Best tubes for the Rocket

[2tone]

When Ken wrote the article the old 84 was still available and maybe in production, Groove tubes used them (unlabeled) and so did New Sensor (unlabeled). I bought the box of 50 from New Sensor, unlabeled, as did Ken, and after that factory closed (likely Saratov?, later the Reflector factory I think made them..As long as Groove tubes had them, ken usually ordered from them because they were "matched" and had a guarantee for replacement...rarely needed with these tubes..

[redshark]

I'm not an expert on that tube. I know it started in the late 70's and the variations.

I told you everything I know. 2tone also added more information.

I was buying those tubes and got all the different versions and I still have them but then I started hunting for the Hammonds.
Many people started selling those without testing, cheap and I got lucky. I bought a bunch and tested them all, matched them and as I said I have only used one quartet since 2010.

[rooster]

Alright, my closing comments. First off, the getter has changed in all Sovtek tubes. It is now a halo getter vs. the solid dimpled getter that started in the very earliest Russian production of the EL84. This halo getter first appeared in 2007. If you can find a Russian EL84 dated prior to 2007 you will find it has the dimpled getter. Good luck.

Which is to say that my labeled 6P14P/6n14n EL84s from 2011/2012, and my New Sensor labeled EL84s BOTH have the halo getter. In other words I am doubtful there is any difference in the construction of these two versions. There is mention of the 'gold plated grids' in the 6P14P/6n14n but, hm, I'm dubious.

Does this new getter mean that the Russian tube is now lost its mojo? Good question and I have no answer today. My two quads are both very quiet in the amp and the either quad seems to be handling the 350 plate volts, no problem. I will make some recordings of the two sets this first week in April, Scout's Honor.

BUT..... Something occurs to me regarding the comments from 2tone (via Ken), redshark, articles in VG mag by Ken, and my own experience with the Rocket/AC30TB circuit. Q: If the mids of the amp is such a concern/focus that we are spending our time and $$ looking for EL84s that create the perfect midrange 'storm', why don't we just add an adjustable Mid control to the circuit? If one tube offers a 'sweeter midrange' vs. another, with a Mid control could we dial in the tube that is lacking this quality? Alright, skip to the chase, I have added this to my build. It is a10K Linear pot that has a fixed 4.6K resistor attached from it to ground. Which means the least I can turn the control (altering the tone stack 'Slope') is 4.6K at the 'OFF position (half of the factory 10K value), at 12 o'clock is the factory 10K, and the max I can reach is 15K (5K over the factory value). Of course there is a 10K lineal sweep form 4.6K to 15K. Does this get us somewhere in the ballpark of the elusive tone that the AC30TB and Ken's Rocket are known for? I'll record this in my sound samples, too. You can be the judge.

[Mark]

Just a quick question while talking about Rocket tubes and tone. I found the rail voltage does have a bearing on the tone of the amp.

What voltage is on the centre tap of the output transformer of your Rocket?

I found the EL-84's are still running hard at my Rocket clones rail voltage of 315VDC. I have been thinking of using a Zener to drop the rail down to 300VDC, this also drops the preamp rail which can result in DC voltage on the guitars volume control, thus scratchy sounding pot.

I have Holland Amperex 12AX7's in my Rocket clone. I find they offer the least coloration. I did have a long plate RCA 12AX7 and I found that amp sounded fatter when pushed.

I thought the valve rectifier was quite important to the amp too.

[rooster]

Mark, yes, I agree the plate voltage is a consideration. In my 350 plate volt AC30TB build, the cathode cap is not 50 ohms but instead 56 ohms. One quad (branded 6P14P) is seeing approx 44mA, and the other (branded New Sensor) 46mA. I do appreciate all voltage concerns in relation to ultimate tone, but I have had maybe 6 vintage AC30s over here loaded with the correct speakers, 350 VDC on the plates, using a 50 ohm cathode cap, and using a variety of power tubes - and they all sounded really good. Knowing that the bias point - if it is too low/cold - will bring out the Class AB characteristics, I started high with my cathode resistor (65 ohms) and eventually brought it down by paralleling a 10 watt resistor of various values until it sounded OK. Which is to say I don't think there is an ultimate plate voltage sweet spot for an EL84 tube. I think instead that a cathode cap that mates to a plate voltage between 315-350VDC and the particular tube's bias needs is where I'm at on the subject. Of course, these Russian tubes I am buying on eBay don't really give me tubes with the same bias point so I am blurring my idea of perfection in regards to bias. Since I am not going to change out the bias resistor for every set of tubes, I comfort myself with the sounds I heard in these vintage Vox amps. They were not perfectly biased and still they sounded really good, all very similar, given the same speaker compliment. Haha, those magic speakers!

You wrote, "What voltage is on the centre tap of the output transformer of your Rocket?" Did you mean PT? I'm confused, but I don't have a fused center tap on my PT if this was your question so I can't measure this quickly. I think I am missing something.

BTW, in my AC30, I dropped the preamp tube voltages down to Ken's preference by using non-stock dropping resistors. This voltage is pretty key I think. I can measure those if you want, for reference. Oh, also I should note that on the 2nd gain stage I used a switch to either add or remove the 25uf cathode cap but I never use it because I think it overpowers the circuit. I would also point out that many Rocket owners here really like that cap in play.... I don't get it. I also use a 220K load to ground resistor off the PI input cap when I build Rockets for customers because it mimics the Normal channel load - and this again reduces the preamp gain. It is also worth noting that Ken created an AC30TB schematic that found it's way into the GT Tube book and in this schematic Ken properly shows the shared (TB channel and the Normal channel) cathode resistor for V1 as 1.5K/25uf. If you were to actually build the Rocket and wanted to duplicate the AC30TB channel, then you would not use the 25uf cap on the 2nd gain stage (just the 1.5K cathode resistor), and change the 1st stage cathode to (approximately) 2.7K/25uf. All of which I find really interesting when you consider how much Ken loved the AC30TB amps. I somehow feel that he made the AC30TB schematic after he created the Rocket because 2tone tells the story of how Ken contacted him after he took delivery of his Rocket and suggested a switched in/out 220K resistor to the PI to duplicate the Normal channel.

I actually wonder if there aren't a couple of different Rocket circuits out there? I will never know, of course.

[Mark]

If I were you I would look for a 25 watt variable axial ceramic resistor and use that in lieu of the 50 ohm resistor. I have a couple of 70 ohm resistors for AC-30 builds. What is Ken's recommended pre-amp rail voltages?

[urlhttps://www.amazon.in/Variable-Resistor-Wire-Wo ... B009017SGY][/url]

[2tone]

Alright, my closing comments. First off, the getter has changed in all Sovtek tubes. It is now a halo getter vs. the solid dimpled getter that started in the very earliest Russian production of the EL84. This halo getter first appeared in 2007. If you can find a Russian EL84 dated prior to 2007 you will find it has the dimpled getter. Good luck.

Which is to say that my labeled 6P14P/6n14n EL84s from 2011/2012, and my New Sensor labeled EL84s BOITH have the halo getter. In other words I am doubtful there is any difference in the construction of these two versions. There is mention of the 'gold plated grids' in the 6P14P/6n14n but, hm, I'm dubious.

Does this new getter mean that the Russian tube is now lost its mojo? Good question and I have no answer today. My two quads are both very quiet in the amp and the either quad seems to be handling the 350 plate volts, no problem. I will make some recordings of the two sets this first week in April, Scout's Honor.

BUT..... Something occurs to me regarding the comments from 2tone (via Ken), redshark, articles in VG mag by Ken, and my own experience with the Rocket/AC30TB circuit. Q: If the mids of the amp is such a concern/focus that we are spending our time and $$ looking for EL84s that create the perfect midrange 'storm', why don't we just add an adjustable Mid control to the circuit? If one tube offers a 'sweeter midrange' vs. another, with a Mid control could we dial in the tube that is lacking this quality? Alright, skip to the chase, I have added this to my build. It is a10K Linear pot that has a fixed 4.6K resistor attached from it to ground. Which means the least I can turn the control (altering the tone stack 'Slope') is 4.6K at the 'OFF position (half of the factory 10K value), at 12 o'clock is the factory 10K, and the max I can reach is 15K (5K over the factory value). Of course there is a 10K lineal sweep form 4.6K to 15K. Does this get us somewhere in the ballpark of the elusive tone that the AC30TB and Ken's Rocket are known for? I'll record this in my sound samples, too. You can be the judge.

yes I think around 2007 is about right. I did notice the getter difference, and then the sound was different too from that point on. The old ones on ebay etc are the ones to get..

[rooster]

Mark, I have made an assumption here based on the documented builds of Rockets in the early days, however I will add that I was always under the assumption that Ken was shooting for 90VDC on V1 but I am very sure that the AC30s that he played and loved were probably around 120VDC. (Haha, ask me if I ever measured the voltages on those AC30s that came through here, no, I did not.) V1 in my AC30TB is 85VDC/.65VDC (Plate target was 90-120). V2 (V1B in the Rocket) is 118/.9 (Plate target was 95-100). V2B (V2B in Rocket, as well) is 191/119 (Plate target was 185-195/118-120). You can see I am low on V1 according to the charted voltages but the amp sounds good so it stays for now. I may have to rethink this V1 voltage because I purchased a BrownBox voltage convertor and frankly, it doesn't help the AC30 too much. It could well be that I end up dropping the V1 voltage too low. This hadn't occurred to me till right now... My PI plates are 181 and 183. My tubes are Amperex Holland I63, V1 and V2, and my PI/V3 tube is an RCA small plate from 1980. I use all .047 Panasonic 630VDC caps where the .047 cap is called for, compliments of a redshark trade.

[ I realize now that my voltage to V1 dropped when I added the Ef86 channel. All three channels share the same supply point. I have subsequently changed my power supply such that V1 is now 90/.66. And the amp sounds better, tighter. Thanks, Mark! If you hadn't asked me about my voltages I would never have revisited this. ]

I will also add that yesterday I purchased an NOS quad of the 6P14P/6n14n El84s dated 1993. These seem legit, arriving from Russia in a couple of weeks. Not known if they are even in the park as to being matched but oh well. I will record these along with the other two quads.

As to the adjustable resistor, damn, that is some pretty downhome construction there, Mark,but maybe worth a try. It is a shame there is not a bullet proof and easy to use adjustable cathode resistor, something that can be approached from the top of the chassis.

2tone, thanks for the confirmation. Hm. let's see how my sound bites sound to you.

[Mark]

I have found you need a fairly dramatic change in cathode resistor values to effect the bias of the output valves. To have the current at 70% of plate dissipation, I needed to double the cathode resistor value to 100 ohms. I believe some of the earlier AC-30 amps used a 80 ohm cathode resistor.

[rooster]

Mark, of the six that I saw, they all had a 50 ohm resistor. I can only assume that this was owed to the tube quality of the time of manufacture, the tube manufacturer's suggested bias point for Class A, and then finally the resultant sound. Using a 100 ohm resistor - with any quad of EL84s I can think of with plates in the 300-340VDC range - I can be sure this amp is into Class AB very quickly. This is not the sound that drew Ken to the AC30TB and not the sound these amps are appreciated for. But it took me a while to appreciate this, I confess.

The all tube Thomas Organ Vox Cambridge Reverb, the 18 watt amp, engineered by the Americans, was built with a 120 ohm cathode resistor with 260VDC plate volts. Ken liked the Berkeley (2/10" speaker version) and said he thought it sounded as good as the 18 watt Marshall. (So here's a reference to a 60 ohm (120 divided by 2) cathode resistor on the AC30 when tubes were still quality). BTW, the Marshall 18 watt used a 100 ohm cathode resistor and saw 310VDC on the plates, quite a different experience. I've played both and although reverb is a magnet of sorts, I preferred the original Marshall.

Well you've probably been down all these roads so I would bore you or try to be an advocate of the 50 ohm resistor. And for the record, I don't use a 50 ohm resistor, as I stated above. My only concern is the Class A vs. Class AB response and sound of the circuit.

[jtaylor996]

I've tried lots of things in my rocket (made by @rawnster), but here is what I like:

Amperex holland 12AX7A, in every position.
Mullard GZ34. I never thought that would make a difference, but by god it does.
6P14P still kicking 6 years later.

I need to go order some backups for all of these, not sure where to get a matched quad of the 6P14P anymore.

[2tone]

Mark, of the six that I saw, they all had a 50 ohm resistor. I can only assume that this was owed to the tube quality of the time of manufacture, the tube manufacturer's suggested bias point for Class A, and then finally the resultant sound. Using a 100 ohm resistor - with any quad of EL84s I can think of with plates in the 300-340VDC range - I can be sure this amp is into Class AB very quickly. This is not the sound that drew Ken to the AC30TB and not the sound these amps are appreciated for. But it took me a while to appreciate this, I confess.

The all tube Thomas Organ Vox Cambridge Reverb, the 18 watt amp, engineered by the Americans, was built with a 120 ohm cathode resistor with 260VDC plate volts. Ken liked the Berkeley (2/10" speaker version) and said he thought it sounded as good as the 18 watt Marshall. (So here's a reference to a 60 ohm (120 divided by 2) cathode resistor on the AC30 when tubes were still quality). BTW, the Marshall 18 watt used a 100 ohm cathode resistor and saw 310VDC on the plates, quite a different experience. I've played both and although reverb is a magnet of sorts, I preferred the original Marshall.

Well you've probably been down all these roads so I would bore you or try to be an advocate of the 50 ohm resistor. And for the record, I don't use a 50 ohm resistor, as I stated above. My only concern is the Class A vs. Class AB response and sound of the circuit.

When I switched my Rocket to high voltage (it was the only one I know of that had 2 voltage taps on the PT), I had to go from 50 (stock value) to 68 ohms to avoid red plating the 84's. Resultant sound was great, no loss in class A tone, but tighter bottom, bolder mids...

[rooster]

2tone, I have to ask, what was the high voltage you ended up with on the plates?

[2tone]

2tone, I have to ask, what was the high voltage you ended up with on the plates?

I ended up a little over 350!!

[2tone]

also my V1 volts went up to around 104, from the 90',s