Maker Pro
Maker Pro

Tube amp acting very strange-- tubes acting as speakers

dietermoreno

Dec 30, 2012
238
Joined
Dec 30, 2012
Messages
238
Tube amp acting very strange-- tubes acting as speakers?

I have a 100 watt tube amp and it did something very strange today.

I noticed that when I disconnected the speaker cable from the speaker output jack and I still had music playing on a line in into the amp, the amp still played music, but it sounded like the music was coming from the tubes!

When I turned up the pre amp gain and the power gain up all the way and my music player gain up all the way, the tubes are playing music that fills the room with sound like a speaker!

Then after 10 minutes of studying this phenomenon I smelt smoke and immediately disconnected everything and turned the amp off. I hope I didn't break anything.

I don't think tubes are supposed to do this. I've got a dynamic mic to play music, but that is less far fetched than tubes playing music.

Perhaps the output transformer is what is doing the music playing like if I connect a dyamic mic to an audio source the dynamic mic will make sound?

That doesn't make any sense that the output transformer would be making sound, because there is no membrane for it to vibrate.

I had my ear against the output transformer and all the sound seemed to be coming from the glowing hot tubes, not the output transformer.

Does this indicate a short of the output transformer shorting to a power tube and the power tube glass with the help of the Edison effect is acting as the speaker membrane?

When I disconnect the audio source, I still hear sound, just I hear static, and when I hold the input cable in the right place it even occassionally picks up morse code even though there is no mixer connected! Shorted power tube acting as a non linear device?
 
Last edited:

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
That was a daft thing to do.

A tube amplifier uses an output transformer which stores energy. If this is not delivered to the speaker, then it has to go elsewhere.
The high voltage may arc over either outside the transformer or inside with disastrous results. An arc can be made to sing.

The high voltages can take the transformer into saturation causing the core to expand and contract in time with the music. With luck it may still work but it has obviously been damaged. I hope that you are rich so you can get the transformers rewound.

If things are obviously not working right, then do not push it to the limit and find the fault.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I smelt smoke

I don't think tubes are supposed to do this.
Those two things make sense.

How you got to that point was your normal route -- making sh!t up and doing stupid things.

When I disconnect the audio source, I still hear sound, just I hear static, and when I hold the input cable in the right place it even occassionally picks up morse code even though there is no mixer connected! Shorted power tube acting as a non linear device?
Anything can act as an antenna, even a guitar. Or even the internal wiring of an amplifier, especially when the input is open.
 

dietermoreno

Dec 30, 2012
238
Joined
Dec 30, 2012
Messages
238
Wow! Thanks for the knowledge!

I never knew that an output transformer could act as a speaker until now. I guess that energy has to go somewhere, in accordance with the 2nd Law of Thermodynamics.

So that "somewhere" that the energy went is possibly "oversaturating" the transformer core to expand and contract in time to the oscillation of the audio signal, in addition to some of the energy possibly went to high voltage arcing shorting the power tube creating a non linear mixer (the cause of the morse code detection when an input cable and or even the internal wiring of the amp acts as an antenna)--creating sound--AND CAUSING SERIOUS DAMAGE TO THE AMP.

So from now on I will always turn off the amp before disconnecting it from the speaker!

I hope it still works. I hope guitar amp power tubes and output transformers are forgiving to idiots.

As Steve said:

*steve* said:
"It is impossible to make anything foolproof, because fools are so ingenious." --RobertHeinlein
"Those who try to build idiot-proof systems always underestimate the persistence and ingenuity of idiots." --anon
"If you make your system more idiot-proof the idiots will build a bigger idiot" --anon


So I suppose that is why tube amps never have headphone jacks and solid state amps can have headphone jacks since solid state amps don't use output transformers.
 
Last edited:

dietermoreno

Dec 30, 2012
238
Joined
Dec 30, 2012
Messages
238
Well I tested the tube amp today (the way it is meant to be used-- with a speaker load connected to the speaker output jack) and it works to play music out of the speaker (instead of vibrations of oversaturating the transformer).

Those tubes sure are forgiving. I don't think transistors would be quite as forgiving (hence my problems with my solid state PA amplifier).

I bet those tubes could survive an EMP.

by forgiving, I mean in terms of shorts. I can have a high voltage arcing from an oversaturated output transformer and the tubes still work, compared to I have one short in the solid state amplifier and it goes out with a mystery to what happened.

Compared to the solid state amplifier will happilly connect to no speaker load and can use the air as the speaker load to transmit audio signal inductance across the room (I did this before with a transmision distance of 10 feet using a 2 foot tall metal stool as the transmitting antenna and another 2 foot tall metal stool as the receiving antenna). Compared to when I tried to do that with the tube amp, very bad idea, I suppose that tube radio transmitters should never use output transformers and must use a series of tubes with overcurrent cutoff circuitry to drain off the excess current from arcing and shorting to protect the transmitter from having a non existant speaker load. Or just use transistors in your transmitter and then the transmitter will happily transmit with no speaker load (of course it has no speaker load, it only has the air as the speaker load).
 
Last edited:

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
So I suppose that is why tube amps never have headphone jacks and solid state amps can have headphone jacks since solid state amps don't use output transformers.

no a false assumption, tho for 4, 8 or 16 Ohm outputs it is usually the case.
When you get into Public Address amplifiers that are designed for feeding multiple speaker systems, then 70V or 100V speaker systems are the normal way. These solit state amplifiers do have tapped output transformers.

Dave
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I guess it's unfair to presume everyone knows that tube amps (for the most part) MUST be run with a load.

However if you notice something strange happening, the best response is to turn it off and try to figure out whether it's a problem before turning the volume up to 11.

Yes, vacuum tubes are very tolerant of the sort of abuse that can kill other electronic components, but only up to a point. They can suffer all sorts of nasty problems which will degrade their performance and/or shorten their lives drastically.
 

Solidus

Jun 19, 2011
349
Joined
Jun 19, 2011
Messages
349
Exactly what Steve said.

In regards to oversaturation of a transformer - it's not typically possible. You magnetically saturate the core and much more than that and trafo go boom, smoke like peace pipe.

The output transformer converts one impedance to another, but it only does with the presumption that the load is connected. Without a load, the power valves will still try to push a current across an effectively infinite impedance, which will cause all sorts of nasty inductive feedback to occur on the primary side to account for this, namely the jump in impedance on the primary. This occurs in high-power transformers all the time, but when you couple plates to that primary, you get burnout fairly quickly. I've seen some fairly nasty radio transmitters where this occurs - the plates and grid structures glow like a heater - they're not supposed to.

Valves are very notably microphonic, that is, their physical plate structure can pick up AF vibrations and "feed back" the circuit if they're in proximity to the speaker. This is a pretty well documented effect.

I suppose (this is speculation) that the adverse overloading of the output trafo could have caused the magnetic field to start vibrating the plates, provided the valves were close enough. To test that, however, would take 10% experimentation and 90% disaster.

As valves do fundamentally have diode-related effects (only passing current in one way), they CAN in very-high-gain stages demodulate AM radio frequencies. This is essentially a very sh*tty crystal radio. The usual cure for this is to increase the control grid resistance (in series to the grid, not the biasing resistor) to reduce the voltage gain of that stage. The effect is also notably worse when a guitar is connected - the pickup coils serve to be a crude inductive antenna. There is one song by Rage Against the Machine that actually has this occur during recording.

Before powering on that amp again, even if it seems to be operating normally, I'd search for the signs of that smoke and get that part replaced soon. The next time that amp is pushed close to failure...I guarantee that failure won't be fun and interesting.

In regards to oversaturation of a transformer - it's not typically possible. You magnetically saturate the core and much more than that and trafo go boom, smoke like peace pipe.
 
Last edited:

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
All transfomers produce sound (see magnetostriction) where the core vibrates. If the laminations are loose, they will flap and the transformer can be very noisy. The higher the flux, the more the vibration.
 

dietermoreno

Dec 30, 2012
238
Joined
Dec 30, 2012
Messages
238
Exactly what Steve said.
I've seen some fairly nasty radio transmitters where this occurs - the plates and grid structures glow like a heater - they're not supposed to.

Valves are very notably microphonic, that is, their physical plate structure can pick up AF vibrations and "feed back" the circuit if they're in proximity to the speaker. This is a pretty well documented effect.

I suppose (this is speculation) that the adverse overloading of the output trafo could have caused the magnetic field to start vibrating the plates, provided the valves were close enough. To test that, however, would take 10% experimentation and 90% disaster.

I didn't hear the sound comming from the output transformer. I actually heard the music comming from the tubes. This music is not radio, it is the audio source connected to the amp. The only radio heard occasionally was morse code, presumably the same signal that I have picked up before from an airport beacon 7 miles away from my house.

The beeping on and off was gone when I turned the amp back on with a load connected to its output. So that would appear that demodulation was happening to an extent much more than normal when no load was connected, specifically appearing to indicated a heterodyne tone when a radio wave energizes the tube, appearing to indicate some sort of feedback non-linear behavior of the tube.

I had never heard any of my amps make any beeping like that ever before, except when I intentionally used an audio mixer to make a feedback loop, but using an audio mixer as the feedback loop didn't work that well because that's not what its supposed to do so it has circuitry to make it not do that, and this time with the tube amp connected to no load unintentionally the feedback beeping was much louder than mains hum compared to with using the audio mixer as the feedback loop mostly the only feedback was mains hum harmonics.

Yes the grids were glowing just like the heater when there was no load connected! Maybe that was the cause of the smoke.

Well if you must know, the experiment was to use the amp to transmit audio inductance a long distance of 10 feet from transmitting antenna to recieving antenna. Yes it worked. Energy efficiency lol, 100 watts transmitted for milli watts to drive a pre amp transistor 10 feet away. The speaker output jack was connected to the antenna instead of connected to the load. It was probably the dumbest thing I have done yet.

I'm scared to think of what happened to tube radio transmitters in the days before transistors.

Well I'm going to go off topic a little bit since the problem is solved.

Well I tried the experiment again today with a solid state amp and I found that if I use on the receiving end a guitar cable that has coils in it intended to prevent tangles and if I connect the coiled cabled to an antenna and to the guitar and then connect the guitar to the receiving amp, the inductive antenna effect greatly increases the gain of the received signal above the mains hum and allows me to transmit farther. Now what was quiet from 10 feet away is ear bleeding loud high fidelity and I can transmit quietly from 20 feet away. So I think I should build such an inductive antenna when I build a crystal receiver so the antenna wire doesn't have to be as long. Maybee when I build a crystal receiver, I could use the guitar, coiled guitar cable, and metal stool as the inductive antenna and then simply connect the long wire antenna to the metal stool to guide in radio waves to the inductive antenna. It works so well for receiving inductance that even when I turn off my transmitter (that was transmitting the audio output of an FM radio station), that when I put my ear up to the speaker on the recieving side, agast I hear the same FM station coming in from the air waves very quietly (it is the closest 50kW FM transmitter to my house only 15 miles away)!
 
Last edited:

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
What you're trying to do is not new science. It's actually been done commercially. One of my first jobs was working as an AV tech. The company I worked for supplied and maintained audio visual aids for schools. We sold and installed two types of wireless headset systems. One was RF and more expensive. The other was pure audio, which I'm going to cover here.

In those days it was a mixed bag of vacuum tube and solid state design with solid state gaining more ground every day. The audio wireless system was nothing more than burglar alarm foil that we strung around the classroom a full 360°. It was usually placed close to where the walls met the ceiling. One end of the loop was connected to the ground terminal of the 8 Ohm speaker jack of a phonograph, tape recorder / player, etc. The other end was connected to the audio (hot) terminal. This effectively threw a near DC short across the speaker jack because the resistance of the loop was well under an Ohm. While this long loop was reactive with inductance it could not be connected to a common 8 Ohm audio output device without modification. The phonos, tape players, etc sold with the headsets had a resistor (<8 Ohms) in series with the hot pin of the output jack. It was there to absorb and negate the short caused by the loop antenna.

The headsets had no visible antenna, as it was encased in the plastic as yet another closed circuit of many loops of wire. What I don't remember about them is whether they were powered or not. The 9V battery was all the rage at the time and powered many small portable devices. It's also possible that the headset employed crystal earphone elements which could feasibly work with no active amplification at all.

This system was not nearly as effective as the RF models, as they had many dead or weak spots in the room.

Chris

EDIT: I forgot to mention that none of the audio sources were high power. Most of these devices were in the 4 to 8W range.
 
Last edited:

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
Most public buildings in the UK are fitted with audio loops driven by dedicated amplifiers. All NHS hearing aids are fitted with a coil to pick up the magnetic field. This works well if installed to the proper specification although some interference can be caused by fluorescent lights.

I rather thought that that the original idea was to use capacitance to transmit the signal. I have not bothered to calculate the voltage required.
 

dietermoreno

Dec 30, 2012
238
Joined
Dec 30, 2012
Messages
238
Oh wow, I didn't know that anyone actually did that commercially (transmitting audio inductance instead of using RF).

My experiment was to eventually work up to using RF so that the range would be longer, I didn't know that anyone actually did this commerically.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Oh wow, I didn't know that anyone actually did that commercially (transmitting audio inductance instead of using RF).

What they are doing is creating a sort of transformer where the loop around the room is one winding, and the coils inside the receivers is another winding.

Note that the wire around the room was called the "audio loop", not "the antenna"

In general, reception is only possible whilst inside the loop.

I also note a dangerous mention of guitars and radio in a recent post.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
Look up Wikipedia 'Audio Induction Loops'.

Specification and installation details can be found in British Standard BS7594.
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
Oh wow, I didn't know that anyone actually did that commercially (transmitting audio inductance instead of using RF).

My experiment was to eventually work up to using RF so that the range would be longer, I didn't know that anyone actually did this commerically.

First off there's no such thing as "transmitting audio inductance". Inductance isn't transmitted.

Secondly RF energy, ergo 'antennas', have little in common with Inductive Audio Loop technology.

In an Inductive Audio Loop only a small fraction of the energy in the loop escapes the loop. The manetic field just expands and contracts as it follows the audio perturbations of the current flowing in the loop. They can be thought of as a giant (air core) transformer with awful coupling coefficient and mutual inductance that's unmeasurable. In contrast RF Transmitting antennas function on totally different principles. Antenna / Transmission Line theory and design can fill a text book. In fact they do, so I'm not going to go into it here.

Chris
 

dietermoreno

Dec 30, 2012
238
Joined
Dec 30, 2012
Messages
238
Would it be correct to call the primary winding of the air coil transformer the transmitting transformer and would it be correct to call the secondary winding of the air coil transformer the receiving transformer?

I think I might try this experiment again (with a solid state amp for the "transmitting transformer" of course so that over-saturation land is not reached) to see if a loop of audio cables connected to the transmitting amplifier with the "receiving transformer" inside the loop of audio cables will have better results than my previous random windings design.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Would it be correct to call the primary winding of the air coil transformer the transmitting transformer

No.

It already has a perfectly good name, the Primary Winding.

and would it be correct to call the secondary winding of the air coil transformer the receiving transformer?

No.

It already has a perfectly good name, the Secondary Winding.

I think I might try this experiment again (with a solid state amp

Don't come back to us once you've shorted the output through a length of wire saying "Smoke came out, so I tuned the volume up, and now it doesn't work".

for the "transmitting transformer"

And here you are with fantasy terms again.

Even if I agreed with your point above, in no way does that make an amplifier a "transmitting transformer".

of course so that over-saturation land is not reached)

Yeah, but now you have to make sure you don't short the amplifier out. This is a lot more likely with an induction loop as your output.

to see if a loop of audio cables connected to the transmitting amplifier with the "receiving transformer" inside the loop of audio cables will have better results than my previous random windings design.

Truly, the mind boggles.

I guess you're going to put a guitar in there somewhere too?
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
But an electric guitar pick up coil may just do the job.:):):)
 
Top