ampage
Tube Amps / Music Electronics
For current discussions, please visit Music Electronics Forum.

ampage archive

Vintage threads from the first ten years

Search for:  Mode:  

Re: Added-On Bandmaster Fuse Blows


 :
12/12/1997 1:36 AM
Steve A.
Re: Added-On Bandmaster Fuse Blows
Ray:  
 
    I read Torres' books, too, and didn't have any trouble with the fuse blowing until I went to measure the bias current in cathode-bias mode using the transformer shunt method. Dan had told me to switch the amp to standby whenever I'd switch between fixed and cathode bias, and pentode and triode modes. After blowing the 150mA fuse I added to the amp, I put in a larger-sized fuse and proceeded to blow out the 250mA fuse in my DMM!  
    After checking the circuitry (and doing a few refinements I picked up from various posts here), I once again tried to measure the bias current in cathode-bias mode, this time watching the meter closely. When I turned off the standby switch, the current rose quickly up to 100mA before I chickened out and switched the main power off. Evidentally the large power supply filter caps are charging up through the shunt, causing the current to rise gradually. Since observing that, I'm a lot more careful about the switches when I'm measuring bias current, and haven't blown another fuse.  
 
    Since your amp worked ok for a few months before this problem started up, you might want to double-check a few things in your amp, like the tube bias current and possibly the condition of the big power supply filter caps. If everything checks out okay, you may want to up the value to 250mA, and if that blows try 500mA.  
 
>>> Or do I even need this particular protection scheme?  
 
    Torres strongly recommends the fuse whenever you rewire the output stage or do similar mods: if there is a problem, fuses are a lot cheaper to replace than output transformers. He does mention that you can remove or bypass the fuse once you are sure that everything is working properly (put in a 5 amp fuse to essentially bypass it)- if that is your preference. I'd try the 500mA and if that blows, I think there may be a problem developing in your amp.  
 
Steve Ahola
 
Today
ampage
archive
And now, a word from our sponsors:


 
12/12/1997 7:51 AM
R.G.
Problems like this are one of the reasons that I was working on an output stage protector.  
 
This thing is still in proto stage, but it works like this. All the output tubes have in their cathodes a power FET rated for over 600V and a 10 ohm resistor to ground. The top end of the resistor goes to a comparator which compares it to a reference voltage.  
 
When the voltage across the resistor exceeds the reference even momentarily, the comparator sets a CMOS latch. This latch disables the gate drive to the power FET's.  
 
Effectively, the current is limited to whatever the reference voltage sets it to, and cannot exceed it even instantaneously. This will even protect from a shorted output tube or lost bias.  
 
This thing works from filament voltage and goes into a 3 by 4 circuit card for a four tube monitoring circuit.  
 
The power FET looks like a 0.1 ohm resistor when turned on and an open circuit when off, so there isn't any effect on the amp's sound. It's always fully turned on when the amp is running.  
 
I'll provide more info if you're interested.
 
12/15/1997 2:32 AM
Steve A.

R.G.  
 
    Sounds like a great idea! Stereo amps have had output protection circuits for many years, so why not add something to guitar amps besides fuses?  
 
>>>When the voltage across the resistor exceeds the reference even momentarily, the comparator sets a CMOS latch. This latch disables the gate drive to the power FET's.<<<  
 
    So once the CMOS latch is set, all of the power tubes would shut down- right? How would you reset it (after whatever problem was found and corrected)- just turn it off and back on, or does it have a reset switch? Would the instantaneous response create nuisance lockouts from voltage surges or spikes, or would you just set the reference voltage high enough so that it wouldn't trip under normal playing conditions?  
 
Steve Ahola
 
12/15/1997 8:13 AM
R.G.

>So once the CMOS latch is set, all of the  
>power tubes would shut down- right?  
Correct.  
 
>How wouldyou reset it (after whatever problem was >found and corrected)- just turn it off and back >on, or does it have a reset switch?  
It can be done either way. The CMOS logic includes  
a reset input as a normal part of the latches. All you have to do is bring out the reset point to a switch.  
 
The real question is whether you WANT to do that or not. This is like a fuse: if it tripped, something was wrong. An internal switch to hold it reset might make more sense. I suppose that a switch to force the thing on could act as a fail safe so that you could patch around a failure in the sense/shutoff unit itself, but the failure rates for the parts inside the unit would make it by far less likely to fail than the amp it's protecting.  
 
 
>Would the instantaneous response create nuisance >lockouts from voltage surges or spikes, or would >you just set the reference voltage high enough so  
>that it wouldn't trip under normal playing  
>conditions?  
 
Voltage surges and spikes are highly unlikely to get into the cathode of a power tube, given the high plate impedance to B+. This thing is measuring the voltage across a 10 ohm resistor, remember. There is generally not enough current in  
a spike to raise the voltage across such a low resistance.  
 
As to how you set the threshold, it can be done either way. If you just use a resistor directly into the sense comparator, it's effectively instantaneous, down into sub-microseconds. If you use a series R/shunt C on the capacitor input, you set a current-time limit, so that instantaneous spikes and trash won't trip it.  
 
This amounts to noting that the sheer mass of the metal inside the tube has some thermal inertia and won't die instantly.  
 
Ideally you would match the RC time constant on the trip line to the meltdown time constant of the tube, and run the trip point ahead of the tube death a certain amount. I think that the ability to set the response speed from microseconds to seconds would cover whatever range you'd like.
 

  Page 1 of 1