Analisys of power amp limitations of tube bias range

[speakerfritz]

Have a few MB-350 amps that have individual bias running kt-90's or 6550's. Amp seems to do well with kt-90 tubes that have Ip numbers between 28 and 31. Tubes with Ips over 32 need full max out bias pot position to fall between the recommended 30 to 35 ma bias setting. Tubes with Ips of 33 or more, can't be dialed in. Once you try a tube with an Ip in the 35 range B+ fuse blows if I let the amp variac up to full power and let the bias go past 37ma's. Have three amps and they all do the same thing. On 6550's no problem using them if i follow the same IP rules. I think the original design called for 6550's and the option to kt-90 was an after thought. 6550's Ip a lot lower than kt-88's and kt-90's. KT-88's work as well if the Ip rules are followed. Problem is when buying new tubes, its not a guaruntee that you can get tubes the Ip between 28 and 31. So, I'm wondering what can be done to increase the range of usable tubes. I have a few dozen new tubes with Ips over 33. Here is the schematic for one of the tube bias circuts as well as the bias power supply section. Thinking of putting in a pot in parallel to the 27K bias supply resister which would let me decrease the negative voltage which should reduce bias current. So I would use the new pot as a board range adjust, then fine tune using existing individual tube bias pots. To prevent something spooky from happening, would probally put a resister in line with the pot I want to put in parallel to the 27K resistor. comments / recomendations?

[HollowState]

Originally Posted by speakerfritz

Thinking of putting in a pot in parallel to the 27K bias supply resister which would let me decrease the negative voltage which should reduce bias current.

Totally backwards as I see it. You want to increase negative bias voltage in order to reduce tube idle current. You can do this by either reducing the 27K ohm resistor in the supply (first picture) or increasing the 27K ohm resistor on the low end of the trim pot (second picture). Or do a little of both in order to get the trim adjust pot near the center of it's range when things are stable. You could also increase the pot value to perhaps 25K ohms which would raise the voltage somewhat because the supply would be less loaded.

[speakerfritz]

Originally Posted by HollowState

Totally backwards as I see it. You want to increase negative bias voltage in order to reduce tube idle current. You can do this by either reducing the 27K ohm resistor in the supply (first picture) or increasing the 27K ohm resistor on the low end of the trim pot (second picture). Or do a little of both in order to get the trim adjust pot near the center of it's range when things are stable. You could also increase the pot value to perhaps 25K ohms which would raise the voltage somewhat because the supply would be less loaded.

thanks for the input. def, first picture would be better so that the change applies to all the tubes. second picture is on a per tube basis and the pots in the second picture are small (about the dia of a pencil eraser, PIA ot desolder and replace). So reduce the 27k resistor to 25K, adding a 400K resister to the existing 27K resister via alligator clips would let me test that. That would net 25.2K

I took a look at a few newer models. I'm attaching some bias circuts from other models in the line and they took the approach of replacing the last resister with a 100k pot.

[HollowState]

Originally Posted by speakerfritz

So reduce the 27k resistor to 25K, adding a 400K resister to the existing 27K resister via alligator clips would let me test that. That would net 25.2K

Reducing 27K to 25K may not be enough. But it's a starting point. You may find that 20K or even 15K is necessary. You'll just have to experiment. Whatever is best to keep the adjustment pots within range with some reserve left. And yes, system wide is better then each tube. I hate those tiny pots myself. They use them because they're cheap!

[speakerfritz]

Originally Posted by HollowState

Reducing 27K to 25K may not be enough. But it's a starting point. You may find that 20K or even 15K is necessary. You'll just have to experiment. Whatever is best to keep the adjustment pots within range with some reserve left. And yes, system wide is better then each tube. I hate those tiny pots myself. They use them because they're cheap!

10/4, planning on trying a few different values than one I find one that will get me mid way on the tube bias bots, will solder it in.

Talked ot manly today and they confirmed that the existing values were to tune in 6550's.

[speakerfritz]

wow...could not find the 27k resistor. This thing has 5 rectifiers circuts along with the associated caps and resisters. You would think once the negative bias caps were found, it would be a walk thru the park. The negative bias caps are the only caps that are wired positve terminals to ground. It was easy to find them. But finding the 640 ohm and 27K resistor that hangs off them was an no go. They must be on the back side of the PCB or I think the resisters were previously replaced and thats why I can not find them by color code or direct measurements. Looks like I am going to have to stick with tubes that have an Ip of 29 +/- 2ma's. bummer....I have 32 KT-88's and 24 6550's that have an Ip of 32 or more.