Great stuff Mark!
I also read that Ken hated the fact that his favorite NOS el84s were too expensive. New production tubes weren't up to the job, neither sonically or robust enough.
I haven't built one with B+ as low as that. I do like around 300 more than 350 or so. Now I know why.
Need some zeners.
roberto wrote:We are around 15V @100 mA for the EZ81, while we are approximately at 45V for a 5Y3.
Looks like the EZ81 is a nice little rectifier, not much drop and rated for a 50mf filter, where 100mA is sufficient I'll start considering it over a 5Y3.
I now have a 4PDT switch so I will do a comparison between 100 ohm and 1K ohm to hear if there are any sonic differences.
So far 1K is looking to be the kindess value to the Rocket. My calculations when using 1K ohm screen grid resistors make me think the screen grids are dissipating 3 watts when a signal is applied to the amp.
I'm still looking to see if the way I'm measuring the screen grid dissipation is accurate.
Yesterday I put the 100 ohm and 1K ohm resistors on a 4PDT switch. The idea being to find out if there really is any audible difference between thee two values.
As far as testing goes, I used my 82 1957 reissue Strat, and a R0 Lespaul.
I didn't use Celestion speakers I used Kendrick Blackframes in a Fender Bassman cab as they aren't as loud as Celestion speakers and the bass and mids are more balanced than the Celestions.
In short there is a difference and it is debatable as to which one sounds better. Unfortunately I liked the 100 ohms better than the 1K ohm screen grid resistors.
I heard a difference in the mids, when I stood above the speaker box I didn't hear any difference at all, when I stepped back from the cab the difference was noticeable.
I think the 1K's were a little thinner sounding and a bit "clearer" sounding than the 100 ohm's. If you were to use the amp on the verge of break up the 1K resistor might be better. To my ear the 100 ohm resistors sounded more natural with both guitars.
Try them both, neither are horrible sounding, they are just different.
Its a USA Strat, S/N: V000440. It is a great looking guitar but more importantly a great sounding guitar.
It's not original.
It has been refretted by Greg Fryer (made red specials for Brian May).
The finish was bad on this guitar, it has been resprayed in nitro using 50's factory techniques.
The pickups were rewound in enamel (as it originally was on the guitar) by Slider and sounds so much better than the originals ever did. I went back to the enamel wound pickups over formvar as I though the guitar sounded better with enamel wire pickups. It sounded bigger than the formvar wire pickups.
I still have the original bridge, but I have a Callaham bridge on the guitar now as it sounds chimier and more vintage like with this bridge.
There are a few of them around, they are going for $2K+. I'm not ready to sell it yet though.
The Japanese guitars from this era are very nice too, I played a Tokai 57 reissue and it was heart breaking as I'd already purchased this guitar. The Tokai was awesome, I wish I could find it again.
Mark wrote:
I think the 1K's were a little thinner sounding and a bit "clearer" sounding than the 100 ohm's. If you were to use the amp on the verge of break up the 1K resistor might be better. To my ear the 100 ohm resistors sounded more natural with both guitars.
"The Rocket has 15H of inductance between the screen and the plate nodes where the Liverpool has a 1k ohm resistance. 15H has reactance of almost 1k at 10Hz, nearly 10k at 100Hz, and nearly 100k at 1kHz. I suspect this helps significantly to keep the screens in check, making the larger screen resistors unnecessary."
Forgive me for dragging out this old thread, but I'm new here and I've been doing a bit of research into screen grid resistors. My understanding is that a choke exhibits reactance only when passing an AC signal. As the choke in the Rocket (and in my EL84-based home build) is in the power supply, it is only seeing DC (apart from perhaps a small amount of ripple) and therefore cannot have any reactance. Or am I missing something?
It’s a fair point.
Lots of inductance requires lots of turns of wire, so the resistance of all that wire should also be taken into account, typically around 500ohms.
Then as it warms up, its resistance will increase a fair bit more. The resistance temperature coefficient of copper is about 4%/°C, and a 20°C rise or more seems reasonable, so it could end up not far below 1k.
guitardunc wrote: ↑Sat Jul 24, 2021 9:48 am
As the choke in the Rocket (and in my EL84-based home build) is in the power supply, it is only seeing DC (apart from perhaps a small amount of ripple) and therefore cannot have any reactance. Or am I missing something?
Of course there is no reactance to pure DC. But there is ***BIG*** 120Hz AC ripple on the B+ at the first filter node. This ripple can easily be on the order of 5 to 10 volts! A 15H choke offers 11,310Ω to a 120Hz signal. That's the beauty of a choke... Zero resistance to the dc current but 11.3K resistance to the 120Hz ripple. Really smooths the B+ a lot.
Just use a scope or meter set to AC volts and look at the B+ before the choke and compare the AC ripple to that seen on the other side of the choke. You may be surprised at the difference.
sluckey wrote: ↑Sat Jul 24, 2021 11:44 am
… That's the beauty of a choke... Zero resistance to the dc current but 11.3K resistance to the 120Hz ripple. Really smooths the B+ a lot…
My reading is that the issue at the heart of this thread is limiting screen grid current.
Of course, a low ripple on the HT feed to them is great, but it’s a different thing. eg if choosing a choke for smoothing the HT feed to the screen grids, it’s counterproductive to ignore those with higher resistance. Provided their current rating is sufficient, one whose resistance is 500ohms may be preferable to one whose resistance is 50ohms.
And will be cheaper and smaller
Re screen grid resistor values, 100Ω will help prevent oscillation, but it won't do much to limit screen current. Higher values (470Ω, 1k) will obviously have more effect.
A larger choke (say 30H vs 4H) is likely to have a higher DC resistance, and that will help to limit screen current. Higher inductance might limit the peak current seen in a transient since the screen current will initially be supplied by the screen cap, causing its voltage to drop.
Thanks for your thoughts, guys. Even though I'm using a 20H choke, I've decided to go ahead and replace my 1W 100R screen resistors with 5W 1K0 flameproofs. It'll be interesting to hear if there's an audible difference.
Tone Lover wrote: ↑Mon Aug 30, 2021 4:54 am
I’m pretty sure you would only be able to tell if you had them on them on a switch and flipped back and forth. Bill
