6550 tube bias characters

[speakerfritz]

so I have been testing 6550 tubes. here's what I found so far. Russian 6550 and kt-88's run cooler than asian and other regioal tubes. Apparently sovtek WA's and A's have a cooler operating range than say the WE branded ones. I have about 50 6550's all in all. the WA's come in at about an Ip of 27 to 31, the A's about a 31, sevelanta's come in about 30 to 31. I test the Ip's using a bias probe on my mercury 2000 tester. I think the tester plate voltage is only 250V. Interestingly, after I grade a tube by Ip, and move them into my amps (500 to 600V plate voltage) they come out pretty close on bias range. Eample, my amps recommend 30 to 35 ma's of bias . If my tube tester bias probe method grades the tube at 33ma ot higher, I can't use that tube in my amps....they won't bias down to the 30/35ma range. Finding tubes that can bias in my amps between 30and 35ma's isn't easy. KT90's will work, only sevelenta KT-88's will work, valve art and other chinese kt-88's wont, EH el-34's will.

 

[kevinkr]

Svetlana?

 

[dgta]

That simply means that you need a higher (negative) voltage range in your bias supply (assuming your amp has fixed bias). Most designs are capable of simple modification to extend that range a bit. Post the schematic and I'm sure someone can help you.

 

[speakerfritz]

That simply means that you need a higher (negative) voltage range in your bias supply (assuming your amp has fixed bias). Most designs are capable of simple modification to extend that range a bit. Post the schematic and I'm sure someone can help you.

Thats correct...a modification to the bias supply via one of a few resisters. I was prepared to undertake this but with schematic in one hand, and the amps exposed on the table....was not able to locate the resistors in question. either the color coding had faded, or someone came in and changed the resistors . The next time I try to hunt down the resistors in question, will probably focus more on finding the bias supply by voltage. The amp uses a PCB with traces on both sides, so it could be the case that the main bias resistors are on the back side unviewable with out removing the PCB.

 

[dgta]

You DON'T want to locate the resistors by color code alone. There could be other resistors of the same value in that area. Nor by voltage.

It's best to locate components by actual connections, i.e. start from the grid pin, follow the wire or PCB trace to the connection of the coupling cap and bias resistor, etc. THEN when you're reasonably sure you found the right resistor, verify value. If it's the wrong value, double check that you traced correctly.

 

[speakerfritz]

You DON'T want to locate the resistors by color code alone. There could be other resistors of the same value in that area. Nor by voltage.

It's best to locate components by actual connections, i.e. start from the grid pin, follow the wire or PCB trace to the connection of the coupling cap and bias resistor, etc. THEN when you're reasonably sure you found the right resistor, verify value. If it's the wrong value, double check that you traced correctly.

Thats what my plan was....double sided PCB's made that a dead end. I would hate to remove the PCB, it would involve removing both trannies, both power supply caps, and the wires to 3 selectors and 8 bias trim pots are very short and very thin.....it could get hairy.

I just bought a box of 25 6550 tubes which were claimed to be new. got the tubes and they were in fact all new. tssted and matched them and out of 25, got 18 that fall between an Ip of 27 and 31. which is good for my bias pot ranges. an Ip of 33 on my test equipment would bias a tube in my amps with the bias pots turned all the way down to 37.ma. the recommended bias range for my amps is 35ma.