|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
63.225.146.247
In Reply to: Re: proper biasing posted by Zenjabil on December 21, 2005 at 05:12:17:
The new output section sees about 1/10 the number of tube failures, plus it has better low level detail and is smoother sounding too. You would also pick up some authority as well.
Follow Ups:
No ...Not really.. I'd have to spend more money upgrading the amp, than I paid for it in the first place. I just want as much info as possible to allow me to set these up properly. I think what Z said is probably the case: When I put new tubes in , I do preheat them for a couple of days before hitting them with the B+, howevewr the tubes that were/ are good are of questionable age and stress.. because, as he said , when a tube section fails, the rest of the tubes have to pick up the extra current that the failed section used to draw. I wish there was some indication when a tube section first fails/ IOW some visual change in the bias current / I could check it on a regular basis , and if it changed abruptly, I could do a complete visual tube inspection at that time... If I were to use an external digital meter , instead of the built in one , how should I hook it up ? I should probably replace every power tube from scratch and go from there ? Maybe if I got a sweet deal on a trade-in of these [ they are in mint cond.], I'd get a brand new set of M60's :)))))
I think Ralph or Bill can fill you in on the details, but as far as I recall the amplifier itself can be used as a tube tester. You do have to check each tube separately. For the amp to behave as you would like, i.e., for the bias to fall as an indicator that one section of one tube has failed, would require some very complex circuitry that might negatively affect the sonic quality, if it could be done at all.
Not that I'd like to disturb the bone stock nature of my amps, but just to throw around some ideas... I was thinking that installing 2 test point terminals on each tube [ one for each section ], seems easy enough... It would either read around 20ma per section for a good tube.. and read zero for a tube section that is dead and drawing no current. Or another half baked idea that I had was putting a LED in series with each section , If any LED's either flicker or extinquish , than that would indicate a dead or dying section? Although putting a LED in the signal path would probably degrade the sound? I'll probably never modify these amps in any way , but was just thinking for fun. Kevin Davis
All the tube sections in each phase of the output circuit are in parallel with each other and driven by a single driver tube section. One of the later mods to the output circuit includes the installation of very low value resistors, one in series with each plate or cathode of each output tube. (I think Ralph finally settled on installing them on the plate side of the circuit.) If you have these resistors installed in your amplifier, and if you can detect a voltage across one of these resistors, you could calculate current using Ohm's law. Since the tube sections are in parallel, you would unavoidably be measuring the total current, but if there is no detectable voltage across a given plate resistor, that would indicate that the tube under test is a "dead" short. So you cannot use this method to tell whether one tube vs another is current hogging, but it might work to tell you that a given tube is kaput. I am feeling insecure in writing this without having looked at the schematic in months, so if Ralph or anyone else thinks I am wrong, please do jump in and say so.
> low value resistors, one in series with each plate or cathode of each output tube.
> if you can detect a voltage across one of these resistors, you could calculate current using Ohm's law.
> Since the tube sections are in parallel, you would unavoidably be measuring the total current,I don't know the circuit, but if I understand your description correctly, you would not be measuring total current. Only the current through the resistor being probed.
> but if there is no detectable voltage across a given plate resistor, that would indicate that the tube under test is a "dead" short.
No voltage would indicate no current thus an open....
(but I don't know the failure mode of tubes)
A dead short (plate to cathode) would probably burn the resistor
and if not, the voltage would be high.> So you cannot use this method to tell whether one tube vs another is current hogging,
Thank you for pointing out the errors in my thinking. Tubes can fail in either mode, as an open or closed circuit, but that does not usually happen except that the 6AS7 has an internal fuse which can blow and give you an "open", in which case one would measure no current across its plate (or cathode) resistor. With regard to my statement about what you would see if you measure V across one of the plate or cathode Rs in the case where the tube is functional, Thievenen's (sp?) Law would suggest you are correct.The reason for the introduction of the plate or cathode resistors in the Atma-sphere circlotron was indeed to equalize the current draw for each tube under dynamic conditions. One might argue that the very low value of the R used (don't remember but ca one ohm) would not be sufficient to have a significant effect on this phenomenon, given that each tube has a plate resistance of (don't know for a 6AS7, but greater than 100 ohms). I have no opinion because I have never installed this option into my Atma-sphere MA240s which use 6C33C output tubes.
Why isn't the 'fusible link' mentioned or shown in the poop sheet?
Or is it simply a weak spot in the current path that 'acts' as a fuse? ....... mike
The instructions for the amplifier kits years ago had a copy of a data sheet from Tung-Sol for the 6AS7. It does not mention a fusible link either. There is a diagram of the Tung-Sol tube and it is shaped like the Sovtek or Svetlana. But I bought, in ignorance, RCA 6AS7 and it is smaller, just an inverted U-shape without the heat dissipating structure on top. Those tubes arc pretty quickly, as I learned. I looked at them and at the 6SN7's and their links look no stronger or weaker than the ones we're talking about. But the two RCA tubes I tried as emergency replacements, for all their arcing, still have their links intact.The original update installed 1 ohm resistors at the plates and 150 -180 ohm resistors at the grids.
The 150- to 180-ohm resistors at the grids are "grid-stoppers" and have nothing to do with the subject of this thread. Grid-stoppers prevent hf oscillation caused by an interaction related to the Miller capacitance of the tube itself. Typically grid-stoppers are needed where the tube has a very high Gm. It's those one-ohm resistors, which in a circlotron could be installed either at the plate or cathode, that are there to ameliorate imbalances between the tube sections that are in parallel with each other.
I think for current sharing the 1 Ohms need to be at the cathode. As you said before, the high output impedance of the plates should make a 1 Ohm resistance almost moot. (I'm relating to SS cicuitry (all I know anything about)) so may be missing something......mike
Thanks for the information. I'm very ignorant and try to keep my comments to just what I have experienced with my amplifiers. The rest of the time I lurk and learn. If resistors at either plate or cathode can have the same effect, can you say why they are now put at cathode rather than plate?
When I got the resistor upgrade from Atma-Sphere for the M60 kits I had built, the resistors were for each grid and for each plate of the output tubes. I was told one advantage was being able to measure across the plate resistors so you could match tubes.Recent posts here from Ralph suggest they no longer use resistors at the plates, but at the cathode instead. I would think you could also use the cathode resistor to measure tube performance.
As for finding bad tubes, wouldn't it be easier just to pull out the tubes one by one and look at them? That's how I found my latest one -after a heart-stopping zap in the speaker gave me a hint. It took only a few minutes to check the tubes and examining them is a nice zen-like exercise that at the least renews one's respect for tube design and manufacture.
> one advantage was being able to measure across the plate resistors so you could match tubes.Makes sense
> they no longer use resistors at the plates, but at the cathode instead.
Makes sense too. Helps swamp out variations in the tubes.
> I would think you could also use the cathode resistor to measure tube performance.
Right.
> wouldn't it be easier just to pull out the tubes one by one and look at them?
I don't know much about tubes, but I would think plenty could go wrong with no obvious physical manifestation at all.
> a nice zen-like exercise that at the least renews one's respect for tube design and manufacture.
I'd use a meter accross the cathode resistors, and ogle my spares for mystical insight ........... :) ........ mike
> they no longer use resistors at the plates, but at the cathode instead.
> Makes sense too. Helps swamp out variations in the tubes.I think that was also what the plate and grid resistors were meant to do.
> I don't know much about tubes, but I would think plenty could go wrong with no obvious physical manifestation at all.
The tubes in question have fusible links. The blown links are what the original post is about. Whether they are blown or intact can be determined by just looking at them.
> I'd use a meter accross the cathode resistors, and ogle my spares for mystical insight ........... :) ........ mike
For this common failure visual inspection is a lot easier than disconnecting the amps and opening them up for testing.
While zen-like tube gazing may be good it is seldom done. The positive experience of examining the tubes and gaining respect for their complex manufacture happens at the best moment to counteract the negative effect of having a tube go bad. With my budget I have had to come up with philosophical approaches to the cost of tubes.
Bill Layer said something about designing a tube tester devoted to the specific needs of Atma-Sphere users, but I haven't seen any mention of it since. I found a schematic for a tube tester. It's complex but maybe could be pared down to just octal base for a start.
?
(this is the right tube, right?)
?
(this is the right tube, right?)
> ?
> (this is the right tube, right?)
Not 6sn7. 6as7 the output tube. Go here:
If you're only worried about catistrophic(sp?) failure then OK ....
Geez ........... Quite a 'forgiving' circuit ! :)
It's a bit discouraging to learn that to cure tube failure one has to spend hundreds on an upgrade. I built the amps from kits and then spent time and money tearing out the wiring to install the resistor upgrade which you once offered. Now you recommend spending more to undo that. I can see both sides of the issue, and respect you for even offering upgrades. But replacing tubes is cheaper, so I'll have to stick with that.
This post is made possible by the generous support of people like you and our sponsors: