With known dead tubes, I don't see the point. Why not just break
the envelope and take a full color photo of the damaged goods?
Nobody rebuilds tubes (except Klystrons and the like), not even
the WE300B. If its that precious you can't break it, I don't want
to be held responsible anyway.
If the knowing is worthy of the sacrifice, and genuinely need to see
something that can't be determined by breaking in, fine. Just don't
hold me responsible for what happens... I don't need nobody mad
at me over something gets lost or damaged. Assume its a goner.
the envelope and take a full color photo of the damaged goods?
Nobody rebuilds tubes (except Klystrons and the like), not even
the WE300B. If its that precious you can't break it, I don't want
to be held responsible anyway.
If the knowing is worthy of the sacrifice, and genuinely need to see
something that can't be determined by breaking in, fine. Just don't
hold me responsible for what happens... I don't need nobody mad
at me over something gets lost or damaged. Assume its a goner.
so in the final analysis
So lets suppose the circuit I am using is specifying a 300 Ohm Cathode resistor for each tube is it better from a longevity point of view to run the kt88 given that the svetlanas for instance, do differ in their ratings(the kt88 having higher plate dissipation and max Gscreen Volts).
Can the KT88's run happily at the settings for the 6550. Screen Volts are 300Vwhilst plate volts would be in the region of 420 to 450V?
Thats the vexing question and feel free to pick holes in my reasoning.
Thanks
Nick
So lets suppose the circuit I am using is specifying a 300 Ohm Cathode resistor for each tube is it better from a longevity point of view to run the kt88 given that the svetlanas for instance, do differ in their ratings(the kt88 having higher plate dissipation and max Gscreen Volts).
Can the KT88's run happily at the settings for the 6550. Screen Volts are 300Vwhilst plate volts would be in the region of 420 to 450V?
Thats the vexing question and feel free to pick holes in my reasoning.
Thanks
Nick
I'm running 6550 TungSols at +450V on both anodes and screens, i.e UL. One of the reasons my tubes last so long is I build in a 3rd order HP anti blocking filter to the amp input
(i.e -3dB at 15Hz). This avoids dangerous input surges which can damage screen grids and their resistors below o/p tranny cutoff.
This may explain why my 807's still function well after 50 yrs stage service.
richj
(i.e -3dB at 15Hz). This avoids dangerous input surges which can damage screen grids and their resistors below o/p tranny cutoff.
This may explain why my 807's still function well after 50 yrs stage service.
richj
SpreadSpectrum said:
It would be no suprise if some KT88's are cloned 6550's.
Many are around with identical looking plate structures, and to cap it, the distortion spectrum on both are similiar, except the JJ KT88 version which is considerably higher.
Which leads me to suspect the JJ KT88 version is actually a sweep tube design from what clone ?
I will dig out the 3rd order filter.
richj
I've tried both New Sensor re-issues, the TS 6550 and GEC KT-88 in my Simple SE. They look identical internally, the only visible difference being the shape of the glass envelope. The GECs may have some differences in metallurgy, but sound quality is fairly close. I prefer the 6550s for a more neutral, uncolored sound (although the coloration of the GECs was very pleasant) and superior soundstage (more holographic, 3D presentation), and better bass response. You can feel the bass in your body more with the 6550s....
The Simple SE uses 220k grid leak resistors, do I need to chage this to 100k for running the TS 6550s? Its been running for around 400-500 hours with no issues so far.
The Simple SE uses 220k grid leak resistors, do I need to chage this to 100k for running the TS 6550s? Its been running for around 400-500 hours with no issues so far.
You'll get away with high values ye lower B+, but when you pile on the power with heat, a high value grid leak can end up with a runaway red anode, i.e grid going towards + and leakage not in control.
If paralleled p-p pairs, one must observe the grid leak value per tube. Sorry driver has to sweat more.
richj
I have a stupid question.
In cathode bias, it is very clear what the grid leak resistance is by looking at the grid leak resistor. But how to calculate that in fixed bias?
Is the grid leak resistance = sum(resistance from the grid to ground)? For example, in the classic Dynaco st-70, it will be 270K + (bias pot (somewhere in 0 - 10k) + 10K) / 2 (2 tubes) ?
Thanks,
In cathode bias, it is very clear what the grid leak resistance is by looking at the grid leak resistor. But how to calculate that in fixed bias?
Is the grid leak resistance = sum(resistance from the grid to ground)? For example, in the classic Dynaco st-70, it will be 270K + (bias pot (somewhere in 0 - 10k) + 10K) / 2 (2 tubes) ?
Thanks,
pchw said:
Generally the bias pot will be a small percentage of the total resistance and can be ignored, if it is not take the thevenin equivalent resistance of the mid point of the pot plus the thevenin equivalent of any adjoining resistors feeding the bias pot and add them to the grid resistance. In most cases if you are talking about total thevenin equivalent resistances of less than 10% or so you can just ignore them.
I just got one of my amplifiers in for service from a client, this one is currently equipped with KT88EH that were installed 8 months ago, they all matched closely at that time and now they do not. Total hours probably in the range of 400hrs.. Plate supply is 525V, plate and screen current combined is 55mA per tube. The amp came in with a blown fusible resistor on one of the screens - I've had amplifiers of the same basic design run on a set of EI KT90 for 4 yrs without these sorts of problems. (although I did have 1 that exploded pretty spectacularly early on)
I've had problems with the JJ KT88 as well, and so far of the tubes I have used those with the best long term track record have been Svetlana 6550C.. (8yrs and counting)
What's your recent experience and what would any of you recommend for this amplifier?
I still support some of my old products, but this amplifier is 15yrs old, has had a lot of use, and has been completely rebuilt twice in that time. (It's had 3 owners) Now I just want it to work for a while..
I've had problems with the JJ KT88 as well, and so far of the tubes I have used those with the best long term track record have been Svetlana 6550C.. (8yrs and counting)
What's your recent experience and what would any of you recommend for this amplifier?
I still support some of my old products, but this amplifier is 15yrs old, has had a lot of use, and has been completely rebuilt twice in that time. (It's had 3 owners) Now I just want it to work for a while..
kevinkr said:
This is a common problem. Most tube testers cannot test at 500V so all vendors test at 250V B+. Some will go higher when asked to do so. Any slight discrepancies will highlight at 500V with the errors magnified.
A tube when put in a well designed circuit should hold Ik within a few % over it's life time. My TS 6550 and Sv KT88's do exactly this.
Mains/heater variations is just one off a palette of variants that can greatly effect emission and B+. Take your measurements in steady conditions.
richj
How many hours of life time do you estimate for SVKT88 for 25-30W idle dissipation? The JJ KT 88 got 3 amplifiers to blow along them when they grid leaked big way at over 2000 hrs. And some manufacturers keep insisting on large grid resistor values and auto bias just because it worked on their bench. Just to drive with 12AX7 and the like. The TS 6550 reissue and GL KT88 reissue are the best I have used by the way. The TS just has softer-darker tone. The GL is best overall. And I do include vacuum and manufacturing quality first, in saying this.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.