Hi,
Forgive my ignorance but as far as tube technology goes, I have absolutely no experience. I just completed the Firefly tube amp from Doug Hammond and as far as I can tell, this amp sounds great.

I noticed that one of the tubes, the 12AU7 (just before the transformer) gets really hot, too hot to touch; either this is normal and I am over-reacting or if this is not usual then  I should take another look at all the wiring, voltages, etc.

Is the temperature of the tube only related to the filament heater or is it affected by the high voltage as well (I have a little over  6.5 vac on the filament)?

thanks for your help,

Normand (ncc)

After a bit more research and reading different sites, I think I just answered my own question. As far as the firefly amp is concerned, the 12AU7 does get very hot (too hot to touch) but this should not be a concern.
I decided to still install a small 40mm computer fan powered by the 6.3 volt with a voltage doubler rectifier (Duncan's PSU).

All is well.

ncc

Hi NCC,

Tubes get hot from the glowing filaments. You can *see* if a tube is running too hot by looking at the plates, if they glow red (you may have to turn of the light to see it) they run to hot!

Joep

I once saw an old tube stereo at an afternoon yard sale. Curious, I reached inside to pull a power tube. Just had to know if they were 6V6's or 7591's.

Funny how long it takes pain to go from your finger and thumb tips to your head.

Yea, the amp was on.

Tubes are about hot enough to iron flat your fingerprints for a good spell...

Quote from: JoepHi NCC,

Tubes get hot from the glowing filaments. You can *see* if a tube is running too hot by looking at the plates, if they glow red (you may have to turn of the light to see it) they run to hot!

Joep

I don't think this is really right. My Marshall amp always glows red, and also my guitar teacher, you see if the tubes work when they glow red...

True, but only the filaments (the small spirals inside the tube) should glow, but not the plates (the metal outside).

Bye,

Joep

oh yes, of course...

If you're gonna commit some evil crime and fingerprints are a pain, burn your fingers on a vacuum tube lol :mrgreen:

We we're playing a Gig, "Joe" [names have been preserved to protect the Innocent...of what?] noticed his amp fading, especially the attack and asked me what I thought...I looked in the back of it [A Fender Bassman] and noticed one of the output tube plates was solid Orange, and the amp itself was producing a slight heat Odor.
 He wished there was some way to cool the tubes.
 I recanted an article I'd read about 2 foot long Radio Broadcast tubes which were water cooled...hanging from sockets in a 'bathtub'.
 The short story: ... A super heated, powered up tube gets immersed in  a glass of cold water !!! ....POW/GzzzxZZ...fuze...
 Kind of dangerous, but it might have worked had the tube been allowed to cool first...don't try this at home...always carry a spare [preferable something 'nice] when travelling out of state.

I've tried the cooling technques, the best one I've found is to get the tube to bias correctly.
 Tubes in PCB...I would have to recommend point to point wired sockets or at least a cooling fan....one overheating is enough to cause PCB trace failure PROblems.

Read the Tube Amp FAQ at GEO.

The best way to cool tubes is to move air across them. Even a little air is a huge help. If there is no airflow, a tube will get hot enough to fry eggs. A minor amount of airflow will keep even hard-driven output tubes cool enough to touch with your fingertips.

There are two major subsets of tube use: small signal and power. Preamp tubes like the 12A?7 series are almost always reasonably OK as long as outside air can get to them. Power tubes need free air flow or they will get too hot, have short lives and die.

A good thing to do with amps is to mount a small fan running on lower than its specified voltage where it can blow cool air into an amp. A 240Vac fan run on 120Vac is about right, as is a 12Vdc fan run on 7-8Vdc. This can potentially double your tube life if the amp has a poor air pathway, like especially AC-30's for example. You will need to open it up periodically and vacuum out dust that deposits from the higher airflow.

Things with only preamp tubes just need air openings.

No amount of cooling will save a tube with too much standing current caused by improper  biasing. You have to get the biasing right for the conditions that the tube operates under.  The Tube Amp FAQ has a section on multiple methods of biasing.

Thanks for all the responses guys.
In the end, I did put in a small fan as suggested and I believe
it will help over the long term.

I started the thread by stating my ignorance; but I should add that
I expects the tubes to get hot and the filament to glow red. I know that much!   :?
But what I wasn't sure of is that two out of three tubes are cooler, i.e. touchable and the third one is not. This is a Firefly amp (Doug Hammond's design). There are other differences between the warm tubes and the hot one which explains what's happening. The first two tubes are 12AX7, they are used in the  pre-amp section, there voltage is much lower.
The hot one is a 12AU7, higher voltage, used to drive the output.

I managed to learn a little more again today.
take care,

Normand (ncc)

err... it`s not the voltage alone, it`s the power dissipation:
some 12at/aY/aU tubes are getting hotter than 12aX-es, coz more current is flowing "thru their plates", be it even @ only 170V,
compared to lo-current @ 320V 12aX circuits...

I have really cool permanent callouses on my fingertips from changing hot tubes over the years, and it comes in handy when working with hot components because I can't feel 'em!

DON'T cool tubes with a fan, tubes like to be hot! Like was said though, make sure the bias is right with no cherry-plating or anything but putting a fan directly on the tubes of an AC30 or older Marshall can affect the sound...I've heard it myself...there's been some great threads on this over at the plexi palace forums

 take the back panel off the AC30 instead, and change the 50 ohm cathode resistor to the old original value of 82 ohms...that's the reason most of the post-late-62 AC30's were notorious for "blowing up", because the tubes were being pushed beyond their current limits, and coupled with the poor ventilation, could literally catch fire inside...

 I saw it happen with a fender tweed high-power only a few months ago...the two inner tubes sockets and all the green cloth heater wire insulation went into flames

though there is a nice tool made from heat-resistant rubber for service-people pulling "hot tubes":


Tube Glove

it should be mentioned however, that the tube`s filaments are more sensitive towards getting broken through mechanical shocks, when hot...

The 12au7 in the firefly gets pretty hot to the touch. As long as the plates aren't red striping, it's okay. Anyone building their own tube amp should learn how to bias the output stage. There is excellent info at Randall Aiken's site. Google "aiken amps" and check the advanced tech section.

The firefly is pretty easy to bias. It is Class A cathode biased. Since it is a dual triode output stage, you don't have to worry about the contribution from the screen. Divide the current by 2 since the cathode is shared by 2 triodes. Multiply this by the difference between the plate and cathode voltages and you have the power dissipated across the plate in one triode. Check the 12au7 datasheet for the max dissipation allowed. If it is too high, try a larger cathode resistor and calculate it again.

Doug

Hmmm... that's interesting. My Firefly's 12AU7 tube doesn't get that hot. I can put my finger on it for quite a while. Did you use the Hammond 269EX transformer?

Did you measure your voltages and see what you have going on in there? When I initially hooked up my transformer incorrectly and I was getting 585V at B+1 it got very hot, but now that it is running on 300V, it's just fine....

Just some thoughts.  

QuoteDON'T cool tubes with a fan, tubes like to be hot!

I must respectfully disagree. A common thread in all the manufacturing data I've ever seen is that tube envelope temperature is a first order predictor of short life - the higher the glass temperature, the shorter the life.

Obviously, if you never turn them on, they last forever, but the trick here is to get the tubes to live long enough to be useful in actual operation. I think we would all be disappointed in re-tubing every gig - or every ten minutes, which is a distinct possibility with no airflow.

High temperatures boil off any residual gas adsorbed in the envelope, contribute to gas current, and can lead to thermal runaway in otherwise good tubes. High gas current tubes don't sound particularly good to me, but there could be someone who likes the tone.

The temperature of the glass is not necessarily related to the internal working temperature too directly. Between the plates and glass is a pretty high vacuum. The plates get rid of heat only by radiation. The change of absorption/reflection of IR at the glass surface changes trivially with temperature. It is true that if you get the glass up near melting, it will re-radiate and further raise the plate temperature, but the tube is in a bad way then. But the glass temperature is not a big effect on what goes on between the cathode and plate. That's what good vacuums are for.

QuoteLike was said though, make sure the bias is right with no cherry-plating or anything but putting a fan directly on the tubes of an AC30 or older Marshall can affect the sound...I've heard it myself...there's been some great threads on this over at the plexi palace forums

Everything in a tube amp affects the sound. The most common comment that my amp-tech friend here in Austin gets is "My amp burned out during a gig last night. It really sounded great just before it blew out. Can you fix it so it sounds like that all the time?" I have a personal momento of that - my tech friend saved me an EL34 that got hot enough to melt the glass and the vacuum actually sucked in a dent in the glass before it vented.

If you believe that you can hear a difference in tone depending on the temperature of the outside glass, there is no particular reason for a fan blowing directly on the tubes - what you do need, though, is some amount of air flow over the tubes. The amount of flow isn't large, which is why I mentioned two ways to get slow flow with available fans. Sure, fan vibration in the chassis might be transmitted as microphonics or modulation of the signal by mechanical movement of grids relative to cathodes and plates - but you oughta not mount fans so they act as shakers anyway. A smooth, slight, quiet flow of air is what you want.

I won't quibble about whether you or anyone else can hear a difference in tone with fan cooling - there are people who will swear that they can hear the difference in sound depending on the oxygen content of the copper wire carrying sound signals, and the metallic composition of the solder in joints the signal goes through. They might even be right. One quite-famous Austinite swears that he can hear the brand of his effects batteries, which direction his guitar cords are plugged in and the order in which the screws were put in the back of his speaker cabinet. Maybe he can.

I can tell you that I have tested it both ways in a blind test with a few pretty critical musicians and no one has ever been able to do better than 50/50 at just guessing whether they were using an amp with a fan or without unless they could actually hear the fan running. If there was no fan noise, they were 50/50 on whether the tone was better or worse for both fan and no fan.

It *was* funny that I got 100% good guesses when I had a fan running out of sight and *outside of the amp(s) in test*. I loved that one.

Like RG says, the life of the tube is directly related to the temperature inside the glass envelope.

I put a cooling fan in all my tube amps that didn't come with one from the factory. I have never had any problems with them causing noise of any kind. I have one that is mounted directly the speaker cabinet and one that is suspended by springs. Both work fine.

If you want to see the spring mounted one, follow this link:
http://www.diyguitarist.com/Images/Tr-rear.jpg

And here is the one mounted directly to the cabinet:
http://www.diyguitarist.com/Images/BD_RView.jpg

The trick seems to be running them at 1/2 to 2/3 of their rated voltage. They run slower, are quieter, and create no problems that I am aware of

I haven't had any problems with the firefly. But I used expandable metal grille in the front and back of the box, so it gets plenty of air. I wouldn't necessarily recommend installing a fan unless you have a known heat problem.

Doug