Tube Rolling w/Crack

[Paul Birkeland]

Yeah, you can read up about interpreting load lines.  The lowest distortion for a triode will be a load line of infinite impedance (C4S), while acceptable performance generally starts at around 3x plate impedance.  You can also look at the load line you've drawn and examine each chunk of the load line that has been sliced up by the curves on the graph.  If the segments are pretty equal in length, especially close to your quiescent operating point, then you'll probably be OK.  If the segments get really short on one end and really long on the other, you will hear that. 

[xcoolhandx]

6GU7 is a NO-GO,someone mention running "6GU7"on crack along with E80CC...all em I getting is some "hum"..waist of time

[Loon]

Pin assignment is different

[xcoolhandx]

Pin assignment is different

thanks,hope this helps
http://www.ebay.com/itm/201083377532

[Loon]

Looks promising.  Keep us updated.

[rlyach]

I just received my Tung Sol / Chatham 5998 and couldn't believe the difference from either the RCA or Svetlana 6AS7G. I then looked closely at the cage structure of each tube and noted a very significant difference. In the 5998 the plate is much longer than the cathode and grid, which will collect electrons that are even emitted on the sides of the cathode. Further, the 5998 topology will also shield one triode from the other, reducing cross talk. The 6AS7Gs I had have the cathode and grid about the same length as the plate. I assume this was to facilitate uniform electron travel, but this also allows some (probably small but also probably not insignificant) cross talk between the triodes. Is this correct? I guess I was struck at the significant difference between the tubes. I have drawn a representation of the two tubes for reference. I hope I got it right. Anyway, this difference may help explain why the 5998 sounds so much better than the RCA 6AS7G configuration.

[Paul Birkeland]

You could play a strong 60Hz tone into the left channel of your Crack and measure the 60Hz output on the right channel with the 6080 and 6AS7 to get a feel for whether or not there is any crosstalk difference between the two. 

FWIW, the spacing is pretty tight between all the elements in one half of a 6080 compared to the distance between the two structures themselves. 

[rlyach]

PB,

I used audacity to create a 60 hz tone on the right channel and played it through my DAC which has the capability to reach +4dbu (pro audio 0VU). Then I sent it to the amp and listened with a pair of headphones. I definitely heard the tone on the quite channel. I then used tape to isolate the ring (right channel) on the TRS plug to hear just the cross talk. The signal was quite a bit louder with the RCA 6AS7G than the 5998, although it was still present with the 5998. I am not sure where the cross talk is coming from but in order to hear it you have to have the amp up at 80% volume with with a +4dbu signal. At -10dbV I can't hear anything with either tube. That is where the amp is spec'ed to run. Now I will look for an oscilloscope to actually do some more measurements. One more thing, I noticed that the 5998 was stronger then the 6AS7G that I have. All in all I like my current combination of the 1958 French RT 12AU7 long plate (balanced) and the 5998. I am really enjoying this amp.

P.S. You are correct on the spacing. The plate separation on the 6AS7G is about 2.5mm for each triode, and about 6mm between the triodes.

[Paul Birkeland]

You should be able to get an AC voltage reading with your meter for the channel that's supposed to have the 60Hz and the channel that isn't.  This should give you a good starting point for actually measuring the crosstalk.

Using the scope will let you use something like a 1Khz tone, which will be quite distinct from 60Hz noise introduced through the mains.  (IE better accuracy)

[rlyach]

I don't have my multimeter here in my office so I can't do the measurements yet, but I do need to make a correction. I traced down the cross talk to the signal source and not the amp. When I generated the tone on a stereo channel I left the other channel with no data. That was a big mistake. I repeated my initial testing by generating a 60 hz tone on the right channel and silence on the left channel. Now I get no audible crosstalk with either tube even with the volume all the way up. My Tesla T1s have an SPL of 102db so it appears that if there is any crosstalk difference between the tubes because of their respective topologies, it is moot. This seems more reasonable to me.

[junkimchi]

What is the proper procedure to follow when rolling a tube? Really don't want some of my tubes to be blown out.

[Doc B.]

Pretty simple, just be sure the amp is off when you swap tubes, and stay with the types that are recommended for the circuit.

[Exaliftin]

I am sensitive to sibilance on a lot of recordings and was wondering if there are any tubes that can help smooth out any harsh sounds. I have the crack speedball with modi dac. Thanks

[JamieMcC]

I am sensitive to sibilance on a lot of recordings and was wondering if there are any tubes that can help smooth out any harsh sounds. I have the crack speedball with modi dac. Thanks

Mullards in either one or both tubes positions would be worth a try. Both are nice sounding tube also 

[Joshewah]

I found that the RCA clear top was smooth and less silibant than the new production tube my kit came with. It's an affordable tube too, can usually be found lightly used for about $10 online.