100 ohm resistors for filament ground reference burning up

[sbirkenstock]

Hi everybody,

I put 100 ohm resistors for the filament ground reference in my amp project.
Once in a while they just burn up. They look like sparklers for a second. (Amp project is still open).
Amp sounds fine otherwise.
What would cause that?

Best regards,

Stephan

[ToneMerc]

Hi everybody,

I put 100 ohm resistors for the filament ground reference in my amp project.
Once in a while they just burn up. They look like sparklers for a second. (Amp project is still open).
Amp sounds fine otherwise.
What would cause that?

Best regards,

Stephan

Does the PT have a center tap filament winding? If so, are you disconnecting it when deploying the 100R resistor scheme?

TM

[Jana]

Having the CT of the filament supply connected to ground along with the 100R resistors shouldn't cause them to go sparky.

I would suspect something else--maybe a tube is arcing internally and you are getting a high voltage going across the 100R.

[TUBEDUDE]

The heater isn't elevated is it?

[d95err]

What type of resistors are you using? 2W? 1/4W?

[Malcolm Irving]

...
I would suspect something else--maybe a tube is arcing internally and you are getting a high voltage going across the 100R.

Yes. Not sure what output tube(s) you are using, but many of them have the plate pin right next to a heater pin. This is a prime area for an arc-over if the power tube is driven to extreme clipping. (It's due to the 'fly-back voltage' caused by the output transformer's leakage inductance).

The 'sparkler' effect is consistent with a sudden high voltage across the resistor.

[Malcolm Irving]

In another thread somewhere, it was mentioned that a famous guitarist jams pieces of cardboard between the tube-socket lugs of his power tubes to prevent arc-over. But as someone else posted - that sounds like a quick way to get yourself an electric shock!

[Phil_S]

Ouch. Cardboard is just thick paper. It can absorb moisture. It is not a reliable insulator.

[Malcolm Irving]

Ouch. Cardboard is just thick paper. It can absorb moisture. It is not a reliable insulator.

Yes, definitely not recommended. If this is the problem here, I think there is a way of using diodes to give some protection.

This thread could be a starting point:

http://www.diyaudio.com/forums/tubes-va ... ction.html

The posts in that thread by 'tubelab.com' and 'Rod Coleman' are relevant.

[Malcolm Irving]

Reading that thread again reminds me that another possible cause of arcing is a (maybe intermittent) loss of connection to the load (the speaker).
If the connection goes open the voltages in the output tubes and output transformer go very high.

[Phil_S]

I put 100 ohm resistors for the filament ground reference in my amp project. Once in a while they just burn up....
What would cause that?

Sounds like an intermittent short. Look for it. Could be a wire whisker, insulation problem, something else? A magnifying glass might be helpful.

[Malcolm Irving]

Sounds like an intermittent short. Look for it. Could be a wire whisker, insulation problem, something else? A magnifying glass might be helpful.

Yes, agreed. Somewhere there is probably a high voltage intermittently connecting, or arcing, on to the heater circuit.

[Leo_Gnardo]

In another thread somewhere, it was mentioned that a famous guitarist jams pieces of cardboard between the tube-socket lugs of his power tubes to prevent arc-over. But as someone else posted - that sounds like a quick way to get yourself an electric shock!

Hm that famous guitarist might get himself even more famous for something he didn't plan on.

I have succesfully used GE RTV silicone caulk in rare instances where such arc prevention was necessary. And there is a special insulating paper called "fish paper" that might be used in such an application but in general I wouldn't do any thing but make sure there's nice clean tube socket with no greasy dusty residue.

BUT your problem may be due to an arc track that already exists on one of the output tube sockets, or perhaps another location. Frequently arc tracks are found on the tube itself, and an arc on either tube or socket can leave carbonised residue each on the other. Most often seen between pins 2 and 3. Best solution is replace the tube or socket.

Those with Dremel grinders and steady nerves can sometimes carve away the afflicted parts, not recommended for the faint of heart.

[sbirkenstock]

I am using five 12ax7 and a pair of 6L6.
All JJ´s at this point of time. Don´t want to blow old tubes before everything is stable.
The 100 R go to ground. No elevation (tried it, but it´s only easy when using cathode bias, my amp is switchable, it did not make much difference)
I use 1/4 watt metal film resistors.
The transformer does not have any center tap for the 6.3V filament.
It´s a mercury magnetic transformer for a Vibroverb.

[Leo_Gnardo]

I use 1/4 watt metal film resistors.

Although theory tells us 1/4W should be enough, practice has proven it isn't. Fender used generally 1/2W carbon comp, scarcely ever seen one worn out but plenty burnt out when output tubes short internally.

I'd bump up the wattage rating on your balance R's and maybe seek out some CC resistors. Need not be exactly 100 ohms, anywhere from 75 to 270 ohms seems to work just fine. It is nice to match the resistors, say within a percent or 2, a couple seconds work with your ohm meter. Carbon comp have the advantage of being able to handle brief overloads without failing. Some of the modern resistors seem to work better as unplanned fuses.