Why do people design outside of specs of EL84?

[Davidscott]

I am hoping someone can answer this question that has probably been asked a million times before!
I have looked at a lot of tube amp design schematics that use EL84 power tubes. Every one I have seen, show the B+ (voltage on the plates of Push-Pull) much higher than the specs show. In looking up specs on this tube I see they rate them from 250V to a max of 300V. Yet I have seen many circuits showing way over 300! even as high as 400V! Does anyone really know why this is? and what would be the results if I build an amplifier that actually ran the EL84 in the 290V range? I am working on an amp right now that has 323V and 322V on its two Tubes. Thinking of lowering this. BTW, its biased at around 29mA.

Please help me understand this. When I have presented this question on other forums, i get the answer (its fine, these tubes can handle a lot). But if that were true, how come the tube MFG do not spec these to a higher voltage!
And my other concern is the temperature of these tubes. I measured it to be between 308F to 316F. that seems really hot! Yet the bias point puts it in at around 68% dissipation at 8W per tube. way below max Wattage for this EL84.

Thanks for the help.

[JMPGuitars]

The vague answers are correct. But let's consider it a little further.

What are the datasheets for? Guitar amps? HiFi? Old electronics? Corporate protection?

Is Va (anode voltage) really the most important thing? Or is it really a question of what the tube is actually doing (and overall output)?

Good tubes are generally a lot tougher than what the datasheets show, and you can see this by reading...the datasheets! Look in the back where they track anode characteristics, and you'll see some EL84 anodes get tested up to 600V. Newer sheets might not show that because production maybe isn't as good, or maybe companies prefer to not be liable.

Tube dissipation is probably the most important factor for managing healthy tubes. With our cathode-biased 18-watters, we target 85% tube dissipation. Go too high or too low and you'll have issues with noise or tube longevity.

The output current is controlled by balancing different aspects of the tube. Vg, Vs, K, etc. all play into this. Driving the current beyond what the tube can sustain is what you need to avoid.

Here's a new production 7189 tube I tested on my FUtracer tube tracer:

7189 Tube 2-1_plot_1.jpg

7189s are supposed to be mil spec EL84 tubes, that are rated higher. I run EL84s through the same tests also. You can see I drove the anode current over 200mA there. In other words, the tubes can handle what we're throwing at them in these guitar amps.

Thanks,
Josh

PS. Always pay attention to notation when discussing this stuff. 29mA is not the same thing as 29MA. 29mA is equal to 0.029 amperes. 29MA is 29,000,000 amperes. That would wreck a tube and then some.

[Bieworm]

That’s great to know. I would like to add another detail..
some brands are by experience known to be feeble, and others will handle voltages that will light up a dark room redplating. I avoid JJ EL84 at all cost! Mil spec Sovteks will be far more forgiving.

[Davidscott]

thank you Josh. that is a very good perspective! and does help answer some of my concerns. A little background on me. I am no expert on tube amp design. I went to college and studied electronics engineering. I have been an engineer now for over 30 years. In the modern design world, we always (for the most part) follow the specs. We rarely knowingly design outside the limits of MFG spec sheet. Doing so many times can bite you. And I can think of a couple times in my career where it has, when someone did not read or interpret the specs correctly. So its concerning to me when I read tube specs, and then I review schematics of modern amplifiers and see they are not following those rules! But as you stated, seems the tubes are much more robust than they are given credit for.

As a follow up question: how well would an amp (for example Fender blues jr, or Marshal xxx) perform if it were to be designed within the middle of specs and not on the extreme upper limits. Would that amp perform just as well? or better? Or worse? I wonder this, and I always try to design to the middle of the specs. But am i being to obedient? As long as you follow the basic rules the tubes will work. So why push the limits?

[JMPGuitars]

I totally understand where you're coming from. Tubes are a different beast though, especially for guitar amps. Consider any device you may work with. Some may have very strict specs, others may list their specs depending on specific variables.

A tube's limits and performance are based on variables. You know for safety, the thing that matters most is dissipation, since that gives you the sum of the variables. We can see from calculating the bias that guitar amp circuits do (usually) consider the real world limitations of the tubes, but they intentionally push limits to some degree.

Your 29mA dissipation isn't even close to the typical or limitation values in some of the sheets. I assume you were listing Ia which is often ~50mA in typical characteristics.

The most important factor to understand guitar amps is that even our clean signal needs distortion, or it sounds sterile and unpleasant. If you want a truly clean guitar sound, you get a solid state Jazz Chorus or a modeler based on one. I'd bet you'd still see some distortion if you put it through the appropriate tools.

In designing tube guitar amp tone, you want to push the tubes into distortion. The trick here is to balance the ideal tone without killing the tubes too quickly. So we push to 85% dissipation with heightened Va, with Vs < Va and we might tweak from there. Those are the basic rules for our guitar amp tone. Other circuits will go to different places, but if you review enough, you'll see some common themes (especially if categorized by bias type).

Designing for guitar amps is different than HiFi. In HiFi you might consider the typical characteristics and limits of the tube more strictly since the goal is near zero distortion. HiFi design is where the engineer's brain will likely be more comfortable.

You have to break the rules to make some noise.

Thanks,
Josh

PS. The sentence "In designing tube guitar amp tone, you want to push the tubes into distortion." is probably the best and simplest answer to the thread title.

PSS. To answer your follow-up question more directly, designing in-spec with essentially cold tubes will give poor tone.

[Davidscott]

Awesome response Josh! Thanks for the education. Something for me to process...
I want to start a new thread because now you got me questioning some of my math.... stay tuned.

Dave