Been toying with the 100W amp (above) primarily in the output and NFB areas. The OT I had was an ultralinear Hammond (1600 series.) Being that it is my first build, I decided to hook it up as described on the box, in UL mode. Worked great, sounded great. Being that I am never satisfied with something that works great or sounds great, I decided to add a DPDT switch to switch output section from UL to pentode mode (tapping the HT supply right after the caps & choke) Needless to say, it opened the amp up tonally speaking quite a bit... for about five minutes until it burned one of the tubes! When I got in it, I found that the plate resistor on the blown tube had drifted from 50 to 4.5k, and another tube (in opposite phase section from blown tube) had a completely open plate resistor. anyone have any thoughts on what could have caused this? My competing theories are: blown tube causes resistor(s) to fail vs. bad resistor causes tube to fail vs idiot should have left a perfectly fine sounding amp aloneI'd go with the third one were it not for the massive, albeit temporary, improvement in tone.

I replaced the faulty resistors and tube and it seems to hold up to all sorts of sine waves very well taking the screen voltage from the PS. (also seems to clip on a scope a bit differently than it did in the UL config) I took out the switch, as I am leaning towards the non UL sound anyhow. I was switching it in SB, but in retrospect I am sure the caps were nowhere near discharged.Will plug a guitar in and see if i can fuck it up again.thanks-

As far as I understand, there is no reason you can't run an amp in either ultralinear or pentode mode. This is probably a dumb question, but did you power the amp off before throwing the switch? The transformer could generate a big flyback voltage if any of the taps were in series with a switch that went open. In theory, I think that could blow some shit up.

madraso\_ wrote:"presence" is added. Maybe some kind of HF/NFB thing going on as you (Ben) suggest... rewired it to UL and it is fine. sounds great. dont really want to fuck with it anymore, but appreciate any thoughts on this? thanks.Your avatar suggests that you own a scope... can you throw that on and check for oscillations?

yeah. it fucked up again. seems that the amp functions fine on the bench, but plug a guitar into it and then plug it into a cab and it shits itself rather quickly -resistor flash and then blown fuse- maybe it objects to my playing.It seems to work with the presence control at 0, then shits the bed when enough presence is added. Maybe some kind of HF/NFB thing going on as you (Ben) suggest... rewired it to UL and it is fine. sounds great. dont really want to fuck with it anymore, but appreciate any thoughts on this? thanks.

Madraso,The only other thing I can think of would be that the plate dissipation in pentode mode in an overdriven condition is potentially much higher than for ultralinear mode. If the output stage is designed to run near the limits of the device in UL mode, you could be pushing it too far in pentode mode.It may be helpful to figure out what is going on if you drive it with a sinewave of a particular amplitude (you mentioned you can get away with that without blowing it up) in both pentode and ultralinear mode, and note the difference in output amplitude. You can use this to estimate the plate resistance for each case. This will allow you to figure out the plate dissipation (you need to know the impedance presented to the output by the OT, as well).

Maybe posting some circuit values or a schematic would help.

Today's job: fix hum on a pair of Hammond nuvistor condenser mics.Re-capped PSU. Hum remains even when PSU is switched off. I should not that these mics are unbalanced output (circuit is cathode-follower type for 50 Ohm impedance). Considered a possble ground-loop, so tried wiring output as hot to pin 2, cold to pin 3, and ground floating at one end to try to stop the loop. Still hums a bit using my Motu interface (on transformerless preamp inputs, there is still a path to ground from hot and cold).Eventually bit the bullet and fitted two isolation trasformers in the PSU. Also removed captive unbalanced output cables and fitted XLR sockets. You can see it's a bit cramped, but the transformers sit on the left, externally, and are mumetal-shielded. In case you are wondering what the valve in the top-right is, it's a ~200V valve regulator. It's filled with gas and glows purple. Scary.

How have you implemented the switching from UL to pentode? Screen resistors on the output tube sockets that are connected to a switch that selects either the UL tap or the voltage from a PS supply node that's been dropped from the plate voltage via a choke or resistor or what?Are you sure the speakers you're using are okay? Are you sure the output tubes are okay? Are you sure the plate voltage it's supplying when in pentode mode is within spec for the tube type you're using?Pentode mode sounds way better than ultralinear, for guitar anyhow. Don't give up on pentode mode. If anything, tape up the UL taps and take the switches out, and make it pentode only. It'll sound better.

Scott-Damn you for reminding me how much better it sounded few minutes that it was working in pentode mode. Now I feel compelled to dig some more!!!the switch was doing exactly that, a DPDT that looks like a pair of these:--- UL tap from OT--- screen resistor (1.5k 5W)--- node directly after chokespeakers are fine, tried two different cabs just to make sure. tubes ... well it cooked a perfectly fine kt88 with no red-plate warning the first time i tried it so now I am using an old set that does work, and are well within tolerance in their parameters. A voltages are in the 520's, less than the max for a JJ kt88. plate dissipation at bias is set to just around 28W. Rodabod-the output section setup like a vintage Sunn 100W with 4 kt88's running into a Hammond 1650 series 100W UL OT (my voltages tend to run a shade higher than SUNNs.) I can post a schematic, or a link to what I am working off of. Since this is my first attempt at an output section, I tried not to stray too far at the beginning. http://sunn.ampage.org/site/schematics/modelt.gifBen-I will have to put it on the scope again and see what it does in pentode. I had it on there before, and did not see anything out of the ordinary but I am not always the most observant...Megathor-This might be something, although there is no red-plate warning to indicate too much dissipation, it just goes. I had set the bias for 28W and thought that should be fine. I can try a bit less and see if that does it. I like your idea of comparison- thanks!and in general, thanks guys! Will update after a few more fuses blow;)