smashed pcl86 screen resistor

[gontek]

Hello,

I build push-pull guitar amp, more or less based on attached schematic, I used one 12ax7 in fender style preamp, and pcl86 triodes as a phase inverter,
My screen resistors are 1K 2W, yesterday while plaing hard using a power brake - which is just 8Ohm LPAD - one of my screen resistors was burned, and one pcl86 died (now it burns down new screen resistors). I thought is was old pcl fault,  replaced the failed tube.
I repeated the experiment with the same result, smoked another two resistors, and probably tube,

what could be the reason for smoking screen resistors? could it be LPAD related?
My voltages are:
315V pentode anodes
313V screens
fixed bias set for ~20mA

Kind regards

[tubeswell]

Hello,

I build push-pull guitar amp, more or less based on attached schematic, I used one 12ax7 in fender style preamp, and pcl86 triodes as a phase inverter,
My screen resistors are 1K 2W, yesterday while plaing hard using a power brake - which is just 8Ohm LPAD - one of may screen resistors was burned, and one pcl86 died (now it burns down new screen resistors). I thought is was old pcl fault,  replaced the failed tube.
I repeated the experiment with the same result, smoked another two resistors, and probably tube,

what could be the reason for smoking screen resistors? could it be LPAD related?
My voltages are:
315V pentode anodes
313V screens
fixed bias set for ~20mA

Kind regards

Screen Resistors smoke because too much current.


Too much current in a tube is caused by either
1) bias being too hot causing excessive screen current under heavy signal conditions , (in which case rebias) or
2) running tube way beyond recommended parameters (too high a voltage etc) into an improper load resulting in permanent tube damage (see 3 below) - running into an attenuator can do this (or exacerbate things), or
3) when amps with Long Tail Pair inverters are overdriven, this can cause bias problems when output tubes on one side redplate under heavy signal conditions. Running into an attenuator can exacerbate this condition (because the circuit is being pushed hard).
4) short between the tube electrodes that causes current to take an unwanted path from source to sink (from damaged tube that maybe was previously overheated causing structural parts like screen grid support rods to melt and warp bringing screen into contact with g1 or cathode).

[gontek]

Bias is quite cold,
I will order new set of tubes, also I increased screen resistors resistance, and power rating

Thank you

[Merlin]

That's what happens when you play a fixed-biased amp hard. The difference between idle, and maximum current, is much larger than in a cathode biased amp, since fixed-bias has no self-limiting ability. A resistor with a bigger power rating should solve your problem, yeah.

[gontek]

I probably found the problem,
there was bad oscillation when on a higher gain, It can be heard on clean, but not really while pushed hard with tube screamer and powerbrake
Oscillation came from 2nd stage preamp grid leak resistor to close to B+ wire

Thanks for advices

[Willabe]

Oscillation came from 2nd stage preamp grid leak resistor to close to B+ wire

Glad you found it.

Could you be a little more clear? What B+ wire?

[Latole]

Bias problem.
Adjusting the bias at rest is misleading and if you play at high power, which is the case, the bias must be cooler.

What is the maximum power of this tube? 315V X  20ma = 6,3 watts

[gontek]

Bias problem.
Adjusting the bias at rest is misleading and if you play at high power, which is the case, the bias must be cooler.

What is the maximum power of this tube? 315V X  20ma = 6,3 watts

max plate dissipation for pcl86 is 9W

[gontek]

Oscillation came from 2nd stage preamp grid leak resistor to close to B+ wire

Glad you found it.

Could you be a little more clear? What B+ wire?

it was B+ wire feeding the plate of 12ax7, I mouted grid leak directly on tube socket,
separating elements by ~1/2 inch, killed oscillation

[Latole]

Bias problem.
Adjusting the bias at rest is misleading and if you play at high power, which is the case, the bias must be cooler.

What is the maximum power of this tube? 315V X  20ma = 6,3 watts

max plate dissipation for pcl86 is 9W

IMO Bias may be too hot .

[gontek]

Bias problem.
Adjusting the bias at rest is misleading and if you play at high power, which is the case, the bias must be cooler.

What is the maximum power of this tube? 315V X  20ma = 6,3 watts

max plate dissipation for pcl86 is 9W

IMO Bias may be too hot .

maybe you're right, I will try lower bias,
pcl86 is popular, cheap (in Poland) tv tube, so I don't worry much ;)

[Willabe]

it was B+ wire feeding the plate of 12ax7, I mouted grid leak directly on tube socket,
separating elements by ~1/2 inch, killed oscillation

I believe you, but that shouldn't be a problem. 

[Latole]

Bias problem.
Adjusting the bias at rest is misleading and if you play at high power, which is the case, the bias must be cooler.

What is the maximum power of this tube? 315V X  20ma = 6,3 watts

max plate dissipation for pcl86 is 9W

IMO Bias may be too hot .

maybe you're right, I will try lower bias,
pcl86 is popular, cheap (in Poland) tv tube, so I don't worry much ;)

I have these issues few times

[gontek]

it was B+ wire feeding the plate of 12ax7, I mouted grid leak directly on tube socket,
separating elements by ~1/2 inch, killed oscillation

I believe you, but that shouldn't be a problem.

You're right, It was not grid leak, but input 68k grid stopper

[Willabe]

it was B+ wire feeding the plate of 12ax7, I mounted grid leak directly on tube socket,
separating elements by ~1/2 inch, killed oscillation

I believe you, but that shouldn't be a problem.

You're right, It was not grid leak, but input 68k grid stopper

That's what a grid stoppers for. It's supposed to be mounted as close to, with in reason, the tube sockets grid pin as possible. It shouldn't be a problem either.