Pushing EL36 ??

[Tesla88]

Hi ,
back from holidays i resumed an amplifier originally intended to work with a couple EL503 in the output stage.
After many test with various tubes i ended up with EL36 due to better performance in terms of output power and efficiency , so plate load seems ok and i 'm getting more than 50W sine power and about 200mA of current at max signal , B+ at 350V DC . Amp starts clipping at 60W circa.
Bias is about 30mA each tube

Is it safe to run EL36s (Pa max is 16W) tubes at this power levels ?

Is there any relation between Pa and Pout from a safety point of view ?

I think most is related on class of operation , and here i'm working near class B.

Anyway , i run amplifier at max power for at least 5-10 min and no redplating in darkroom ...

 

[JonSnell Electronic]

In a word Yes.
The 70% point (class A) is 32mA for that pentode with 350v Anode supply.

 

[stevenrotterdam]

Here you can read a lot about this.

CHAMP ELECTRONICS

 

[Tesla88]

So is ok to obtain 60W from 2 EL36 (Pa max 16W ) ?

I Ask cause looking at EL34 , EL503 and EL36 plates dimension and i'm scared of pushing the small EL36 plates...

 

[Tubelab_com]

You state that the amp idles at 30 mA per tube on 350 volts. Thats 10.5 watts per tube, so you are OK at idle.

You state that the amp draws 200 mA at 50 watts output. That means that you are applying 70 watts of DC (350 * 0.2) to the amp to get 50 watts out so the amp is dissipating 20 watts at 50 watts out (70-50), which is 10 watts per tube, so that point is OK. A Class A amp will dissipate the most power at zero output. Most class AB amps will dissipate the most power somewhere between 1/3 and 2/3 of full power. I would do a check at 25 watts output and if you are under 16 watts per tube, you are probably OK.

If the amp will be used for average music listening where the amp rarely sees clipping, you are OK since the AVERAGE power output for most music is at least 10 db below the peaks. This means that 99% of the time the amp puts out less than 5 watts, unless it is a guitar amp or used for loud EDM.

 

[Blaireau]

EL36

The datasheet from philips says 44,5W with 7.2% distortion in B class push-pull. So the output power should be weaker in AB mode. Are you sure you've really measured a 50W output power ? http://frank.yueksel.org/sheets/030/e/EL36.pdf

 

[JonSnell Electronic]

If you overdrive the valves, it will give you red hot anodes and that will reduce automatically the electron flow. (compression) and therefore the power output plus heavy distortion. As long as this doesn't last for long, no lasting damage should occur.

 

[artosalo]

The datasheet from philips says 44,5W with 7.2% distortion in B class push-pull...

Yes, but with Ua = 300 V. The op told he has 350 V.

 

[gcwills]

In my experience, EL36 (6CM5) valves are capable of high output power in spite of the datasheet reference to Pa max of 12W. As already mentioned, John Chambers at Champ Electronics has demonstrated high output power from these valves. They are low impedance devices, and capable of high current at lowish HT.
My experiments with these valves can be seen here: A new take on the Playmaster amp

 

[Tesla88]

Amp is intended for guitar use so overdrive is possible , i'll test the amp at various power searching for the worst conditions as suggested by tubelab .
Is good to test the amp at max distortion for long times to see what happen ? In these conditions current at center tap is about 250mA.

B+ is 360V at idle , 340V at max undistorted output.

Previously i used EL34 and 6L6 having more Pa max .With 6L6GC no way to obtain more than 20-25W ...
EL34s biased at 50mA gives 40W with a current of 250mA , clip is about 50W with more than 300mA input meaning more than 80W of input ... and a lot of power at clip.

With the EL36 (Philips Miniwatt) i read 21 V rms @ 8 Ohm load with a good sine wave on scope and the big OT goes wodn to 30 Hz with 40W of undistorted power 🙂

 

[Tubelab_com]

Is good to test the amp at max distortion for long times to see what happen

If that's what it's going to see in use, then test it that way.

Running a guitar amp hard into clipping imposes a different set of conditions on the amp than would be seen in a HiFi amp. Testing a guitar amp with a load resistor well into clipping actually reduces the dissipation in the output tubes, because they begin to operate as switches. If the amp is driven hard enough and the OPT is reasonably good, the output will become a near square wave, and the tubes will run remarkably cool.

However, things change with a speaker connected, especially if the speaker's cone resonance falls inside the range of a guitar. A guitar that uses standard tuning goes down to 82 Hz, and many guitar speakers have a resonant peak higher than this. The speakers impedance may be 25 ohms or more throughout much of the range where "power chords" are played. The tubes are switching between cutoff and saturation into an elevated reactive load. This can drive the peak plate voltages into the kilovolt range causing cheap tube sockets and OPT's to break down. This is a real problem with the common audio tubes that have the plate on pin 3 and the heater on pin 2. I have seen an arc at the socket pins. This will not be a problem with the plate cap type tube that you are using, but be aware that the plate voltages can go higher than the accepted 2 X the B+ voltage.

Note: Don't scope the output tube plates directly. Use a voltage divider on your scope unless you are absolutely sure that it can eat transients of 3X the B+ or higher. I use a series string of 10 X 100 K 3 watt resistors from plate to ground with the scope across the bottom resistor, but I am working on big amps on a 650 volt B+. I have seen 2KV when operating into a speaker with a guitar doing a feedback sustained power chord.

The other issue with a guitar amp is the duty cycle. A HiFi amp sees a few random excursions to near clipping, and the average power output is less than 10% of full power. A guitar amp can see long play times and some (OK, most) players are stringing together notes faster than they decay, so the amp never gets a break. It's easy to see a 50% or higher duty cycle. A guitar amp may dissipate far more heat than the same sized HiFi amp and many guitar amp makers (like Marshall) put them inside a well insulated box (wood and tolex) with poor ventillation. Most of the heat from a tube is infared radiation. It will heat up nearby surfaces and they in turn need ventillation.

In my experience, EL36 (6CM5) valves are capable of high output power

Most TV sweep tubes (line output) are more efficient, and more conservatively rated than the typical audio output tubes. I have never seen the EL36, but compare the cathode size to a 6L6GC. It's probably bigger. A bigger cathode gives more emission, and therfore a lower voltage drop across the tube when driven to saturation. I have been "testing" some XL500's (13GB5) and I can get 150 watts from a pair on 550 volts with a 3300 ohm load. I am still using the first pair of tubes, I haven't blown them up yet!

 

[DF96]

it will give you red hot anodes and that will reduce automatically the electron flow.

By what mechanism? Hot valves often increase the electron flow by outgassing causing positive grid current thus upsetting the bias, leading (sometimes) to thermal runaway.

 

[Tubelab_com]

By what mechanism?

The melting tube going to air, or the gas induced arc over blowing up the equipment......sorry I just had to do it.

 

[DF96]

As long as this doesn't last for long, no lasting damage should occur.

The melting tube going to air, or the gas induced arc over blowing up the equipment.

Could you two be talking about slightly different equipment operating regimes?

 

[Tubelab_com]

Someone brought up the subject of "red hot anodes." This is a subject that I am familiar with. I have explored all the possible anode colors including white! Any time you see red (or worse) on a receiving tube, metal ions are being released that contaminate the vacuum and contribute to unwanted grid current. Eventually there will be enough ions causing enough grid current to upset the bias causing the tube to runaway. Runaway leads to blown parts, or melted tubes. Of course sometimes I might have explored the red zone on purpose.

The first picture shows an early pair of Chinese 6L6GC's. They are dissipating 44 watts and glowing red. These tubes are over 10 years old now, have been pushed to red several times, and are still very much alive. They were never really great tubes, so I use them for "testing" the limits of new designs.

The second picture shows a pair of 98 cent 6BQ6GA's being tested well beyond their limits. You could read by the light (note the reflections in the OPT's) and feel the radiant heat across the room. They are no longer alive!

The third picture shows an extreme case of gas. I bought two 211's at a hamfest for $5. The seller said that they were "weak". This one blew up the amp. This is "gas" inside the tube that ionizes glowing blue or purple. It creates a near short in the tube which took out the mosfet that drives the grid. The plate voltage in that amp is 1100 volts.

 

[Yvesm]

. . . Of course sometimes I might have explored the red zone on purpose.

At least one purpose is to show such amazing pictures 😀

Yves.

 

[gcwills]

Most TV sweep tubes (line output) are more efficient, and more conservatively rated than the typical audio output tubes. I have never seen the EL36, but compare the cathode size to a 6L6GC. It's probably bigger. A bigger cathode gives more emission, and therfore a lower voltage drop across the tube when driven to saturation. I have been "testing" some XL500's (13GB5) and I can get 150 watts from a pair on 550 volts with a 3300 ohm load. I am still using the first pair of tubes, I haven't blown them up yet!

I should have also mentioned that the EL36 will drive much lower load impedances than say a 6L6GC or an EL34. My amp saw a PP pair happily driving a 2K p-p load. As Tubelab says they have a substantial cathode and I guess this is another way of saying they are lower impedance devices.

 

[Tubelab_com]

At least one purpose is to show such amazing pictures

Sometimes it's just to see what happens.

When I was 16 years old I got a job in a TV repair shop. I got the job by telling the shop owner I would put up TV antennas, but as soon as the guy realized that I really could fix TV's I was fixing them while he put up antennas in the hot Miami summer......all for $1.05 per hour. That was minimum wage in 1968, and what my friends made at McDonalds.

One day the boss was out and I had a sick TV. Without thinking I pulled out the horizontal oscillator tube without turning off the TV and stuck it in the tube tester. The phone rang, and by the time I smelled something hot, the horizontal output tube had melted and the vacuum had sucked much of the glass envelope in around the plate. After the TV had cooled off I had discovered that the tube had become stuck into the socket, and I broke it trying to remove it. I have been trying to duplicate that tube for years without success.......unless I use a propane torch.

Can you make metal envelope of a metal 6L6 glow red? I can. Can you make a big enough plasma ball inside a tube to melt the glass and have it bulge OUTWARD due to positive pressure.....without using a microwave oven? I can.......but these are stories for another time, another thread....and a YouTube camera.

 

[gcwills]

Interesting story about the line output valve meltdown. I have repaired several Marshall DSL amps that due to leakage in the PCB material cause a similar meltdown in EL34s. Here is a picture of one example of this:

 

[Tubelab_com]

I have seen the funnel of death before, but never in an EL34. It's pretty common in a 6AQ5 because the glass is so close to the plate, and those tubes are prone to hot spots.

I have been working on a sweep tube amp. I have one channel breadboarded and have been playing with medium sized tubes (24 watt plates) since that's what I will probably use in the final amp. I think I will try some smaller tubes this weekend to see what happens. Tonight I saw 175 watts flow from a pair if 1960 vintage 6CB5's. I have a need for a 40 watt sized guitar amp head, but it must be small, so some small tubes will get tested.

So far I have been posting in this thread:

http://www.diyaudio.com/forums/tubes-valves/237808-show-me-your-screen-drive-circuits.html

I think that it's about time to start my own.