Gyrators. Op-amp or transistor.

[ddpawel]

Hi,
Is it better to use gyrator circuit built around transistor or op-amp? I mean use in high-gain effects. Does this have an impact on noise, on the sound quality?

[earthtonesaudio]

Op-amp: potential for higher Q, save at least one resistor over the transistor version, lower distortion.
Transistor: potential for lower noise, may allow for easier layout, probably sounds better when it distorts than a distorting op-amp.

[ddpawel]

I just want to understand differences between eg. Maxon SM-9 and Boss MT-2 gyrators. The first one use only op-amps (every gyrator and buffer), the second one uses more transistors. Dime Distortion uses only op-amps. DOD Death Metal - transistors.

Actually I'm doing layout for Boss XT-2. I need some knowledge about it.

http://www.tremolo.pl/Efekty/Distortion/BOSS%20XT-2%20xtortion.jpg

It's noisy and harsh distortion, that uses both type gyrators. Before clipping it uses 3 in mix configuration
-op-amp 2975 HZ, Q=2.9
-op-amp 596 Hz, Q=10
-transistor 1039 Hz, Q=2.5
and after clipping another two also mix of op-amp and transistor.
-op-amp 4015 Hz, Q=2.6
-transistor 37 Hz, Q=1.6

Can you advise any changes in the types of gyrators in this circuit?

[earthtonesaudio]

Amazing coincidence, you're talking to one of the other ten people who will actually admit to owning one of these rare "gems".  I would not advise changing the type of gyrator, unless you're making a new circuit board and one way of doing it is easier to build.  The difference in sound is small.


Anyway, I have modded mine extensively.
See this post, reference schematic and part substitutions:
http://www.diystompboxes.com/smfforum/index.php?topic=77886.msg641813#msg641813

In stock form the pedal is sort of like this:
"input buffer... midrange boost... gain stage... clipping stage... bass/treble boost... notch filter"

What I've done, essentially, is turn the mid boost into a mid boost/flat/cut filter.  Also I replaced the clipping LEDs with one Si and one Ge.  I still think the pedal is a little too bass-heavy and I'd like a little more output with the gain at minimum, but it's much more pleasing to my ears now than it was in stock form, and MUCH more versatile.

One of my mods, the 100k in parallel with R13, reduces the input impedance of the effect even in bypass.  Most people would call this "tone suck" and find it undesirable, but I like the way it alters the taper of my guitar's volume control.  Use that mod at your own peril.

[ddpawel]

I don't mod mine. I am making a new Layout with variable Punch and Contour Frequencies for more control and including Yours and MonsterPedal mods   I did some research before i started.

There's a problem with transistor gyrators, they are not well described in scientific publications. I wondered why some people benefit from opamp, and others from transistor and never found information about it.

[earthtonesaudio]

Cool, thanks for the Monster Pedals mention.  I had not heard of them before.  I like that someone else has some mods for the XT-2.

The key differences between the performance of transistor gyrators and op-amp gyrators are similar to the key differences between op-amp and transistor voltage followers.

-The op-amp can sink or source current; the transistor with resistor load can only do one of these.
-The op-amp has 100% negative feedback which improves linearity, distortion, and dynamic impedance characteristics.  The negative feedback for the transistor comes in the form of degenerative emitter/source feedback from the load resistor.  The problem for the gyrator configuration is that current from this node is also routed to the input (positive feedback).  This increases distortion, decreases input impedance, and increases output impedance, which alters the Q and center frequency and generally makes it a less ideal filter.

However when it comes to guitar effects and distortion in particular, sometimes "ideal" is not the best-sounding choice.  I would highly recommend experimenting with both types to see if you can hear a difference, and if so choose the one you like best.

[sault]

At the risk of sounding like I'm tooting my horn, I'll toot my own horn a little. I just wrote an online program within the last few weeks to calculate the frequency response of a gyrator depending on how you change the components around.

http://awasteofsalt.com/gyrator

I wrote it specifically for the Metal Zone, but you can still input in whatever resistor/cap values you want and save them as a user preset, so you can load the values, play with 'em, go back to the original, etc. I've used it to analyze the Boss Blues Driver, for instance. The results don't always make a ton of sense (a bass boost at 100 hz with a Q of 20???) but the math is sound.

It's written in Javascript, so you'll need that enabled. No downloading required. It shows normalized values (doesn't account for gain), and only the most basic graphic of the frequency response... but the benefit is that as you change component values, it changes the response in real-time. Tested in FF, Chrome, and I think even IE6, so it should work for just about everyone.

If you do take a look at it, or especially if you end up using it to tweak your circuit values, I would love to hear about it!


Saul T

[teemuk]

Personally, I'd say the overall impact of the device choice will likely be of minor importance. Yes, OpAmps are likely more linear and less noisy, but all in all that might not matter too much. I think either one is mostly chosen for reasons such as:

- Existing inventory: basically which one's a more profitable device to use . If you're already stocking a pile of cheap transistors but only a few moderately expensive high-quality OpAmps then the choice may be obvious.

- Layout issue: This is actually quite interesting since some research made on subject claims that - on the contrary what one might initially expect - the transistors are then ones being more difficult choice layout wise. Of course experiences may vary and sometimes the overall layout just happens to offer that OpAmp section and required board area conveniently in near vicinity. I have done few circuit designs with gyrators and sometimes I simply swapped from BJT design to Opamp - or vice versa - because the layout for the other one proved to be more ideal. Sometimes this decision swifted again when the layout advanced.

- Optimal use of components: Say, you have a circuit with four gain stages and one gyrator. Basically, you can cover the gain stages with a single quad OpAmp or with two dual OpAmps. Is it worthwhile to add in another OpAmp just to cover the gyrator section, or would you rather use a BJT. Let's assume another case: Three gain stages and one gyrator. Now, after covering the gain stages you're likely left with one unused OpAmp section; do you leave it unused and realize the gyrator with a BJT or do you try to save up little?

Naturally all these choices more or less interact with each other and I'm willing to bet that most designers actually focus more on them than on the fact of creating the ultimate gyrator with low noise and high Q. I think that stuff gets more important in precision applications, not in some cheap lo-fi guitar pedals.

[ddpawel]

Cool, thanks for the Monster Pedals mention.  I had not heard of them before.  I like that someone else has some mods for the XT-2.

The key differences between the performance of transistor gyrators and op-amp gyrators are similar to the key differences between op-amp and transistor voltage followers.

-The op-amp can sink or source current; the transistor with resistor load can only do one of these.
-The op-amp has 100% negative feedback which improves linearity, distortion, and dynamic impedance characteristics.  The negative feedback for the transistor comes in the form of degenerative emitter/source feedback from the load resistor.  The problem for the gyrator configuration is that current from this node is also routed to the input (positive feedback).  This increases distortion, decreases input impedance, and increases output impedance, which alters the Q and center frequency and generally makes it a less ideal filter.

However when it comes to guitar effects and distortion in particular, sometimes "ideal" is not the best-sounding choice.  I would highly recommend experimenting with both types to see if you can hear a difference, and if so choose the one you like best.

Have you tried to change hard clipping diodes to soft ones for overdrive sound? Or create something like Boss HM-3 type of clipping. That could be interesting and will help reduce harshness.

[earthtonesaudio]

I only changed the diode type, not their location in the circuit.  However, I think that the stock design probably has a lot of op-amp clipping going on, so simply changing the LEDs to diodes with a lower forward voltage will allow more range of clipping sounds, both softer and harder.
Softer because the current limiting resistor going to the diodes allows the signal to spend more time in the "knee" of the diode's conduction curve, and harder because the lower forward voltage will square-clip on smaller signals than is possible in the stock configuration.

The main drawback to this is the output level is decreased.

[DDD]

I read some opinions that gyrators are not good for the guitar circuits. They add some artifacts to the sound that are hard to define.
Also I've tried about 4 or 5 tone controls with the transistor and OpAmp gyrators, and have to say they do they work great, but I didn't like them alll. Why so - I don't know, but it seems to me that  resulting sound is not "wide" but kinda "chewed".
Today I do prefer to avoid gyrators at all.

[BubbaFet]

Sault- very cool use of java script. Good work! You've given me some ideas
for some things I'm trying to accomplish at my day job, unconnected with gyrators.

[ddpawel]

I read some opinions that gyrators are not good for the guitar circuits. They add some artifacts to the sound that are hard to define.
Also I've tried about 4 or 5 tone controls with the transistor and OpAmp gyrators, and have to say they do they work great, but I didn't like them alll. Why so - I don't know, but it seems to me that  resulting sound is not "wide" but kinda "chewed".
Today I do prefer to avoid gyrators at all.

Good gyrator is not bad 
Last year I made a copy of the DOD FX56 (3 gyrators) and heavily modified DOD FX70C Corrosion (original 5 gyrators and 1 additional for mid control).

DOD FX70C is really good sounding effect (something between DOD Grunge and MT-2). On lowest gain and mids maxed and set around input gyrator frequency (I can change center frequency of 3 gyrators) it's a really good Q-Zone type effect. On low gain my distortion can boost distortion channel of my amp really well.

So, gyrators are not always bad. If you understand how to modify the original pedal and that type of sound is what you need results will be good and you will be satisfied.

[ddpawel]

I only changed the diode type, not their location in the circuit.  However, I think that the stock design probably has a lot of op-amp clipping going on, so simply changing the LEDs to diodes with a lower forward voltage will allow more range of clipping sounds, both softer and harder.
Softer because the current limiting resistor going to the diodes allows the signal to spend more time in the "knee" of the diode's conduction curve, and harder because the lower forward voltage will square-clip on smaller signals than is possible in the stock configuration.

The main drawback to this is the output level is decreased.

I have done some simulations using XT-2 circuit. It looks like there will be some op-amp clipping but with overdrive diodes and distortion diodes at the same time signal should look more tube-like.
But it's only simulation.