Quality answer to dangers of Tube amps running under no load

Thanks for that info--interesting and helpful to know.

Back near the end of '17 I had taken my Cary amp out of mothballs and I was on the phone with the tech guy trying to clarify something and I told him I had it hooked up to a load and he said something to the effect of not to worry about doing that, but I always still do.

A tube amplifier not connected to a load should never be powered up.

According to the instructions from Cary on biasing the SLI80 HS, "Some other tube amplifier designs will oscillate or go into overload if they are used without a speaker or resistive load attached to the speaker output terminals. Cary Audio tube amplifiers are inherently STABLE DESIGNS and may be operated without a load on the output terminals for adjusting bias or while in burn in."

I was just going to post the same about my Cary amp(s).

Still would never power up without a load.  

 "Some other tube amplifier designs will oscillate or go into overload if they are used without a speaker or resistive load attached to the speaker output terminals. Cary Audio tube amplifiers are inherently STABLE DESIGNS and may be operated without a load on the output terminals for adjusting bias or while in burn in."

I am still going to remain in the better-safe-than-sorry camp.

I don't really agree with that explanation, it applies more to an ideal transformer. If there is no load on the secondary there is no magnetic field from the secondary current to reflect so the primary becomes one big inductor. An inductor's impedance is frequency dependent, so when the power tube's current hits the higher impedance the anode to anode voltage surges. The higher surge voltage does the damage.

A more immediate threat is if the speaker cable is suddenly disconnected with a current through the secondary. The magnetic field collapses and the stored energy in the primary and secondary has nowhere to go and causes arcing from the induced voltages. Same as a car's ignition coil operation, only the collapsed magnetic field has a relief path across the spark gap. The tube amp OPT has no such safety valve unless you put a few diodes in series with the B+ to ground. But the problem with that is diodes tend to short and that would cause a short through the transformer. 

As a precaution, what I do with spade terminal speaker cables is mount them facing down on the amplifier binding post so if they should become loose the weight of the cable won't  pull it off the post. 

I can tell you this.  Cary Audio burns in all the tube amps on several racks and powers them on for 3-5 days with no load.   They are biased and left to burn in for days without any load on them.  

@cavscout19d I appreciate you sharing the detailed explanation of the ‘how and why’ tube amps need to see a load upon power-up. I found out the hard way when a blew up a beautiful old Crosby radio because I didn’t know the speaker needed to be wired up. I accept @gs5556 correction concerning ‘ideal’ theory vs. actual implementation. I would be very curious to learn how Cary builds their tube amps so that speakers don’t need to be connected. I did run across a video recently in which a tube amplifier incorporated a very special RCA jack that I think was shorted until the RCA plug was inserted. It was in a video in which XrayTonyB explained how ‘multiplex FM’ worked. The video featured an H. H. Scott FM de-multiplexer that de-coded the FM Stereo signal so that a mono FM tuner could feed a stereo amplifier, but I digress. 

I get a lot of amps back that customers powered up forgetting to hook up speakers or a load. Usually due to excitement. Depending on the topology, the damage can be different. In this situation, the cathode resistor goes up in flames. 
I hot swap speakers all the time in push-pull amps which has never been an issue as it’s only a few seconds. 
i recommend not powering up a tube amp without a load. Not worth the risk. If need be, grab a couple of 8 ohm 50 watt resistors to have around. 

Solid state amps can also do this.  My Sunfire Signature sub is now not working because it powered up without a speaker connected.  The amp had been separated from the speaker cabinet.

My "Universal Switcher" unit (for use of comparing any two components/speaker systems) has provisions to switch in 261 ohm, 30 watt resistors for use with tube/whatever gear so that even when the relays are open (milliseconds) during switching operations, there is a load on the amp.

what I do with spade terminal speaker cables is mount them facing down on the amplifier binding post so if they should become loose the weight of the cable won't  pull it off the post. 

@gs5556 , that is a good idea.  It's kind of like the (unwritten?) mechanic's rule that if you are installing a bolt and a nut and the maintenance manual does not actually spell out directionality, to install it with the head of the bolt pointed up so that if the nut was to back off the bolt wouldn't immediately fall out.  

Well.  
 

- I have no output transformer. So does this still apply?

- Four months ago, I powered up my left monobloc. There was a bright flash and a loud pop. It blew the 15” diaphragm in my Altec 604Cs. A tube had apparently arced at powering up. Now, I disconnect my speakers before I turn the amps on, then reconnect them as soon as they start warming up. 

- the tubes were all NOS, so maybe it was just that tube blowing but I’m afraid to leave the speakers connected at startup  

Thoughts? 

If you have the speaker unplugged, the load is now nearly infinite ohms. That gets multiplied by the transformer to an even higher number of ohms (hundreds of thousands or even millions) that the tubes are trying to work against. The result of this is that instead of making hundreds of volts, the tubes now try to make thousands of even tens of thousands of volts. That much voltage can destroy the insulation in your output transformer, and can cause tubes to arc and self destruct.

@cavscout19d Just so you know, this statement is false on account of the fact it violates Kirchhoff's Law. That very high Voltage has to come from somewhere; to do that with an inductor you would charge the inductor with a current and then disconnect the source of the current. As the magnetic field collapses, a very high Voltage can be generated. This is not the situation in a properly functioning tube amp with no load. 

If the tube amp has no load, its perfectly fine (unless unstable, which is a different issue) so long as no signal is applied to the input. Since there is little or no load on the output, there is a very high impedance that the power tubes are driving, which is to say they can't make a lot of power. 

The Dynaco ST70 manual states that its fine in this condition so long as no signal is applied to the amp. Our OTLs are also fine in this state. I don't think I've very seen more than one or two tube amps where this was a problem, with the exception of an amp I worked on that was capable of oscillation with no load, keeping in mind I've worked on hundreds of tube amps over the last 50 years.