Power off question

[catalin gramada]

Hi. Please let me know how much are endanger the power tubes when cut power off please? Have an amp I wanna get rid off stb. switch from pop noise reasons (it not really need it) but have lots of concerns when cut the voltage off. There is a bank of hefty 2000uF power caps installed, the bias voltage will drop very fast. Will be the power tubes in danger or excessive stress please ? I consider the power tubes will dissipate the max power in respect with a rate of bias voltage drop off but still not very clear for me...
I think to use a power switch contact to disconect the shunt bias resistor to ground This will avoid to create a path to discharge the bias capacitors and will raise the bias voltage to max negative value it can provide. No good ? Thx.

[Jon Snell]

2000uF on the HT seem too much to me!
Schematic would help but if you switch the screen grids, no thump because the bias voltage stays stable.

[catalin gramada]

Thank Jon. Is a 4 x kt120 300w amp. There is a heavy thump- the driver is supplied from the same node as the screens so big transient at startup . Anyhow the ideea of raising shunt bias resistor from the ground is brilliant. The max bias voltage last "forever" in bias caps leaving the power caps to discharge as slow without endanger the power tubes. It is a power switch circuit used in old Dynacords. Simply smart.

[vintagekiki]

See how this problem is solved by Dynacord (switch S3^I and S1^III)

[catalin gramada]

Thanks. Yes, that I talking about.

[Enzo]

SOme wiring might remove the condition but not the "problem." This isn't a problem. You might find a small fraction of a second where bias and B+ are not exactly in sync, but it is over in a tiny period of time, nowhere near long enough to harm anything. As soon as power is removed, not only do supplies start to discharge, but the heaters go off and tube ability to conduct current drops off extremely fast. Consider, of all the tens of thousands of amplifier products out there over the decades, how many had some special circuit to hold up bias supply at power off? And of all those amps, how many were infamous for killing power tubes?

[vintagekiki]

If you are an opponent of standby sw, the solution is on / off delay timer.

http://rubli.net/HT_delay/htr214xp.html
A Start-Up Delay For Vintage Amplifiers

https://www.diyaudio.com/forums/tubes-valves/89799-time-delay-relay.html
B+ time delay relay

https://sound-au.com/project39.htm
Soft-Start Circuit For Power Amps

https://sound-au.com/articles/soft-start.htm
Soft Start Circuits For High Inrush Loads

1)
https://www.audiophonics.fr/en/softstart-protection-c-418.html
Softstart & Protection

https://linearaudio.nl/sites/linearaudio.net/files/AX1402JanDidden.pdf
Tube Amplifier High-Voltage Delay

2)
https://diyaudiostore.com/products/soft-start-speaker-turn-on-delay-and-dc-protector-boards?variant=5940139268
https://cdn.shopify.com/s/files/1/1006/5046/files/SSSchematic.jpg
Soft Start & Speaker Turn-On Delay

[The Dude]

I'm with Enzo on this. It's simply not an issue. I've seen plenty of vintage Champs and Princetons that still have good factory original tubes in them.

[Helmholtz]

Why not just increase the bias filter cap to "match" the huge 2000µ B+ filters regarding time constant?

[The Dude]

...... Yes, or replace the 2,000uF overkill caps to a more reasonable spec for a tube amp.

[vintagekiki]

See how this problem is solved by Carver

https://ampslab.com/SCHEMATICS/CarverSilver7A.gif

[Chuck H]

...... Yes, or replace the 2,000uF overkill caps to a more reasonable spec for a tube amp.

I like this option better than bumping the value of the bias cap. I haven't done bench tests but in cad sims I have noticed that excessively large bias caps can take a surprising amount of time to stabilize.

[vintagekiki]

#10

Replace caps is not a solution because it eliminates the consequence, and the cause still exists.

[The Dude]

VK, Could you please use regular quotations, like everyone else so that a person doesn't have to scroll back and find your reference?

I'll again reiterate, this isn't really a problem, so there's nothing to be solved, IMO.

[g1]

Increasing bias filter capacitance will create a new problem of slow ramp-up of bias voltage, won't it? So you will be missing bias at power up rather than power down.
And does it not also create problems with LF response?