Hello everyone,
Very interesting topic which I have enjoyed following. It is not clear to me what the cathode resistor network's purpose is as its' implemented on the KT88(s). Specifically R-34-37, all 22 ohm resistors (see post #316). The two resistors one from each cathode to ground I understand their purpose. What is the purpose of the other two 22 ohm resistors that are connected in series running between the two KT88 cathodes for? Either I need to learn something here, or there seems to be a schematic error? Can some one please enlighten me?
Thanks, Doug S. (aka Mickeystan)
Very interesting topic which I have enjoyed following. It is not clear to me what the cathode resistor network's purpose is as its' implemented on the KT88(s). Specifically R-34-37, all 22 ohm resistors (see post #316). The two resistors one from each cathode to ground I understand their purpose. What is the purpose of the other two 22 ohm resistors that are connected in series running between the two KT88 cathodes for? Either I need to learn something here, or there seems to be a schematic error? Can some one please enlighten me?
Thanks, Doug S. (aka Mickeystan)
It SHOULD be, but I have been using ordinary 1/4 w metal film in my builds. Nothing has oscillated yet and they are less noisy.
I can't say that I understand that one either. I use one 10 ohm 1% 1/2 watt resistor in the cathode lead of all of my fixed bias amp designs. They are useful for setting the bias current, and may save the OPT if a tube arcs by acting as a fuse.
Heavens.
Tubelab, exactly where are you watching with the spec. anal. to detect these oscillations? I cannot see those continuing up the circuit, and am puzzled as to what amplitude those might be to cause interference - thanks!
Eli's ears are not as good as they used to be
They kind of 'came with the territory' as this circuit started out. They have no use now; probably came from a then different circuit. In that they will somewhat counter a d.c. reading over R34, R35 for bias adjustment, they should be taken out. Also, I would recommend that R34, R35 be made 10 ohms as per Tubelab's use - current NFB here by the use of any significant series cathode resistance can compromise the UL operation characteristics. As said, also handy as protection for the KT88s (as are small G2 resistors!).
Nope, nothing so elaborate. It was just my mistake on the schematic. The original schematic had the ground on the right side, effectively reducing the resistors to 11 ohms. Two birds, one stone.
[edit] my turn. typing at the same time.
Is there an advantage to using a parallel pair there? (other than to double their power handling?) On SS amps, this is done to reduce inductance in ww or metal film resistors (which are spiral grooved) on the output device emitters. Same here?
[edit] my turn. typing at the same time.
Is there an advantage to using a parallel pair there? (other than to double their power handling?) On SS amps, this is done to reduce inductance in ww or metal film resistors (which are spiral grooved) on the output device emitters. Same here?
Last edited:
On the matter of resistor types carbon composite are an order of magnitude noisier than metal film, carbon film and wire wound, in that perspective the 3 last mentioned looks suddenly very similar in their noise figure.
Resistors as fuse is not a bad idea, but then NEVER use c.comp resistors for such purpose, when they start to fire it's very hefty and they can destroy more than the "fuse resistor" idea would save, just an advice.
I can though understand when some guitar amp gurus use c.comp resistors but that's a totally different thing, they are "distortion artists" with respect and we aren't going to build a guitar amp are we?
oh and about tubes picking up RF garbage, this concerns also SS, and it's not only radio stations, mobile phones etc that can upset a circuits bias point and give a sublime harsh sound (It's not always a matter of oscillation!), even CD players at least in the past some had tremendous digital noise, this is reported by other too so it's not only my experience but I do remember 20 years back a CD player I had on top of my receiver would totally floor the FM receiver and I didn't understand why some days I couldn't receive radio stations properly (88-108 MHz region!) until I by chance switched the CD player off while listening to the radio and suddenly noticed a dramatic difference. There are so many ways RF/digital noise can enter an amplifier, for instance our lovely speaker cables are perfect antennas, long and un-shielded, well for some 25 years back in my youth when I was very fond of RF-tech stuff I did also build couple FM transmitters too so I am a bit familiar with the HF stuff.
Cheers Michael
Resistors as fuse is not a bad idea, but then NEVER use c.comp resistors for such purpose, when they start to fire it's very hefty and they can destroy more than the "fuse resistor" idea would save, just an advice.
I can though understand when some guitar amp gurus use c.comp resistors but that's a totally different thing, they are "distortion artists" with respect and we aren't going to build a guitar amp are we?
oh and about tubes picking up RF garbage, this concerns also SS, and it's not only radio stations, mobile phones etc that can upset a circuits bias point and give a sublime harsh sound (It's not always a matter of oscillation!), even CD players at least in the past some had tremendous digital noise, this is reported by other too so it's not only my experience but I do remember 20 years back a CD player I had on top of my receiver would totally floor the FM receiver and I didn't understand why some days I couldn't receive radio stations properly (88-108 MHz region!) until I by chance switched the CD player off while listening to the radio and suddenly noticed a dramatic difference. There are so many ways RF/digital noise can enter an amplifier, for instance our lovely speaker cables are perfect antennas, long and un-shielded, well for some 25 years back in my youth when I was very fond of RF-tech stuff I did also build couple FM transmitters too so I am a bit familiar with the HF stuff.
Cheers Michael
Last edited:
There are only a few items left to resolve to finish this off:
Then I can start drawing the 6GK5 version, while someone "smoke tests" and tunes this version.
..Todd
- The global feedback component values (R16, R17, C13)
- The mains fuse rating
- Name
Then I can start drawing the 6GK5 version, while someone "smoke tests" and tunes this version.
..Todd
Attachments
Last edited:
Oh thanks dhieber,
I use Adobe Illustrator. Schematic capture applications just frustrate me when it comes to drawing restrictions. But then Illustrator doesn't consider electronics, so all error trapping must be done by committee.
I've used it for designing DIY PCB's too, but since it doesn't reverse annotate or correlate to the schematic (no ratsnest), I've decided it's too error-prone of a method. You really need to use a PCB package to make sure the traces match your schematic. It took me about 6 revisions to correct all the wiring errors on this board.
..Todd
I use Adobe Illustrator. Schematic capture applications just frustrate me when it comes to drawing restrictions. But then Illustrator doesn't consider electronics, so all error trapping must be done by committee.
I've used it for designing DIY PCB's too, but since it doesn't reverse annotate or correlate to the schematic (no ratsnest), I've decided it's too error-prone of a method. You really need to use a PCB package to make sure the traces match your schematic. It took me about 6 revisions to correct all the wiring errors on this board.
..Todd
Attachments
Last edited:
I have two "sniffer probes" that I bring in close proximity to each tube and semiconductor in a running amplifier. One "probe" is simply a piece of coaxial cable with about 2 inches of the center conductor exposed (shield braid stripped back). This end is covered with heat shrink to avoid accidental fireworks. The other probe is a "coil" of about 10 turns wrapped around a ferrite core of unknown origin. This is connected to a cable and covered with heat shrink. The idea is to "listen" to the offending circiut without electrical contact. I have seen many cases where oscillation ceases as soon as a capacitive scope probe is connected into a circuit.
The plain sniffer is used most often and responds to frequencies from a few hundred KHz to over 1GHz. The coil based sniffer responds to low frequencies and can be used to find ringing power supplies and noisy switchers.
Some amplifiers especially ones using the 5842, 6EJ7 and other high GM tubes may oscillate so rudely that they wipe out the reception on a nearby TV set. Usually the oscillation is much milder. Often the amp may only oscillate for a small fraction of the audio cycle, often at the transition into or out of clipping, or as one tube goes into or comes out of cutoff in a class AB P-P amp.
Sniffing the amplifier with the spectrum analyzer can often find these cases, and lead you to the offending circuit. The analyzer will of course pick up all of the RF pollution that we have in our environment like radio and TV stations, cell towers, and unintended emissions like the noise leaking out of this computer. Usually an oscillating amplifier 1 inch from the probe is much stronger. If you see a signal that you believe is suspicious, switch the amp off and see it it goes away.
I am missing a reply from me to the previous post, also have been off for a few days.
Has this topic been transferred elsewhere/terminated? - not that I can find. But I have now done the complete amplifier (mock-up) as a 'chopping block' for further commentary - but would not like to waste time by posting on a discontinued thread!
Has this topic been transferred elsewhere/terminated? - not that I can find. But I have now done the complete amplifier (mock-up) as a 'chopping block' for further commentary - but would not like to waste time by posting on a discontinued thread!
Johan,
I think people have paused to catch their breath and (possibly) gather parts. Your mock-up has to be of considerable value.
In the last schematic, I drew the LTP CCS terminating to ground, but I also see it being run to the negative bias supply in some schematics. Does it need a more negative terminus?
I'm reading steadily out these tube things, but I have a long way to go.
..Todd
I'm reading steadily out these tube things, but I have a long way to go.
..Todd
Attachments
Last edited: