Some ideas about tube amp and emulation/simulation

[rodriki1]

Hi people.

The first - I am here again to question about tube amps.

For years i have been reading on the web about
tube amps, and its comparation with SS amps.

It sounds like an endless discussion. Considerable confusion.

I am not a professional in audio.

But i know almost every guitarrist want to have
a good distortion of tube amps.

I do not have one tube amp.

All i know, after years of readings, maybe can be expressed
in a few lines.


- EVERY AMP GIVES SOME TYPE OF DISTORTION
 AND TUBE AMP GIVES GOOD DISTORTION FOR GUITAR
- tube produces distortion with low harmonics
- tube produces distortion with EVEN and ODD harmonics
- tube gives large headroom (very good for audio)
- It is easier to implement tube amps compared to SS
- Good amps demands excelent electric source
- the output transformer affects the sound
- tubes are high impedance input
- tubes have capacitances that affect the sound
- output power tubes acopled with transformers give only odd harmonics
- With overload, EACH stages of tube amp give some amount of distortion. At the end the distortion
 is very complex.
- When overloaded tube can have some sort of DC bias shifting
- Math associated with audio amps are complex and few people likes it
- Triode tube can exibit three halfes power law curve
- IMPORTANT: MID booster for guitar will do an amplifier distortion more assymetric.

Besides everything else, reading dimitri ideas about triode emulation
i got really interested.
I did contact to him. He is very gentle and intelligent and the idea is very interesting.
I would to thanks him and the other guys who share ideas here.

I would say that i have tried a simple FET PREAMP with low source resistor and
a proper drain resistor in computer simulation (CIRCUITMAKER).

I got distortion of second harmonic followed by less third harmonic.

In practice, things sounded pretty nice for a simple fet preamp.

Some guys have made attempts to get one aproximation of tube amps with FET transistor.
The reviews say that they do not exibit the same tone and distortion. Even though they are
good in tone and distortion.


I must be very careful here........


FIRST QUESTION.

Somebody out there has tried or projected "SUCCESSFUL" FET emulation (sorry by this expression) for a tube amp ?
(this may sound crazy for some of you) Or just the tone copy


SECOND QUESTION.

Is it someway (technically) possible to simulate an entire tube amp in distortion ?
Without tubes of course.


For example:

If i like JTM45 - Fet or Op amp we can get even the tone???


If these question can not be answered, I WILL GIVE UP searching for emulation of tubes.

I HAVE SOMETHINGS TO SAY ABOUT TUBES, JUST TO JUSTIFY MY QUESTIONS.

- TUBES AND GOOD OUTPUT TRANSFORMERS ARE REALLY HARD TO FIND HERE IN BRAZIL.
- TUBES WORK WITH HIGH VOLTAGES (that i dislike)
- TUBE AMP IS A BIG BOX WITH PARTICULAR TYPE OF SOUND (EVEN THOUGH IT IS GOOD)
- ONE GOOD TUBE AMP HERE IS VEEEEEEERRYYYYYYYYY EXPENSIVE.  

I must warn that i know ideas from:

John Murphy duty cycle from CARVIN about SX100,SX200,SX300 amps.

Stephan Moeller with VOX ac30 sim.

Gabe Velez with fet buffering bipolar junction transistor.

Nad soft clipping schem.

Pritchard and his patents.

Runoffgroove. (fet simulations with tone and distortion DIFERENT FROM ORIGINAL)

R.G. - Saying its not apliable fet for simulation of tubes

AMZ (MINI BOOSTER - CASCODE FET)

Peavey patents.

Zener Clipping - led clipping


Lots of related patents.

((((((Thanks to all these guys. and ARON)))))))


GOOD LUCKY FOR ALL (KEEP STUDYING)

rmfidelis@gmail.com

http://geocities.yahoo.com.br/rmfidelis/TubeSimulator.html

[aron]

I think the problem is just saying "Tube amps are better than solid state" is too generic.

For what type of music? For what musician?

[jmusser]

I kind of went through this the other day, by asking about using germanium type effects through a SS amp. The general concensus was that the germanium type effects were the ones used at the time, (maybe because that's what they had access to) and had them tuned to the various Vox 30 and Marshall amps of the period. I'm not sure that you can't get some of those tones through a solid state amp, if you tailor them for a specific amp. I don't believe that you can use a germ effect with an emulator, and get the Vox 30 with the germ effect tone. I know, that you can get a real nice tube emulation with the 4049 chip, which gives you the softer clipping, and not the harsh unnatural (less organic) Si tone. I plan to try the English Channel (Vox AC30 Emulator), which I think is a FET tube emulator, and also Mark Hammer's new 4049 emulator. Since my ear is not yet sophisticated enough to hear all the different nuances of the various amps and put a label on them. I can tell you that when I first plugged in the ROG "3 Legged Dog", I couldn't believe how close to a tube amp, it made my SS Crate sound like. As expensive as the tube amps are, emulation may be all I'll ever be able have, but since the 3 Legged Dog, I'm a believer!

[rodriki1]

Hy JMusser

Thanks for your more precise affirmation about 4049.

I did some good experiences with it too.

Other day i said to Will that i think
amp design is about math. Not beliefs.

Do not take me wrong (or anyone).

Just consider that i am talking about AMP DESIGN, not personal
life subjects (were beliefs, values, feelings guides us).

(here at the forum things happen too fast and sometimes we can say
some things not well behaved)

Considering about ENGLISH CHANNEL:

People here have said that fet emulation
has its own tone and distortion, different from its original tube amp.

MICHAEL IBRAHIM have posted ideas about fet bias to get
more acurate amp tone emulation.

Using CIRCUITMAKER things sound like heaven.

I did some tests but real thing (testing on simulation) is not
so well behaved. I got oscilations, and some noise indeed, where i had
to change a little.

But computer simulations is so far one good idea.

Thanks again for your reply.

Good luck.

[dominatrix_volleyball]

You might not be able to knock it out in one fell swoop. It's probably too complex of a problem. But if you break it up into smaller, simpler problems, and then try to get them to work together, it might not be so overwhelming.

Something that stands out to me about the LXH2 Fender simulator is the huge rows of diodes across the 5th stage. I don't know if that's the most important aspect... but try to find the core ideas in all of the stuff you're looking at. Figure out HOW and WHY they work, until you can work with them like ingredients to mix and match.

If you look at ROG's 3 Legged Dog, it looks like a pretty basic cascading of 4049 stages after an FET. The front end of it looks VERY similar to a lot of their amp-based circuits. This says two things to me... 1) you can replace the front end with something else... or 2) you can take their front end as an ingredient into your own recipe. Also, you can look at what can be done with 4049s, and see what some refinements can do.

Hemingway once said "The best part of your book can be something you were only lucky to hear in passing, or it could be the wreck of your whole damn life, and one is as good as the other." It makes me think: Don't stay married to one idea... do what you gotta do and try to be open to as many angles as possible.

Anyway.
I mostly came out of lurking to say I hope you don't give up.!

[petemoore]

I've played SS amps that sound alot like a clean tube tone or very good
 I've played SS amps that crunch alot like a tube amp
 and ones that have both
 I've not played an SS amp that 'works' like a tube amp.
 If an SS amp is modded to be ''like' a tube amp it will be inherently different, under various conditions these differences will become more apparent, for instance driving the input or output to distort.

[rodriki1]

As i said before my english is not enough to thank you.

I feel more air in my chest.

A new friend also did some excelent observations like you did.

For example: Reading for the tenth time

Tubes Versus Transistors
 Is There An Audible Difference?
 Russell O. Hamm
 Journal of The Audio Engineering Society Presented September 14,     1972,at the   43rd Convention of The Audio Engineering Society, New York Sear-Matthew's Audio Systems, 212-874-013720 W 84th St., New York, New York   10024

I got some some fresh clues.

Of course this article is not the end of discussion.
There are somethings they do not mention.

But light is always welcome.

I will try to keep my mind more opened.

If i got some new idea i will report.

Thanks a lot for you all again.

Rodrigo

[GFR]

Rodrigo,

For some deep and detailed info on SS emulation of tubes, I suggest that you read all the patents granted to Eric Pritchard. He goes into extreme detail with his designs.

I never played one of his amps, but I've heard that the PRS amp designed by him was very good.

[rodriki1]

Hi

Thanks again for answers.

About Pritchard I have read almost all his material.
Very clever and helpful but too much complex for me at that time.
And i can not work with patents, unfortunately.

And now i have seen others work too. The subject is very complex.

Stay at will to send me emails maybe to keep portuguese conversation
with somebody about electronics. I used to read all articles from
CCDB. I implemented some of his ideas. GENIOUSSSSSSSS.

I never have the intent to bother someone.

Hey Petemore. You always helping us.

SOME WORDS like "WORK" can make me get the clue.

Really NOBODY maybe have ever got one project of a TRANSISTOR AMP "WORKing" like tubes amps do.

You all are right about tube amps.

Let me increase the "LIST of things of how tube amps work".
Just for curious.
Maybe some of  You are tired of knowing it but for me it is interesting.
Maybe that could be usefull for stomp box UNIVERSE.

- Tube amps have low negative feedback. This avoids a lot of trouble.
- Tube amps WORKS like an "NATURAL" (HAPPY ACCIDENT)
  COMPRESSOR (a property of output penthodes with output transformer).
- When overloaded tube amps give an EDGE (harmonics higher than third)
  effect that is translated for the brain as loudness. the compression
  is not heard because of edge and the amp sounds louder.


I have heard a simple demo of JCM2000. Just guitar and amp. Even from my computer speaker, it was amazing.

Sound is fat with natural treble.

Thanks again. GOOD Lucky for all of you in the USA (we are watching here).

[JimRayden]

Basically, when driven hard, tube amp starts to "squish" the top and bottom of the sine wave, while transistors seem to chop it off. At first, the "squish" just sounds like it's louder or compressed. Then it starts to show the signs of overdrive.

That leads to one more point for your list:

- Tube amps SLOWLY deform from one state to another (clean, compression, overdrive), while transistors more kinda "jump" into distortion.

Or at least that's what I've learned. Correct me if I'm wrong.

----------
Jimbo

[Joe Davisson]

What's up!

While I haven't found a perfect solution, there are some tricks to taming bipolars/JFETs that work pretty well. I'm not trying to put my ideas above and beyond everything else, just trying to explain (repeat?) what I've learned:

1. Bipolar switching speed is proportional to the input current. And so is the tendency to hard saturate.

2. JFETs would be very close to tubes, except for the gate/source diode. While normally reverse-biased, input signals above roughly .3v can put the diode into forward-conduction, shutting the JFET off, however briefly or slightly. This effect produces buzzy output, especially at higher gains.

The solution to both 1 and 2 is to ensure that the input signal is very, very low. Unfortunately, a higher-gain pedal would require way too many stages to be practical or low-noise.

3. Transistors act in a non-linear fashion.
4. Transistors invert the input signal.

3 & 4 gave me the idea for the input diode (ala Vulcan). Now the non-linear reaction of the input diode tends to negate that of the transistor. With a JFET, it also prevents forward-biasing of the gate/source diode.

The best form of the above idea is to use a JFET-Vulcan stage, but with more "normal" source/drain resistors (~1k/10k), and metal-film bias resistors to reduce noise.

If trying to do something like a power-pentode, a zener diode may be used with a darlington transistor. Beware that the input diode trick only works with single-supply, NPN design, so output transformers are still required unless an output capacitor is used (not recommended).

[Mark Hammer]

I suppose all of us should remember that what some find desirable about "tubes" is actually something that can only be found in a Class A amp and not a Class AB.  For others its the opposite.  It would be a mistake to think that what we hear and like about tube amps comes only from the preamp section.  One of the implications of this is that what person X thinks of as a great tube simulation may sound sub-par for person Y, because they each have different assumptions about "amp tone" in their thinking.

There's a whole lotta stuff between the input jack and the speaker on an amp.

[rodriki1]

Hi guys

First i would like to say it is a honor to have you guys
here in discussion. You Joe and Mark always have good
ideas to share.

I will try VULCAN as soon as possible.

I liked the discussion about PIGGYBACKING too.
Found at this forum.

This sounds to be a good way to get good sounds.

I keep thinking if it would be possible anyway to make
a sort of DYNAMIC PIGGYBACKING at BJT.

Mark is right.
Some like the taste of preamp and others the power amp.

Or like R.G. said some do not know tube amp sound at all.

The idea here is that we can get in an easier way the
sound type of tube preamp.

Preamp clipping is not so much "dynamic".

The SAG effect and SQUISH occurs at power amp section.

SAG occurs when ouput penthodes suck to much current from the source.
The plate voltages drops, the gain too and the clipping increases.

The squish is a Automatic Gain Control that happens when self realimentation decreases the gain of penthodes by the supression grid.

(for better understanding www.aikenamps.com , this guy has good
theory to share about tube amps)

And the transformer affects the way the SPEAKER "moves".
As i read before, the transformer (with low damping factor amps)
leaves the speaker more "loose", and this can be felt.

I am looking for a good compression/sustainer mixed with some good
overdriver to try to get the "FEELING OF TUBE AMPS".

The SUSTAINER EFFECT of tube amps sound to be a good and desired property (together with not fizzy distortion).
The sound is ALWAYS loud.

Thanks again.

I will try VULCAN and the others ideas with diode as soon as possible.

See you later

Rodrigo

[Connoisseur of Distortion]

i have accidentally stumbled across a design that recreated what appeared to be soft clipping, but a considerable amount of it.

using a simulator, i layed out a push-pull input stage driving a 1:1 audio transformer. by biasing the transistors correctly to clip on ONLY HALF OF THE WAVEFORM, it would create a clipped signal on the output if the transformer.

biasing a single transistor stage to perform desireably is (in this situation) a pain, so i simply put a single diode to ground on each output from the push-pull.

if you are interested in trying this, please note that the idea is for each half of the push-pull to clip ONE HALF of the wave form, then the transformer will find a "common signal" (?) using the corrupted waveform and the pure waveform.

FYI, i have NOT tested this, but the simulator created what looked like a sine wave with a sinusoidal, though definately smooshed, output.

[rodriki1]

Hey

Connoisseur of Distortion

Could you send me a layout of the idea?

Thanks anyway for the help.

See you later