• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

RS1003 in SE triode-UL monoblocks

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
You need drive tube with internal resistance less than 1,8kOhms.(18/10 kOhms)

you mean 1.8k is Zout or Ri of tube driver stage ?

my 6AQ5 has 1.9k Ri at triode connection, but its gain is a bit low 9.5. So I need one input stage.

the tube with high gain and high transconductance and low resistance so far.... is expensive :cool:. only 6C19, 6C45P is still quite cheap, but I don't like their aggressive sound.

Do you have a equation of calculating Ri at triode connection for penthode tube ? (when we have no information of Ri in triode connection from datasheet)
 
Last edited:
All of that trodes are very good (linear, very low internal resistans), but gain (real) is low, and You need 2 stages!
E810F/7788 (no E801F:D) is excellent tube, and gain is ~28x, EF184/6J7 with gain 28x, E180F/6688 wth gain 29x, 12HG7 with gain 24x, 11HM7 with gain 24x are all perfect!
Also russian 6S45P-E triode is very good with gain 25x (Ua=400V, Rk=100 Ohms NO-baypass and Rload=10kOhms), and very low distorsion and linear caracteristic.
 
6S45P-E have not "aggressive" sound with NO-BAYPASS katode resistor!

OK, I will try this method and compare E810F vs 6S45P with no bypass.

what difference of bandwidth in case of bypass capacitor or without bypass ?

I feel having loss some low frequencies, thin sound, a bit less dynamic without bypass cap (specially high value Rk>1-2k) but I don"t know why. so using this cap is routinely for me :D
 
Last edited:
for your experience, when do you take place bypass kathode ?
6S45P-E was originaly construckted for VHF-UHF high gain amplifier
With baypass capacitor(s) increase gain at very high freqence (>50MHz) and tube easy go to short "ringing" (self-oscilacion) at peak level of input audio signal. If You have osciloscope or spectral analizator with 200 or more MHz bandwith, You can easy see "overshoot" at begening of every start test sqere signal (only at high level part of signal). This "overshoot" moved tube working point, and increase gain in very small part of interval, or put tube in self-oscilating when input audio signal is in peak. That result with short expanding of audio signal, or You hear it "aggressive".
Without baypass capacitor, gain at very high freqences (VHF-UHF region) decrase, and tube work all time in linear region. You can stoped oscilation ("ringing","overshoot") with neutralization, very shot wires of components, more blocking capacitors (at heater wires especialy!), soldering parts directly at socket, using "feritte-roads-pearls", etc.
All tubes with high transconductance have more/less simmilar problems like 6S45P-e (E810F also!).
 
All tubes with high transconductance have more/less simmilar problems like 6S45P-e (E810F also!).

I've got this problem when I've been exchanged Plate choke load (high gain) to Rload (low gain ) with E810F. Sometimes, I can hear the treble that has overcut or distortion.

but, we can get low impedance with C bypass. It is definitely a advantage for the driver stage. Without C bypass, internal resistance will be increased significantly. remember that all of Ri in datasheet of any tube is from measuring tube in case of doing C bypass.
 
I must use SHADE-feedback to reduce gain and distorsions.

Hello Davorin,
F3a/RS1003 is a great tube and your amp looks nice. But with your Schade implementation you will increase distortion radically. A triode with bypassed cathode isn´t the best solution. If you want it to work, you should go for unbypassed cathode and pentode-connected EL83 driver. This will probably reduce distortion two to four times wrt triode-strapped driver. If you insist on triode-strapping and UL you must remove the cathode cap to lessen distortion somewhat but it will still be very high.
 
Last edited:
Remove cathode decoupling in the output stage ? Say hello to current feedback, sluggish sound, (1+µ)*Rk times lesser damping and wimpy bass response.
Additionally, the fact that this so called Schade-feedback isn't really great in distortion reduction has been discussed here a few times already. But, to each his own
 
Remove cathode decoupling in the output stage ?

NO!

Whoever talked about the output stage? Read again! This is all about the driver.

About Schade-feedback could you please give an explanation of why one shouldn´t get reduced distortion with local feedback around the output stage? A good start is to study the last pages of Schades paper from the late thirties. The Ua /Ia will give a hint. With some basic knowledge one should realize there are certain demands on the the driver as the output stage is almost working as an(in simple terms) I/V-converter.
 
Last edited:
I misread that, but it doesn't get better with that configuration either. A pentode with unbypassed Rk has an immense high output resistance which isn't optimal in driving a big triode with its miller capacity.
This gets even worse with that "Schade Feedback" which is simply a parallel derived and parallel applied (voltage) feedback which has the big problem to further reduce the input impedance of the output Tube. You can imagine that as an "inverse bootstrap". The "Schade-Feedback" works against the anode voltage at the driver tube and effectively reduces the load resistance of the driver (which lessens its amplification). Everybody knows that distortion rises when you decrease load resistance because the nonlinear inner resistance of the tube contributes more. On the other hand, this adds lots of K2 to the spectrum which some people like very much. That's it.
 
At RS1003 plate is ~230Veff (Pul=22W, Rl=2400 Ohms, UL-mode). Shade feedback resistor (150kOhms) and parallel EL83 plate resistor (9kOhms) and RS1003 grid resistor (150kOhms) are voltage didvider for feedback and voltage ~12,3Veff is in oposite direction of voltage at EL83 plate. With ~1Veff at input, and 25x gain of EL83 is ~25Veff at plate, but with negative feedback voltage at RS1003 grid is ~12,7Veff, enough for full drive. This is only -3dB feedback, but very good to reduce total gain of amplifier.
I have very old classic distorsion meter (but very good and precision), and can't shaw distorsion caracteristic at computer, but negative feedback reduce distorsions to ~1/2 value without feedback.
At triode-mode negative feedback is very small, only -1,5dB.
 
A pentode with unbypassed Rk has an immense high output resistance which isn't optimal in driving a big triode with its miller capacity.

Think you misread again;). I fully agree concerning triodestrapped output stages. But we are talking Schade not triode output stage.

The pentode(or MOSFET ref. MJK) with unbypassed Rk is the best you can ask for driving a Schade as that one wants a current source driver. Being it a nonideal current source can make it even better as the nonlinearities of both the Schadeod pentode and the pentode driver can partly eliminate each other through distortion cancellation with the right combination of tubes. This can make for a really good result in the end. But as your discusssion seem to be about triode drivers you are absolutely right, it will not work well at all.

The advantage of Schade is that you get triodelike characteristics but pentodelike effiency, with a slight disadvantage that some pentodes/tetrodes can produce somewhat higher uneven harmopnics closer to clipping. UL will not even be necessary as Zout will be on par with the triodestrapped version. Combining the UL and Schade can theoretically better THD and Zout performance further. Maybe this will also lower the uneven H´s.

Davorin,
Forget about your trioded EL83 giving acceptable results into a Schade output stage. You can guesstimate the Schade Zin to 2-3kohm. Parallell with Ra this gives your trioded EL83 a steep, far from ideal loadline. No problem to check that in the Ua/Ia-curves.

Also see the_manta´s explanation in his post above.

It will be an easy task for you to pentode-couple and remove the cathode caps if going the Schade route.

If you want to get an acceptable result with Schade you must go pentode, FET or cascode driver. The result will most certain be even better with a gyrator instead of the lowish Ra.
 
My katode baypass capacitors are very good, but yes, I'll try in future with fixed bias.

your winding 40V 50mA in the power transformer is the best choice for negative bias voltage.

At triode-mode negative feedback is very small, only -1,5dB.
why don"t you use global NF ? I think it is simplifier than Shade, maybe better... :)

for filter high freq a bit, Zobel circuit (RC grounded in the line of feedback or RC paralleled with primary winding of OPT) will be the choice.
 
Last edited:
Very interesting, thank you !
Buuuuuut, the circuit, Fig.39 doesn't really relate too much on what was said here.
Here it is again:
An externally hosted image should be here but it was not working when we last tested it.


The purple part is the feedback. It is parallel derived and series applied, increasing the input resistance and lowering output resistance.

With an IT it is actually possible to use a triode driver.

The only purpose for the interstage is the feedback (series applied). Due to common ground issues a single tube stage with that kind of feedback isn't possible without an IT.

There is a significant difference to the feedback of the RS1003 amplifier on the first page of this thread (parallel applied) which has all the cons pointed in my thread above.

Or does the "Schade" denomination simply mean that the feedback is taken directly from the next anode and not becoming phase shifts ? (As he notes in the upper part on page 178)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.