for - unlike most dual pots.

But watch out if you try to float the internal circuitry and just leave the metal case safety grounded, it does not work. It is essential that that case and shielding around the tuning gang all be at the same potential or supply frequency leakage from the power transformer secondary injects straight into the oscillator and causes major amplitude modulation.

..... Phil

On 01/05/2015 06:17 PM, Phil Allison wrote:
dave wrote:


Look on the bay for a Heath IG-58 (several variations). They are
very low distortion for the price. (Typically 0.04 to 0.02 % THD,
depending on level and frequency). You can get a guaranteed one for
$125 or less. Or wait for a bargain and grab one for $50.


** You meant a Heath IG-18 right?

Yeah - that looks pretty good, except for the lack of a decent
frequency dial.

My generator has a 3inch aluminium disk that turns with the frequency
pot, marked across 290 degrees, turned by a 6:1 reduction drive using
ball bearings.

I also have a ancient, valve Wien Bridge generator that uses a large,
dual tuning gang coupled to a 30:1 reduction worm drive - as you
might imagine, frequency adjustment is very silky and resolution is
genuinely infinite.

Lotsa bench generators use dual tuning gangs for frequency, cos they
give ultra smooth, noise free adjustment that lasts for - unlike most
dual pots.

But watch out if you try to float the internal circuitry and just
leave the metal case safety grounded, it does not work. It is
essential that that case and shielding around the tuning gang all be
at the same potential or supply frequency leakage from the power
transformer secondary injects straight into the oscillator and causes
major amplitude modulation.

.... Phil

Yes. I misread my memory. Thanks. I have an early prototype HP200AB with
a big knob, which is great for sweeping arrays for resonances but too
fuzzy for absolute distortion testing. I am using my Sound Technology
generator because it's 10 times quieter than the Heath. I have a crappy
little Tek function generator which is great for scifi sound effects. My
iPod Touch v3 has a sine wave generator.

"Yeah - that looks pretty good, except for the lack of a decent frequency dial. "

I have considered attaching one of those cheap lanel mount vrequency counters. Won't help it adjust more smoothly but the accuracy has to be better than any dial.