Tube Rolling w/Crack
- Thread starter Dr. Toobz
- Start date
I am using:
hd-600 - crack - marantz sa8001= not a lot of volume to play with, I suspect the marantz has a high output.
hd-600 - crack - dac - mac optical out = a little bit more volume play, but not much.
I have no ability to control volume from the marantz and don't really want to control the volume from the digital side. Padding sounds like the way to go.
Is it possible to tame the impact of the extra gain with a complimentary driver tube? In addition to the 5998, I have a mullard cv4003 on the way.
hd-600 - crack - marantz sa8001= not a lot of volume to play with, I suspect the marantz has a high output.
hd-600 - crack - dac - mac optical out = a little bit more volume play, but not much.
I have no ability to control volume from the marantz and don't really want to control the volume from the digital side. Padding sounds like the way to go.
Is it possible to tame the impact of the extra gain with a complimentary driver tube? In addition to the 5998, I have a mullard cv4003 on the way.
Tubes are amplification devices. Tube design and application control the gain. 12Au7's are medium MU (gain) devices along with the other compatible tubes for the Crack. So there's not much you can do with the tubes. I've noticed the 12BH7 and E80CC tubes have a little more gain than a ECC82/12AU7 but that's not going to help you. I haven't tried padding the volume pot and I'm not sure if there are any negative impacts. Hopefully someone will chime in with more or better recommendations.
Yes, I understand that a tube is designed for amplification, and thanks for the input on the 12bh7 and E80CC. Having the extra gain, I would stay away from those until I pad the pot. I get to full volume at about 830-9 o'clock with the stock 6080 and 12au7 - (but I'll take my questions on padding to another thread).
Im running a 12BH7 and I thought they were actually lower "gain" than 12AU7 (around 17 versus 20, from memory) although I cant recall noticing a whole lot of difference. Maybe it's the way the drivers are used in the circuit. One of my sources is a Tjoeb 99 CDP which supposedly has an output in the 3.5 volt range. Im not so sure it's that high but it's definitely higher than my 2 volt digital source. My main phones are Beyers though which have a lower sensitivity than the Senns. My AKG 601's are lower still. My Quarts are higher than the Beyers and dont get a ton of use but I have enough play in the pot with those as well. I think most of the volume control range issues discussed on the forum have been with Senn phones. I assume that's because their higher sensitivity but it could just be their popularity, or a combination of the two. Then again, listening level has a major impact as well as, not just the source, but the source material. My CD's and LP's can be all over the place. But I tend to listen louder than I probably should at times 
Anyway, below is a link to a pre-attenuation method on the Goldpoint site. It is only 2 resistors and signal passes directly through only one of them. I would have no problem with using it if needed. Wouldnt cost an arm and a leg to implement using quality resistors. Of course, you need 4 resistors, 2 per channel. Anyway, resistor values listed are based on the value of the pot and desired amount of pre-attenuation/padding. You could experiment with cheap resistors to find out how much you need and then swap in some better resistors. Vishay RN's would probably be a good, budget concious choice but the sky is the limit just like anything audio ;D
http://www.goldpt.com/mods.html
Anyway, below is a link to a pre-attenuation method on the Goldpoint site. It is only 2 resistors and signal passes directly through only one of them. I would have no problem with using it if needed. Wouldnt cost an arm and a leg to implement using quality resistors. Of course, you need 4 resistors, 2 per channel. Anyway, resistor values listed are based on the value of the pot and desired amount of pre-attenuation/padding. You could experiment with cheap resistors to find out how much you need and then swap in some better resistors. Vishay RN's would probably be a good, budget concious choice but the sky is the limit just like anything audio ;D
http://www.goldpt.com/mods.html
Paul Joppa
Moderator
The output tube is used as a cathode follower, so the gain of that stage is always a bit less than 1.0. Into a high impedance, the 6AS7 might have a fain of 0.65, vs. 0.8 or so for a 5998. For lower impedance phones, it's a little more complicated but in all cases the difference in gain between tubes will be small, maybe 1 or 2 dB.
Thanks for the helpful link. I was able to put in a 6as7g and I actually find it more tolerable at the same volume levels than the 6080. We'll see about the 5998, I'm definitely interested in putting the pad as suggested. Since I read electric schematics at a first grade level (the lines are wires right?), I'll give it a shot, but will post to make sure my implementation is proper.
Sh7eleven said:
Me too, in terms of schematic reading skills. This ones easy though. Resitor 1 (RP1) just goes inline between the signal wire from the RCA input jack and the pots input. Resistor 2 (RP2) would solder from the input of the pot to the pots ground (you can just twist one lead of each resistor together and that connects to the pot input ... it's the same, electrically, as connected those 2 leads seperately).
For testing, you can actually just clip some inexpensive resistors in temporarily with some small alligator clips (Radio Shack) to determine the correct amount of attenuation. You can twist one of the leads of resistor 1 to one lead of resistor 2 (maybe apply a small amount of solder to keep them together). Clip the signal input wire on the free lead of resistor one. Clip the two resistors twisted/soldered leads to input of the pot. Then clip the free lead of resistor 2 to the pots ground. Just do this for each channels input, left and right.
This is what I did when I was experimenting with the attenuation just in case I needed it. I did it with an unmounted pot and took measurments just to get an idea of how it worked. You are attenuating the input signal some so you do lose some overall, maximum "volume" compared to no attenuation. But if you arent using anywhere near the full range of the pot, you will never miss it. Also, you dont need to find the exact resistor values. Just for a rough example, if you needed a 46.6 K resistor and could only find a 47K, that would be plenty close enough.
I only have an hour or two on it at this point but I did like what I heard. I immediately noticed the increased clarity that an acoustic guitar came through and the bass seemed very focused. I had a cv4003 arrive yesterday. I'm eager to listen to the pair, but I don't want to make too many changes too fast. Otherwise I have a tough time sorting through how a new piece changes the sound.
Sh7eleven said:
Agreed, spend some time with the 5998 then add the CV4003. I dont have great ears but the differences between 5998 and 6AS7 or 6080 are pretty easy to distingush. For me, the differences between 12AU7's are more subtle. I really have to spend a good amount of listening time with a particular 12AU7 before trying another one. So you are definitely on the right track making slower changes. Glad you like the 5998 so far. Really nice tubes.
Interesting observation to report.....tried out a 6922 with about 400 hours on it, and the sound is very different than the burnt-in 6CG7! I can't decide which one to keep. The 6922 apparently has extremely low distortion at that operating point, or at least compared to other tubes I've tried as drivers. The sound is very modern, crisp, clean, and has great decay and space around the notes. The only downside is a more forward soundstage. On the other hand, the 6CG7 sounds more like the "tube" stereotype. It has richer harmonics and is noticeably less clean and detailed, but wins in the depth department with a slightly larger soundstage.
What is the operating point for the 6922 that you're using?
Running properly, the 6H30Pi would be a fun one to try. With a second speedball PCB, the 6BX7 is an interesting choice also, though you'd need to shove 20ma of current through it with 5v on the cathodes for it to work well.
Running properly, the 6H30Pi would be a fun one to try. With a second speedball PCB, the 6BX7 is an interesting choice also, though you'd need to shove 20ma of current through it with 5v on the cathodes for it to work well.
I'm using the regular operating point, wired for the different pinout. This seems to be the same one as used by the Seduction, or at least the plate voltage (75V) and use of LED bias (via Agilent HLMP-6000) are similar. I'm also using the Speedball C4S to load the driver tube.
Going back and forth, I still can't decide which to chose, as the sound is almost as different as solid state versus tube. One is clean and unveiled, the other is warm and lush. A happy medium may be keep the 6922 and slip the 6080 out of the cathode follower position in favor of a warmer-sounding tube, like the Russian 6N13S.
Going back and forth, I still can't decide which to chose, as the sound is almost as different as solid state versus tube. One is clean and unveiled, the other is warm and lush. A happy medium may be keep the 6922 and slip the 6080 out of the cathode follower position in favor of a warmer-sounding tube, like the Russian 6N13S.
I'm another Crack/Speedball user firmly in the Tung Sol 5998/WE 421A camp. As Laudanum pointed out, they are not the same tube. If the data sheets can be believed, the WE 421A's amplification factor is a bit higher than the Tung Sol 5998's, and so is its transconductance. They are close, however, and for many listeners both sound better than the 6AS7G/6080, no matter who manufactured the latter. These tubes' basic electrical and operating characteristics genuinely make a significant difference in the Crack's sound. I have Tung Sol and Cetron 7236s as well, and they fall pretty much in the middle: Better than the 6AS7/6080, not as good as the 5998 or 421A. And looking at the 7236's data sheet, its MU and transconductance are lower than the 5998/421A, but significantly higher than the 6AS7/6080. The 5998 and 421A definitely impress me as bringing out the most detail in a musically necessary, not analytical or dry, way. From what I've picked up in this forum and elsewhere, I believe it's probably the effect the 5998 and the 421A have on the Crack's output impedance. These tubes lower the circuit's stated 120 ohms enough to reap damping benefits, especially with the 600 ohm DT 880 phones that I use with my amp. It's a shame that a kind of modern "tulip mania" has made so many NOS tubes--the WE 421A, the Tung Sol 5998, and now even the 7236--so expensive.
I am also firmly in the 5998/WE 421A camp and have not used anything else in a long time [Crack & Speedball with DC heated tubes]. I personally really disliked the 7236 in this configuration and with all headphones that I use.
Your mention of "tulip mania" now including the 7236 is great news for me. I have a quite a few and tube prices always seem to be depressed during the summer months. Late fall early winter might bring some windfall resale profits for me on those ... and for that matter make other people happy.
Your mention of "tulip mania" now including the 7236 is great news for me. I have a quite a few and tube prices always seem to be depressed during the summer months. Late fall early winter might bring some windfall resale profits for me on those ... and for that matter make other people happy.
Supply and demand, I guess, the 7236 gaining more and more attention as 421As disappear, 5998s become ever rarer and more expensive, and amps like the Crack become more popular. Seeing 7226s selling for between $40 and $60 apiece recently really impressed me with how fast tube price acceleration can happen: I bought a beautiful PAIR of NOS 7236s just a year ago for about $30. Five years ago, I bought NOS 421As for about $85 apiece, NOS 5998s for $25. Makes me wonder what's going to happen when the 421A and 5998 supply is effectively exhausted. I don't see anyone ever producing modern versions. Well, maybe SOMEBODY, but they'll cost $500/pr. (The boxes, however, will be really nice. 
)
)Paul Joppa
Moderator
The demand will die down when they become unavailable or too overpriced; then a few years later the hoarders who missed the bubble will let them go for cheap because nobody wants them anymore. Seen it before. At least with tulip bulbs, you can eat them!fipple8 said:
