Maker Pro
Maker Pro

Tube amp acting very strange-- tubes acting as speakers

Solidus

Jun 19, 2011
349
Joined
Jun 19, 2011
Messages
349
Well, so will taking a 100' braided cable, jacking it up 50' and putting a variable capacitor across the ends.

Varying the make-up of the jury-rigged antenna will vary reception, but I think it goes without saying that you will not have reproducible results.

As it is not a purpose-built radio receiver, what you will end up receiving depends on components in the amplifier, how they are oriented, what your antenna is, how objects in the room the amplifier is in are oriented, and the time of day due to atmospheric RF propagation.

If you're interested, how this works is that any length of wire that is exposed can serve as an HF detector.

Now, HF travels through an audio amplifier differently than audio. As most amplifiers have capacitively-coupled triode cascades in their pre-amp stages, the high-frequency signals induced in the cable cannot pass into the grid and modulate the amplifier in the normal way. Instead, it travels through the grounding of the jack to the cathode of the first stage, where the diode effect of the tube rectifies it. At the plate you have only the positive side of the carrier, more specifically the positive envelope of whatever frequency the carrier was modulated at in that specific moment.

At that point it is as good of audio as any, and can proceed to be amplified in the usual audio manner.

Treatments for this are to place a <500 ohm resistor in series with the input line to the grid, or to place the same resistor as a shunt to ground on the jack itself. This is a resistance that most modern amps (usually at 100K, 220K, or 1M grid-load resistance) wouldn't blink at, but for extremely-low-magnitude signals such as what you're evidently trying to receive, serves to nearly infinitely attenuate them, or at least to the point of insignificance.
 

dietermoreno

Dec 30, 2012
238
Joined
Dec 30, 2012
Messages
238
Thanks for the explanation of how the diode effect on audio pre-amp tubes works.

Its so simple as adding a resistor to block low magnitude signals from having a path to ground.

Silly Leo Fender, why would he not think to put a resistor in every guitar amp for 5 cents higher production cost. (Well actually, the tube amp in question is a Peavey.)



I have another question: Is it a bad idea to connect a line level signal to a hi-Z input of an amplifier?

Perhaps I didn't mention because I assumed it was insignificant: when the guitar amp started smoking while my friends were at my house, 2 mics were connected to a mixer to boost the signal and a line level audio signal was also connected to the mixer, and the high output of the mixer was connected to the hi-Z line in jack of the guitar amp since the guitar amp didn't have a low-Z line in jack.


I think of it this way: The crystal set uses Hi-Z headphones. Imagine if you connected those Hi-Z crystal headphones to the speaker output jack of your 100 W guitar amp. After your ear drums are blown, you would be electricuted. So is this correct logic in why you can't connect a line level signal to a hi-Z input of an amplifier?
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
So is this correct logic in why you can't connect a line level signal to a hi-Z input of an amplifier?

You can if the H-Z input is preceded by a resistive attenuator also known as a Pad..

Chris
 

Solidus

Jun 19, 2011
349
Joined
Jun 19, 2011
Messages
349
Thanks for the explanation of how the diode effect on audio pre-amp tubes works.

Its so simple as adding a resistor to block low magnitude signals from having a path to ground.

Silly Leo Fender, why would he not think to put a resistor in every guitar amp for 5 cents higher production cost. (Well actually, the tube amp in question is a Peavey.)



I have another question: Is it a bad idea to connect a line level signal to a hi-Z input of an amplifier?

Perhaps I didn't mention because I assumed it was insignificant: when the guitar amp started smoking while my friends were at my house, 2 mics were connected to a mixer to boost the signal and a line level audio signal was also connected to the mixer, and the high output of the mixer was connected to the hi-Z line in jack of the guitar amp since the guitar amp didn't have a low-Z line in jack.


I think of it this way: The crystal set uses Hi-Z headphones. Imagine if you connected those Hi-Z crystal headphones to the speaker output jack of your 100 W guitar amp. After your ear drums are blown, you would be electricuted. So is this correct logic in why you can't connect a line level signal to a hi-Z input of an amplifier?

Tubes are high-input Z devices because the control grid of the tubes is driven to a very high voltage relative to the current dissipation. A typical preamp may be biased at -20V but the current dissipation (its draw) may be microamps.

Exact numbers depend on the amplifier. Amplifiers are mainly configured for guitars to take advantage of one or two different modes of distortion generation - the most common being plate saturation. A clean guitar amp with headroom would run a higher bias, designed to give the input a lot of room to fluctuate before driving to either mode. Metal/heavy rock amps are biased lower to take advantage of what is known as plate-starving distortion - where the input causes the plates to be temporarily 90%+ deprived of electrons. This induces a very characteristic harmonic edge to the sound.

The first preamplifier stage is not configured for a line-level input. Line-levels typically are amplified to have too high of a voltage swing and much lower impedance than the tube would like. However, in most situations, I doubt this would cause anything other than much reduced efficiency and much greater distortion to the signal.

As to the source of the smoke, I highly doubt that had anything to do. Having designed very high-power, very-linear RF amplifier designs optimized for over 1kW, the only point at which grid dissipation becomes a factor is at, well, those sorts of power levels, and it only occurs normally where the grid itself is feeding the output. In a preamp tube grid dissipation is most likely an unheard-of worry, and it would only result in a blown tube, and that's only running that very grossly over-ratings. Tubes are used for their tolerance to abuse.

However, if the preamp input was straining the system and causing drive to go near an unsafe level, combine that with an unloaded secondary (no where for the power to go) I suppose this may have blown some components.

Without seeing the system (photos would help) and having a much more laid-out explanation of every variable, it's pretty much impossible to determine what the source of failure is.
 

dietermoreno

Dec 30, 2012
238
Joined
Dec 30, 2012
Messages
238
Just for your information, the tube amp in question has the pre-amp tubes replaced with transistors. I'm not sure if that is relevant so I didn't say it earlier, but now that you gave a page long response on the workings of pre-amp tubes in guitar amps, I think it might be relevant.

Is that a problem in itself, since transistors are less forgiving of excessive current than tubes, and that is why an excessive ouput was generated that damaged the amp compared to if tubes were used in the pre-amp gain section it might not have happened?

The amp was connected to a load when my friends were using it with line level sources connected to its hi-Z input.

Previously when the tubes were acting as speakers, that was when I had the amp connected to no load AND connected to an audio mixer which the audio mixer was cranked to max volume and the amp was cranked to max volume.

When the amp was connected to a load when my friends were using it with line level inputs and excessive mic feedback, that is when it started smoking and when I turned it back on it was motor boating.

Actually, I think I might have a thread already on the amp motor boating. I don't remember and I don't know how to look for my threads in the UCP. I have created so many threads that I don't even remember creating all of them.



Edit: Found it. Its this thread.

I think maybe a moderator should combine the 2 threads together because the cause might be the same (the cause being connecting a tube amp to no load until over saturation and then over driving it to an excessive level).
 
Last edited:

Solidus

Jun 19, 2011
349
Joined
Jun 19, 2011
Messages
349
I don't think I need say that if things start smoking, it's not a good idea to run the power on it again before a full systems check.

The notion of pre-amp transistors changes the whole game...well, maybe not the whole thing, but a lot of it.

Depending on the type, transistors can be affected strongly by EMI - that is, electromagnetic interference. FETs (field-effect transistors) use control modulation in a manner that is analogous to tubes - that is, they are voltage controlled devices that use a voltage potential at the gate to electrostatically manipulate the main current conditions.

Could be whatever you blew was a suppressive element, now you have oscillations.

A look at a schematic (find it - Google) would answer that better and shed some light.

You need to distinguish/discern between current and voltage. Current does not flow unless drawn. It doesn't matter if you go into your circuit breaker, parallel every cable so you have more than 150A able to supply. If you only connect a 120V 120W (1A) lamp, that's all that circuit will move.

Transistors, because they are devices that rely on slight imperfections of electron distribution in a silicon wafer, are much, much less tolerant to overvoltage conditions. Transistors in a nutshell work with one side being electrons, the other side being holes. Electrons flow to holes but a hole can't flow to an electron. That's a diode. Add a modulation mechanism (for FETs, a silicon zone electrically isolated from the junction) and you get a transistor. Drive it too hard, and you destroy those very sensitive regions. A lot of transistors can be destroyed during handling due to static.

Tubes are 3,4, or 5 metal plates in a vacuum envelope. Steel plates don't care how much you put across them as long as they stay cool enough. If you drive it too hard, you simply reduce life, if anything at all. A lot of amplifier designs over-rate the tubes to deliberately exaggerate tonal qualities of the setup. I've seen tube designs that run the power section at a supply voltage of 150%+ the maximum datasheet rating and the setup works.

Tubes are still used in amplifiers for three major reasons:
1) They are tolerant of significant abuse.
2) They have power levels much, much higher than semiconductor designs of the same size.
3) Tube blows, unplug & replace it. Transistor blows, either you desolder it, or in many circumstances, its failure may trash the circuit.

On an offhand rambling, I suppose that difference could be related to size - that is, the actual transistor element is usually microscopic, while tubes are very macroscopic devices. Their electrodes are large. I suppose if you were to build a gigantic transistor - say, the size of a vacuum tube - you could possibly overrun it safely. It's easier to squash a cockroach than a rat. Speculation for another time.

I need you to find me the schematic for the amp, as well as take a reading of the approximate voltage of the line output at the level you were driving it at. That will give us all a better indication of what exactly occurred.
 
Last edited:
Top