Tube Rolling of Schiit Saga Pre-amplifier

[gvl]

Maybe they were bad.

Many tubes in Darkvoice hum, what's strange the hum often tends to go away with tube burn-in. Yes, it's actually a thing.

 

[SIY]

Hum isn't a sign of tube wear. In some circuits, it originates from the heater, so some tubes will hum and some don't. There's ways in circuit design to greatly reduce this variability, but who knows if such steps were taken here. My guess is no.

 

[Snafu]

Just my own experience: i have Copland CTA301 mkII pre amp (i have had it 15 years or so) and when E88CC/6922 tubes wear out first signs are hum and sharp buzz (not loud but you can hear it next to speaker). When replaced with similar tubes this hum/buzz goes away so it is some how related to tube wear (how exactly, i don't know). With new tubes amp is as dead silent as good tube amps usually are.

I replaced E88CC's with 6SN7-variations (with adapter) and haven't had worn out/failing tube yet. CTA301 is known to be harsh for lower quality / longevity tubes that can't take abuse. But obviously tube<>amp match is different with different tubes and amps.

cheers

 

[SIY]

Generally, when a tube ages, gain goes down and overall noise can increase. This, of course, assumes a competent circuit.

I’m not clear on why you’d want to substitute a 6SN7 for an ECC88. The former has lower mu, lower gm, and higher rp, as well as needing to be run at significantly higher voltages. A typical ECC88 circuit will have 90v on the plate, whereas a 6SN7 will need to run at least twice that for linearity and gain.

 

[Snafu]

Pre amp uses E88CC/6922 tubes, slightly different than ECC88. You can swap E88CC to ECC88 but not other way around. CTA301 is designed for E88CC.

Main reason was, as usual, "because i can and they look awesome" but i also wanted to solve amps tendency to wear out E88CC's. Amp was checked by manufacturers service and everything was fine, except it ate 6922's... So i tried 6SN7's and it worked very well. imo sound got better but floor noise definitely went down a bit. Everything works now fine, sound is good, low noise floor and no need to replace tubes every 2-3 months.

Also, 6SN7's are cheaper than "quality nos" 6922/E88CC's. Nowadays rather buy newly made guaranteed 6SN7 variants than "nos"-tubes, after all these "nos" tubes have tens of years of unknown history measured by ebay seller... And volume knob is for gain differences

Not sure if this makes technically perfect sense but as i said it works.

 

[SIY]

The reality is that lots of the same tubes have been labeled 6922, ECC88, and E88CC. Officially, there’s some differences in heater current, but in reality, it’s a complete crap shoot.

If you’re really lucky, you’ll stumble into some genuine CCa. If you’re unlucky, you’ll get some 6ES8 that have been relabeled...

 

[Snafu]

^ Yes, true about labeling and classifying them correctly. Main reason i moved on from 6922/variants was that it was really difficult to get reliable tubes no matter how they where labeled or how much they cost.

It kind of pisses you off when you have nos set of Philips superduper$$$ tubes and they go bad in a few months... happened more than once.
There might be something wrong with my Copland's autobias but i have heard that this amp is harsh for tubes and it's more like feature than fault...

But, at the moment, all good and with tubes "if it works don't fix it" is golden rule

 

[trl]

[...] Somewhere else I read about an integrated amp to have selector for high or low NFB/damping factor, low giving more distortion and warmth. This would be a nice feature, just like with Schiit Saga pure/tube. Just for fun!

Not sure how many realized that John Atkinson also did measurements on this amp: "Fig.6 was taken with 100% global feedback; the actual distortion is below the noise floor until the power reaches 50W or so. Repeating the 8 and 4 ohm tests with 0% feedback gave the same clipping power, but with about 20 times higher distortion below clipping (figs.8 & 9)" - source: https://www.stereophile.com/content/ch-precision-i1-universal-integrated-amplifier-measurements.

100% feedback

vs.

0% feedback

Now that explains why NFB devices from AudioGB just don't measure so very well. Now why would I want to pay money on a NFB device if I can get something cheaper and much better measuring, and of course...I can add later some free VST plugins and create my own distortions.

 

[rebbiputzmaker]

The reality is that lots of the same tubes have been labeled 6922, ECC88, and E88CC. Officially, there’s some differences in heater current, but in reality, it’s a complete crap shoot.

If you’re really lucky, you’ll stumble into some genuine CCa. If you’re unlucky, you’ll get some 6ES8 that have been relabeled...

You have to be able to understand the tube codes to know, yes labels are pretty meaningless. Ultimately if you know the construction you will know what is real vs re-screened junk.