[Whitigir]

Are you certain you cannot roll the T8000 pre amp tubes?

Nobody stop you

 

[kevin gilmore]

well i'm sure I can roll tubes, but I understand what is necessary to do so.

also the filament power supply is rated to 1 amp max, and some of the siemens Cca tubes for example would run right against that limit.

servo schematic posted. you really need 3 voltmeters to adjust this, but 4 is better.

if the gain of the tube is even 10% off spec, or the differential current is 5% off, that tube would not be able to be adjusted.

i'm sure mikahil would be willing to make you a universal adapter

 

[paulchiu]

ok. i have two Fluke 87V, maybe I will order 2 more before the SRM-T8000 arrive. Pretty sure the stock tubes are not the best.
Have you heard the unit Kevin?

well i'm sure I can roll tubes, but I understand what is necessary to do so.

also the filament power supply is rated to 1 amp max, and some of the siemens Cca tubes for example would run right against that limit.

servo schematic posted. you really need 3 voltmeters to adjust this, but 4 is better.

if the gain of the tube is even 10% off spec, or the differential current is 5% off, that tube would not be able to be adjusted.

i'm sure mikahil would be willing to make you a universal adapter

 

[JimL11]

Are you certain you cannot roll the T8000 pre amp tubes?

The 6DJ8s are not separate preamp tubes, they are the first stage devices of the amplifier, which is DC coupled from input to output. See KG's two posts, the two tube sections need to be closely matched for the servo to work. Of course you can do anything you want, but if you don't have the instrumentation to confirm section matching, don't be surprised if the servo goes out of range and you can't zero the output voltages. For some reason, amps don't work well when their output voltages are sitting at +/-350 volts. A Fluke 87 won't measure tube gain, which is what needs to be matched. Just pretend it's an all transistor amp - you don't roll transistors.

 

[kevin gilmore]

Yes for many hours see above

 

[Whitigir]

The 6DJ8s are not separate preamp tubes, they are the first stage devices of the amplifier, which is DC coupled from input to output. See KG's two posts, the two tube sections need to be closely matched for the servo to work. Of course you can do anything you want, but if you don't have the instrumentation to confirm section matching, don't be surprised if the servo goes out of range and you can't zero the output voltages. For some reason, amps don't work well when their output voltages are sitting at +/-350 volts. A Fluke 87 won't measure tube gain, which is what needs to be matched. Just pretend it's an all transistor amp - you don't roll transistors.

No-one stops anyone from rolling transistors though if you know and can and or have the meaning to do so

 

[JimL11]

ok. Pretty sure the stock tubes are not the best.

Yeah, but a less-than-the-best-sounding tube that allows the amp to function normally is going to sound a heck of a lot better than a "better-sounding" tube that causes the amp to malfunction. Do what the hell you want, if it works, great, if the amp blows up, don't know if Stax will cover the repairs, since you did something that you were warned not to do. Just saying....

 

[Whitigir]

Yeah, but a less-than-the-best-sounding tube that allows the amp to function normally is going to sound a heck of a lot better than a "better-sounding" tube that causes the amp to malfunction. Do what the hell you want, if it works, great, if the amp blows up, don't know if Stax will cover the repairs, since you did something that you were warned not to do. Just saying....

Yeah, that would be the exact scenarios

 

[VandyMan]

I recall someone mentioning earlier that changing the tubes voids the warranty.

 

[Whitigir]

Updated schem from KG

 

[paulchiu]

I recall someone mentioning earlier that changing the tubes voids the warranty.

You are going to have to change those tubes eventually. Typically 5000 hours, right? Rather inconvenient not user replaceable.

 

[VandyMan]

You are going to have to change those tubes eventually. Typically 5000 hours, right? Rather inconvenient not user replaceable.

found this:

[should confirm but I recall Suzuki-san saying Stax will take care of tubes servicing, user isn't apparently supposed to fiddle inside this amp (some other Stax amps have internal switches for instance).

 

[Whitigir]

You are going to have to change those tubes eventually. Typically 5000 hours, right? Rather inconvenient not user replaceable.

That is when and why you would need a dealer for

 

[kevin gilmore]

The tubes are run at 10% power. So they are going to last a long time. So it could be 10k hours. Longer than the electrolytics.

adjusting the servo, balance and offset is a tricky thing. Any one that has ever done this on a diy-T2 (or the original) knows exactly what i'm talking about.

spice simulations really do indicate that tube sections have to be within 5% at a per plate current of 3.75ma or the servo clips and when the servo clips, there is a differential voltage gain that is significant. plus substantial dc offset.

It would be really easy to build a small test circuit for measuring tubes, saves a lot of trouble. A 100V, -15V power supply, one current source (one transistor,2 resistors, 1 led) and 2 plate resistors and 2 voltmeters

of the 50+ tubes I tested (all pulls from Tektronix scopes et all) only 3 are within 5%
of the 20+ NOS tubes I tested, only 1 is within 5%

I understand why stax designed this unit this way. Lots of decisions relating to the size of the box, the amount of power dissipation, available Japanese semiconductors etc. My designs don't have any of these constraints.

 

[paulchiu]

The tubes are run at 10% power. So they are going to last a long time. So it could be 10k hours. Longer than the electrolytics.

adjusting the servo, balance and offset is a tricky thing. Any one that has ever done this on a diy-T2 (or the original) knows exactly what i'm talking about.

spice simulations really do indicate that tube sections have to be within 5% at a per plate current of 3.75ma or the servo clips and when the servo clips, there is a differential voltage gain that is significant. plus substantial dc offset.

It would be really easy to build a small test circuit for measuring tubes, saves a lot of trouble. A 100V, -15V power supply, one current source (one transistor,2 resistors, 1 led) and 2 plate resistors and 2 voltmeters

of the 50+ tubes I tested (all pulls from Tektronix scopes et all) only 3 are within 5%
of the 20+ NOS tubes I tested, only 1 is within 5%

I understand why stax designed this unit this way. Lots of decisions relating to the size of the box, the amount of power dissipation, available Japanese semiconductors etc. My designs don't have any of these constraints.

Great stuff. Thank you Kevin.