Silly question but if I replace one of the resistors in a CRCRC filter with lets say a 6AU4 damper diode (25 V drop). Will that RC stage still behave effectively like one of the filter stages or does it not? I understand an L can replace any resistor. But what about a damper diode? Assuming I needed to drop 25 volts vs dropping it with a resistor, did I lose a filter doing this?
The diode's forward incremental impedance is roughly 50R, so there would still be some AC filtering,
in addition to its static DC voltage drop.
in addition to its static DC voltage drop.
According to this http://www.r-type.org/exhib/aas0004.htm the conduction cycle is not meant to exceed 15% but used as you suggest it would be 100%.
Cheers
Ian
Cheers
Ian
I think that 15% only applies to the peak current of 1150mA, but not to the idle current of 190mA.
In a traditional CRCRCRC... where stages multitap to the voltage they need is where i'm thinking. Just use a damper to drop the preamp stages voltage in the last filter instead of using a resistor. The higher voltage stages tapped off earlier in the chain. Of course I'd have to work the whole train on the breadboard to see if I' hitting the preamp voltage properly when powered.
Last edited:
Any damper diode that can not take the full B+ DC current 100% of the time, that all the stages of a preamp needs . . .
is either a very wimpy/busted damper diode, or the preamp is using something like 845 tubes for each preamp gain stage.
By the way, the plate current draw of a single triode of a 6SN7 preamp gain stage does not draw the amount of current to cause full voltage drop of a damper diode. At least . . . at that low current through the damper diode will cause its plate resistance to be a little higher than with full damper diode current rating.
It still will be lower plate impedance than if you just use a resistor to drop the same voltage.
Therefore, you will get more ripple filtering using the resistor than using the damper diode.
The only "advantages" of using the damper diode there are:
Delayed B+ to the following stages (not needed in 99% of preamp designs).
The Damper Diode . . . Looks Pretty.
Just my opinions.
is either a very wimpy/busted damper diode, or the preamp is using something like 845 tubes for each preamp gain stage.
By the way, the plate current draw of a single triode of a 6SN7 preamp gain stage does not draw the amount of current to cause full voltage drop of a damper diode. At least . . . at that low current through the damper diode will cause its plate resistance to be a little higher than with full damper diode current rating.
It still will be lower plate impedance than if you just use a resistor to drop the same voltage.
Therefore, you will get more ripple filtering using the resistor than using the damper diode.
The only "advantages" of using the damper diode there are:
Delayed B+ to the following stages (not needed in 99% of preamp designs).
The Damper Diode . . . Looks Pretty.
Just my opinions.
I agree, there is no real benefit in using a damper tube - from a ripple reduction point of view quite the opposite.
Cheers
IAn
Cheers
IAn
6a3, I agree slow B+ is really the only advantage of putting a damper in the train. I was just off on a tangent.
It would make sense to use a damper at the very start of the train after C1 to slow the B+ everywhere, especially if T1 is a toroid with big C1, the damper becomes like a big shock absorber 🙂 Then just use resistors to adjust the taps to the stage voltages after you measure what each stage draws. I learned something.
It would make sense to use a damper at the very start of the train after C1 to slow the B+ everywhere, especially if T1 is a toroid with big C1, the damper becomes like a big shock absorber 🙂 Then just use resistors to adjust the taps to the stage voltages after you measure what each stage draws. I learned something.
I use a PY81 damper diode between the full wave rectifier and the first capacitor as a delay in my Aikido headphones amp so that all the tubes (especially the tubes in the regulated power supply) are ready for action when the voltage comes on. The PY81 takes about 50 (old production) to 30 seconds (newer production) to conduct fully so it works very well as a delay.
Attachments
The most important use of a damper diode is that they get rid of the hash of silicon diodes. Some of the data sheets recommend that a damper diode has warmed up before current is being drawn but if you have them heating up at the same time as the rest of the power tubes then you limit the maximum voltage on your capacitors and the inrush surge that capacitors create when using silicon rectification.
If you use NTC's on the primary of the mains transformer that get switched out after the amplifier has switched on and get back into circuit with a power failure (if not using an UPS) then you can use a fuse that is very close to normal power consumption. This may help saving your output transformers in case a short develops in one of your output tubes.
If you are using a capacitor input filter, the most important thing to reduce hash and harmonics is:
1. Full wave, solid state rectifiers, connect the center tap Directly to the minus lead of the input cap, and connect the + leads of the solid state diodes Directly to the plus lead of the input cap.
You must keep the large maximum current transient current loop Very Short. It has all the hash and harmonics, because it has an extremely fast rise time.
2. Bridge, solid state rectifiers, connect the bridge minus Directly to the minus lead of the input cap, and connect the + lead of the solid state diodes Directly to the plus lead of the input cap.
You must keep the large maximum current transient current loop Very Short. It has all the hash and harmonics, because it has an extremely fast rise time.
Of course, if you choose to use a choke input filter, and use a choke that has very little distributed capacitance across the windings, the current is much lower, and integrated over a longer time than a cap filter. The choke input filter does not have a fast rise time, and has less hash and harmonics.
The amplifiers I designed use solid state rectifiers, choke input (whenever possible), and have less than 100uV hash and hum into 8 Ohms.
Just my opinions and experience.
1. Full wave, solid state rectifiers, connect the center tap Directly to the minus lead of the input cap, and connect the + leads of the solid state diodes Directly to the plus lead of the input cap.
You must keep the large maximum current transient current loop Very Short. It has all the hash and harmonics, because it has an extremely fast rise time.
2. Bridge, solid state rectifiers, connect the bridge minus Directly to the minus lead of the input cap, and connect the + lead of the solid state diodes Directly to the plus lead of the input cap.
You must keep the large maximum current transient current loop Very Short. It has all the hash and harmonics, because it has an extremely fast rise time.
Of course, if you choose to use a choke input filter, and use a choke that has very little distributed capacitance across the windings, the current is much lower, and integrated over a longer time than a cap filter. The choke input filter does not have a fast rise time, and has less hash and harmonics.
The amplifiers I designed use solid state rectifiers, choke input (whenever possible), and have less than 100uV hash and hum into 8 Ohms.
Just my opinions and experience.
I'm a bit surprised about what you wrote about the recommendation in datasheets of some damper diodes to let them warm up before current is being drawn from them. I never saw this recommandation before but ofcourse I didn't study the datasheets of all damper diodes.
Could you please mention a type for which this recommendation applies?
EL506,
Thanks for making my think again about power outages.
One important thing about power outages, is if the power outage is just momentary, and then come back on . . .
it causes a Hot Start.
The output tubes filaments/cathodes are still hot.
The rectifier tubes filaments/cathodes are still hot.
The damper diode filaments/cathodes are still hot.
So, there is no B+ delay here, and no delay of output tube current.
Other than some kind of series transient limiter (and guess what, they already are at the running-state resistance during the hot start),
The best reduction of inrush current is the choke input filter (at least the inrush transient current is lower than for cap input filters).
I use series fast blow and slow blow fuses my the primary circuit.
A typical example is a 1 Amp fast blow in series with a 0.6 Amp slow blow.
This gives protection for both larger than designed inrush current, and protection for larger than designed long term warmed up current.
Thanks for making my think again about power outages.
One important thing about power outages, is if the power outage is just momentary, and then come back on . . .
it causes a Hot Start.
The output tubes filaments/cathodes are still hot.
The rectifier tubes filaments/cathodes are still hot.
The damper diode filaments/cathodes are still hot.
So, there is no B+ delay here, and no delay of output tube current.
Other than some kind of series transient limiter (and guess what, they already are at the running-state resistance during the hot start),
The best reduction of inrush current is the choke input filter (at least the inrush transient current is lower than for cap input filters).
I use series fast blow and slow blow fuses my the primary circuit.
A typical example is a 1 Amp fast blow in series with a 0.6 Amp slow blow.
This gives protection for both larger than designed inrush current, and protection for larger than designed long term warmed up current.
Apologies, I do not remember where I read it. (I've been looking all afternoon.) I remember that I came across it in some obscure location when I started last year to investigate the use of damper diodes for rectification / slow B+.
It was either in Tomer or one of the many tube manuals. Basically it had to deal with that if the damper diode had to be fully heated up before full current was drawn by the rest of the circuit. (otherwise there was cathode stripping).
If the damper diode had a similar heating up time as the rest then it was not a problem. Many damper diodes like the 6CL3 are fast heat up (6CL3 is 11 seconds) which is similar to tubes like the EL84 / 7868 / EL506. The problem is when the damper diode is slow heat up.
Some rectifier tubes are specifically made for slow heat up and can withstand that. But quality NOS rectifier tubes are not cheap....
Thank you for your reaction.
Like I wrote, the PY81 damper diodes of early production (early fifties) had a heat up time of about 50 seconds. Because of that, the screen grid of the tube for horizontal deflection, which was the EL81/PL81 at that time in The Netherlands, had to be designed so that it could take/dissipate the B+ on it while the anode voltage was not there yet (for that the PY81 had to be functioning, so had to be heated up enough).
Like I wrote, the PY81 damper diodes of early production (early fifties) had a heat up time of about 50 seconds. Because of that, the screen grid of the tube for horizontal deflection, which was the EL81/PL81 at that time in The Netherlands, had to be designed so that it could take/dissipate the B+ on it while the anode voltage was not there yet (for that the PY81 had to be functioning, so had to be heated up enough).