|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
182.239.117.20
Folk,
I heard a lot of people saying SS plate CCS may sound glare.
A knowledge member once posted the following saying his actual experience, putting a resistor between CCS and tube plate could avoid the glare sound. What do you think of this approach and if it could avoid glare then what is the reason behind ? And what would be the value of this resistor ?
Follow Ups:
I have never liked CCS in any tube design except for the 6418 which is already a bit crazy to deal with.
No matter who's design, mine, theirs whoever it always sounded thin. A nice BAC (big a$$ choke) always sounded better.
Gordon
J. Gordon Rankin
> > No matter who's design, mine, theirs whoever it always sounded thin
Hi Gordon,
Have you studied why CCS would sound thin?
"Have you studied why CCS would sound thin?"
Perhaps a better question is why a choke sounds so good, better than either a resistor or a CCS. I believe the answer might lie in a choke's energy storage, something lacking in both of the alternatives.
Eric,
Sorry for the delay, not up here much these days.
~~~
Really it shouldn't matter that much. Sure you have to account for the change in voltage and any variation of the CCS because of that. But even with minimal ones like a 6GM8/ECC86 at 1ma and has 20V range and only going 1Vac rms at 1K. The difference between something like that and what I was using before a 400H/2ma (running 1ma regulated in the cathode, bypassed) butt stacked choke the difference was staggering.
I pretty much tried 1ma, 5ma, 10ma and then on some amps 40ma, 60ma and 100ma in various designs and some crazy ones I dreamed up.
Thanks,
Gordon
J. Gordon Rankin
Register / Login
Home
AudioAsylum Trader Audio Asylum Thread Printer
Get a view of an entire thread on one page
For Sale Ads
FAQ / News / Events
[ General ] [ Amp/Pre ] [ Speakers ] [ Hi-Eff ] [ Head ] [ Tubes ] [ SET ] [ OTL ] [ Tube/DIY ] [ Vinyl ] [ Cables ] [ Tweaks/DIY ]
[ Critics ] [ Digital ] [ Hi-Rez ] [ PC Audio ] [ Prop Head ] [ Iso ] [ Tape ] [ Tuners ] [ Vintage ] [ Music ] [ Classical ] [ Jazz ] [ Rock ]
[ Video ] [ Films ] [ Chat ] [ Reviews ] [ Inmate Systems ] [ Central ] [ Discount Dr. ] [ Shady ] [ Gallery ] [ Search ]
More Asylums
[ Asylum Support ] [ Forum Rules ] Find:
Tube DIY Asylum
Does a CCS work if placed in series with a plate resistor? (nt.)
203.94.155.130
[ Follow Ups ] [ Post Followup ] [ Thread: Display Email Next ] [ Tube DIY Asylum ] [ FAQ ]
Posted by Anthony Dockrill (A) on June 12, 2003 at 22:19:20
Follow Ups:
Yes, it has advantages. - VoltSecond 16:38:18 06/15/03 (1)
In Reply to: Does a CCS work if placed in series with a plate resistor? (nt.) posted by Anthony Dockrill on June 12, 2003 at 22:19:20:
I see I'm in the minority on the positive aspects of using a series CCS resistor.
A series resistor with the output of a CCS offers protection for the CCS. It is useful to limit current during arcs (tube and transformer), leading edges of shorts (oops, I shorted the output leads again), rapid reductions of the B+ (the B+ arced/ failed/ shorted or another tube on the B+ arced), the tube went into oscillation and other nasty oopses. 5 to 10% of B+ is a good starting point for the voltage drop on this resistor.
A series resistor also helps with stability of the CCS/ tube combination. At high frequencies the CCS appears as a capacitor and possibly a negative impedance if you are using a Hawksford cascode or if you have long leads (> 1 inch) on a standard cascode. This resistor buffers that capacitance (and negative impedance.) 1 to 5 times the current set resistor value is a good start point for this resistor value.
A series resistor helps when the CCS saturates. It softens both the recovery after saturation and at the start of saturation. 5 to 10% of B+ is a good starting point for the voltage drop again.
There are times when we don't need a series CCS output resistor, but they are fewer than what many would expect. It is easier to just design the resistor in, even if it doesn't drop as much voltage as the rules of thumb above. Sometimes the CCS damage from an arc is latent or just degradation in CCS performance, I don't want to risk that. This resistor doesn't have to be an expensive resistor either. A non-bargain basement metal film works pretty good.
Agreeing with somebody else's post, a resistor in the output is not a reliable way to reduce the voltage stress on the CCS. A resistor across the final CCS output stage is useful for reducing the power dissipation in the CCS. If you drop 20% of B+ on the CCS and 40% in the series resistor, when B+ drop 10% (the air conditioner kicked in, the neighbor's playing with big power tools again etc.) you now only have an output swing of two times 10% of B+ instead of two times 20% of B+. Now if the plate voltage bias point shifts up another 10% because of tube variations, we can't get any clean power out at all because the CCS is in saturation.
The negative comment on CCSs I get most often is "I replaced my output choke with a CCS and can't get any power out and the sound sucks." A CCS doesn't store energy like a choke. So the plate voltage can go down below B+, but not above B+. This is made worse by the CCS voltage drop.
CCS like to have voltage headroom. Both BJTs (rated > hundred volts) and FETs like to see several, if not 10s, of volts from base to collector (or gate to drain.) This is from both Early voltage and capacitive effects.
The plate voltage on CCS fed tubes tends to vary more than usual with CCS loading. Even with the CCS current set dead on. This is because the cathode to ground voltage for the tube is fixed (V = Iset * R_cathode) with CCS loading. It sort of looks like a battery from the cathode to ground!
When we have resistance between the B+ and plate, the tube current is allowed to change as the plate to cathode voltage changes. This is because the change in grid to cathode voltage (from the plate voltage variation) causes the plate current to change via the cathode resistor. This local feedback (from the cathode resistor) reduces the variation in plate bias voltage from tube to tube. With a CCS fed tube, the voltage across the cathode resistor is fixed so we don't get the local feedback.
So if you use a CCS, the B+ tends to be
B+ CCS estimate = Voltage cathode to ground + 2 * Voltage Plate to cathode at the tube's bias point + Voltage drop in the series resistor + 10% more.
I had long since forgotten that post, but seeing it again I seem to recall it. VoltSecond was very active 20 years ago; he still posts occasionally but not often. He's a highly competent electronic engineer and well worth listening to.
The best practice may be to use two resistors in series, and take the output from their union.
Hi Paul,
> > The best practice may be to use two resistors in series, and take the output from their union.
May I know why split in two resistors, any further advantage?
One resistor buffers the tube, and the other buffers the CCS.
I have old commercial 12bh4a linestage and it kind of sucks so I plan to hack it with VR tubes , CCs and put a BAS choke on the plate. That ought to cure it :)
The 12B4 is pretty hard to screw up... :) Don't forget its relative, the 12A4 either...the mu is higher.
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Well, it's all debatable. If you ask $16k for a 12bh4a linestage it needs to be little better than plain Jane 12bh4a, and plain Jane 12bh4a is nothing to write about.
You do have a point. I never worry about such things...I just go and build something I think interesting. If I had to worry about an expensive component, I doubt I would get very far...LOL
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
as per Gary Pimm.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Hi Bas,
If putting a 1K plate stopper between
The output is taken from the junction of CCS and 1K? Or the junction of 1k and tube plate?
It depends on what is causing the glare. If it is transient oscillation (as is common with high transconductance tubes such as the 6DJ8 family) then you are buffering the plate and would connect at the other end.
I have got the reply from NAZ,saying to use big R as much as possible, but I intend to use as driver, swinging 100V p to p needed, so what appropriate value the R should be, if it is plate stopper, normally 100 ohm can do
"Hi Eric,
It depends on the operating points you use but only in as much as I would use as high a value as you can that still allows your chosen operating points. Eg, if you were operating the tube at say 300V B+ and wanted a plate voltage of say 120V at say 15mA you could use a 10K plate resistor. This would drop 150V, leaving a 30V drop across the CCS. Of course, this would only be OK for max AC swings of around 20V on the plate but if this needed to be higher, say for a driver stage then you would use a lower value plate resistor to allow greater head room for the CCS..
As I mentioned, the plate resistor is not technically required at all with a suitable CCS but IME it definitely sounded better with the FETs I have tried. The way I look at it this gives you the best of both worlds in that the CCS provides the high impedance load that tubes want for the lowest possible distortion and the resistor shields against the often harsh traits of FETs
I have no theory, and thus no idea what the value "should" be.
I have used 220 ohms at the plate, as a precaution, with 6922s at 70v/4mA. But I did not try without it, so I don't know whether there was a problem without them. The circuit also used 220 ohm grid stoppers as precautions.
Hi Paul,
Then what would be the value for this buffer R, if I intend to use this stage as driver stage, swinging 100V P to P, I was told on such case it cannot use high value of R limiting the swing
That post by Naz is just more evidence of the fallacy of attaching SS devices to the signal paths of vacuum tube circuitry. Anyone who has seriously compared the old school sound of vintage '60s gear to contemporary amplifiers knows the risk.
He wrote to me some years ago that resistor in place of CCS actually sounded surprisingly good. He mostly used CCs in series with output transformers.
Well, JFET and MOSFET supposedly better than bipolar, could also make fantastic zero NFB amp
I also own Ayre top gear amp, such as K1xe, VX5 twenty, smooth and absolutely transparent, delicacy and subtle signal handling even better than triode amp due to no coupling cap and transformer
Maybe we think of the cooperation by tube and FET could achieve something more, no need to be reluctant to SS because nowadays the music source most come from DAC which is made of SS
I find "old school sound of vintage '60s gear" interesting and nostalgic but I wouldn't want to listen to it now days. Too bloated, fat, slow and muddy.
I like clean and clear and fast and detailed and that's what I get from my tubes.
Way better than anything built in the 60's.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
There's nothing bloated or muddy about a Citation V. And it has none of the "scritch" or annoying sibilance-like distortion I hear during extended listening with most SS gear. The last thing I want to do is introduce those characteristics into high quality tube gear that is otherwise pristine and transparent. There's no benefit to this that can't be accomplished by other means, and it simply isn't worth the risk.
Don't have any of that stuff going on.
Given the way you speak of it, you must have tried using some sort of SS and had bad results.
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Fair enough. I can't stand SS amps for mids and highs (the upper ordered HD makes them sound bad). I have a Cit 5 but honestly I haven't really listened to it (I should sell it). I own one Cit 2 and have owned two others. I sold them. I can't stand the way they sound (all those pentodes with their upper ordered HD and all that feedback trying to fix it). That amp ends up sounding like a SS amp.
I built a different amp (different front end circuit using triodes and put the output tubes in triode mode) with the Cit 2 I kept. My cousin uses it. It doesn't sound bad. The output transformers are outstanding. I have a spare pair of Cit 2 output transformers. I built a all triode PP 300B amp using them. That amp sounded good but I like DHT SE better.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
It may sound rough and can't image for crap and it's an effing B class stage but damn it, it's fun as hell on 96db FR speakers.
system.
Some even do it with a plate choke. So essentially the resistor is a plate stopper.
Experiment with the value. Anything from 47R to 1k or so ballpark.
Hi,
Plater stopper, are you implying the glare sound in fact is oscillation coming from SS CCS?
Since there are all kind of opinions. And it can drive you mad unless you do some emperical stuff...and even then...you can go nuts.That said I think a plate stopper is nothing more than a little filter for hf.
Edits: 04/25/24
Hi Bas,
If putting a 1K plate stopper between
The output is taken from the junction of CCS and 1K? Or the junction of 1k and tube plate?
nt
Or, for a lower output impedance, the output can be taken from between the Rset (current set resistor) and the source of the bottom MosFet.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
,
.
Have Fun and Enjoy the Music
"Still Working the Problem"
.
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: