amp simulator

[neebster]

I wanted to start a more significant audio project than what I'd done before, so I have been cobbling together various circuits with the intention of creating a """quasi""" tube amp simulator. The goal is to have something that sounds decent and performs the functions of an amp with reverb, while being more portable than an amp. Kind of like a sansamp, but with saturation, reverb, and a cabsim.

So, ultimately I want the external layout to be like this image:



I have whipped up a schematic. It's... big.




There are a few things I'm not sure about:

Some (most) people seem to put a diode on the leg of each mosfet in their clipping stages. But I have seen this without the diode, and I am under the impression this makes for a softer clip. I kind of want my "mosfet clip" to be softer than my "zener clip" option.

Also, I totally copied the tone stack of a fender amp. Considering that this is now inserted after buffering stages, I'm thinking I should maybe lower the potentiometer values (and subsequently change the cap values).

I'm vacillating between wanting and not wanting to use the mosfet-SSR's for switching. It is really handy to be able to control bypassing from both a switch on the enclosure and a future control pedal footswitch. I don't know how much noise this is going to add. Again, this is supposed to be tube-ampy, so a little noise is not a deal breaker.

Saturation will be interesting, I hope it works. My only real concern is that either the transformers get hot or I overload the opamps that drive them, but I did select somewhat higher power opamps for that stage. I just don't know what type of (lack of) resistance will develop after the core saturates.

Someone told me this was "everything and the kitchen sink", but really it's all stuff you would expect an amp to do.

After I get some feedback around the internet, and fix whatever stupid mistakes I'm making, I'll start planning the layout.

[bowanderror]

I haven't seen transformers driven in the way your using them in the Saturation section. What's the idea there?

[iainpunk]

welcome to the Forum,

the main problem i have is reading the schematic tho, you used the square block symbol for the chips instead of the more easily readable triangle symbol for the opamps.

I haven’t seen transformers driven in the way your using them in the Saturation section. What’s the idea there?

looks like DC bias current partially pre-satturating the transformers.
9/10 transformers in amps have the anti-saturation gap, so i woulnd't worry about transformer saturation in general, since it also sounds kinda bad in most situations, since the clipping threshold is lower at low frequency's, making it ''fart out'' more easily.

i think most of the circuit parts look fine, from an engineering perspective.
from a pedal/amp building perspective tho... its a nice idea, looks good on paper, but such builds are hard to trouble shoot due to being really complex, hard to get sounding just right, and possibly/probably way to sanitized, losing the ''feel'' of a real amp.

if i were re-desinging parts of the circuit, i'd put in some discrete transistor gain stages in the ''overdrive section'' this gives you some of that ''analog feel'' which occurs mostly in class A preamp stages. opamps are super linear, which makes them feel 'forced to create distortion' instead of letting the distortion occur naturally due to open loop operation of BJTs, Jfets or Mosfets.

am slightly puzzled why you chose a instrumentation amp for the input, such amplefiers are mainly useful for balanced signals or summing/subtraction functions, not for simple input stages.

there is also a whole bunch of stuff that can be greatly simplefied, like a bunch of not-so-necessary (inverting) buffers.

that ''peak'' LED indicator is also really odd, it compares your input to a 6.6v and 8.4v and only turns OFF the LED when its inbetween those DC voltages, i doubt that's what its meant to do.

hope these pointers help you design it further,
Cheers, Iain

[garcho]

The goal is to have something that sounds decent and performs the functions of an amp with reverb, while being more portable than an amp.

What is the purpose of this device? What I mean by that is not what the functions are (drive, verb, etc.) but in what context do you plan on using this device?

I'm asking because one thing about amps is that they actually make sound. How do you plan on hearing what you're playing if this is to be a more portable option than an amp? Is it for going direct into a desk or PA or something? Are you showing up somewhere that has something to plug in to? This pedal design of yours will require a large metal enclosure, as well as a power supply, maybe not all that extra convenient. Do you live in NYC or somewhere inconvenient to drive, or not have a car? I'm familiar with that kind of thing. Maybe just get a tiny amp that's easy to carry, some of them sound really good! For non-quiet gigs, you would need a dedicated monitor mix. But some of those little tube amps sound better than many bigger fancier more expensive options in a studio or bedroom, or a café.

Will the main use be headphones? You might want to drive the op amp output with a pair of transistors. You're already adding a kitchen sink, might as well include a real headphone amp.

Regarding the power supply, how do you plan on getting 15 volts out of the regulator when you're giving it 12? How do you plan on getting -15V when you give it +12? Is that a typo? It's a moot point because linear voltage regulators don't add voltage to the input, they only subtract (in the case of "negative" voltage, subtraction would be considered adding additional negative gain).

[marcelomd]

Some (most) people seem to put a diode on the leg of each mosfet in their clipping stages. But I have seen this without the diode, and I am under the impression this makes for a softer clip. I kind of want my "mosfet clip" to be softer than my "zener clip" option.

MOSFETs connected like that wil conduct via the body diode. It wil sound +- like a pair of ordinary silicon diodes.

Check this article by Jack Orman: http://www.muzique.com/lab/zenmos.htm

[iainpunk]

Regarding the power supply, how do you plan on getting 15 volts out of the regulator when you're giving it 12? How do you plan on getting -15V when you give it +12? Is that a typo? It's a moot point because linear voltage regulators don't add voltage to the input, they only subtract (in the case of "negative" voltage, subtraction would be considered adding additional negative gain).

switching buck/boost converters!
come to think of it, hope he likes switching noise..

also, U79... the two caps are the same value, which is a no-no and one cap has its polarity wrong way round.

cheers

[garcho]

switching buck/boost converters!

Right, I saw that for the 12, 9 and 5V inverters (good point about the noise, that's a lot of switching! none of them are in sync!), but there are two linear voltage regulators, LM2940 listing 12V as the input and 15V as the output, the other LM2990 listing 12V for the input and -15V for the output. Some pretty cool linear voltage regulators! 

EDIT: If you're not a bot, I'm sorry to offend you but considering it's his first reply and he hasn't responded in 3 days, maybe we're just getting thrown a bone by Behringer's AI guitar design software. Maybe he's debugging it on the forum 

[neebster]

Lol, not a bot, just been working.
Thanks all for the pointers. I'm still fairly new.

The idea is that I'm wanting to sort of emulate an amp, doesn't have to sound like a tube amp in all settings, but would be nice if in certain settings it sounds like a cranked champ or something.
Headphone amplification was going to be done on a separate board. I was running out of space.

I'm reworking the design now to simplify a lot of things and fix the errors others have noted.

The instrumentation amplifier idea was because I had previously designed (with some help) a piezo preamp which used one. It's quiet, and handles a really high input impedance.

But yes, I can see I made some mistakes here.

As for the power, I was trying to avoid heatsinks, but these damn chips are so big when I look at the pinout that I think I will go with some other power source. I quickly run out of board space.

The saturation section is because I've heard recordings of saturation "color" boards using those same transformers that sound pretty good. I was thinking that the clipping stage with a bit of the transformer saturation mixed in might round things out.

edit: thanks for all the replies!

[garcho]

Welcome to the forum!

Glad you're not a bot, there's been a small invasion lately.

Can you get away with just +/-12V? You can up the 9V stuff, lower the 15, and use 6V for vcc/2 and 5V stuff?

[iainpunk]

but would be nice if in certain settings it sounds like a cranked champ or something.

the 'champ sound' is really the sound of asymmetric single ended operating gain stages. some chaps do have feedback on its power section, some don't and changing the power section's feedback really changes the character of the distortion.
when i build my DIY amplifier, i also aimed for champ tones, but solid state instead of tubes.
i like the 5C1 champ's sound the best, it features a single pentode and a feedback-less power stage also featuring a pentode (beam tetrode. basically the same thing in all but name). to capture the pentode sound, i used Darlington transistors, not the same exact sound, but closer than jfet transistors due to the less linear behavior in the middle of the curve and more gain.

if you like Triode tube sound more, you should look in to JFET or MOSFET designs, i recommend the moonLNDer by Johnny.Reckless. but there are tonnes more out there, that aim to recreate tube amp sounds, (mostly Marshall's), if you want a more ''cranked fender champ'' kind of sound, i'd suggest bigger coupling caps for more low end feeding in to the transistor gain stages to have it ''mudd out'' like the real ones did, giving you that broken bluesy rumble.

if you want great input impedance, use a Jfet input buffer, or mosfet input opamps, like the TL072.

i think i found the problem with your ''peak circuit'', one of the 10k's goes to ground, instead of -15v.

cheers, Iain