Octave effects and amp damage?

[GregS]

I was trying out a new bass octave effect pedal today with my home designed/built bass amplifier. The amplifier uses 4 6550 power tubes and puts out around 160W. My homemade speaker cabinet is a 4 x 10 sealed non-ported cabinet with each speaker rated at 150W for a total of 600W. I have two of these amplifiers which have been in regular service for years without any problems to speak of. I play a 4 string bass in standard tuning.

About 20 minutes into my trial with my new octave pedal one of my power tubes fried. I did a google search to see if any others had had a problem like this with an octave pedal. Not much turned up but there was a case where someone had this same problem with two different amps. It occurred to me that the low frequencies generated and fed into the amp from the octave pedal might cause more stress to the power section of the amp as it would be conducting higher average current over time. Same thing for the voice coils of the speakers. The low E would be generating a 20 Hz signal.

Anyway, I checked that my bias voltage was still there (it was and was normal) and replaced the fried tube and tried again. I also had my multimeter set up measuring the average cathode current of the tube. I played for a full hour without anything breaking a sweat. The cathode current would increase a little when I played hard in the low ranges but it did the same way with the octave pedal turned off as well.

My conclusion is that the octave pedal is a little harder on the amp and I happened to have a tube that wasn't 100% and couldn't handle it. I don't think my speakers are at any risk as they are in a sealed cabinet and have a combined 600W rating which is approximately 4 times the maximum power the amp puts out. However, I've dug up stuff on the internet about the risk these sub octave effects pose to speaker cabinets. I'd like to get input on this subject in general from people with greater technical knowledge which is why I'm posting this subject on this forum.

I'd be interested to hear people's thoughts and wisdom on this subject.

Thanks,
Greg

[Enzo]

A comment: Everyone wants to find something to blame. They want every failure of an amp to be BECAUSE of something. They also like to cite past performance. This worked for 20 years, why would it stop now? Well, life isn't like that.

Everything works until it doesn't, simple as that.

But it happened when I as PLAYING??? Of course. Most failures occur when something is ON and in use.

You had a tube failure. Tubes, new or old, can fail at any time. Like a big bug can fly into your windscreen at any time. They don't only do it on Tuesdays or only when the Everly Brothers come on the radio.

You may have an FX pedal that makes 20Hz, but that doesn't mean your amp can produce it. Never assume. I'd be willing to wager that your tube would have failed even without the FX. It simply ran out of time on this earth.

[J M Fahey]

Agree and add.

Besides your particular failure, which neither I find related, octave pedals are nonsense (sorry) in a Bass.

Starting with the sad but true fact that standard speakers do not reach even 40 Hz flat , what we hear when playing the lowest open string is its harmonics and our brain reconstructs the fundamental, so sending them even lower frequencies simply does not work..
For various reasons,all conspiring against us:

1) below lowest frequency, sealed speakers drop at 12dB/oct ; tuned a scandalous 24dB/oct

2) they require 4X the excursion per octave we go down.

3) The longer coil needed further lowers efficiency.

4) we hear less at such low frequencies .

One amp which incorporates a bass octaver is the Ashdown ... famous for turning those oh so popular 2x10" bass cabinet cones inside out like a glove, no kidding.

EDIT: this is the measured frequency response of an 8x10" cabinet, upgraded with Eminence Alpha 10 instead of the original ones:



look at the catastrophic rolling down below ..... 100Hz !!!!!

Look at the curve of the original Eminence BP102 "Ampeg 10"speaker" which matches the above one , the World's most famous Bass speaker drops like a rock below 100 Hz !!!!!!!:

http://www.musiciansfriend.com/acces...w-bass-speaker

[pdf64]

My (sketchy!) understanding is that as frequency is moved down, the magnetic circuit 'headroom' decreases, and that when it saturates, the primary inductance (which facilitates the primary impedance) collapses. At that point the impedance transformer action will probably fail and primary impedance becomes the primary resistance, eg for this amp, a few ohms rather than a few hundred ohms.
That might push the power tube plates into overdissipation and bring forward an impending failure mode?
Perhaps repeating that current measurement test comparing 20Hz and 40Hz at full power would indicate whether the above was the case?

[GregS]

Some clarification is in order. I'm well aware of the fact that speakers and cabinets can't reproduce very low frequencies well. That isn't the point here nor what I was aiming to do. You can put frequencies all the way down to DC into a speaker which isn't good for it as it can heat up the voice coil. You won't hear them but they may be doing damage which is the point here.

My (sketchy!) understanding is that as frequency is moved down, the magnetic circuit 'headroom' decreases, and that when it saturates, the primary inductance (which facilitates the primary impedance) collapses. At that point the impedance transformer action will probably fail and primary impedance becomes the primary resistance, eg for this amp, a few ohms rather than a few hundred ohms.
That might push the power tube plates into overdissipation and bring forward an impending failure mode?
Perhaps repeating that current measurement test comparing 20Hz and 40Hz at full power would indicate whether the above was the case?

This is exactly what I was thinking myself and why I posted this thread - to generate discussion on the possibility of octave effects damaging equipment.

Yes, tubes do fail but from my own experience this isn't common unless they're either very old, new but defective, or being worked very hard. If you get past the first 10 hours or so with a new tube it should be good for a long time. Also, I bias my amps conservatively and don't push them hard at all. The probability of a tube that's only a year old and only seen light moderate use suddenly dying for no apparent reason is very low. Not impossible. My thinking was that it was a borderline tube that the octave pushed over the edge. However, as I said before, with my multimeter hooked up (after replacing the failed tube) the average current through the tube did not seem to increase any more when the octave pedal was engaged compared to when it wasn't. So it could have been a random coincidental failure.

[gui_tarzan]

I did some amateur testing with what I can detect at low frequencies (since I play mostly bass and there were previous discussions about this) and after conferring with a friend that used to use and buy/sell high end audio gear he said the same thing that JM said - we can't actually "hear" the lowest notes of the range, but rather we hear the harmonics and our brains fill in the rest. So when someone claims they're hearing 20-40hz I have to wonder if they've done the research.

As far as it damaging a tube, I'm no engineer and I don't have the decades of experience that some of our fellow members do but I guarantee it's more likely the tube just failed like Enzo suggested. These aren't precision instruments or parts, they are hand-made and there is always the potential of a failed part - ten hours or more aside - because of that. They're not absolute, every tube is slightly different in quality and durability. My gut says you had a bad tube, I've replaced many recently and they ranged from brand new to thirty years old and everything in between.

As far as low frequencies causing the damage to the tubes, even with a higher current draw for more extended periods, I think that's like chasing a ghost. It's likely not anywhere close to the actual reason for failure. If the tube was weak construction-wise it would stand to reason the stress killed it. But if it were the low frequency itself why would it not take out the other tubes? I think the answer is right there, just a bad tube or other part that took it out.

[R.G.]

Power tubes are mostly internally current limiting, in the sense that it is very, very hard to make them conduct more current than a given "max". This is one reason that tube output stages are so tolerant of being loaded by less than the ideal plate to plate load. For most tubes, shorting a speaker output will not damage the tubes; at least not for a long while.

As for running out of magnetic headroom - sure, you can saturate a transformer primary by feeding it too much volt-time integral (that is, too big a voltage at too low a frequency). But even there, the tubes have a strong tendency to self-limit in terms of current. There is always an issue with holding a tube at high current and high voltage across it for too long, and having it melt down. But that is what it is - heat damage. The plate goes red, then yellow, then something melts and the tube dies.

But tubes are very tolerant of this, too.

You didn't say HOW the tube died. Did the plate melt? Did the *glass envelope* melt? (I've seen that!) Did the heater go open? Did some elements short? Does the tube just not have enough emission? That would be an important set of knowledge for gaining insight into what happened. Is it possible that this socket has an intermittent grid bias?

Then there is the issue of what AC signal actually got fed to the tubes. Most guitar amps don't respond down to 20 hz flat in the electronics. Preamp voicing usually rolls off the low bass well above there. So it's possible that even if the pedal fed the amp 20 Hz, the amp may not have fed this to the power tubes.

I don't know which, if any, of these were at play. But to actually find out what happened, these kinds of things have to be checked. A very instructive thing to do would be to put in an new tube, instrument the amp so you could watch and record the actual voltage and currents in the output tubes under similar conditions. This could tell pretty quickly if they were even getting near the tube power limits.

[GregS]

I did some amateur testing with what I can detect at low frequencies (since I play mostly bass and there were previous discussions about this) and after conferring with a friend that used to use and buy/sell high end audio gear he said the same thing that JM said - we can't actually "hear" the lowest notes of the range, but rather we hear the harmonics and our brains fill in the rest. So when someone claims they're hearing 20-40hz I have to wonder if they've done the research.

Thanks for your reply and input to the thread. To clarify, I did not claim to be hearing 20-40 Hz. I said the octave effect could track down to the low E on the bass and therefore possibly input a signal as low as 20Hz into the amp. Whether anyone can hear it or not is irrelevant to the topic of this thread.

You didn't say HOW the tube died. Did the plate melt? Did the *glass envelope* melt? (I've seen that!) Did the heater go open? Did some elements short? Does the tube just not have enough emission? That would be an important set of knowledge for gaining insight into what happened. Is it possible that this socket has an intermittent grid bias?

It came on very suddenly. I noticed the sound fade and when I looked at the amp the entire tube was glowing a very bright orange. I shut the amp off immediately. Nothing seemed to be very hot - not any hotter than normal, no melting of anything, no burning smell. The heater still works and there appears to be no obvious damage to the internal components. However, the glass now has a silver residue on one side that looks exactly like the same thing on the top surface of the glass. I can't quite remember the reason the top surface of tubes is like this but believe it has to do with some element being released during or shortly after manufacture. I tried the amp again with the failed tube and found it was only conducting 12 mA at idle instead of the normal 52 mA I had biased it at. The three other tubes were still conducting 52 mA. I don't think it's likely the socket has intermittent grid bias as it's a new high quality socket.

[QUOTE]Then there is the issue of what AC signal actually got fed to the tubes. Most guitar amps don't respond down to 20 hz flat in the electronics. Preamp voicing usually rolls off the low bass well above there. So it's possible that even if the pedal fed the amp 20 Hz, the amp may not have fed this to the power tubes.[QUOTE]

Good point. That makes sense.

I don't know which, if any, of these were at play. But to actually find out what happened, these kinds of things have to be checked. A very instructive thing to do would be to put in an new tube, instrument the amp so you could watch and record the actual voltage and currents in the output tubes under similar conditions. This could tell pretty quickly if they were even getting near the tube power limits.

Actually I did exactly this and found that the tubes were nowhere near power limits. The average cathode current rose slightly over idle when playing but nothing major and no difference between the octave effect being on or off - which supports what you said about most of low frequency response being filtered out in the pre-amp.

So hopefully that information I provided about the tube failure will provide more insight into what happened. I'm curious to know.

[J M Fahey]

Well, glowing a very bright orange. points straight at lack of bias.

If only that tube did so, then it was not a general bias failure , so possibilities are:
1) socket problem, pin 5 lost contact.
****Might***** be related to the octave pedal, in an indirect way: extra vibration shook the tube more than usual.

2) pin 5 wire was improperly soldered inside the hollow octal contract at the tube base (I have seen this on cheap chinese tubes, to the point of having to resloder them myself) , similar mechanical/vibration considerations as above

3) weak tube construction, some internal part worked loose by vibration and grid lost bias, either because it got internally open or it shorted something inside.

4) what Enzo says, which happens all the time: some part fails at random , you were using a red T shirt, and think it failed because you were wearing a red t shirt.
A "ridiculous" example but which I hope conveys the idea that many times coincidence does not mean causality.

[pdf64]

' The average cathode current rose slightly over idle when playing but nothing major'

Unless it was just at low gain setting, that seems strange.
What happens to the cathode currents at full power (compared to static conditions)? With and without the octaver?

[Justin Thomas]

The shine in the top of the tube is the getter, for trapping gases that are released over the normal life of the tube. That you now have something similar on the side of the tube makes me think one of your grids got vaporized! I've seen tubes redplate, but never turn orange... that's pretty darn hot if the plates were orange! Was there any increase in hum before the tube blew?

About the bias - don't THINK "unlikely because it's new & high quality." KNOW for sure! Even if the bias is reaching the solder tab, are you sure it's making it to the tube pin? As the tube is pretty well dead, maybe play "Dissect-A-Tube?" And, dim bulb tester! No sense in torching costly power tubes. Can you post a pic of the dead tube? I'm just curious.

Justin

[gui_tarzan]

Thanks for your reply and input to the thread. To clarify, I did not claim to be hearing 20-40 Hz. I said the octave effect could track down to the low E on the bass and therefore possibly input a signal as low as 20Hz into the amp. Whether anyone can hear it or not is irrelevant to the topic of this thread.

I wasn't referring to you Greg, sorry if you took it that way.

[J M Fahey]

To see it from another point of view (although it leads to a similar conclusion) , post mortem autopsy or not, there's something we already know: the tube was glowing orange.

As justin says, it is a lot of heat ... which in due time means a lot of power is being dissipated.

And power is V * I .

Since voltage is predefined by the PSU, only variable left is intensity.

Tube current (at a given plate voltage) depends straight on grid voltage so everything points to either gross lack of bias or destruction of grid function (internal shorting, wire breaking, loss of wire to pin contact, etc.)

I fail to find other plausible explanations.

That the tube afterwards does not pass much current, if any at all, may come from destruction of internal structures ...... but while glowing orange it was passing a lot.

[GregS]

I wasn't referring to you Greg, sorry if you took it that way.

Ok. Thanks for making that clear.

Greg

[GregS]

Well, glowing a very bright orange. points straight at lack of bias.

If only that tube did so, then it was not a general bias failure , so possibilities are:
1) socket problem, pin 5 lost contact.
****Might***** be related to the octave pedal, in an indirect way: extra vibration shook the tube more than usual.

2) pin 5 wire was improperly soldered inside the hollow octal contract at the tube base (I have seen this on cheap chinese tubes, to the point of having to resloder them myself) , similar mechanical/vibration considerations as above

3) weak tube construction, some internal part worked loose by vibration and grid lost bias, either because it got internally open or it shorted something inside.

4) what Enzo says, which happens all the time: some part fails at random , you were using a red T shirt, and think it failed because you were wearing a red t shirt.
A "ridiculous" example but which I hope conveys the idea that many times coincidence does not mean causality.

1) There definitely was more vibration happening than normal so that is a possibility.
2) Possible. I'd have to take the base apart to confirm that. From what I can see inside the glass the connections are still intact there.
3) I'd say likely. For the record the tubes are Electro Harmonix 6550's. I'm not sure where they're made.
4) Most likely random although extra mechanical stress from vibration might have contributed. I concluded from my simple test that the power tubes are not dissipating any more with the octave on than with it off. That was my original suspicion.

Greg