Rectifier?

[Yew]

Hey guys, Some of you might know that im doing an engineering degree. Recently in a lacture we came across the Rectifier. Instantly the words MESA and Boogie popped to my head.

How would using a rectifier effect my tone? Could you Incorperate It into a boost/Overdrive pedal, or even on the guitar its self?


Thanks guys
Image;

[ijustwannastrat]

I don't understand how it would.... A rectifier only turns Alternating Current into Direct Current ( Edison-1, Tesla-0)

Of course, why would Mesa have two or even three rectifiers in their amps...

Only this guy knows...

[Yew]

Hmm, Ive had a better idea.

Opamps, and make a pedal from them. I can go High/Low pass Frequancy band cuts/boosts, Buffering, Negative feedback gain controls and even change the type of amplifier circuit ^^

Could make quite a funky pedal

[flateric]

A rectifier cannot be used as an effect! As previously mentioned, it's used in tube amps to recombine the positive and negative signals, once amplified, back into a positive-only signal so the valves can handle it.

[ashcatlt]

Pretty sure the rectifier tube is part of the power supply section. Converts the AC wall power to the DC heater voltage for the tubes. Don't know for sure why you'd need two of them, but...

Something very similar to that diode bridge is used in some effects. Used as a clipping stage in series after a good bit of gain you'll get a sort of "octave-up" fuzz. If you wire it in between a couple transformers with two inputs you get a ring modulator.

[thetragichero]

basically the "dual" and "triple" rectifier is a marketing ploy
i like the bit of sag i get with a tube rectifier, for what it's worth

[Yew]

I was just wondering if it would change the way the speaker reacted to the signal if is was only positive...

[flateric]

Wouldn't that just make it half as loud?

[ashcatlt]

Dec 9, 2010 15:40:43 GMT -5 Yew said:

I was just wondering if it would change the way the speaker reacted to the signal if is was only positive...

This would be a VERY BAD THING!  Voltage across the speaker corresponds to velocity.  When one end of the voice coil is more positive than the other it moves one direction, when they swap over it goes the other direction.  

The rectified signal is like if you get in your car, put it in a forward gear, accelerate to some speed, then slow back to a stop, and then repeat.  You'll end up moving down the road.  The unrectified signal is like putting it in drive, accelerating, slowing to a stop, throwing it into reverse, accelerating, and then slowing back to a stop.  Done carefully, you'll end up back where you started.  You really don't want your speakers to move down the road!  Luckily, if this is in a stompbox, the DC component won't get as far as the speakers since you're pretty much assured to hit an AC coupling (DC blocking) cap at some point along the way.  

So what will it do?

Well, I had thought to wait for somebody to ask,n but I guess I'll just go ahead and elaborate a bit.

I we look at yew's picture up there we'll notice that this thing isn't actually outputting a flat DC voltage.  What it's actual doing is "folding" the part of the wave which swings below the reference voltage (nominally 0V) so that it now swings above.  To get anything approaching DC out of this we'd need a BAC (three words, three letters each, first is Big, last is Cap) across the output.

Without that cap we've now got a new waveform which oscillates around a point halfway between the reference voltage and the top.  The amplitude of the wave is half what it had been.  

It's no longer a sine wave - nor even symmetrical - but the peaks and troughs come twice as often.  The fundamental frequency is doubled.  There's your octave up.  The asymmetry is your fuzz.

But that's not quite the whole story.  Yew's picture illustrates the theory.  A theoretical diode conducts in only one direction.  That is only when the anode is more positive than the cathode.  

But a real diode will only conduct when its anode is enough more positive.  There is a threshold voltage below which there won't be any conduction.  

In practice, there will be a finite period of time during each cycle - from just before to just after the "0 crossing" - where the output voltage will suddenly drop to the reference voltage and stay there.  This will tend to "square off" the pointy little troughs in the output wave. 

How long it stays at that ref voltage depends on the relationship between the full swing voltage and that "turn on" voltage.  If the input has been amplified sufficiently it'll sound like distortion.  If the input is relatively small it'll start to act like a noise gate.  It's worth noting here that our guitar signals don't usually stay at any steady voltage swing. The swings are very big at the attack and get much smaller as it goes into sustain.  Worse, it usually kind of goes up and down out there in the sustain.  This is something to think about when deciding how much gain to pump into this thing.  Do you want your held notes to just suddenly fart out?  Some folks do. 

Now, you may have noticed where I said "nominally 0" above, and then talked about reference voltage.  Where is that thing where I talked about all voltage being relative?  Anyway, the only thing that matters to a diode is the difference between it's anode and cathode.  It really doesn't care how these voltages relate to the rest of the circuit or anything else in the universe.  The input can just as easily oscillate around the bottom of the battery (0V) or the top (9V) or anywhere in between, as long as it's wired correctly. 

[sumgai]

ash is correct on all but one account. Rather than just spill it out there, I'll wait and see if anyone catches the error. It is a bit subtle, to be sure.

But even so, a +1 effort if I ever saw one. ;D




sumgai

[flateric]

Dec 7, 2010 12:29:14 GMT -5 Yew said:

Hey guys, Some of you might know that im doing an engineering degree. Recently in a lacture we came across the Rectifier. Instantly the words MESA and Boogie popped to my head.

How would using a rectifier effect my tone? Could you Incorperate It into a boost/Overdrive pedal, or even on the guitar its self?

Going back to your original post Yew, I assumed that you linked the word rectifier with the good reputation Mesa Boogie amps have for overdriven tone. Ash's fascinating insight into the daily life of a speaker cone is a good explanation of what might happen, but it is clear that a recifier is primarily an assembly used in valve amps to modify polarity, and not produce overdrive tone; this is done by driving valves hard beyond their normal amplification headroom and inducing clipping in the signal. (Now I will get pasted for my oversimplifications).

[sumgai]

No pasting necessary.

♭eric's explanation of ash's explanation is also good, and even closer to guessing at what was going through yew's mind, at the time.....

Even so, there's a fly in his ointment too. The Real Meal Deal:

Dec 9, 2010 18:14:25 GMT -5 ashcatlt said:

To get anything approaching DC out of this we'd need a BAC (three words, three letters each, first is Big, last is Cap) across the output.

DC is the absence of an alternating polarity, which ash did state. However, DC is not defined as a "straight, flat, no variation" voltage (current), it only means what it says - all electrons flow in only one direction. Which means that we can have DC that pulsates in any fashion whatsoever, and it will still be DC. Hence, the rectifier does remove one side or the other of an AC signal, and the output is DC - not pure, mind you, it's highly irregular, but nonetheless, it fits the classic technical description, as according to Hoyle.

So a BAC's job is to filter out some, or as much as possible, of those variations, making the DC as regular as possible - this is what we classically call "regulated DC", the voltage varies within a tolerance that we can accept. Said tolerance is dependant of course on what the circuit needs, and how much we're willing to spend. Good ball-park figures are:

Low-power transistor and IC circuits (pre-amps and effects, things like that) usually run between 5 and 15 volts, so a tenth of a volt is acceptable;

High-power signals often run closer to 50 volts, sometimes more - 10% regulation is usually sufficient here.

(Note my oversimplification here - I've ignored the aspect of current demands, but I've made a general point that is correct at the layman's level.)

The point is, ash was fully correct in describing a diode's action, then made a completely normal mistake in stating (he may have been thinking correctly and simply mis-typed)) that a rectifier also needs a BAC in order to produce DC.

BTW, that "car going down the road" analogy was very good, I can honestly say that I wished I'd thought of that! ;D

Dec 10, 2010 5:47:10 GMT -5 flateric said:

...... it is clear that a recifier is primarily an assembly used in valve amps to modify polarity.

I hope by now that two things are even more clear. One, valves/tubes are not the only things that use DC, and therefore they don't have an exclusive on rectification. I think I covered this just above.

The other thing is that I have a penchant for reading all the way through the meaning of each and every statement posted in these forums, and where it might cause later confusion, I step in and "lend a helping hand" - or so that's how I see what I'm doing. If others are not so overjoyed at getting the straight scoop, then they're welcome to open a debate/dialog/whatever, and we'll work our way to an accomodation with which we can all be happy. Not that I expect it, you've all been too kind to me over the years, but I don't want any late-comers to have any doubts.


HTH



sumgai

[JohnH]

I'd agree that diodes in speakers circuits are generally a bad thing!, though I have heard of back-to back zener diodes in series, being used as a limiter to protect tweeters.

I have also heard of people from the 60's (probably also now in their 60's) using them for a fuzz effect, while simultaneosly studying the effects of recreational pharmaceuticals.

Nobody owning up?

As to what will happen - I would say that the voltage across a speaker is most closely related to the position of the cone, rather than its velocity. By feeding a speaker with a rectified signal, either via a single diode or a bridge rectifier, it is really getting a 'dirty dc' input. The cone will be pushed a distance one way, as when you put a battery across the terminals, but it won't keep going, (and cause a Marty McFly effect). It will move until the average push from the magnets and coil equals the elastic pull of the cone trying to resist, and then it will vibrate around that position.

I am not sure what such a spiky signal will do in terms of reverse emfs and possible amp damage. Hence, some one else can try it first.

The whole thing with Mesa and their choice of rectifier circuits is more subtle than this, and assumes that there is a smoothing capacitor to take out the ripple from the power supply. Diode and tube rectifier circuits behave a bit differently. Diode rectifiers tend to maintain a very consistent voltage, while a tube redtifier results in some sag of the supply voltage when current demand is high. This drop in voltage slightly compresses the sound on loud notes, and some times this is considered to be desirable.

John

[flateric]

Nearest thing I get to recreational pharmaceuticals is taking ibuprofen after an afternoon in the garden - who would think I'm a middle aged rocker playing bass guitar in a punk rock band? Heh!

[ijustwannastrat]

So we can all agree....

Recifier in the wrong spot will sound more like a Rectum-defier, and that it could shoot your speaker out of your cab.... I can see a new terrorism niche here...

[thetragichero]

brings "battle of the bands" to a whole new level?

[ashcatlt]

Dec 10, 2010 15:34:22 GMT -5 ijustwannastrat said:

So we can all agree....

Recifier in the wrong spot will sound more like a Rectum-defier, and that it could shoot your speaker out of your cab.... I can see a new terrorism niche here...

Well, it looks like JohnH doesn't completely agree that it'll shoot the speaker cone across the room.  I agree that it was a bit of simplification on my part ignoring the elastic force of the surround, but I'd maintain that the cone will at least try to keep moving in that one direction.  It's at least theoretically possible that the magnetic force could exceed the structural limits of the speaker.  Even if it's able to push the cone far enough for the voice coil to leave the gap it's likely the speaker will never work again.

Note also that we've assumed that the speaker is wired so that positive voltage will push it out away from the basket.  This might not always be the case.  I don't know if there's any convention on this (?), but slamming the thing back into the gap can't be good for it either.

[sumgai]

ash,

There're two reasons why John's correct - once the physical limit is reached, and the cone tries to rebound back from that limit, it can't - it attempts to move forward again, and it can't.... (lather, rinse, repeat). The net effect is a mechanical form of hysteresis.

Not to mention, let alone forget, there's a BAC (and it ain't a cap, but instead it's a coil....) in there somewheres. That has a theoretical limit as to how much current it can absorb when the magnetic object contained within can't move. At that point, it heats up, and mechancially distorts, which gives rise to what you said about sefl-destruction. But as the cone moves (in either direction), there's a certain amount of "back EMF", or IOW, it generates an non-trivial current (voltage) within that same coil. That back EMF tends to dampen the cone's excursion distance as well, which acts as a double-whammy on Marty's Excellent Adventure.

All in all, it's a neat idea, and it probably can be done under the right (?!) circumstances, but not ordinarily so. Me personally, I'd like to see it happen, just for drill. ;D




sumgai

[jcgss77]

Yew, I think you should experiment. Make yourself a pedal, and incorporate the rectifier in a circuit you design/modify. You just might end up with a quite original effect! As another Nut said to me, doing something which is not conventional is a step in the Nutz direction! Yeah!

[ashcatlt]

So, I'm going to try this soon. (Yeah, soon... )

The idea I've got is to put a diode bridge after an opamp gain stage and see what I get out of it. If it turns out to be something even vaguely interesting, I might end up making it permanent. Of course I'll record samples and post them here.

The thing is, I'm pretty sure it's going to end up as a specialty thing. I don't see myself using particularly often. So I'm thinking of building it in as an option in the Rat clone that I've "working on" for a couple years now. The idea was to put in a switch that would take out two of the diodes and leave it looking like the parallel-diode-to-ground clipping section of the Rat.

When I originally conceived the idea, I thought it would be a simple matter of shorting across the series diodes in the bridge with a DPDT. Now that I've looked at it, though, I see that this won't work because the two diodes that are left point the same direction. That'll give an extremely asymmetrical overdrive, which might be cool, but...

So, a challenge: What's the fewest poles that will change this diode bridge into a pair of diodes pointed opposite directions between hot and ground, and connect the ins and outs?

[JohnH]

One thing that may keep this idea on the workbench is that sending an audio signal through a bridge rectifier will require it to overcome the forwards voltage drop on each half cycle before any sound signal gets through. So instead of decaying to a cleaner sound, like parallel back-to-back diodes do, there'll only be sound on the peaks of signals, like a really baddly biased transistor amp. To get something nearer to good, if the pedal is made to run at a higher than normal voltage, so it has a good peak to peak swing, and use Schottky diades instead of standard Silicon, for their low fowards voltage. Then it might be interesting, and should amount to an octave-up fuzz effect with inherent noise gate!

cheers

John

[ashcatlt]

Sorry, no clips just yet because I haven't messed with it much, but I did get it passing sound today. Honestly, with there's so much gain in the opamp stage that it clips pretty hard, and then I added clipping diodes after the diff amp because otherwise the output swings as close to 9V peak-to-peak as the opamp will allow, and that's way too loud for my equipment.

So, it's hard to say exactly what the rectifier portion is doing, other than gating the sustain pretty severely. I'm able to get some usable sustain from a L'il Killer HB, but the SCs fart out pretty easy. There isn't really any noticeable "octave up" effect, but then I've only tried it with my Nashville tuned mini, which is already up an octave.

Here's more or less what I put on the breadboard, though:

[sumgai]

No comment.

[jcgss77]

So I am assuming that running the whole signal through the rectifier will kill the sustain? Interesting...maybe this would be a good effect for a bass then, for when sustain is not needed? I hear people say that neck-throughs have too much sustain, and they use a string dampening system. I think I will be slapping one of these on the BB and see what it does. Seems like it will work better with lower gain. I wonder how it will sound with an LM386 distortion. Interesting...

[thetragichero]

i will try a rectifier pedal on the breadboard once i get the fuzzface and ampeg scrambler clones finished

[Yew]

Try puting a pot as one of the resistors in the opamp, should vary the amount of negative feedback, and hence vary the amount of gain...

[ashcatlt]

I'm thinking you're talking about varying the gain of that first opamp stage. (?)

If'n yew read JohnH's post above, and read closely my description, I think yew'll see why I don't think that'll help. I tried a smaller resistor (187K) at R5 and the gating was just too much. It wants diodes with a smaller forward voltage, and possibly more clean gain, which means a bigger supply voltage. The signal needs to get in and out of that "zero crossing" region as fast as possible.

I think what I've got sounds pretty fun and horrible as is, though some filtering might be in order. I've recently come across the idea of op amp based rectification, which looks easier to accomplish. What I've seen says that it needs op amps which perform well near the rails, and I'm not sure the tl072s which I've got quite qualify.

[Yew]

Okay, i was just trying to remember what my electronics lecturer said., this is bad, as it was only a few weeks ago, and i have to do an assesment in the next few weeks