Dynaco ST-70 and tube life or outputs tubes and rectifiers

6,

perhaps you can clarify Dave's statement in the post #16 link from Peter. The implications of this are that it DOES 'double' the current capability.

"If I may build on Stormy's original concept however, consider this: If you install diodes that can handle the PIV requirements all on their own, then install them as originally directed, but additionally, run a jumper between pins 4 and 6 of the rectifier tube socket. Now the sections within the tube are paralleled, which doubles the effective cathode area being used at any given time. Double the cathode area provides double the current capabilities. It's like getting the benefit of two rectifier tubes out of one tube. In this way, both the PIV AND current limitations of the tube (either modern or NOS) are enhanced."

For the record, that's an excerpt from Dave's post #31 in that thread link. I'm not sure how to do this 'snip-it' thing and get the pretty blue box. But, as I read it, by doing this you 'double' the current capability of the tube. However, as this is a 'cycle' thing...... I can see how it might be construed as being still limited to the 230ma or what ever the 5AR4 is rated for. But, that begs the question; Are you only getting 115ma out of the tube at any one point in the AC duty cycle? This is getting clear as mud for me.........

Got it, thanks! WC

6DZ7 said:

You get the full current out but each section is conducting 1/2 of normal, twice the cycle rate, together, instead of being in alternation.

And add the resistor for the voltage drop, which of course can be customized to get the B+ you want. Doesn't have to be exactly what the 5AR4 or any other gives.

But of course the peak voltage needs to be respected. The cap voltage rating is always a "working" voltage rating so be up to that shouldn't cause so much anxiety as gets posted. A quality cap gets hammered in testing for 1000's of hours at that and higher. So a few seconds during power up is nothing. I believe most caps are normally tested to withstand 150% of working voltage for a specific period. And have a higher short term peak rating as well as the WVDC.

Click to expand...

A quality cap. This makes me wonder why 75% of the upskirt shots on AK have the cheapest lytics available in them. Don't be Vague, specify Sprague.

BTW, the real weak link in the 70 is the power transformer, but that's easy enough to change with the repros out there with color coded wire even. I guess it could change the sound by not sagging on a heavy bass note or something, but the voltages and all are the same. I bias my tubes at 35ma, and expect them to last forever. If I still had tube rectification, I'd get an Amperex or Mullard from a trusted person and be done with it. They last for ever. I like the idea of using the 6U8A, it's a favorite of mine and tends to fly under the radar.

Let me add to Peter's question by addressing the issues of the diodes doing some of the rectifying. I guess that even though each diode now 'rectifies' to DC.....they are still doing so on opposite sides of the AC 'out' of the PT's HV winding.....so there is still an alternating pulse here? I'm back to stuck in the mud again too. Glad to see Peter still questioning this as well. Thanks Pete! WC

Peter, I'm also in that same camp. Seems as if the overall rating being, I believe, 250ma, then each plate is capable of 125ma? OR...... is that the full 250 on each plate but at alternating intervals....such that as I believe Dave was suggesting, now each plate at 250 can combine to conduct a full 500ma? Going back to the 'cycle' thing, if the diodes are taking the pulses.....and can carry the current (particularly when using two each per leg).... I'm now not sure of what Dave was saying about 'effectively double the current'. Seems to me it either 'is' or 'is not' double the current. If as 6 says, it's double the current but at half the duty cycle (or how ever he put it)..... then it really isn't double the current at all....which makes me wonder why Dave would imply it is? Guess where this is lacking the in-depth understanding comes into play. But, as Peter says: questions are good! Sometimes, it's just understanding the answers that isn't so easy!! WC

6,

Ya, ya, ya........ I'm sure you're right. But, I still am confused by it, so help me 'get' it fully. In your post you said: "...full current out of each section.." so..... 250 ma? Then you said: "....conducting 1\2 of normal....." and then "... at twice the rate, together, instead of in alternation." So, that still sounds like 250 X 2 (for each section) X 1\2 (the 'conducting part) then X 2 (twice the rate).

250 X 2 = 500 X .5 = 250 X 2 = 500

Starting to see why this is confusing to those of us who don't fully understand it all? Again, not trying to be difficult...... just trying to really get a grip on the concept here.

Don't want to have to add a stop-watch to my tool collection! WC

6, so the math is more like 1\2 X 2 = 1 X 250 which now makes some sense. I'm still a bit fuzzy on the reason why the concept of 'double current' is even being discussed in this. It also seems to negate Dave's prior statement. WC

I tend to be of the opinion that power supplies should not "sound" like anything. Its DC power. The only thing it should be doing is providing clean voltage at the proper level and at sufficient current levels to not sag. If it does otherwise, you "hear" it, and its not a good thing.

Listen to a tube amp with a weak rectifier sometime. They have mushy bass response. Hang a voltmeter on the B+ line, and you can see it dip every time the sound level peaks, which goes along with that mush sound. Make that not happen and suddenly instead of "warm and tubey" you get "tight and controlled".

If you look at fancy high end tube amps, you see big caps and big chokes. Its not just for show.


All that said, I kinda like vacuum rectifiers, but mostly for visual reasons. I tend to leave them alone if they aren't causing trouble but I don't really think I'd go out of my way to design an amp for myself that used one. They're expensive, not conducive to a low impedance power supply, and they make for a more expensive transformer.

Agreed!

Interesting discussion!

The term "duty cycle" (as it relates to rectifier tubes) relates to both plate dissipation, and cathode capability, so let's take them one at a time.

Taking plate dissipation first, and keeping it simple then, the concern raised is that if a half wave rectifier tube is made to conduct continuously (as in passing pure DC current), then it's plate dissipation (duty cycle) is doubled over that of passing an alternating (AC) current, where the tube only conducts half of the time. That is absolutely true.

Plate dissipation in a diode is the product of current flow times the conducting voltage potential (drop) between the plate and cathode within the tube over time. Considering each section of a full wave rectifier tube in isolation then (basically a 1/2 wave rectifier), if current that normally flows only half the time is now made to flow all the time, then the plate dissipation of the section is doubled, because now electrons are striking the plate all of the time instead of half the time. However, if both sections are paralleled, then the voltage potential between the plate and cathode is now half of what it normally is in each section, because both sections are conducting all the time now. Going back to the plate dissipation formula then, with the sections paralleled, each section can now pass the same current flow ALL of the time, yet plate dissipation will remain the same.

Relating this to a GZ34 then (and excluding charging currents), each section is capable of providing 225 ma into a load using a cap input filter, which produces a 17 volt drop between the plate and cathode -- as long as it is conducting only half the time. When both sections conduct in alternating fashion, then the full 225 ma is available all the time. Relative to each section then, this means that each plate is capable of dissipating 225 ma X 17 volts = 3.825 / 2 (since it is only conducting half the time) or 1.9125 watts.

So what happens when the sections are paralleled, and made to conduct all the time? Well, each section passes its current flow of 225 ma, but with only 8.5 volts dropped across the elements, the plate dissipation for each section becomes .225 ma X 8.5 volts, or 1.9125 watts -- the same as before. So the plate dissipation of each section remains the same as with conventional operation, except that with both sections capable of dumping 225 ma into the load all the time, the tube can now pass 450 ma if the load demands it.

But being ever skeptical, you just don't get something for nothing, right? There's got to be a catch, and there is.

The duty cycle -- as it relates to the plate -- isn't the problem. But it is at the cathode. Oh the dual cathodes will pass a total of 450 ma, because with each individual cathode only having to pass 225 ma, that's what they were designed to do.
But they were designed to do that only half the time, not all the time. So the real impact of increasing the duty cycle with the proposed connection is not whether the tube can reliably pass double the rated current flow -- it's for how long it can do it. As you would guess, it can do it for half of the normal life if operating at a 225 ma total current flow in a conventional setting. Cathodes can only emit so many electrons for so long. Relatively, if you double the electron flow from the cathode, either instantaneously or over time, then serviceability is cut in half.

Now, relating all of this back to the ST-70 (where it came from), that amplifier under normal quiescent conditions draws about 215 ma, or very nearly the limit of the rectifier tube just idling. By pre-rectifying and paralleling the sections of the GZ34, the tube is operating at about 1/2 of its new 450 ma capability. Now that 450 ma capability cuts tube life in half as discussed above -- if it was actually operating at that current level. But since it's only operating at (about) 1/2 of the new capability, tube life is increased right back to where it was in the first place before all of this started -- except that the tube can now handle the current flow much more easily, with less voltage drop.

I hope this helps!

Dave

Unless someone comes up with some kind of 'But......', I get this! Moreover, I appreciate the detail in the response. Seems to me this is a wise 'mod' to make, particularly on the ST-70. Going with Gadget's comment about Power Supplies, sound, sag, 'cleanliness', and proper levels....seems to me that sacrificing a bit of 'life-span' on some of the replacement rectifier tubes out there to get the results of this application is a small price to pay....particularly when a pretty decent 5AR4 can be purchased for about $20 these days.....and with the dual-diodes AND the common cathode cathode...... this seems like a very wise thing to do to improve the top-end performance of the amp. WC