NTC THERMISTOR as SOFT START for TUBES

[Leakstereo20]

Originally Posted by Wavebourn

Here is mine: 12.6V rectifier and stabilizer for filament, 12V relay, and 20 Ohm wirewound resistor. Never blows up in peaces. The PCB contains also -80V bias rectifier.

Would you mind make a schematic drawing of the nice simple PCB of yours?

Thanks in advance

[cyclecamper]

Originally Posted by roline

I use CL80's in all applications. Also started to use Damper diodes to soft switch the B+ from the first cap to the rest of the B+ filter chain. It allows for slow start for a slow ramp up of the B+ without the high peak and dip as the outputs come on line. It works great!

Please explain what you mean by these "damper diodes to soft switch the B+ from the first cap to the rest of the B+ filter chain."

In the past I had considered putting a diode on the input side of the inductor on a CLC filter, in the hopes it would prevent oscillation of current thru the inductor between the caps on either side of it back and forth and back and forth... Sorry to ask such a basic question (it's been 45 years in software since my minimal electronics training), but how do I calculate the additional voltage drop across the diode in the forward direction and would it be significant? Can somebody recommend a hefty diode for the job (500 V, 500uF cap, 1.3H inductor, 2500uF cap on a 180-watt tube amp)?

Otherwise I was considering adding two other ways of damping the low-frequency oscillation: a resistor in parallel with the inductor (ruining its sharp q) and a resistor in series with a cap (across the output) to burn up & slow down the voltage changes (but it requires a big cap to damp optimally).

[Wavebourn]

Originally Posted by Leakstereo20

Would you mind make a schematic drawing of the nice simple PCB of yours?

Thanks in advance

Here you go:



The relay is powered from output of this regulator.

[trobbins]

For a fixed bias amp, I power a protection relay coil from the raw filtered bias supply, to provide a mechanism for bias voltage failure. The NO relay contact is used to connect a low resistance (or just bypass link) across a soft-start resistor in series with the secondary HT winding.

Additonal functions can be integrated by using a FET to switch the relay coil, and use a long RC time constant FET gate drive signal from the bias voltage to give say a 20-30 second delay on relay energisation, which can alleviate B+ voltage overshoot (especially if ss diodes are used and output stage bias current is needed to keep B+ lower than filter cap voltage ratings or tube anode/screen ratings).

The above focusses on the long delay needed for output tube heaters to allow bias current to rise at the same time as B+ rises. I'm happy to use a suitably chosen NTC to alleviate transformer in-rush, and then when the protection relay kicks in to also help alleviate B+ capacitor in-rush or diode stress if they are of concern.

[kenpeter]

Originally Posted by simon7000

Which is very close to optimal, as amplifiers that do this seem to last much longer with less service issues.

The relay does not see much current so it does not need to be large but should have a flash suppressor capacitor across the contacts from the turn off surge as the inrush limiter will be cold at that point, so there can actually be more voltage across the contacts at turn off.

When using a relay you can also use a resistor, but high surge current rated resistors are not common and a bit pricey.

The downside of the less expensive inrush limiter is that they do fail and sometimes with a flame, so only put them in metal chassis away from other parts.

And don't put flaming inrush resistor on the return path side of primary.
Don't size resistor so many ohms the real fuse on hot side never blows.
Don't make relay controller so dumb it can't or won't close relay when
there's a short (and fire) that needs the fuse blown.

I'm just sayin' because I saw this recently in a SMPS, and it irritated me.
Some dweeb is surely gonna put out that fire with ... use imagination ...

Damn engineer wouldn't spend a few bucks to replace with an NTC
that wouldn't have caught fire, and could have blown the main fuse.
It really should have been on the hot side, but an NTC would at least
made his junk a little bit safer without redesigning the whole board...

No, he put that relay so it "wouldn't need" an NTC. Yeah, what...?
Me no engineer - me too dumb - thinking not job - shut up and test.

[roline]

I use a CL80 on the 110 side of the transformer for inrush current control for the first cap. It allows for a smaller fuse to be used. Then C-DIODE-C-L-C. The diode bridge feeds the first cap, Now I place a damper diode as a pass device to another cap that feeds the subsequent filter and B+ bias string. I've done variants of this on many amps and also used it on a slow start for my RIAA and Preamp ( 12AL5 in this application).
The damper diode is the tube in the middle behind the 6550's.

[Leakstereo20]

Originally Posted by Wavebourn

Here you go:



The relay is powered from output of this regulator.

Thank you very much Wavebourn.
The circuit looks great. I like the fet regulated heater supply.
Could you say something about the delay-time? The +B 300V from the gate of the fet?

Rgds.

[Leakstereo20]

Wavebourn, you have made several tubeamplifiers for PA according to your site.

Soft-start for longer lifetime of capacitors and tubes is one issue, "overload" is another.

For home hifi it would be nice if the amplifier shuts down (or partially down) by overload ....may be depending on the amount of mA measured across the cathode resistor? ...like a DC sensor for solid state amplifiers.

Do you have any suggestions to a simple tubeamp overload protection, which does not affect the sound? ....in several tube designs with good reputation old tubes "runs" with bias and leaves us a red glowing plate and a tube melt down.
Can we do anything to avoid this?

[AuroraB]

Rick...
seeing your pictures,- have you looked at overvoltage suppressors?
For serious and sensitive installations, it is quite common to install a 'coarse' suppressor setup at the power intake, and another 'fine' suppressor in the relevant consumer line circuit ( going to your ampliifers and stuff) ....

[hpeter]

Originally Posted by Eli Duttman

When the amp. employs "fixed" bias, put the thermistor(s) in the rectifier circuitry. The idea is for the bias supply to turn on and protect against cathode stripping, before B+ is applied.

dont understand why tubes (as being recommended in general), should be heated first and later anode supply turned on
tube heaters when cold emits no electrons, so leaving them connected to circuit under full voltage cannot make them break.
Just piece of metals in vacuum
In my opinion, leaving everything normally connected is better than heating them up and later turning anode supply on. Risk of current spike through tubes

Vacuum rectif. has softstart "built-in", but the voltage sag is not nice.