Cathode bias el34 power transfromer question

[Electric Altar]

Hello,
This is my first real post here after reading for a few years.  I have built an amp with a preamp based on an Orange OR120 and power amp from the matchless Chieftain, cathode bias el34.  I love the way the amp sounds and have been using it with my band but am concerned at the heat generated by the power transformer, it gets pretty hot.  I can hold my fingers on it for a bit but its hotter than any of my other amps.   
This is the first amp i have ventured on my own to put together all previous amps i have built i stuck to a known design.  I have included a schematic with voltages and my transformer specs in hopes more experiences eyes can give me some guidance. 

A schematic with voltages should be attached.
Please let me know if there is any other information i can provide.  I used transformers intended for another build in the 35 watt range. 
Many thanks in advance for any suggestions or information.  This forum has been a great wealth of information, i really appreciate this kind of community and its contributors.
-Brian

[shooter]

I did something similar, but used the Clubman PA section, I build it with the *idea* of going to possibly 80W so I used a WPT100 for my PS iron, play for hours, it just yawns

[HotBluePlates]

It would be handy to know the voltage across the 110Ω cathode resistor for the output tubes. And maybe also the voltage at the screens (pin 4).

EDIT: I should have paid more attention to your attachment! 

[mresistor]

Looks to me like you are running about 23 watts per tube which is considerably above recommended 70% of max plate dissipation.   25 x .7 = 17.5 watts. You might try increasing the cathode resistor to say 130R or so and then see what the dissipation is.

[mresistor]

You can use this bias calculator that is pretty easy to use.  http://www.tedweber.com/webervst/tubes1/calcbias.htm

[sluckey]

Cathode biased amps usually don't follow the 70% rule of thumb. It's very common for cathode biased amps to run at 100% of max plate dissipation.

[HotBluePlates]

Looks to me like you are running about 23 watts per tube which is considerably above recommended 70% of max plate dissipation.  ...

319v plate - 17v cathode = 302v plate-to-cathode
17v cathode / 110Ω = ~155mA -> /2 = ~77mA cathode current per tube

321v screen B+ node - 254v phase inverter B+ node = 67v across 22kΩ dropping R = ~3mA phase inverter & preamp current
341v plate B+ node - 321v screen B+ node = 20v across 1kΩ dropping R = 20mA - 3mA preamp current = 17mA screen current / 2 = 8.5mA screen current per tube

77mA cathode current per tube - 8.5mA screen current per tube = 68.5mA plate current per tube
302v plate-to-cathode * 68.5mA = ~20.7 watts dissipation.

Since this is a cathode-biased amp (which can never work reliably far into class AB), I'd consider this very acceptable plate dissipation, especially if you don't see redplating.

And at 200mA, I bet the PT is well-sized to cope with this amp's power draw. I think your amp is fine, even if it does seem hotter-than-typical. IIRC, the Matchless Clubman 35 I once owned acted similarly and with no issues.

[Electric Altar]

Awesome, thanks HotBluePlates!! I thought it was ok, but the nagging concern of a meltdown or something was on me with it being hotter than my other amps. 
I am really liking the tone and volume of cathode bias el34 in pp, it gets me the perfect amount of crunch and volume to hang with my drummer.



Since this was my first attempt at conjoining different schismatics if anyone has any suggestions or points of concern based on my schematic i would love to hear any feedback.

Thanks again guys!!!

[mresistor]

Cathode biased amps usually don't follow the 70% rule of thumb. It's very common for cathode biased amps to run at 100% of max plate dissipation.

Thanks Sluckey - when am I going to remember this?   

[mresistor]

HPB - In the weber bias calculator ( of which I totally missed the note about cathode bias again! ) the note says the assumption that screen current is 5% of the cathode current. In using that I come up with 73ma of actual plate current. In your calculations you have determined a different, and more accurate? figure of 68.5 ma of plate current per tube. Do we really know the exact amount of screen current without measuring with a current meter?   IOW  why does Weber state this assumption? 

[sluckey]

Do we really know the exact amount of screen current without measuring with a current meter?   IOW  why does Weber state this assumption? 

You can calculate the actual screen current because you have all the voltage drops for all the B+ nodes.

Weber states an assumption because their calculator doesn't have a schematic with voltages to be able to accurately determine the actual screen current.
 

[kagliostro]

Have you considered if the construction allows enoug ventilation inside the case / chassis ?


Franco

[hesamadman]

Cathode bias runs hot. If you dont have proper ventilation, the aluminum chassis is going to get hot and transfer a significant amount of heat to other metal objects touching it. My last EL84 build was transferring heat to the metal fender style power switches. They were EXTREMELY warm. A little bit of venting solved that issue quick.

[HotBluePlates]

... In your calculations you have determined a different, and more accurate? figure of 68.5 ma of plate current per tube. Do we really know the exact amount of screen current without measuring with a current meter? ...

Before I looked closely to realize all needed info was on the schematic, I asked for voltages right at the screen pins. The idea there was we have the voltage of the filter cap feeding the screens, and know the resistance of the screen resistors. So we could use Ohm's Law to know the screen current.

As Sluckey said, the schematic has voltages for every power supply node, and the node which powers the screens power no other circuit. So I was able to subtract the current found for the preamp/phase inverter from total current flowing through the 1kΩ dropping resistor (leading to the screen node) to find total screen current.

There is some room for error, as my math assumes each output tube draws exactly the same screen current and cathode current, where we truly only know total current of both tubes for each quantity. But unless the amp is redplating on one of the tubes, I bet the math isn't far from reality.

Separately, the 70% "rule" is only a convenient estimate anyway. Class AB means operation between class A (typically 100% dissipation at idle) and class B (theoretically 0% dissipation at idle, typically some amount more than 0%). So the true correct idle current depends on whether the output stage is being worked closer to class A or class B.

Unfortunately, we almost never are given complete information on the design idle bias (meaning both bias voltage and idle current, so we know how to adjust for varying tube samples). On the other hand, "close enough" very often is close enough in most amps.

[mresistor]

Thanks for the answer guys,  appreciate it . 


Also, for the OP if your cabinet is not well ventilated (look at some marshall cabs with that top vent over the tubes) AND your rectifier tube is also close to the PT, and you are running the volume up towards 7-8-9  the PT is going to get pretty toasty.

[PRR]

> weber bias calculator ...the assumption that screen current is 5% of the cathode current.

That's a broad assumption.

In practice: 6L6 and friends will have low Ig2. Sometimes as low as 2%. EL34 and friends will have high Ig2. 10% and more.

EL34 data for self-bias shows a typical 150mA plate 23mA G2. 173mA total of which _13.3_% goes to screens!

So "5%" is just a general round number so as to make the chart "simple". It will get you in a tolerable zone. Since it is a Fix-bias chart aiming at 70%, which might as well be 60% or 80%, little 7% (5% to 13%) deviations are no big thing.

[Paul1453]

Are you running the 5AR4 or GZ34 rectifier tube?

If you have the seemingly more common 5U4 in there instead,
you would be using the full 3A for 5VAC your PT is rated for.

If you are running a 5U4 I suspect this may be the reason the PT is getting hotter than you want it to.

Your 6.3VAC supply should be running well under its capacity, so shouldn't be causing excess heat.   

[HotBluePlates]

If you have the seemingly more common 5U4 in there instead,
you would be using the full 3A for 5VAC your PT is rated for. ...

I don't believe this would be an issue.  The high average current draw from the high voltage winding is most likely the cause of the heating.

Of course, if the tubes hang down Fender-style and/or the rectifier & output tubes are positioned close to the PT, that will contribute to heating as well.

[Electric Altar]

The amp has some ventilation over the PT and obviously out the back , i may add more though as it is a smaller size head box... perhaps on the side near the PT. 

To the other question above i am currently using a solid State plug as the voltage drop from the Tube rectifiers was knocking the voltage pretty low.  Though when i did use the tube rectifier the heay of the PT was about the same. 

Thanks to all for the information!!