King Soloman - new bright overdrive!

[Brisance]

After a bunch of simulations, testing and modifications, I present to you, KING SOLOMAN!

It's best used for solos (hence the name, pun intended) and functions as a nice tone shaper, which makes a guitar stand out while removing lows  and the nether mids and with drive turned on also overdrives the signal. Loves a mildly overdriven tube preamp stage after it!

Tone pot rolls off the harshest of highs, while leaving volume largely uneffected (subjectively anyway).

If needed I can also get sound samples today although I'd prefer to take my time on that and get a guitarist, who can play and mic one of the Breket tube amps.

On demand I can also provide a vero layout, which I used for my prototype, although I am in the middle of designing a PCB for the "true" version of it.

Schematic:


Prototype:

Guts(a bit different than the final version):




COMING SOON: images of simulated frequency response (I have no way of measuring the actual), images of the true version, sound samples

[antonis]

Didn't you notice any Op Amp overheating..

(I think that first output is quite loaded at pot's low values so maybe you should raise the 100R to 1000k and play a little with pot and cap values..)

[Brisance]

Hmm, good spot, I didin't concentrate enough on power ratings, only on frequencies, while designing. However a TL072 has max rated output current of 40mA per channel, and the output will not reach nowhere near that, considering the cap will not the biased DC through and has an impedance of 142 ohm at 20kHz and 2,8k at 1kHz, so even blasting it with a maxium output amplitude sinewave @ 20kHz, it will not reach close to half of it.

Besides while I have not touched the IC after prolonged use, after a few hours of jamming with my friends it did not show any damage.

[~arph]

I don't see why it would overheat.. the TL072 has a very high input impedance, plus there is also that 55k path to GND.

[antonis]

The 55k path don't bother me but the 100R resistor in series with 56nF cap to GND at high frequencies does make me think of heating...


edit: My bad  - I didn't notice that there isn't such a gain that could amplify the signal at high voltage level...

[Govmnt_Lacky]

Any reason for the parallel 100uF caps in the power section? Wouldn't a single 220uF be OK?

Also.. is that a 1uF film cap? Why such a large film. Isn't the usual value there a 100nF?

[Brisance]

100uF because I didn't have 220s at hand most other caps are designed around what I have in stock as well and the 1uF is a ceramic, which are small

[antonis]

the 1uF is a ceramic,

I think is better to use 100nF ceramic to serve as IC power decoupling also (as close as possible) and omit the 2n2...

[Transmogrifox]

COMING SOON: ...frequency response...I have no way of measuring the actual...

If you have a computer and a sound card you can get close.  Make a .wav file with very narrow pulse to emulate a dirac delta and play it out to your stompbox.  Then you only need to record the impulse response and perform the FFT in Audacity or other tool of choice.  The impulse response characteristic might be the best for avoiding clipping (as opposed to playing a sine chirp and recording the response).

Yes you get the frequency response of your sound card send and return and cabling, but probably that is negligible and you could calibrate it away if you wanted.

[SISKO]

COMING SOON: ...frequency response...I have no way of measuring the actual...

If you have a computer and a sound card you can get close.  Make a .wav file with very narrow pulse to emulate a dirac delta and play it out to your stompbox.  Then you only need to record the impulse response and perform the FFT in Audacity or other tool of choice.  The impulse response characteristic might be the best for avoiding clipping (as opposed to playing a sine chirp and recording the response).

Yes you get the frequency response of your sound card send and return and cabling, but probably that is negligible and you could calibrate it away if you wanted.

Another way is to fed the input with white noise, record the output in Audacity and perform the FFT.
Its the same thing as in the impulse response but a little easier or "controllable" in my opinion since you dont have to worry about the width nor the high of the pulse. And you can continously control the input to get the best result you are looking for (differents frequency responses for different levels of input )

[amptramp]

You are forcing the first op amp to pull up to 7.5 volts in order to drive the inverting input to 4.5 volts.  Instead of a 33K to ground, have 66K to ground and 66K to the +9 volt rail.  The nearest standard value is 68K.  This may give you more headroom.  If you want variable bite, the 1K resistor connected from the clipping diodes to ground could be transformed into a 5K pot.  If you like extra controls or need variability, that is.  Some people would put a capacitor in series with the 33K in the first stage, but using two resistors gives you some protection against supply voltage variation.  Imagine the supply had ripple on it.  With 470K and 470K on the non-inverting input, you get half the ripple.  If you have 68K and 68K on the inverting input, you also have half the ripple on the inverting input, so the two ripples cancel out.  The only cost is a slight extra current drain.

[PRR]

The 1K--100r high frequency load on the '072 will stress it but not hurt it. High frequency distortion will rise, and there may be mild HF clipping (slewing). While it may be "wrong", it may be a fine distorter.

Driven HARD, that first '072 section will dissipate a bit over 100mW which is well inside ratings.

For battery power you might consider the extra ~~10mA when driven HARD versus battery life. I suspect few players/passages would approach such current, but you could check with a milliAmp meter or voltmeter across a 100r resistor series with battery.

[Brisance]

Thanks for input everyone!


Transmogrifox and SISKO:
Good idea, in fact I am gonna make me a noise generator, I wanted to use it just for evaluating tone circuits by ear, but the audacity idea is stellar!

amptramp:
Some good points here! The 68k and 68k idea seems really nice, not that I felt there was not enough headroom, but the PSU ripple cancelling out makes perfect sense and is genius! About the variable clipping resistor, I will consider it and hook one up for tests, see how much it changes the sound, I wanted to use the KISS mentality on this one though, having as little unimportant controls, as possible, but if it has a large impact on overall sound in different settings, I will definitely use it!

PRR:
Didn't even consider the slew rate dropping, but within normal conditions, as in me and my friends playing on it, if it was evident, then it certainly didn't sound bad. And since it was basically constantly on and played through for multiple hours straight and there was no visible or audible damage. Measuring current draw is another great idea, I think the best way would also be using an analogue multimeter for that, since the needle smooths things out, thus gives a better average as well(and I won't feel sad for the little thing collecting dust as well )

[amptramp]

Gah!  I was out walking the dog this morning and it came to me that what I had recommended did not provide ripple cancellation because the non-inverting input would have half the ripple on it, the inverting input would have half the ripple on it and therefore, the output would have half the ripple on it.  Still, the sensitivity to ripple is 0.5 whereas the ripple would have been multiplied by the gain before.  I am going to have to think about how I would design this stage to eliminate ripple.

[Brisance]

One of the Ideas I had was having a series inductor on voltage input before the caps, besides I scored a bunch for free recently. Maybe 100uH? These are the size of a resistor anyways.

[anotherjim]

100uH would be fine for stopping HF interference, but you want to get some without the u before the H if you want to cut supply ripple. A bit bigger than a resistors those would be A resistor before your power smoothing  caps would be better, though there will be some load dependant droop - it happens and not always for the bad. 100R would do and there's your current test done too.

[PRR]

> design this stage to eliminate ripple.

R.G.'s site, "noiseless biasing".

If you hold the nose of this first stage at a clean 4.5V, all the rest will follow.

The ripple on the opamp supply pins is non-issue unless you use crappy "DC" (nearly) supplies.

Yes, decent 120Hz filtering needs far more than uH tiny chokes. A hundred uH is still "nothing" at the top of the audio band.

[bluzeyonecat]

Killer build man! Way dig the name! I'm new to all of this but am learning slot just following along. Please forgive the simplicity of this question but I noticed your pots look like they are  hooked up with just positive and negative. I thought the middle lead was what made it variable. What am I missing besides obviously alot? Lol! Thanks for your time.

[PRR]

> I thought the middle lead was what made it variable.

The middle lead ("wiper") is connected to one end. It IS variable.

Actually he could just connect one end and the wiper. But it is common practice to strap the wiper to the un-used end. The main advantage is that *when* the wiper loses contact (dirt, wear) the pot value goes to "max" instead of "infinity".

Fer example, his "Gain" pot. With the left and and the wiper connected, pot effective value runs from zero to 100K, gain runs from unity to 21. If the wiper loses contact, pot value goes to "infinity", gain goes to "infinity" (probably several thousand, LOUD.) With wiper strapped to the un-used right end, when wiper goes bad, gain only goes to max 21.

Same for "Tone". When wiper goes bad, sound gets very dull, but not "dead". You can finish your solo in a mellow mode, instead of playing air-guitar.

[bluzeyonecat]

Lol! Gotcha! Thanks man.