what would cause the screen grid to catch on fire?

[MitchK]

Ok well it wasn't quite that bad.

here's what I'm doing. I'm putting in a quad of kt88s into a 5150, and because the original sockets are all too close I've made a little riser to mount the tubes on so they are spaced apart a bit better.

here's what I've done

took the pcb board totally out, removed all the sockets, remounted to sockets to my riser, connected wires from the sockets to their spots on the pcb board, and then separately i've made two additional modifications. I've put in an adjustable bias, this was actually already in there, but I just put in a better quality potentiometer. the other thing was that I put in a switch to bypass the two diodes in the power section I've been told these are fairly superficial and removing them can have a positive impact to the tone so thats why I put the switch in to be able to bypass them and switch back and forth and compare the difference. so basically I left the diodes exactly where they but attached a wire to bypass them and put a switch in the wires path.

so what happened was that when I turned the power to the amp on, with no power tubes, without taking it off standby this is just the power switch at this time, there was a crackling in the speakers and a bright yellowy sparking, not really like an arcing more of a burning. I took a look and this happened on pin 4 of the tube thats right by where the OT connects to the pcb for those who are familiar. What on earth could cause this? at this point the only thing I could think of is that I may have accidentally left some flux paste on the contact, could it have just combusted because of the electricity? I don't really want to power it back on to find out without seeing what all you great folks think.

[Amp Kat]

Well theres no doubt that you have a short in the power supply somewhere or something touching something it's not supposed to. There is a resistor I think around 400 ohms that takes the place of a choke in the 5150 and that connection goes to the screens. It is also where the screen voltage is derived from and should be around 500 with the stanby on and drop to about 450 when off. Check from the end of that resistor and pin 4 and see if you get the total screen resistor value there. Three more things is that 1.) If I'm not mistaken those screen resistors are only 100 ohms and those KT-88's aren't going to like them very much so I'd install some 1k's. 2.) If you decide to run EL-34's you will have to tie pins 1 & 8 together and the last tube socket on the left looking at the front to the back is isolated from ground so you'll have to jump a wire from ground to 1 & 8 or it will be open. 3.) That 5150 OT isn't going to support 4 KT-88's at high volumes for very long at no matter what impedance setting as they are unbalanced and not very stout and underrated IMO as far as wattage is concerned allready. You'd do much better runnnig just two and set the bias accordingly. Trust me it will be loud enough to keep up with almost any 100 watt amp out there.

[MitchK]

Well theres no doubt that you have a short in the power supply somewhere or something touching something it's not supposed to. There is a resistor I think around 400 ohms that takes the place of a choke in the 5150 and that connection goes to the screens. It is also where the screen voltage is derived from and should be around 500 with the stanby on and drop to about 450 when off. Check from the end of that resistor and pin 4 and see if you get the total screen resistor value there. Three more things is that 1.) If I'm not mistaken those screen resistors are only 100 ohms and those KT-88's aren't going to like them very much so I'd install some 1k's. 2.) If you decide to run EL-34's you will have to tie pins 1 & 8 together and the last tube socket on the left looking at the front to the back is isolated from ground so you'll have to jump a wire from ground to 1 & 8 or it will be open. 3.) That 5150 OT isn't going to support 4 KT-88's at high volumes for very long at no matter what impedance setting as they are unbalanced and not very stout and underrated IMO as far as wattage is concerned allready. You'd do much better runnnig just two and set the bias accordingly. Trust me it will be loud enough to keep up with almost any 100 watt amp out there.

yeah thats the thing for quite a while I was running it with just two kt88s, sounded fine and all I just thought i'd do the entire quad. to be honest i did mostly out of wanting to fuck around with my amp I was thinking about getting a mercury magnetics OT to replace the stock one, I've heard good things,. I don't know if they are rated any higher though, and also I've been told time and time again that the circuit won't drive more than 120 watts anyways?

oh and finally when you say set the bias accordingly, do you set the bias at half of what it would normally be with a full quad?

as for the the state its in right now, when I read from pin 4 of this one socket that sparked up to the end of the choke resistor I get 500 ohms. 400 for the "choke" and 100 for the screen resistor. Its really hard to think of where it could be coming from, essentially all I did was take off the sockets and then reconnect them to the board with wires so that I could space them apart. obivously something else happened in the process but I can't figure out what. on the PCB since the sockets are no longer on the board, I did have to drill two holes so that I could re-mount the pcb in the chassis. there were actually two holes in the pcb already, I just made them a bit bigger. so now I have the pcb mounted on a screw in two opposite corners. the screws are grounded of course but I'v checked and checked again to make sure that the do not come in contact with the circuit. Unless its arcing through the pcb to the screws. they are in close proximity for sure, but I figured I'd see singe marks were it happened but I only see that on pin 4 which is far away from the corner. everything is also all glued up so there is no actual direct contact.

[MitchK]

just made a slight discovery. I can't seem to get a reading from two of the screen resistors. the outer sockets are fine I read pin 4 to the choke and get 500 ohms, but when I read form either of the middle sockets 4th pin to the same spot I get infinite resistance. when I read the resistors themselves i get the same thing, I'm reading from the underside though so I'm going to take the pcb off again and check them out from the top side.

could this actually be the problem? it really doesn't seem like it because its only the one outer socket that sparking up not the these middle ones. any thoughts?

[aletheian-alex]

Ok well it wasn't quite that bad.

here's what I'm doing. I'm putting in a quad of kt88s into a 5150, and because the original sockets are all too close I've made a little riser to mount the tubes on so they are spaced apart a bit better.

here's what I've done

took the pcb board totally out, removed all the sockets, remounted to sockets to my riser, connected wires from the sockets to their spots on the pcb board, and then separately i've made two additional modifications. I've put in an adjustable bias, this was actually already in there, but I just put in a better quality potentiometer. the other thing was that I put in a switch to bypass the two diodes in the power section I've been told these are fairly superficial and removing them can have a positive impact to the tone so thats why I put the switch in to be able to bypass them and switch back and forth and compare the difference. so basically I left the diodes exactly where they but attached a wire to bypass them and put a switch in the wires path.

so what happened was that when I turned the power to the amp on, with no power tubes, without taking it off standby this is just the power switch at this time, there was a crackling in the speakers and a bright yellowy sparking, not really like an arcing more of a burning. I took a look and this happened on pin 4 of the tube thats right by where the OT connects to the pcb for those who are familiar. What on earth could cause this? at this point the only thing I could think of is that I may have accidentally left some flux paste on the contact, could it have just combusted because of the electricity? I don't really want to power it back on to find out without seeing what all you great folks think.

HOLY COW! Do you mean CR200 and CR201? Don't bypass those! They run from the plates to ground to ssave the OT in case of spikes... not terribly useful, but if you REMOVE them, it is fine... if you BYPASS them with a wire, then you are short circuiting over 500v directly to ground.


unfortunately, you may have fried either the reanny, the PCB, the wiring harness, the supply caps, or EVERYTHING!

[stokes]

Alex is 100% correct on this.If they are suppressor diodes they have absolutely no effect on your tone and if you connected them to ground,well we just keep coming back to my point about the basics,Mitch.Monkeying in areas you dont fully understand will get you in trouble.You say you bypassed these diodes,if you did actually bypass them you put your ground potential at about 500vdc,not good for everything in your amp.

[MitchK]

THANK YOU! you may have just saved my amp. I have not run the amp with the bypass (didn't plan until on it until I got some feedback and i'm glad I did just that). I'll be taking that out now of course, and probably won't bother with it at all, I just was looking at things the wrong way and I see where I was going wrong now. I know I probably look like a fool but I don't particularly care, there is a reason I didn't turn it like that.

as for the original problem though, any thoughts? even if those resistors are gone I can't really see how they would create the problem. There are no power tubes installed, so the "dead end" of that wire just comes before the 100 ohm resistor rather than after, if that makes sense?

I just can't think of where to begin, why would the arc, the sparking, happen just right there on that one pin with no power tubes installed? I would think that means that this is where the short is happening, but it doesn't seem to arc to anything else just in that one spot? I understand it could be a manifestation of things happening elsewhere in the circuit but like I said I just don't know where to begin aside from taking it apart and re-doing it all.

[Steve Conner]

It sounds like you didn't leave enough clearance between (what used to be) the screen pin of the power tube socket, and some other grounded thing nearby. When you turned the power on, the high voltage on the screen pin arced to that other thing, and the resulting short to ground burnt a couple of your screen resistors out.

I suggest redoing the wiring between the original PCB and your new riser board, making sure that wire carrying the plate and screen voltages is proper stuff with 600V rated insulation, and that there are no loose strands of copper poking out at the ends where it's joined to the board. You need to see about 1/4" of clearance all around every high voltage lug on the sockets.

And let this be a warning to anyone else thinking of joining the Jack Kevorkian School Of Amp Modding!

[stokes]

Jack Kevorkian School Of Amp Modding!
Good one,Steve

[MitchK]

ok so good news, well you'll probably take it as bad news, but good news for me anyways, I just redid everything on that socket, turned the amp on and no snap crackle pop. SO my next take is to put some power tubes in and fire it up, the problem is the grids on the middle sockets are going to do nothing for me since their resistors have decided to stop working. My question is I should be alright to "fire it up" (with me this actually becomes a sort of tragic pun) for a test run with just two tubes in it, in the outer sockets where the resistors are ok, no?

[ecc83]

VALVE! amps

To misquote the great Basil Faulty, "perhaps we should let them all die"!

ecc83