Soft, Medium, and Hard Breakup Power Tubes

[mwelch55]

I noticed that Guitar Center and Musicians Friend offers their Russian Power Tubes in Soft, Medium and Hard breakup.  Does anyone have experience with this?  Do the Soft breakup sets really breakup earlier? 

I was thinking that might be a way to get power tube breakup without harming my ears.

[Ed_Chambley]

The difference is extremely minimal.  The idea is supposed to be "soft" breakup, the tubes saturate more easily.  Sometimes it is advertised the other way around.  All tubes act differently.  A power tube will begin breakup when the voltage from the previous stage exceeds the operating conditions of the power tube.


Said another way.   If a power tube is breaking up, the signal from the Phase Inverter or previous stage is Single Ended begins to push the tube(s) past normal operating conditions.  So for any tube to breakup (saturate) the voltage swing (signal) has to be large enough.  When not large enough, you must increase gain or simply put, signal level.


Some tubes do this easier like a EL84 and El34 Pentodes, but just how hot a tube tests generally doesn't mean much.


It is much easier to boost your signal, and cheaper too.  Boosting the signal creates a condition where V1 gain is more which makes the next stage receive a hotter signal and the amplification factor is larger.  The amp will breakup sooner and according to the amp, sometimes a lot sooner.


What amp are you referring to?

[mwelch55]

The preamp will be like the Chupacabra or Fortin Cali Mod.  The power amp is going to be LTP PI pushing Tungsol 5881s with 380 plate volts, 6.6K primary impedance, and cathode biased.  I might put in a switch to use cathode or fixed bias.  I was hoping to get something similar to Soldano HR25.

[92Volts]

As Ed_Chambley says, "soft" tubes saturate (reach maximum power draw) sooner.

99% of guitar amps cannot drive the grid above 0v because of the phase inverter/coupling cap design. So the amount of signal that a power tube can take before saturation is directly determined by the negative bias voltage required at idle.

So "softer" power tubes require less negative bias to draw the same amount of current. I'm pretty sure that's all they test for.

This typically means they'd draw less current at 0v (max signal) too, basically being "weaker" tubes in general... but it's likely not a perfect predictor of undistorted power.

[brewdude]

The preamp will be like the Chupacabra or Fortin Cali Mod.  The power amp is going to be LTP PI pushing Tungsol 5881s with 380 plate volts, 6.6K primary impedance, and cathode biased.  I might put in a switch to use cathode or fixed bias.  I was hoping to get something similar to Soldano HR25.

I presume the Chupacabra is a high gain pre-amp, which generally implies (in my way of thinking) that PA distortion is less favorable and ends up being mushy and flabby.  Disclaimer: I'm no expert.

[mwelch55]

I realize that a high gain preamp is challenging in many ways.  I tried a high gain preamp on my first amp and then when it didn't work out, I modified it until it sounded great.  I spent a lot of time planning this one and I think it will work out.  If the power amp turns out not to be what I expected, I'll make some small tweaks and get it working like I want.  I am also considering EL34 output tubes, but will try 5881s first, since that was the design.

The attached schematic is what I had in mind.  The preamp voltages are not actual voltages in my amp, they are the voltages I am aiming for because that is the voltages in a working Chupacabra (Ceriatone) amp.  It is still a work in progress.

[PRR]

> The attached schematic...

Are we happy with the way the BJT clippers are wired?

On the face of it, one should be other way round. No?

[mwelch55]

I have never used that clipping configuration before (two BJTs).  There are several schematics out on the internet with the BJTs wired that way. 

A schematic that is supposed to be the Atomica has similar wiring except it uses different transistors.  Here is that part of the Atomica circuit.

[mwelch55]

I might not even use zener or BJT for clipping.  I'll attach it to the circuit with clip leads to see if I like it before I actually drill holes for the switch.  The schematic is a draft and may change slightly when I get the amp finished.

[PRR]

> BJTs wired that way. 

I'm pointing at the Base-Collector junction, where the others are Base-Emitter junctions.

B-E breaks down at 7V.

B-C breakdown is usually over 40V. In this case the part specified is a 120V part, and breakdown may be 140V. That's a big difference from 7V or 20V.

Interestingly, in this circuit, it does not matter NPN or PNP, base to base or emitter to emitter.

It may be a valid option. Clipleads are good. But at first glance the collector junction isn't what was intended.

[mwelch55]

Thanks PRR,

You've definitely given me something to think about

[mwelch55]

Ed was right about the difference between soft-medium-hard breakup is minimal. 

I bought a matched set of tungsol 5881s from Antique Electronic Supply.  There was no soft-medium-hard breakup option from AES.  At 340v on plate, the bias current was 54 ma.

I also bought a matched set from Musician Friend that were "soft" breakup.  In the same amp without changing any settings, the bias current (340v on plate) was 50 ma (not much difference).

When I listened to them, I couldn't tell any tonal differences.  To my ears, they sound the same.  My take is that the soft-medium-hard breakup ratings are just marketing BS.

Here is my schematic.  The amp sounds great (and very quiet for High Gain Amp).

[mwelch55]

When designing this amp, the B+ power supply unloaded was about 380v.  In previous amps that I have built, the B+ drops about 20 volts when the tubes are installed.  On this one, the B+ dropped about 40 volts.  Is that normal?  In my previous amps, I have used 6V6 output tubes and this amp uses 5881.  Does the 5881 tubes load the power supply more than 6V6?  Is that the reason my B+ dropped so much?

Mike

[Ed_Chambley]

Yes, 5881 draw more current than 6V6.


Voltage drop of 40 seems reasonable.