Bias Supply protection

[M Gregg]

OK so here is a question,

The bias supply...if it fails the amp could be toast

So how do we protect the Bias supply and still have a fail safe amp..

One idea is to use a fail safe bias with the dropper resistor in the supply side then should the pots fail the tubes will shut down.
Another idea is to put these droppers as close to the supply as possible so any possible short is protected by the dropper resistor.

If we fit a fuse and it blows......

So the question is ideas for best protection?

Regards
M. Gregg

 

[Merlinb]

Fuse in series with the power valve cathodes.

 

[M Gregg]

Fuse in series with the power valve cathodes.

Yes,

However what happens if the rectifier on the Bias supply goes short (burns out the Tx bias winding?)

So is a fuse a bad idea in the bias supply...?

Regards
M. Gregg

 

[dgta]

Merlinb gave you the answer.
A fuse in the cathode of the power tube (each tube) protects against excessive current, no matter the source or reason. So any short, open or any other fault in the bias supply is irrelevant.

Yes, a fuse in the bias supply is a bad idea.

 

[jackinnj]

If you were using "Red Light" a la SYwouldn't the led string go open if the max current is exceeded for a few seconds?

 

[dgta]

Depends on the LEDs. There are LEDs that can take 100mA indefinitely and his power tubes might not. But that arrangement does not have a grid bias supply and the OP's question was about bias supply failure.

 

[M Gregg]

Merlinb gave you the answer.
A fuse in the cathode of the power tube (each tube) protects against excessive current, no matter the source or reason. So any short, open or any other fault in the bias supply is irrelevant.

Yes, a fuse in the bias supply is a bad idea.

Well,

Just the power tubes is not the only answer, how can you protect the bias supply winding from short...if this burns out then the whole power Tx is scrap unless the bias is from separate small Tx. ( Over rated diodes is one way)..

Yes the power tubes would be protected by a fuse..so great we have saved the tubes but a short in the rectifier and the power Tx is scrap?

(Yes I know this is very old Gnd..just wondered if anyone had any other ideas to protect the bias winding)..The problem is that there is often not enough current in the bias supply to make it rugged against short or blow an over rated fuse..So it looks like perhaps over rated diodes, good insulation, followed by dropper resistors is still the only way..

A fun link (I have no connection)

http://www.diy-audio-guide.com/fixed-bias-protection.html

Its interesting how many people just give the bias supply a passing thought..its something you have to do...however if it fails...the fix it bill can be huge...

Regards
M. Gregg

 

[Vinylsavor]

Hi!

A very simple method would be a relay switched B+ with the relay being triggerd by the bias voltage. Either have the B+ on a spearate transformer which is switched on through the relay. Or have the relais in the ground connection to the center tap.

If the bias voltage fails, the relay opens and B+ is drained through bleeder resistors and power tubes.

If the bias voltage doesn't come up at turn on, the relay stays open

Best regards

Thomas

 

[DF96]

Not foolproof, but I have a front-panel LED which glows green when all is well, but red if HT is present without negative bias. I also have bias LEDs on the chassis when are out for correct bias, green for too little current, red for too much current.

 

[trobbins]

I often use the relay technique. For old amp restorations that originally used a half-wave rectifier, then going to full wave provides the additional oompf to power the relay coil (24V to 48V coil gives reasonably low current draw), and gives lower ripple to boot. I've also put a series resistor with the bias winding to smooth out the ripple even more, and that could effectively provide reasonable protection for the bias winding. The relay should really only switch HT AC

 

[M Gregg]

Not foolproof, but I have a front-panel LED which glows green when all is well, but red if HT is present without negative bias. I also have bias LEDs on the chassis when are out for correct bias, green for too little current, red for too much current.

Hi DF96,

It would be interesting to see the circuit you have used for your bias set/protect..any chance of a quick sketch?

Regards
M. Gregg

 

[M Gregg]

I've also put a series resistor with the bias winding to smooth out the ripple even more, and that could effectively provide reasonable protection for the bias winding.

So you have a resistor in series with the bias AC? (value)

Regards
M. Gregg

 

[DF96]

For the bias setting I used a bi-colour LED (red one way, green the other) between the output cathode and a potential divider across the HT. I was using mixed bias, so (from memory) something like -16V on the grid and +18V on the cathode. The divider presented +18V so the LED has zero volts across it when the output bias is correct.

The front-panel LED was a bit more complicated, but essentially the green diode was driven from the bias supply and the red diode from HT. I adjusted the currents so when all is well I get a greenish-yellow. I haven't got time right now to find and scan in the circuit. Will do it later or on Monday.

 

[trobbins]

Any chance of using a series resistor with the bias winding will depend on how much 'head room' is available for the bias voltage without a resistor. If the existing bias was full-wave and only just had enough voltage to manage a moderate tolerance of valves then there would be no oportunity to use a series resistor. If a relay coil is an added load then that will limit the amount of any series resistance, especially if say only a 24V coil was used - and the design then needs to usually involve a dropper resistor for the relay coil, and just allowing the relay to stay on at say -25% line voltage.

In a recent 6L6GC quad restoration that originally had half-wave and 55V winding, I went to full bridge and a 1k series (but no relay loading).

 

[kevinkr]

In the case of a half wave two diodes in series can provide protection against a shorted diode. Specifying diodes with high PIV ratings will prevent failure due to transient events. Using discrete diodes as opposed to bridges may help. (The only rectifiers I've had fail here have been bridges for some reason - always conservatively rated and usually heat sinked.)

A resistor from wiper to the most negative filtered voltage can protect against pot failures.

A fuse or fusible resistor in the cathode circuit may provide additional protection. Conventional fuses should only be used on low plate voltages, high voltage cartridge fuses are required at normal plate voltages. When the link vaporizes an arc may be initiated, certain sand filled types in ceramic bodies may work ok above 400V, but why risk it - check first.

 

[jackinnj]

How about a depletion mosfet pair in series with the B+ feed to the output transformer center tap. On resistor sets the maximum current which will flow through.