If it were up to me I would ditch the 0D3 in favor of one of Kevin's shunt regulators. With a CCS fed shunt reg. you would not need the second LC. This is the arrangement that I use in my personal line stage which like yours is a transformer coupled LTP with a CCS "tail".
The series resistor is needed for the regulator tube to regulate. The 0D3 changes the current going through it according to the amount of voltage that needs to be dropped to come out with 150V. When voltage needs to be dropped it increases current draw which means more volatge is dropped across the resistor. I hope that's clear. You can also use a CCS instead of the resistor which improves performance.
Maybe using the 13H choke as L1 might improve filtering and passive regulation a bit.
Hi Guys, nice to meet you.
To John:
Well, I am on a learning trip. Actually I do have Kevin's Ccs/shunt reg here and wired it this afternoon and will compare this to the OD3. I have as well the simple ccs and the current sink from Kevin here and will compare theor sonics. and finally a version just with a resistor.
The chokes are very high quality amorphous chokes custom build from 50Ae, and they outperformed anything I had before...unbelievable transparency.
So, I will as well play with the feeding circuit. Currently as well everything is tube rectified with 5c3s and dual mono, which made a huge difference in the passive version.
So, I will play with all of these configurations and report back if you like.
To Kaputt:
Have you looked into the schematic of the karna ? The resistor is not in series with HV like in a normal OD3 setup. It is going from HV+ to OD3 to ground. And as you say, the CCS in front of the whole setup determines the voltage and current of the OD3 already. So: WHY is this resistor where it is ?
BTW: With a 0,1uF cap dead quiet...now I need to play with cap quality and size at this position...any experiences with that ? What is the smallest working value ?
To John:
Well, I am on a learning trip. Actually I do have Kevin's Ccs/shunt reg here and wired it this afternoon and will compare this to the OD3. I have as well the simple ccs and the current sink from Kevin here and will compare theor sonics. and finally a version just with a resistor.
The chokes are very high quality amorphous chokes custom build from 50Ae, and they outperformed anything I had before...unbelievable transparency.
So, I will as well play with the feeding circuit. Currently as well everything is tube rectified with 5c3s and dual mono, which made a huge difference in the passive version.
So, I will play with all of these configurations and report back if you like.
To Kaputt:
Have you looked into the schematic of the karna ? The resistor is not in series with HV like in a normal OD3 setup. It is going from HV+ to OD3 to ground. And as you say, the CCS in front of the whole setup determines the voltage and current of the OD3 already. So: WHY is this resistor where it is ?
BTW: With a 0,1uF cap dead quiet...now I need to play with cap quality and size at this position...any experiences with that ? What is the smallest working value ?
The resistor is simply used to measure shunt current. Value may have been selected in the interest of looking for 5V as a ballpark number. Sure, he could have used 100 ohm (as I do), but then you are checking for 1.6V.
I always use a 0.068u cap, and don't bother checking if it helps or not - best to have it there. 0.1u might be cutting it a little close, but as long as you check for oscillation with a scope.
It's also a good application for a polypropylene cap, or other high frequency capable cap.
I always use a 0.068u cap, and don't bother checking if it helps or not - best to have it there. 0.1u might be cutting it a little close, but as long as you check for oscillation with a scope.
It's also a good application for a polypropylene cap, or other high frequency capable cap.
This is a different series resistor. 330R is far too small to act as part of the regulation. It is either to allow current measurement, or to counteract the negative slope resistance of the regulator so that a ahunt capacitor can be added without oscillation taking place.kaputt said:
DF96, oscillation is an interesting aspect....so with the 330ohm resistor in place: Would you position the cap from HVreg to groud or from Anode to Cathode of theVR tube ?
My initial thought was: Maybe as well a protection: What happens if the Triodes for whichthe HV gets regulated fails or is not warmed up yet ? The current source would still try to feed the circuit with 60mA, no ? As the main consumer of current is not there, the 60mA would try to go over the shunt element, so the VR tube, no ? This would damage the VR tube as it can stand only 40mA in the case of the OD3...But calculating 60mA*330ohm= 20V (approx.) voltage loss vs. 5V when the main consumer is there. Is this voltage loss not doing something: Reducing the switch-on voltage enough to not switch the VR tube on when the main consumer is not yet ready ?
At least in my circuit with a raw voltage of 186V unregulated this might be the case..., no ?
My initial thought was: Maybe as well a protection: What happens if the Triodes for whichthe HV gets regulated fails or is not warmed up yet ? The current source would still try to feed the circuit with 60mA, no ? As the main consumer of current is not there, the 60mA would try to go over the shunt element, so the VR tube, no ? This would damage the VR tube as it can stand only 40mA in the case of the OD3...But calculating 60mA*330ohm= 20V (approx.) voltage loss vs. 5V when the main consumer is there. Is this voltage loss not doing something: Reducing the switch-on voltage enough to not switch the VR tube on when the main consumer is not yet ready ?
At least in my circuit with a raw voltage of 186V unregulated this might be the case..., no ?
And to think I started this 9 page thread trying to learn how to use the regulator. If one goes back to the beginning few pages of the thread everything is spelled out. By the way my regulated 6SN7 line stage is still working strong with the dual regulators
yesterday i used russian gas tube stab. as small bias supply, fed from 100k 400vdc; and with 470n paralell -- no oscilation
this is not like neon bulb, which would blink like crazy (people should put strawman to rest)
this is not like neon bulb, which would blink like crazy (people should put strawman to rest)
If the resistor is there to prevent oscillation, then the cap should be from the top of the resistor to ground. The purpose of the resistor is to isolate the reg valve from the cap.Blitz said:
If the VR is not switched on then it draws no current at all. If it is switched on then it could draw too much current, so you have to ensure that this cannot happen. A 40mA VR should not be fed from a 60mA current source, unless there are careful interlock arrangements.
i have seen here a prejudice very often, that gas regs are noisy, and is impossible to add too high capacitor in paralell (because of oscilation risk).
if it does fit application, no reason to avoid it like plague.
big cylinder cathode area, is responsible for this regulation stability.
quite different from wire like electrodes on neon bulbs
if it does fit application, no reason to avoid it like plague.
big cylinder cathode area, is responsible for this regulation stability.
quite different from wire like electrodes on neon bulbs
It usually is impossible to add too big a capacitor in parallel because of the risk of oscillation. However, sometimes you will be lucky. The snag is, if you use a bigger cap than the datasheet says then if you need to swap the VR tube then you cannot be certain that the replacement will be stable. A smaller cap is fine because the discharge inside the VR is slow to react to changes - and I guess this could be current-dependent.
I'm sorry it lits a little bit, the total voltage before the VR tube stage B+ 364VDC, the target for the load is 150VDC and power consumption 10mA plus 10mA for the VR so total 20mA. This mean 10K7 resistor, I used 12K that I have on my stash that's giving near 18mA is enough and this is the thing the VR isn't lit strong?
I changed the resistor to 8K2 that's giving 26mA plenty of for my load and now lits more and sounds better.
I changed the resistor to 8K2 that's giving 26mA plenty of for my load and now lits more and sounds better.
Some will light brighter than others. One of my 0D3 is only visible in a darkly lit room unless you run it at higher (30-35 mA) current.
The RMS noise voltage is theoretically sqrt(2 q V^2 delta_f/I) over a bandwidth delta_f, where q is the elementary charge (1.6022E-19 C), V is the DC voltage across the tube and I the DC current through it. Below a few hundred hertz the noise density increases above the theoretical value. See J. F. Dix and K. B. Reed, "Cold cathode discharge tubes - Impedance and noise properties", Electronic Technology, vol. 39, no. 1, January 1962, pages 31...37.
