Tubes?

[turbo]

Thanks Pese 
i was looking for that.

Marco.

[z_p_e]

There has been much talk about tubes vs. transistors regarding negative feedback (nfb), speed or transit time, distortion or purity, and perhaps a few others.

At times I think folks read too much into what SM has said regarding these things, and taken literally, this may lead them down the garden path. I am not saying SM is wrong. I am saying folks may not be separating the wheat from the chaff.

Let's start with Transit Time. Indeed, tubes achieve about 1ns in this regard...considerably better than SS. This fact alone is reason enough to use tubes first. Evidently, the TPU requires fast response times, and I agree. If someone can show proof that nfb increases transit time, please do. However, my current understanding is that nfb has no effect on the transit time of an amplifying device such as a tube, BJT, JFET, or MOSFET. If anything, it may actually have a beneficial effect due to it's curative properties regarding input impedance etc.

Distortion or purity is a no-brainer actually, but many make a big deal out of this. First of all, distortion is really only a concern when dealing with anything other than square waves or pulses.The following is only to appease the "sine wave" advocates.

Common misnomer: Tubes are cleaner than transistors...Wrong! They simply have different spectral characters....that's all!

Another: MOSFETS are similar to tubes...Wrong! The closest SS device to a tube is in fact a JFET. A MOSFET is an enhancement device (most of them), but tubes and JFETs are depletion mode devices.

A simple class A JFET amplifier stage will in fact have a very similar transfer curve and distortion spectrum to one designed with a tube triode.

Negative Feedback

Now here is where many cans of worms can be opened, but I wish only to address a couple. Right off the start, I must again state that this is really only an issue when dealing with non pulse-like wave forms. To once again appease the sine crowd, we continue.

Negative feedback is too general as used in SM's material. NFB can be applied locally around each device, around a few devices, around all devices from input to output, or all combined, both local and global feedback.

Many tout low feedback designs being better, and in terms of TIM distortion, I agree. How is low nfb achieved then? It turns out that the biggest problem with nfb is enclosing many devices in an overall loop. Many devices in series typically means a lot of gain, and that means instability challenges are ahead. "Low Negative Feedback" typically means local feedback only is used....none globally. All designs use nfb to some degree, even if unintentional! For example, if SM designed and built his own tube amplifier or generators, and he used triodes in some or all of the stages, then he was using nfb, even though he may not have been aware of it. Triodes, because of their design and inter-electrode interactions inside the tube, inherently have nfb built in! Look it up if you don't believe me.

All single device amplifier stages use some nfb, even inverters. A piece of wire from cathode to ground still exhibits some resistance and inductance, and hence there will be nfb around the device....it's called degenerative feedback. So regarding triodes, there are actually two forms of nfb being applied simultaneously... and that is without trying.

SM states that SS device amplifiers require tons of nfb to make them work, and because of this they are dirty etc. The amplifier SM is talking about is a typical one, but does not apply to all designs. It all depends on how the amp was designed, and where the nfb points are applied. Perhaps the amp has no global nfb? The point is, is that low nfb designs are possible using all gain device types. Some however are better suited to low nfb designs than others....BJT's being the worst.

Keep in mind that SS or specifically, BJT's are really only "dirty" once they are approaching saturation. But then so do tubes become dirty as well...just in a more gentle fashion. A JFET amplifier is the closet one will come to achieving the same gentle saturation that tubes exhibit, but one has quadratic (squared) type transfer function, while the other exhibits a "1.4 factor" transfer function. Guess which one goes with which?

SM has stressed many times that tubes are the way to go when starting out, and we should believe him. However, in order to make his case stronger, perhaps there was some embellishment of the facts.  Also, we must remember that much of what SM has conveyed, was done so metaphorically, and so it is quite possible that we all missed the real message (if there was one) regarding how much nfb transistors use vs. what tubes use.

In my opinion in terms of the above discussion, most of it is irrelevant. I believe most if not all of what SM divulged was meant to be taken in the context of pulses, so distortion and nfb do not come into play. Timing and precision are the key issues, and is what was referred to when "exact" and "pure" frequencies were mentioned. But that's another post perhaps.

Cheers

[turbo]

Darren,

Let's keep it simple.
Tubes will do the job far more easy, they do NOT require any feedback to achieve catalyst sequence.
They are stable as they are.
Solid State requires feedback and without it, the unit wont even start or just shuts down becuase the control system is affected too much.

Steven told he had a hard time with the SS control part and it took him years to devellop them.
everytime the signals were what they suppose to be, it just did not work.

most of us are in silicone valley because we are more familiair with it.
I just decided to take the road he walked, so this means from tubes to ss.

Marco.

[z_p_e]

Let's keep it simple indeed.

Negative feedback and distortion is irrelevant. Colpits or other sine wave oscillators are irrelevant. I am not the one discussing these things in previous posts about tubes and what needs to be designed to use them in the TPU

Why?

Because pulses are used in the TPU, not sines imho. Tubes can be used for pulses too!

[dutchy1966]

Let's keep it simple indeed.

Negative feedback and distortion is irrelevant. Colpits or other sine wave oscillators are irrelevant. I am not the one discussing these things in previous posts about tubes and what needs to be designed to use them in the TPU

Why?

Because pulses are used in the TPU, not sines imho. Tubes can be used for pulses too!

Hi Darren,

Can you maybe clarify why you think the TPU uses pulses and not sinewaves?
I clearly remember SM mentioning Bushwackers and Freedomfuels posts. In my opinion Bushwacker was mentioned for his research into the HOPE device, which uses AUDIO SINE WAVES as the input.
This was one of the few occasions SM dared to directly point us at something imo......

regards,

Robert

btw, just for the sake of completion.....Freedomfuels post is about setting up a magnetic vortex to the ionosphere, which was the other direct reference SM gave.