The Rocket build thus far

[Mark]

Got back on the Rocket build today, I thought I'd put those 1 ohm resistors on the cathodes of the EL-84's.

I still can't get over seeing 49mA's on the resistors with a 50 ohm cathode resistor. I did add resistance in series with the 50 ohm to see what I'd get.

First off I increased the 50 ohms by 10 ohms. I measured 44mA's. Next I replaced the 10 ohms with a 33 ohm resistor (83 ohms in total). This brought the current down to 37mA's. For the sake of my curiosity I replaced the 33 ohm resistor with 47 ohm resistor (97 ohms in total) the current through the tubes was now roughly 31mA's. Still a long way from 26mA's which is 70% plate dissipation.

I'll have to check my 18 watter and see was is happening there.

Any thoughts or suggestions?

[img766]http://i149.photobucket.com/albums/s69/ ... e456d7.jpg[/img]

[matt h]

(deleted)

[Mark]

Thanks for your reply Matt.

Assuming 91 ohms resistance and that your 31mA figure is per tube, this means your Vk is at/around 12V, right?.

Correct I have 12VDC at the top of the cathode resistor. The cathode resistance was 97 ohms, the rail voltage was in the order of 328VDC, and the cathode current was 31mA's. Which is 316VDV*0.124A=39.2 watts
These valves are rated at 12 watts a piece and 70% of that is 8.4 watts. Meaning my maximum should be 33.6 watts.


With the 50 ohm resistor I have a voltage rail of 315VDC, and I have 9VDc on the cathode resistor. The valves are running at 49mA's. This is roughly 60 watts which is serious.

http://www.webervst.com/tubes1/calcbias.htm

[Bob S]

Just my experiences with this.
I built a couple with 307-310 plate voltage & came to the same conclusions.
50 ohm shared cathode is hard on the tubes.
The Rocket, as designed by Ken, works fine if you pay close attention to the B+ voltages.
I read somewhere about him being very picky about the output tubes.
Made me think twice about using expensive nos tubes.
6p14p

[redshark]

Mark I don't think the 70% rule applies to cathode biasing!! That is a common practice with adjustable bias applying a negative voltage.
I think my rocket dissipates 12.5 watts per tube which exceeds the rating but don't forget Saratovs are rated @ 14 watts and not 12.
In case you are using NOS EL84's I think Ken prefered valvos for their longevity.
Yes, the rocket runs the tubes @ the limit but it sounds awesome!!

If you don't feel comfy about the 50 Ohms cathode resistor then use Saratov 6P14p's and forget about it!!

[matt h]

(deleted)

[Mark]

Hi Matt

In my experience the screen grid current will be in the order of uA when no signal is applied to the output tubes. I've found the screens only draw significant current once a signal is applied to the control grid. I will check though.

I think I'll need to do a bit more reading on all of this, and I'll check my 18 watter and see what it is drawing.

[martin manning]

Look on an EL84 data sheet for two tubes in push-pull. Zero signal g2 current is ~10% of anode current. With signal, it increases to ~25%. Typically plate dissipation is higher on cathode biased power tubes. If you use cathode current you can put it up around 100% knowing that it is really ca. 90%.

[Mark]

After reading the posts I realize I'm assuming too much in my posts, so I decided to put it all in a spreadsheet (which is attached). I hope this has all the relevant info, if not not please advise me.

Agreed Martin screen current was 10% of plate current roughly. It was much higher than I expected it to be. A lesson for builders making a new amp have the datasheet on hand.

The plate dissipation is still really high, between 10 and 12 watts. I'd need Saratov valves and a higher cathode resistor. I'm also wondering if the screen resistors should be more like 220 ohms instead of 100 ohms. Of course the next issue is will it dramatically effect the tone of the amp?

[JoeCon]

Mark

I've been running a 2 EL84 Rocket for over 6 months now at around 14.5w and it sounds great. JJ tubes Edcor xformers and Mark Fowler layout. I thought the voltages were too high but no problem so far. Go for it...

[randalp3000]

I've been gigging with my AC30 type amps at between 110-120% dissipation. 60ohm cathode resistor with 340v on the plates and I get between 45-47mA. I do run 1k's on the screens however to keep them from glowing when pushed hard.

That's the way trainwreck, vox, dr z, morgan, mojave, and komet all do it(50ohm cathode, 100ohm screens).

Let er rip.

[Mark]

Thanks for the replies. I can't help but feel the Rocket is a bit of a pipe bomb with the 80uF filter cap and the output stage running super hot. Granted that I never considered Vox amps to be unreliable, but then again I never owned one either.

I also find it funny all the various comments around saying the AC-30 isn't a class A amp, maybe it's not, but it is trying to be a class A amp.

The amp works okay but the longevity is the question in my mind. I looked at this thread below in the Vox section of this forum, it is the mods for the Brian May AC-30, the mods I find curious are the two 120 ohm cathode resistors (60 ohms in effect) are replaced with two 100 ohm resistors (50 ohms in effect). The 100 ohm screen grid resistors are replaced with 270 ohm resistors, one part of me says if it's good enough for Brian May.....

https://tubeamparchive.com/viewtopic.php?t=20221

I have heard an AC-30 with a 62 ohm resistor and 220 or 270 ohm screen resistors and it sounds fine. I'm told the purple box in the You Tube clip didn't add a lot of gain, it tightened up the sound of the amp somewhat.

http://www.youtube.com/watch?v=k0XOCGGsT8o

I will also read Kevin O'Connor's take on the AC-30, but when I questioned him about it some years ago, he did concede that he didn't build this AC-30. (That is how I remember the email at any rate.)

Still thinking about all of this.

[matt h]

(deleted)

[Mark]

Matt thanks for the reply, but I'm not too sure what you are trying to tell me, I'll try and break it down.

In short- bias resistors/bias voltage/static current don't exist in a nutshell. Supply voltage (and increase of sag through choke for plate supply or not counteracted by size of filter caps and/or induced voltages through choke) all create effects downstream of what the OT is doing. It's not quite fair to say "good enough for brian may!", it ends up being an apples/oranges comparison. (at least oranges/grapefruit, not even tangerines).

As far as the choke goes, I've heard Mark Huss say how he likes this circuit. I wouldn't think of a choke as a source of sag, if sag were to occur the choke will try and provide some regulation with the energy stored in it.

I suppose more importantly I don't know what that has to do with changing the 100 ohm screen resistors to 270 ohms?

This also requires dinking with the drive voltage to compensate if the bias voltage is changed. If the screen voltage is also dinked, you're also dealing with new curves.

I don't understand why I'd need to change the drive voltage to compensate for the bias. I agree the rail will increase with a higher cathode resistor, but the differences in voltages aren't all that great, I've updated the Excel spreadsheet to show the differences in voltage and current.

I don't know about the class A thing, I just haven't looked into it. Agreed the operating point of the valve will be the chief factor and the Z primary will be a part of that.

[Bob S]

That's some interesting information Mark.
Brings up all kinds of points.
From balanced PI to output tubes to Zpri matched halves.
Balancing 4 catthode biased output tubes.
I must re-read Merlin's section regarding balanced output.
Hmmm.
Thanks for sharing.