Limiters

[gbono]

Why don't MI amps use limiters? It doesn't seem to be a very common feature on older amps (duh) but what about newer designs or is it just adding too much cost?

[Brian H.]

Some do

I have a Peavey TNT115BW 160 watt bass combo that has Peavey's "DDT". This is an output clip-limiter.

[Enzo]

Tons of them have limiters, often more than one. It is not used as a "feature," it is just a clip preventative.

Where you won;t often find them is in guitar amps. WHy would you want one? Solid state amps have current limiting in the power amps, and you might want some peak limiting even in those power amps - clipping sounds awful - but a tube amp sounds good overdriven. Why squish it?

[Mark Hammer]

A lot of Crate amps have limiters built in that use feedback from the power-amp chip output to control an optoisolator on the input to the chip.

But Enzo is correct in that this is not provided as a defeatable or controllable feature. It's just there.

[gbono]

Limiters

Okay, I guess I'm coming from more of the consumer electronics point of view here. A/V receivers have soft-clippers, peak and RMS limiters and other forms of non-linear processing used to protect the loudspeakers as well as prevent the electronics from making awful noises.

I can see how these processes may not fit into the “sound” most musicians are trying to achieve - to bad since I also see many over cooked voice coils.

Tube rectifiers have staggering voltage drops associated with their use and provide a sort of soft clipper effect that seems to be desirable. Why do many guitar players use solidstate effects pedals BEFORE their tube amplifiers and claim SS devices sound bad ???

[tubeswell]

Why do many guitar players use solidstate effects pedals BEFORE their tube amplifiers and claim SS devices sound bad ???

The SS effects pedals used in front of the amp, still end up going through the amp and having all the benefits of the softer clipping, and non-linear signal amplification/harmonic distortion added from the tubes in the amp. Okay so to my mind a tube-driven tremolo, or a tube driven stand-alone spring reverb, or a tube driven overdrive sounds better than a pre-sampled digital imitation of an SS pedal, because the tube device is more spontaneous/touch-sensitive (for a start).

But simply not everyone can afford tube-driven stuff, so the ready-made one-size-fits-all SS market is standing by ready for orders from 'many' guitar players.

[Enzo]

Hi gbono

I think you need to collect your references a little.

I don;t hear many guys claiming all solid state devices sound bad, I do hear them claim solid state amplifiers sound bad. Many guys love a tube screamer in front of their tube amps. I think ther are some good sounding SS amps. And one would run a tube screamer in front of one for the same reasons as in front of a tube amp - different tone availability.

Not all agree.

Our tube amps are generally left to clip when they want because it is not the hard clipping of a solid state power amp.

Clipping limiters should be invisible to the sound in an amp until their threshold is crossed. In a solid state amp, clipping is obvious, but I'd have to say the PV DDT is pretty inobtrusive. WHen you wouldn;t otherwise be clipping, limiters should not alter the sound.

I also see many over cooked voice coils.

Sure, don;t we all? But that is not for lack of limiters. An amp limiter has no idea of the specs of the speakers it powers. That 300w amplifier can have all the limiters in the world and still burn a 100w speaker to a crisp. And amps that go to DC won;t be held back by limiters.

[Steve Conner]

The overdrive of a tube amp is musical enough that you can think of it as an inbuilt limiter.

I don't know whether to be happy or sad about the optoisolator trick, because I used it too, and thought I invented it. Does anyone know what part number Crate used? I was using the H11F1.

"We think solid-state sounds evil, but love Tube Screamers" is one of the great contradictions of electric music. The Tube Screamer is basically a diode clipper, the kind of thing that people yank out of their master-volume Marshalls as if it was a cockroach nibbling on the PCB. And yet guitarists love it.

I've been really impressed by Peavey's DDT. The first time I saw it working, I was doing sound for some local metal band, outdoors on a football field, with one of those Peavey 600 watt powered mixers and two 15-and-a-horn cabinets. Given the nature of the gig, I was worried about burning out the horns (especially since they weren't mine!) but I soon realized I could crank it as hard as I liked, and it just sounded like there was a compressor on it. It actually made the sound better. Which wouldn't have been difficult, mind you. Way to go Peavey!

[teemuk]

I’d say it’s adding too much cost. All the better solid-state guitar amps tend to have those, though. There’s also two kinds of them, the “transparent” kind, that just limit the nastiest clipping and is there mainly to protect the speaker elements, and the “overdriving” type that adds some nice compression, and soft clipping. Peavey’s DDT is a good example of the former, T-Dynamics of the latter. The Vox “Valvereactor” circuit limits too. These kinds of circuits are quite ideal to transistor amps because they prevent nasty power amp clipping. Many tube amps naturally tend to have this feature as an inherent characteristic - not all of them, though.

But let's think... Ignoring the crude light bulb in series with speaker -types, Thomas Organ’s Vox amps featured a limiter already in the mid 1960´s and the famous Lab Series amps also had a limiter. Music Man amps have a limiter and so do most of the newest Traynor amps. GMT and Gallien-Krueger even built solid-state amps that had soft clipping power amps (although it's generally easier to do that soft-clipping thing in the preamp). I'm talking about transistor amps, of course.

The Optoelectric limiter I’ve first seen in vintage Gibson and Epiphone tube amps. You can even find a patent for this circuit filed in 1962 by Robert Grodinsky et al. Before that those amps already used a tube-based compressor.

In fact, if I pinch my memory I can make a pretty big list of guitar amps with limiters. …But yes, compared to bass amps. generally the cheapest guitar amps (meaning maybe about 90% of all of them, especially if SS is in question) tend to lack that feature. They also tend to lack multi-band graphic EQs and many other useful features. Put some money to the table and you'll definitely find limiters and other neat stuff.

As a final note, I must disagree with Enzo's remark about the current limiting. Yes, technically it is limiting (no argue in that) but if the amp is designed properly the current limiting will engage only when the amp is about to "blow up". The "SOA protection" is somewhat a better name for that type of circuitry since it clearly indicates that the current limiter should never do anything under normal operating conditions. (In late 1960's and early 1970's there were big issues because they did and this gave a bad rep for these circuits). But what I'm saying basically is that clipping can be a normal operating condition, whereas a short circuited load definitely is not. There's a big difference.

[Mark Hammer]

Steve,

The H11F1 is a great little chip/optoisolator for its intended use. because the FET in there is SO responsive and fast, however, you need what amounts to a ripple-free rectifier circuit driving it, or else any lingering guitar strums would tend to produce an unpleasant burr (unlike a more pleasant Scottish burr) that many hear/describe as distortion. Crate uses a Vactrol VTL/5C7 unit, which is an LED/LDR, as opposed to LED/FET, combination. The LDR has a certain degree of sluggishness inherent to it, which has the same consequence as a well-filtered rectifier by providing a certain degree of immunity to ripple.

[Steve Conner]

Hi Mark,

I wasn't using it as a voltage-controlled linear gain element, but a voltage-controlled soft clipper. The idea was to detect when the power amp started to clip, and adjust the H11F1 so it lopped off the offending peaks by saturating.

Unfortunately the tone was more David Beckham than Sean Connery

[Mark Hammer]

That makes sense at a conceptual level. I guess the trick is to match the temporal properties of the control element (FET in this case) to the parameter you want to control. If it's millisecond-to-millisecond control you're after, the H11F1 is a good choice, if the phenomenon you want to alter takes a second or two to pass, an LDR-based device/circuit is a better choice for the reasons noted.