5E3 Overheating/Bias Issue?

[Chodrick1]

My current 5E3 Tweed Deluxe build has had major overheating issues since I first assembled it. (Mojotone Kit). And I have a couple of questions for anyone more knowledgeable on the amplifier than I.

I first thought to check the bias of the amp, thinking this was the prime cause of the overheating, I did the calculations and I got 41mA on the output tubes each. This is almost 120% dissapation. This seems like a major issue. However I dont know how to correct this. I have the standard 270 Ohm bias resistor, I figured that if I replaced it with a 330 Ohm resistor this would put me back down to about 96% which sounds about right for a Deluxe.

I've read on some forums where people say that you don't want to mess with the biasing resistor and instead replace the 5Y3 rectifier tube with a NOS 5Y3, seeing as the modern construction of these 5Y3s are not true 5Y3 and will place a higher B+ voltage than in the original circuit. I took some measurements, and my B+ Voltage is 393VDC. This is larger than what the standard reading of 370VDC should be. I know my power transformer isnt the cause of this because I am recieving 375VAC where it should be about 380VAC, which I believe is well withing tolerance.

What do you guys think? Im still learning about amplifier construction and I love the sound of the amp now, but it is running so hot. I mean withing 3 minutes you can't even touch the rectifier or power tubes due to the heat, and the entire chassis is hot within 20 minutes, so I feel this needs to be dealt with before I do anymore tone tweaking with the amp.

I feel I should also state that I have also tried a solid state bridge rectifier using 4 1N4007 diodes just to test out the sound using a SS rectifier, and the overheating still occurs.

Thanks so much guys. Any input is helpful.

[eschertron]

What was the B+ with a diode bridge rectifier? I'd expect a lot higher than either tube recto.
What are the plate and cathode voltages? You didn't mention what numbers were used for the dissipation calculation.

[Enzo]

You didn't tell us the plate voltage or cathode voltage. At 41ma a 270 ohm cathode resistor gets me about 11 volts. SUbtract that from whatever the plate voltage is, not the B+ voltage. You multiply the voltage from plate to cathode - the voltage across the tube - by the current for wattage.

[pdf64]

...it is running so hot. I mean withing 3 minutes you can't even touch the rectifier or power tubes due to the heat, and the entire chassis is hot within 20 minutes...

It is completely normal for the glass bulb of power and rectifier tubes to run way too hot to touch.
Does the chassis get too hot to touch?
I think a vintage made 5Y3 may be a good idea.
Doesn't need to be NOS, consider looking for used working ones.

I prefer to use B+ rather than plate voltage measurements, especially for beginners; probing the plate can set off oscillation that screws up the idle voltages, and blunt probe tips can slip off tube socket terminals, particularly if pressing hard to break through surface crud.
They're then liable to short things out.

[mozz]

^^^^ get a vintage made 5Y3.

What brand 6v6's? What brand 5y3? Go with a old 5Y3. What is the plate voltage, what is the cathode voltage? Where did you come up with 96%? That is way too high for a push pull output. Get your plate voltage to about 375.

[Enzo]

I see your point about plate versus B+, however, the OP is using his numbers to calculate 120% dissipation and such, I think he has to learn to take readings sometime, and getting the actual numbers might eliminate excess worry. I'd hate to see a lot of parts changing and hand wringing , only to find the whole thing was based on faulty numbers. How was that current measured? Has the 270 ohm resistor been checked for value? What tube spec was the 120% based upon? etc.

[Chodrick1]

Sorry, I got ahead of myself there. I took some measurments and did the calculations like I had been taught and here is what Ive come up with. I hope this helps.

Also I forgot to mantion that I have place 470 Ohm 2 watt screen resistors to each 6v6, Im not sure what effect that would have but I have included the voltage drop for each tube.

I have JJ 5Y3 and 6V6 tubes. All included in the kit from Mojotone.

Rectifier voltages:
Measured Voltage:
Secondary - 375VAC
B+ Voltage - 393VDC

Calculated Voltages:
Secondary - 384VAC
B+ Voltage - 370VDC

Bias Resistor 22.89 voltage drop
22.89V ÷ 270 Ohms = 0.085
.085A or 85mA
85mA ÷ 2 = 42.5mA per tube

389.4VDC on V3 Plate
22.8VDC on V3 Cathode
.0425 × 366.6 = 15.58W
15.58 Watts
1.265VDC across screen resistor

388 VDC on V4 Plate
22.8VDC on Cathode
.0425 × 365.2 = 15.52W
15.52 Watts
1.32VDC across screen resistor

393.5VDC B1
352.3VDC B2
278.6VDC B3

[mozz]

JJ'S are a 14 watt tube so you are still a little bit over the max. A 300 or 330 ohm resistor will drop it down a bit more but you really should be looking for 60-70% dissipation on the output tubes so you may have to go even higher. I always ask what are your filament voltages? If they are too high a bucking transformer will drop all your voltages.

[Chodrick1]

I am getting 6.83VAC when checking the filaments from leg to leg. Not unusually high I would say. Is there anyway to limit this? I thought 95% dissipation was outrageous since the only other amplifier I have owned was a Deluxe Reverb, and you dont want it to exceed 70%, but on the Tweed Deluxe I saw that 95% is about average. I could try a Mojo756EXsp, which has an output voltage of 330VAC, but that is quite the step down, and I dont think I want to spend an extra $105 for one.

[mozz]

Where are you reading 95%? That's way too high. 70% is the starting point, you can adjust from there. It's not single ended nor is it a class a. I also like my filaments at 6.3vac. Look up bucking transformer, it will drop all your voltages.

[Chuck H]

You can use a string of 5W zeners to drop a bunch of volts too. The 5W's are available in the convenient DO design and can be arranged on a terminal strip high on the legs. This allows for better heat dissipation and I haven't had a problem staying below 10V for the 5W units. Like this:

[Chodrick1]

Thanks Chuck. Ill have to try that. What diodes should I use for this? If I remember correctly, they should drop about 9V per diode? so I could use 2 or 3 of them and drop it between 18 and 27 Volts. This would for sure help.

[Chodrick1]

Sorry, it seems that I mispoke.when I was taking about 95% dissipation, I meant 95% of max safety dissipation. Which given my circumstances, the maximum safe dissipation would be 32mA to 33mA. Still well below my current 42mA

[mozz]

14 watt tube, you want to dissipate 70% of that. That's a little less than 10 watts per tube. 365v plate to cathode, 23 v cathode, that is 105%. You need 25ma to start.

Weber Bias Calculator

[Chodrick1]

Im using the Rob Robinette bias calculator and its showing that I have around 15 Watts for each tube, this is outside of range on these JJ 6V6s.

I tried another suggestion I was given, where I place a 2 Watt 220 Ohm resistor inbetween the 5Y3 393V output and my B+ voltage.
All I had on my bench was 2 470 Ohm 2 watt resistors, so putting them in parallel will get me close enough to 220 Ohms. I ended up with 235 ohms.
It dropped my B+ from 393V to 377V which is more in range with what I require.
Unfortunately I am not sure if I like that amount of sag out of the rectifier.
I think I will try to aquire 5 or so 1N5347 Zener Diodes from Mouser. Hopefully that will keep the amp from sagging as bad. For now the resistors will do. I does delay the overheating issue by a significant amount.