Tube Rolling w/Crack

[nick-seattle]

This is a man who found a way to get whatever tube he found into the Crack.  Congratulations are in order.

I wish that would be the case. The only tubes I can use in the amp are all limited by gain up to 60 work fine, but anything above doesn't work, so no 12ax7 or 5751 rolling (although Sylvania JAN 5751 worked, but JHS 5751 wouldn't  )

[Doc B.]

Are you checking the voltages with each of these different tube types? The nature of the direct coupled circuit is such that the bias on some tubes might be less than optimal unless some other changes are made.

[nick-seattle]

Are you checking the voltages with each of these different tube types? The nature of the direct coupled circuit is such that the bias on some tubes might be less than optimal unless some other changes are made.

Hi Doc, I'm in a monkey mode right now - I see and I do, but have no idea how to calculate things. It's been suggested to I should replace LED with Silicon Diode, but I have no idea what value that diode should be or if there are other changes I should make O_o

[Doc B.]

In general I think one needs to be careful when publishing tube rolling reviews where you plug a tube in that was not intended to be used in the circuit. A 12AU7 and 12AX7 are not really interchangeable. Yes, they have the same pinout - you can put Vanagon wheels on an SLK, too -  but they are designed for different jobs. I'm not saying don't try it, and in most cases no harm will be done. But I think it is important to note in the review that a given tube might not be operating at it's best without some changes to the circuit (or that it might not really be the best choice for the job at hand) and thus it might not be getting the fairest representation.

With this in mind I'll suggest that with the proper changes to the tube socket and heater voltage and current requirements a 6SN7 will typically work in place of a 12AU7. A 12BH7 will too, if one makes sure there is enough additional heater current available. As one gets further away from the 12AU7 spec, for example 12AT7s, 12AV7s, 12AZ7s and 12AX7s, one may need to alter the circuit anywhere from mildly to fairly radically. Also be careful with 6DJ8s, they aren't designed for as much plate voltage as a 12A_7 family tube and thus they may be subjected to too high a plate to cathode potential in some circuits.

[nick-seattle]

Understand - putting a tube that the amp is not designed for is not a fair test. Yah. I've been using my other 3 amps to do the rolling with 12at7 & 12ax7 (McIntosh 275) and 12at7 & 12au7 (300B based SET amp & my hybrid tube amp). Using Crack to roll only 12au7 family

[nick-seattle]

And after publicly stating that I don't use 12at7 in Crack... I just did Crack was the closes amp and I needed to try a new tube - it worked extremely well. I then double checked the results in other amps with very similar outcome.

Check it out - this one is out of this world:
http://tubemaze.info/sylvania-12at7-gray-plates-2-mica/

If you find it - buy it. If you don't, let me know and I'll buy it!

[Laudanum]

I'm using Svetlanas. Really like the airiness they produce and mid base tightness, but at the same time maintain warmth. Or maybe because I'm from Russia and was born the same year they were born

I haven't tried many 6as7 tubes to be honest. Structurally the majority of them look similar, so why bother. However, WE, Mullard, Tung-sol are on my list to try. I haven't had much luck rolling power tubes - they produce practically no sonic difference, so my investment is in drivers - so many to try. If you go to tubemaze.info, you will find all my rolling experiences between 4 tube amps - about 60 tubes and counting. Submit your review of Mullard 6080 there - let's make it fun.

Im really surprised that you dont hear a difference between power tubes.  I dont have a whole lot of different ones and I agree on one point that I think I mentioned ... I cant hear much between the GE 6AS7 and the RCA 6AS7.  I have 3 or 4 of the RCA's and, I think 2 of the GE's one with copper shield, one without.  All ST shape tubes.  The GE's have a smaller bottle but the internals are very similiar.   There are subtle differences that I wouldnt swear on hearing and I highly doubt that I could I pick between these two in a blind test.   But that aside,  between those 6AS7's I mention and the Raytheon 6080's that I have as well as the Svetlana 6H13C and the Tungsol 5998 there are distinct differences between most of them.  Now to be fair, the 5998 and probably the 6H13C are slightly different specs wise.  But the differences between these tubes I find to be more distinct than between most of the different 12AU7's.  Give your drivers some longer listening time and then try a different one, I would be surprised if you couldnt hear differences.   It would only be 3 tubes but I'd bet I could pick between the 6H13C, RCA 6AS7 and TS 5998 in a blind test ... and probably the Raytheon 6080 as well.  I cant say the same for most of the 12au7's I have,  those are mostly a "feel" thing.  And I absolutely dont have golden ears.   Give it a try with the power tubes.

[nick-seattle]

Im really surprised that you dont hear a difference between power tubes.  I dont have a whole lot of different ones and I agree on one point that I think I mentioned ... I cant hear much between the GE 6AS7 and the RCA 6AS7.  I have 3 or 4 of the RCA's and, I think 2 of the GE's one with copper shield, one without.  All ST shape tubes.  The GE's have a smaller bottle but the internals are very similiar.   There are subtle differences that I wouldnt swear on hearing and I highly doubt that I could I pick between these two in a blind test.   But that aside,  between those 6AS7's I mention and the Raytheon 6080's that I have as well as the Svetlana 6H13C and the Tungsol 5998 there are distinct differences between most of them.  Now to be fair, the 5998 and probably the 6H13C are slightly different specs wise.  But the differences between these tubes I find to be more distinct than between most of the different 12AU7's.  Give your drivers some longer listening time and then try a different one, I would be surprised if you couldnt hear differences.   It would only be 3 tubes but I'd bet I could pick between the 6H13C, RCA 6AS7 and TS 5998 in a blind test ... and probably the Raytheon 6080 as well.  I cant say the same for most of the 12au7's I have,  those are mostly a "feel" thing.  And I absolutely dont have golden ears.   Give it a try with the power tubes.

I definitely did hear a difference, but it's not as huge as input or driver tubes. they only one I didn't hear a difference are RCA & Raytheon 6as7 - I bet they were produced in the same place. Really want to get my hands on 5998, but it has to wait right now - they are going for at least $50 a tube for a low measuring tube. It looks like WE 421 & Tung-sol 5998 have the same plate structure, so I would assume that they sound the same - any comments?

[nick-seattle]

Dan's warning about using other tube numbers that are not 12AU7s in Crack is for listening and proper performance.  But if you want to test a tube and it has the right heater arrangement you might find out if it makes sound.  I once used a 12AX7 in place of a 6DJ8 because I didn't have a spare when my 6DJ8 went out.  It didn't sound right but did OK for background music for a few hours.

Dan is warning that you will not get the intended performance, gain, output and distortion using different tubes. 

I do understand the physics that are involved in rolling different tubes - mechanical & electrical But 12at7 does work in Crack and works fairly well especially at low volume. If I'm not mistaken there should be enough bias to allow for a small voltage swings with 12at7. However, I tested the tube at full volume and it still sounded phenomenally - so no matter what physics mighty state, my ears state that this is one of the best tubes that I've ever rolled in Crack (and I have zero interest in it, just want to share my OPINION) and I've tried over 50 tubes (different brands and models of 12au7 & 5814 & 5963 & 6189 ... ) - I'm collecting unusual tubes, so I have all sorts of crazy tubes from the family. 

[Doc B.]

Did you measure the voltages?

[Paul Joppa]

"If it sounds good, it IS good" remains true, and I can't argue with it.

But drawing conclusions is a different issue. For example, a 12AT7 will have a higher plate voltage, driving the 6AS7 to a higher cathode voltage, and (without the Speedball) higher current. Our experiments have shown that increasing the current in the 6AS7 improves the sound - so ascribing the improvement to the 12AT7 by itself would be inappropriate. (With Speedball the current is constant so this effect is less significant.)

The same arrangement would increase the dissipation of the cathode resistor beyond the design condition. With the Speedball, it's the transistor dissipation that would increase. Either way, the components may have a shorter lifetime.

Both of the above are caused by the direct coupled design - sounds good but it's just less robust than cap-coupled topologies.

[Laudanum]

The WE 421 is/is not a Tung-Sol 5998 debate will always rage on amongst very small groups of hobbyists.   Do a google search,  you will find tidbits of info from both sides of the debate.  I dont have a WE and have no plans to spend the money on one.  I'll just have to slum it with the 5998 and two spares that I do have.   By the way, from everything I have found, all of the 5998's are Tung-sol made, no matter what they are branded.   That's not the straight glass 5998A's (which are generally considered a lesser tube, sonically).  I dont know if the A's were made by more than one manufacturer, but the ST shaped 5998 was only made by Tung-sol ... again, as far as I know. 

[Dr. Toobz]

For what it's worth, I did put a 12AT7 in the driver socket one time, just out of curiosity, as I was trying to wrap my mind around the electrical workings of a direct-coupled circuit. I don't have the voltages I measured handy, but I recall the plate voltage being definitely higher than usual, and that was with the standard LED biasing the cathode. I have the Speedball installed, so that probably put the brakes on any large changes. Needless to say, the resulting combination sounded like poo - very low volume and an anemic sound. I take it that the Speedball was doing its job by preventing large shifts in the circuit - so who knows what would happen with the standard resistive arrangement? However, I can't help but feel sticking a totally different tube in there isn't ideal. The furthest I've gone from the standard 12AU7 is an E80cc, and now a 12BH7, both of which are much closer in terms of plate impedance and mu when compared to something like a 12AT7 or 6922. I'm not even sure how the latter would work in this circuit! Wouldn't the high transconductance and low plate impedance result in altered behavior at the grid of the 6AS7?

[nick-seattle]

I have the basic version of Crack (no upgrades whatsoever). 12at7 works and sounds on the level of 12au7 (with Sylvania 12at7wa, it killed the majority of 12au7s except maybe for Mazda - I couldn't hear any sonic differences between these two tubes), but with much more volume. I know it's not suppressed to sound good - but it does (in Crack, in McIntosh, in SET 300b) - and I have experience in what a good sound is supposed to be - I tried all major brands of 12at7, 12au7 & 12ax7 - from Telefunkens selects - at $200 a tube to GE gray plates at $1 a tube)

Not sure about voltages anywhere, but it's a tube, if it works - it will continue working - compared to transistors. I don't see myself upgrading to anything that would involve transistors in Crack - I understand that they might make things better, but it's a transistor I love my Creack - pure!

Tube rolling wise, here are my favorites in Crack: Mazda 12au7, Sylvania 12at7wa, RCA 5962 (red base), Sulvania JHS 5814, Amperex 12at7 france, Amperex 7062, Westinghouse 12au7, Amperex 12au7, Westinghouse 6gu7. Some produce incredible clarity and sound stage, some produce superb warmth - depending on the mood

[Beefy]

I don't see myself upgrading to anything that would involve transistors in Crack - I understand that they might make things better, but it's a transistor I love my Creack - pure!

I could not disagree more with this position. Tubes must have a load and in the vast majority of cases, including the Crack, a CCS provides a *much* better load on the tube than a simple resistor. You get better PSRR, lower distortion, higher gain...... and from every single report I have read (plus my own experience) the Speedball just plain sounds better.

How is it pure to ignore these benefits? Does it make you a "better" tube guy to forego better sound quality by deliberately avoiding transistors?