The Rocket build thus far

[Tillydog]

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.

A choke will oppose rapid changes in the current through it, so if there is a sudden demand from the power tubes, the voltage after the choke will drop for a time before recovering to the level determined by the PT characteristics & resistances in the circuit.

I don't understand why I'd need to change the drive voltage to compensate for the bias.

From the data in your spreadsheet, the no signal grid voltages decrease by ~22% between the 97R cathode resistor and the 50R cathode resistor. Heavy distortion from the output stage happens when the drive exceeds the grid voltage (which won't be static, as we are dealing with cathode bias, but roughly...) so with the 50R cathode resistor, the preamp only needs to put out 78% of the voltage it would with the 97R cathode resistor for heavy power stage distortion. (Actually, the effect will be greater than that, as the higher cathode resistor will cause more local NFB.) The character of the preamp will be different, and the balance of distortion between pre and power sections will be different unless you scale the drive voltages acordingly.

[Mark]

Tillydog, thanks for your reply and input, I appreciate you taking time to help me.

From the data in your spreadsheet, the no signal grid voltages decrease by ~22% between the 97R cathode resistor and the 50R cathode resistor. Heavy distortion from the output stage happens when the drive exceeds the grid voltage (which won't be static, as we are dealing with cathode bias, but roughly...) so with the 50R cathode resistor, the preamp only needs to put out 78% of the voltage it would with the 97R cathode resistor for heavy power stage distortion. (Actually, the effect will be greater than that, as the higher cathode resistor will cause more local NFB.) The character of the preamp will be different, and the balance of distortion between pre and power sections will be different unless you scale the drive voltages acordingly.

I probably shouldn't have included the info on the DC present upon the control grids as I haven't eliminated the possiblity of this being caused but one pair of tubes.

I'm not too sure what you mean by the drive exceeding the grid voltage.

I assumed you mean the signal voltage on the control grids exceeding the cathode voltage. I read up on Kevin O'Connor's chapter on AC-30's, he mentions the cathode voltage rises by 1VDC despite the cathode bypass cap bypassing the audio signal. Thus he implies there is no significant negative feedback present in the output stage.

As the difference between the two resistors is three volts, I don't see there being any great difference in perceived headroom. Of course I could be wrong.

I stopped and had a think about my 18 watter. It uses a 125 ohm resistor and a shared 100 ohm resistor. In an AC-30 situation this amounts to a 200 ohm screen resistor and a 62 ohm cathode resistor. Though that may not applicable to this situation.

randalp3000, what prompted you to go with the 60 ohm cathode resistor and the 1K screen resistor?

I ask as these are the values used by Nik in the Rock-it and favored by Kevin O'Conner.

Again thank you to everyone who has replied to my post, I appreciate your assistance.

I wonder if the vintage values in the AC-30 are still valid given the quality of valves we have available to us today?

[randalp3000]

I lied, I have two 110R's in parallel so I'm really running 55 ohms.

I noticed on a couple amps that when pushed the screens would start to glow pretty bright. So after some reading around here I settled on the 1k's. Seems to be working well.

Tubes are a crapshoot, I still like JJ's the best. I picked up some 6p14p's but I like the bite of the jj's more but they don't last as long. Some will draw a lot more current than others also. I have a MaxiMatcher which most of the tube dealers use to match output tubes. I need to find some time to measure some in the amp and tester to find the #'s I like so I know what to order.

[Tillydog]

I assumed you mean the signal voltage on the control grids exceeding the cathode voltage.

Yes, that's what I meant.

I read up on Kevin O'Connor's chapter on AC-30's, he mentions the cathode voltage rises by 1VDC despite the cathode bypass cap bypassing the audio signal. Thus he implies there is no significant negative feedback present in the output stage.

The bypass cap probably does prevent NFB as such, but the effect that I was clumsily trying to describe is the fact that the cathode voltage rises under signal conditions, and will rise more for a bigger cathode resistor. (I haven't read KOC's books...)

As the difference between the two resistors is three volts, I don't see there being any great difference in perceived headroom. Of course I could be wrong.

It's 3 volts in ~11 = 22% which will equate to at least an equivalent reduction in headroom.

[Tillydog]

Hi Mark,

To get back to your question (I think! - I hope you don't mind me paraphrasing)

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.

[60 ohms]. I measured 44mA's....
[83 ohms in total]... 37mA's...
[97 ohms in total]... 31mA's.

Still a long way from 26mA's which is 70% plate dissipation.

...Any thoughts or suggestions?

Firstly, I don't think the Liverpool is designed to idle at 70% dissipation. There's a comment in the Rocket Voltage Chart spreadsheet that implies this is the case which I think is misleading, and may well be the root of your question.

Looking at what KOC wrote about the AC30 (thanks ) it seems 200mA total current with a 50R cathode resistor is fairly normal (but not necessarily healthy). He puts the transition from class A to B at something like 22W peak output, which, frankly, I'd ignore and treat it as a class A amplifier and look for bias of ~100% at idle (but I've never built one).

Looking again at your spreadsheet with the 50R resistor, you're calculating an anode dissipation of between 10.9 and 12.5 W (11.9W average), so you're pretty damn close to the 100% right there. Your B+ is a little high, I think at 305V vs 290V designed which will have pushed the dissipation up a little.

If it was me, I'd probably settle on a 60 ohms (or a little more) cathode resistor to bring things back from the brink a little.

With the 97R cathode resistor, things look quite cool (average 8.3W anode dissipation).

In relation to the choke and sag question, I would agree with KOC - since the amp is almost class A, there is little change in current draw with increasing signal, so power supply sag (whether from the choke, or elsewhere) will be minimal.

Did you ever get rid of the noise?

[rp]

w/o rereading the post from the start, it's my understanding that 100% dissipation is fine with cathode bias as long as the plate voltage isn't too high. Seems to me KF was a fan of lower voltage higher current in his builds. I'm guessing, but I don't think he cared much if his amps ate tubes somewhat faster, he was building Ferraris, not F-150s. Fisher wasn't selling 100/1000s of amps a year and so not hedging on the side of reliability and warranty concerns. If your amp flared out he'd likely apologize and repair it free. He also had access to quality old tubes which he spent time rolling before consigning so each was specifically tubed for tone and function, and would personally advise the client on care and feeding and expectations. You knew what you were getting.

Check out tube eating machines like old Matchlesses - high voltage and high current draw. Lots of unsupervised AC30s out there from the 115V days running Matchless style, 340V plates and 50mA. Not good at all, but they're out there (or were ) I think part of the original Matchless design philosophy is that they really liked the sound of stressed over-baked Voxes and built a line of amps around it.

[Bob S]

I don't want to derail the excellent technical information being shared here.
I roadied a bit in the 70's.
Most guitar slingers back then loved 60's AC30's
My main recollection about them was how hot they ran.
Lots of snap, crackle & pop. And smoke.
Ken nailed the top boost channel with the Rocket.
And pushed the output tubes hard.
Output tubes are a consumable.
Tone is not an option.

[martin manning]

Average all four valves in the Rk = 50-ohm case is just about 100% anode dissipation, and the worst is 104%, so they are not way out. If you don't see red plating they are probably ok, otherwise increase RK a bit as Matt suggested. You could also swap valves 1 and 4 to balance the two sides of the output a little better. JJ's are very popular EL84's.

[Mark]

The 6p14 tubes arrived and I put them in the amp. Of course it still runs as hot as hell but the valves are within specs (just).

The screen grids are dissipating 2 watts and I believe this will double with signal applied (according to a past post).

Does any one know the screen dissipation for the 6P14 series tubes. I'm using the EB types, but even so I thought the screens were only good for 2.2watts.

Thus a 220 or 270 screen resistor is starting to look like a good idea.

[Mark]

I started a new post on the Rocket screen grid resistors.

https://tubeamparchive.com/viewtopic.ph ... 024#312024

[Mark]

After some noodling I got back into building the amp. This was motivated by working out the best value for the screen grid resistors.

I have to clean the board and it is still missing a couple of components. It looks okay which good. I decided to use Takman resistors after all the resistor talk on the Dumble side of the fence.

I'm now thinking I'll never hear the difference between the Takman resistors and regular Xicon resistors. Got to try the boutique stuff once I suppose.

[img766]http://i149.photobucket.com/albums/s69/ ... eaefd0.jpg[/img]

[img766]http://i149.photobucket.com/albums/s69/ ... 2dc0a5.jpg[/img]

Still trying to work out the best screen grid resistor value and what zener value to use to get it to sound "right".

[Mark]

I used a Classic Tone power transformer instead of the silver Heyboer 270-0-270 250mA, I was concerned about having too high a rail and sag issues, I was wrong on both counts it seems.

It turns out Ken Fischer himself used this value power transformer. Oh well, I should have read this site and taken notes. My next Rocket will be even better!

[img766]http://i149.photobucket.com/albums/s69/ ... 85924c.jpg[/img]
[img766]http://i149.photobucket.com/albums/s69/ ... 77d4fe.jpg[/img]
[img766]http://i149.photobucket.com/albums/s69/ ... bb6580.jpg[/img]