Tubes as clipping diodes

[earthtonesaudio]

Hey,
I just realized I don't really understand clipping diodes.  I get that they hard-limit the signal, but how do they do it? 

Take the regular back-to-back silicon diodes on the output of an effect.  I understand that 0.7 volts for silicon is where the clipping starts to occur, but what's actually happening?  Silicon diodes drop 0.7 volts across them, so is it that when you exceed that level, that the rest of your signal is just dumped to ground?  Or is something else going on here?

Going slightly beyond this concept, I've wondered about using tube diodes for clipping, and this guy actually implemented it in a circuit:
http://www.diystompboxes.com/smfforum/index.php?topic=65127.0

How does that work?  Can you simply substitute a tube for a Si or Ge clipping diode?  If so, what is the threshold level for the clipping (or what number would indicate that on the tube's datasheet)?

[R.G.]

To understand diodes, we use a series of approximations.

1. A diode is a one-way valve for electricity. Electricity goes through with no resistance at all when the anode is more positive than the cathode/bar, and is totally blocked from going through the other way.

2. Well, that's not exactly true. It does block current one way and not the other, but it doesn't turn on at exactly 0V. There has to be a little voltage forward to get it to turn on, and there's some small resistance internally, maybe a few ohms. So the diode goes from an open circuit (infinitely big resistor) when the voltage is reverse, to a few ohms when the voltage is one-diode-drop positive. The diode-drop is the penalty we pay for the diode knowing when to turn on. For silicon that's about 0.6V. For germanium, less, and for LEDs, more.

3. Well, OK, that's not *exactly* true. Diodes don't instantly go from infinite resistance to a few ohms at one single voltage. They kind of "round the corner". A diode will block conduction; but as the voltage nears the diode-drop voltage, the resistance starts decreasing just a hair, then a bit more, then more, then a LOT more, then bango! it's only a few ohms. The diode resistance changes from nearly infinite to only a few ohms over a band of voltage that we approximate by the "diode drop voltage". A silicon diode conducts microamps at 0.45V, milliamperes at 0.5V, and amps at 0.7V. We simplify by saying some voltage in the middle is where the diode "turns on".

In the case of back to back silicon diodes, they are nearly always fed by a resistor or by a current source, as in the feedback loop of an opamp. Fed by a resistor is the easiest to understand. If the resistor is, say 10K, then as long as the signal voltage fed to the 10K plus diodes is below maybe 0.45V, the diodes are much, much larger resistances than 10K, and no apparent distortion happens. As the signal exceeds 0.45V, the diodes start conducting more, and this makes a resistive voltage divider which causes the diodes to eat up anything larger than - pick your voltage here - 0.6V? 0.7V? Depends on how hard you drive them. The signal above the diode forward drop (which varies, depending on how hard you drive it, right?) is limited by the 10K, and excess current is shunted to ground by the diodes.

[earthtonesaudio]

Thanks R.G.  It's getting a little clearer.

So then relating this to tubes,

I have some 6AL5 twin diodes that I'd like to use as clipping diodes if possible.  Their datasheets seem to indicate that they have very little of the "knee" that I see in silicon diode datasheets.  It's fairly linear all the way from zero volts/amps to way past stompbox voltages/currents.  Does that mean that I can use these as clipping diodes, and that they will have a super linear transition from clean to clipping? 

...Or could it be that the tubes are super non linear at small currents, and the datasheets just don't show it because they don't expect anyone to be foolish enough to try something like this?

[amz-fx]

The 6AL5 appears to have a low voltage drop at low currents. It might be suitable for clipping if you have enough juice to power the heater. I don't know about the knee but it would be interesting to measure it.



Pin 6 is an internal shield between sections and should be grounded.

The 6AL5 was used as an audio noise limiter in many old tube radio receivers.



This isn't a clipper but more like a keying section with the cathodes tied together and a bias voltage placed on them to control conduction. The diodes are in series with the audio which goes in through pin 2 and out via pin 7. This is only a piece of a much larger, complicated receiver schematic.

regards, Jack

[Nasse]

http://www.elektor.com/magazines/2001/october/tube-clipper.54803.lynkx

[amz-fx]

http://www.elektor.com/magazines/2001/october/tube-clipper.54803.lynkx

I have that old issue so I dug it out and looked it over...  it is not tubes used as clippers but instead a starved plate design, though still interesting. Thanks for reminding me of it.

regards, Jack

[earthtonesaudio]

Thanks, Jack and Nasse. 

Fender also has something like this, for their Roc Pro 700 amp.  One of my coworkers put the design into SPICE and the output waveform looks a lot like the typical tube waveform, except symmetrical.  Schem here: http://www.fender.com/support/manuals/pdfs/manuals_elec/guitarpdf/Roc_Pro_700.pdf  ...Of course it requires +/-2Vdc, just to bias the tubes, so maybe not the best thing for a stompbox.  But definitely ripe for tweaking, I think.

I will try out the much simpler versions first, and try to get some pictures of the waveform, or even sound samples.

Thanks again, I think this is cool stuff... It will be even cooler if it responds well to the AMZ "saturation" type of mods.

[Ronsonic]

One of these days I want to try the signal diodes built in the 6AV6. It has a 12AX7-like triode with two diodes sharing the cathode. tt was surely designed for RF reception, but I'm sure it can mess up a guitar just fine.

Ron

[sprog]

I was thinking about this last night, so here I am hitting the search button.....

So has anyone done this yet, & if so, how does it sound - is it worth the time & effort? I bought a junk box of tubes off of ebay & want to try this.

Do the tubes have to have heater power to work? I'm also curious about when the clipping starts to occur - is it a bigger voltage that would have more headroom?

[Salvatore]

Brad made a distortion based on the MXR Distortion with a 12AL5 tube as clipping diodes.

[Caferacernoc]

I was thinking about this last night, so here I am hitting the search button.....

So has anyone done this yet, & if so, how does it sound - is it worth the time & effort? I bought a junk box of tubes off of ebay & want to try this.

Do the tubes have to have heater power to work? I'm also curious about when the clipping starts to occur - is it a bigger voltage that would have more headroom?

Yes, the heaters would need power for the tube to work. I believe most tube diodes have a forward voltage similar to a silicone diode. More than a germanium anyway.

[puretube]

http://www.diystompboxes.com/smfforum/index.php?topic=22821.0

[juancra]

The 6AL5 appears to have a low voltage drop at low currents. It might be suitable for clipping if you have enough juice to power the heater. I don't know about the knee but it would be interesting to measure it.



Pin 6 is an internal shield between sections and should be grounded.

The 6AL5 was used as an audio noise limiter in many old tube radio receivers.



This isn't a clipper but more like a keying section with the cathodes tied together and a bias voltage placed on them to control conduction. The diodes are in series with the audio which goes in through pin 2 and out via pin 7. This is only a piece of a much larger, complicated receiver schematic.

regards, Jack

Hi all!

I have a simple question regarding the use of 6al5 tube as a clipping diode. I'm thinking to add it to a Valvecaster stompbox with other kind of diodes (SI, GE, LEDS of different colours, etc) to make a "clipping module" , selecting each clipping section with a rotary switch. The effect in which I want to try this is powered at 12V DC. Is this enough to feed the 6AL5 or it won't go at such low voltages?

Thanks in advance,
Sebastian

[ambulancevoice]

Hi all!

I have a simple question regarding the use of 6al5 tube as a clipping diode. I'm thinking to add it to a Valvecaster stompbox with other kind of diodes (SI, GE, LEDS of different colours, etc) to make a "clipping module" , selecting each clipping section with a rotary switch. The effect in which I want to try this is powered at 12V DC. Is this enough to feed the 6AL5 or it won't go at such low voltages?

Thanks in advance,
Sebastian

well, in clipping configuration, the diode tube (6al5 here) doesnt need a power supply, although they need a heater supply in order for the tube to actually work, and because the tube is a 6al5 tube (the number in front of the name is the heater voltage, so 12ax7 has a heater v of up to 12v, or 12.6v max etc) it needs a 6 volt heater supply, or a 6.3 volt heater supply
as the diagram indicates, pins 3 and 4 are the heater connections

[cheeb]

Haha puretube; you're the Crosslink Bandit.