Tube amp with no load

[Kico]

I too have heard both sides of this debate over the years and it has never been clear to me which might be correct.:scratch2:

Now, just for curiosity's sake I dug out the assembly/operations manual for my HK Citation V (still have the amp :thmbsp kit to see if there was a reference to this. After the build is complete, it instructs you to insert all the tubes in the amp and connect it to a 117V source, set the bias pots to midrange and let it warm up for 10-15 minutes. No where at this point does it say to make sure you have a load/speakers connected, nor is there any mention of shorting the inputs for stray signals.

However, it also has AC balance pots and the manual clearly states in order to balance the AC to connect the enclosed 16 Ohm 20W resistor across the speaker ground and the 16 Ohm speaker terminal. The amp uses SS rectification and maybe that makes a difference?

 

[Stereo_type]

spkr:

...I can't help to wonder why arch overs and the like does not occur with unloaded iso-taps and other devices that have unloaded windings....even my monster AVS-2000 when plugged in and if not loaded....no cautions from manufacture about loading.....what about unused taps (16 ohm tap) no arch over problems?

This is why I previously mentioned how an understanding of the theory holds the answers and the reasons why.

Those are apples and oranges comparisons of sorts above, which is why there's a difference. You might wish to refer to my previous post which touches on the theory, which provides details. If it isn't clear after that, then this should define the bottom line: It's sort of like an issue of extremes; those are due to the extremely high impedance and high windings count on the primary of the OPT in relation to the relatively few windings found on the secondary, along with other factors.

After that first part of the quote above, a tap is just that, a tap from a larger, continuous winding; unused taps are not really 'unused' since they're all internally connected, so _any_ tap or combo of same is enough to prevent the hazard.

Kico:

...The amp uses SS rectification and maybe that makes a difference?

Actually, no, the PS is not the area that is the cause for concern with this issue; the OPTs and output tubes are the section that can suffer minor to catastrophic damage. The short answer as to why, is that the PS has its own, very different circuit configuration that basically tends to swamp this type of potential issue.

 

[Tom Bavis]

Voltage across an inductor is L*dI/dT. The faster the current changes, the greater the voltage. Drive a pentode into clipping - the rate of change of current can be very high... so the voltage is high as well. With a sine wave, no problem. With a LOAD on the transformer, no problem - the energy stored in the inductor goes into the load, not into an arc. With AC transformers, this can cause arcing at the power switch on turn-off or with switch bounce (and it's worst when a transformer isn't loaded...).

If you've ever gotten a shock when measuring a transformer or choke with an ohmmeter, you have experienced this effect. Only 1.5 volts from the ohmmeter, but you get a spark when you disconnect it...

 

[dcgillespie]

Clear, concise, and dead on as usual, Tom!

Dave

 

[buglegirl]

Hiya,

Yep Tom sums it up better than my

Don't do it will explode.

Excellent explanation. Wish we could sticky some of these reoccurring questions. Each one takes on its own life and it might be confusing to future generations wondering what the heck were we thinking about in the year 2012.

Frannie

 

[Kegger]

You can Believe what you want.

But there is solid technical reasons why a load should be present, yes in some designs
you can get away with it and no issues, but other no, generally if your not running an
signal through the amp no harm will tend to come(I still don't recommend it) but run a
signal through it and the output tubes start conducting more the output transformers
will have to eat that if no load is present, I've actually seen messed up outputs which
were harmed from this very scenario. It is real and can happen..

Just like you can get away with old electrolytic's a lot of times, you can also have the
darn things explode on you an take other things with them, point being yah you might
get away without issue but there are reasons not to do so or harm CAN come.

 

[dcgillespie]

This discussion has a theoretical side to it, and a real world reality side to it. On the theoretical side, the science of when a load is required or not to prevent damage is predictable and well established. But common sense dictates protecting against real world consequences, because life is anything but predictable. So there's your power amplifier running with inputs connected but no signal present, and no load connected. The amplifier is happy as a clam -- except you accidentally trip over the input interconnect and break its ground connection. There (very possibly) went your output transformer, which all the theory in the world won't bring back.

Common sense says always have a load connected to guard against the unpredictable, while theory is best left to explain why it is more important to do so at certain times than others.

Dave

 

[roline]

I've been bitten by a power transformer measuring the resistance with a volt meter and I've witnessed a flaming electrolytic capacitor shoot across the R&D lab at National Semiconductor after setting Nello's beard on fire! ( You should have seen the look on his face!!)
Believe what you want, sometimes reality is stranger than fiction!

 

[speakerfritz]

After that first part of the quote above, a tap is just that, a tap from a larger, continuous winding; unused taps are not really 'unused' since they're all internally connected, so _any_ tap or combo of same is enough to prevent the hazard.

nope....unused taps have unused windings with it's own potential. the 4 ohm tap is the center tap. load the 4 ohm tap and you still have 1/4 of the winding unloaded to the 8 ohm tap, and 1/2 the winding unloaded to the 16ohm tap. Loading a 4ohm tap does not diminish the potential of either the 8 or 16 ohm tap. the turns ratio of the windings between the taps determines the potential.

 

[Stereo_type]

spkrfrtz:

....unused taps have unused windings with it's own potential. the 4 ohm tap is the center tap. load the 4 ohm tap and you still have 1/4 of the winding unloaded to the 8 ohm tap, and 1/2 the winding unloaded to the 16ohm tap. Loading a 4ohm tap does not diminish the potential of either the 8 or 16 ohm tap. the turns ratio of the windings between the taps determines the potential.

Not if any, anywhere are loaded. Having their 'own potential' on taps doesn't matter, as long as at least one is loaded, so that is not relevant for the issues of damage. Also, there are plenty of OPT variations, so calling the 4 ohm tap the "center tap" might sometimes be true, yet is hardly correct for all OPTs in any event. I have plenty of OPTs that have a single secondary winding; no taps at all, anywhere. Others might only have one tap, etc.

positron:

Loading any winding protects all the windings (It would be one very poor transformer which had very poor coupling that would fail.) as it lowers the Q of the resonant circuit that is formed, lowering the spike voltage.

Obviously correct and completely sensible from all aspects, both theoretical and of course, "winding" up in reality, too. That tends to mirror what I previously wrote as well: ...a tap is just that, a tap from a larger, continuous winding; unused taps are not really 'unused' since they're all internally connected, so _any_ tap or combo of same is enough to prevent the hazard.

 

[Kegger]

nope....unused taps have unused windings with it's own potential. the 4 ohm tap is the center tap. load the 4 ohm tap and you still have 1/4 of the winding unloaded to the 8 ohm tap, and 1/2 the winding unloaded to the 16ohm tap. Loading a 4ohm tap does not diminish the potential of either the 8 or 16 ohm tap. the turns ratio of the windings between the taps determines the potential.

Also..

Your missing the point, it's not that the secondary winding is unloaded that can cause
issues for the secondary, but unloading the secondary can cause issues for the Primary.

So loading Any secondary properly, protects the Primary from that potential damage then.

 

[jaymanaa]

What about running a shorted output? Just curious.:scratch2:

 

[Shadowdog]

Will smoke OPT in short order, as if you hook up the dummy load R's they have to be really big 25W (50W better but $) + and not used for long periods of a time as they might burn out. I read hanging them in a 1 gallon paint can full of oil is a low cost way to cool them.

 

[buglegirl]

Hiya,

Stupid question but I gotta ask. Could someone not use the same 100 watt or so light bulb they use on a dim bulb tester as a speaker load. (One per output)

Trying to cut down on the amount of crap I have on my small work area has me looking at cross usage devices.

Frannie

 

[Blue Star]

Hi guys,

Apologies for the late comeback. For what its worth, I have been using my amp with only headphones for last few days without any issues. Its sitting at my office desk without any speakers. It does get warm but no issues so far.

I also have been reading this forum and have not tried to turn on the amp without the phones conected.....

 

[Khodabear]

Hiya,

Stupid question but I gotta ask. Could someone not use the same 100 watt or so light bulb they use on a dim bulb tester as a speaker load. (One per output)

Trying to cut down on the amount of crap I have on my small work area has me looking at cross usage devices.

Frannie

I'm looking forward to the answer to this question :scratch2:

 

[Shadowdog]

Light bulbs of around those sizes bring up the memory of Nelson Pass's simple SE SS amps in the last few years using such 150-200W bulbs on diyaudio for sonic bliss with some Si devices.

Anyway good question & 100W gives about 10.6 ohms so a industrial rated 150W clear might be ideal so you can see if the filament stays intact. Resistance might go down a bit as filament heats up so maybe a 200W would be better. Can't see why it wouldn't work if careful to keep filaments intact by mounting bulb base to soft foam or similar.

Randy

 

[dcgillespie]

The problem with using a light bulb is the temperature coefficient of the filament. Yes when cold, a 100 watt house hold bulb measures just over 10 ohms. But if the filament would stay at that resistance, then the bulb would be dissipating over 1350 watts of energy -- but of course it doesn't. Ohms law dictates that the resistance of a 100 watt bulb designed to and in fact operating from 120 volts has an effective resistance of 144 ohms -- much to high to act as a proper load for a high power amplifier. It might work acceptably on the 16 ohm tap of a low power (say 10 watt) amplifier, but even then, it should be checked to confirm what the load actually is so that accurate calculations of power can be made.

It is precisely because of the temperature coefficient of the filament that virtually all light bulbs burn out when turned on, rather than during operation.

Dave

 

[buglegirl]

Hiya,

Ok .. makes sense.

So forget about the light bulb idea. So its resistors or speakers or taking a chance.

Gotta be a better way to do this than what we do. You would have figured someone someplace in time would know what fries a amp without speakers and developed some sort of gizmo that would automatically shut off B+ or something.

Its pretty plausible that someone could have that prized Fisher 800C playing some mood music unattended and Fido gets wild and knocks a speaker cable out of the speaker.

Just wonder what exactly goes haywire that can be monitored to create a protection circuit.

Light bulb idea was me being lazy. I will just build a load can with resistors in it.

Frannie

 

[Shadowdog]

Dave,

I knew the resistance was less when cold. Got it backwards , but that's why as you say light bulbs usually burn out when you turn them on. All circuits under freezing conditions should preform better, with tubes they would maybe have to be insulated so they work depending on the temp.