The Cold Clipping Stage

[loudthud]

BigTeee's thread got me thinking about the cold clipping stage in the 2204. The same circuit is used as the 3rd stage in the "Iron Horse Collision" Express. A 12AX7 with unbypassed 10K cathode resistor and a 100K load resistor. Below is an X-Y plot of input vs output with a 33K grid stopper and a Y-T display under the same conditions. +/- 30V is the max my generator will put out. At no signal the plate is sitting at about 270V with a 300 volt supply, very near cutoff. The X-Y plot is very linear between cutoff and where grid current begins.

Somehow, I don't see any magic. Hard clipping mostly on one side of the waveform. Kind of mushy on the other side. I can do the same thing with an opamp and a few diodes.

[J M Fahey]

Somehow, I don't see any magic. Hard clipping mostly on one side of the waveform. Kind of mushy on the other side.

Well, *thatīs* its job.
Magic? Canīt use such a big word, but you get what you get, which is audible:
A *lot* of unsymmetrical clipping, which adds to the overall sound, even if it gets clipped again more than once before it reaches the speaker.

I can do the same thing with an opamp and a few diodes.

No doubt.
I use a single diode (to be more precise a LED) to get, say, 80% of its effect in a very simple way.
In my case we are talking SS and proper voltages for them, not H.V. tubes, so things more or less scale up.

PS: and thanks A LOT for "wasting" time and energy designing and performing these useful experiments.
Plus you get twice the prize points for leaving the much trodded, beaten to death path and investigating obscure, not very well understood circuits.

[Chuck H]

Thanks LT. I have scoped this circuit before myself under similar enough conditions. I think the reason it's somewhat revered is that it does impart a more silicootie sound even though it's done with tubes. Most of us here are over the idea that only tubes can make music. I build with tubes because I'm not educated in transistors (though I plan to change that). Still, there is a mindset that wants the "right" sound AND wants it to be tubes. Enter the cold clipping stage.

Actually it's not quite as hard as SS clipping in practice. It may be due to voltage sag across the entire amp, I've never tested it, but done with tubes in an actual amp it does sound a little softer and smoother than the hard clipping of SS devices. I do agree that it's just amplifier characteristics and not magic. Bias 'em cold if you want to clip one side. Bias 'em hot if you want to clip the other side. Bias 'em in the middle if you want to clip both sides. Mix and match as needed, rock on.

[Steve Conner]

Good stuff Loudthud.

Chuck, I have no idea how you would go about getting educated in transistors. Here's the problem as I see it.

Fender and Marshall left us with some timeless, great sounding designs for tube amps. In fact they were so popular, they kept the vacuum tube industry alive. You can pick up Fender's schematic for a 5E3 and order a whole kit of brand new parts that will work. And, tubes are mechanical devices, the bias point and the harmonics that they generate are ultimately down to the geometry of the metal bits inside. That makes them stable and repeatable.

Now, transistors. There are no classic transistor amps. Transistors go obsolete every 10 years, or you get the same part number but a different blob of silicon. And that blob has huge manufacturing tolerances and is very temperature sensitive. You can't plonk the average transistor into a simple amplifying stage like the average tube, and expect to get a manufacturable result. Hobbyists do it, but they bias each stage by hand.

The nearest thing we have to a classic is the old Marshalls that used op-amps. They obviously knew about the difficulties I mentioned, and decided to boot them right out of the ball park. Part of a transistor "education" would be building clones of those and listening to them, dull though it might seem.

My top SS tip: the higher the voltage and current rating of a diode, the softer the clipping knee. That's why Marshall used that big high voltage bridge rectifier in their clipper. So to emulate the cold clipping stage, maybe we could use one 1N4148 and one 1N5408.

[Chuck H]

Chuck, I have no idea how you would go about getting educated in transistors.

I don't actualy mean to try and replace tubes. Just learning basic circuit types and topologies. Basic function, etc. It seems there's no one here on my PNW island (though it's pretty big at 70,000 population) who can repair music gear. All the stores (all three of them, little mom and pop shops) send all their repairs to Seattle. And there's no repair shop for individuals if they look in the phone book. Not knowing at least a little something about transistors is all that keeps ME from filling this void.

I'll leave the finer points of how to simulate tubes with silicooties to others.

[loudthud]

Speaking of diodes, what is a cheap commonly available Schottky diode? I've go some SB540 and SB140 but those are pretty big diodes. Seems like the bigger current ones just turn on faster at low currents. I'm looking for something smaller like a signal diode 1N914 size to use instead of germanium. They look almost identical on the curve tracer.

[Steve Conner]

Small signal Schottky diodes tend to be called "hot carrier" diodes, at least by the radio hams who buy them.

I've used the BAT54, but it's a bit inconvenient as it's surface mount. The BAT42 is a similar part in a 1N914-like package, there is also the SD101 from Vishay, and the 1N5711.

They are astoundingly fast. One afternoon I was bored at work, and I soldered a green LED across a 1N5711 in inverse parallel, leaving the diode leads sticking out full length. This device I glued to a plastic stick, and then waved at the microwave oven in the canteen. I had to lever the door open a crack, but I got the LED to light.

In case my boss is reading, I did this at my last job, and no, that was not how I lost it.

[J M Fahey]

I had to lever the door open a crack,

You might have lost much more than your job.
Hard boiled eyes salad anybody?
Oh well.
Thanks God you were lucky

[Chuck H]

Eyes don't get hard when you boil them... Don't ask

[J M Fahey]

It was just a poetic analogy.

[teemuk]

I can do the same thing with an opamp and a few diodes.

...Is what the designers of the 1000 and 1 "tube amp emulation" circuits have been saying for decades.

[Gtr_tech]

I'm not a fan of that typical 800 second stage. I realize its a signature sound, but it just sounds "gonky" to me. I change that stage to a 2.7k bypassed with a .68 or .82u. Smooths things out and makes it sing better.

[Steve Conner]

And what if they were actually right, and we were just imagining that tubes sound better? The cynic in me says that the two main advantages of tube amps are: They tend to have a better speaker because they're more expensive, and the microphonics make them sound more lively. Oh and three, they're very easy to design.

BTW, I have a horrible secret. My eyeballs exploded on that fateful day and I've been posting in Braille ever since. Don't play with microwaves, kids.

[J M Fahey]

I imagined so.
But please share a little secret with us:
we can easily find Braille keyboards; and printing, even graphics, is not that difficult either, just use an old Epson dot matrix printer, without the ribbon, the image gets "printed" on the paper sheet as tiny pits, easy to touch-recognize ..... but where oh where do you get Braille monitors?

[Enzo]

Chuck, you don't have to "know" transistors to service SS stuff. You have to know troubleshooting, and repair techniques.

WHat goes wrong with stuff? Jacks and controls break, solder breaks. WHo cares if there are transistors behind them? TDA7293s work or blow fuses, pretty much, same with LM3886 or TDA2040. Got a mixer with a dead channel? Follow the signal path. If you find an op amp with no signal coming out, it is probably bad. WHo knows what voodoo goes on INSIDE the op amp?

And when that PV 400BH board comes along blowing fuses, look for shorted transistors. You don;t have to understand a light bulb to know it is blown. SHorted transistors are bad, understanding them or not. And when it gets past that point, look up one of the various step by step discussions of the thing on this board.

I learned on tubes in the 1950s. Things I did with transistors were not from understanding, they were from following magazine articles. In teh 1970s I had to learn relay logic. Got good at it. Then digital stuff came along and I had to learn TTL, CMOS, and some microprocesor. It wasn;t until after all that that I came at transistors sort of from the other end... like you would do. And finally, it wasn;t until I actually took the job of repair tech at a large music store that I really learned about transistor stuff. I think it worked out for me anyway.