OK.
Left channel. Pot only goes up to 70K in circuit. Voltage at 70k = 294v. Voltage at 0k = 269v
Re-checked pot - it only goes up to 70k in circuit
Right channel: Pot went up to 135K in circuit, Voltage = 215v. At 0k voltage = 212v.
Repeated high-resistance. This time pot only went to 70k, voltage =214.
BTW, I only have 0.22uF as first cap, so I can tailor voltage upwards when needed.
Left channel. Pot only goes up to 70K in circuit. Voltage at 70k = 294v. Voltage at 0k = 269v
Re-checked pot - it only goes up to 70k in circuit
Right channel: Pot went up to 135K in circuit, Voltage = 215v. At 0k voltage = 212v.
Repeated high-resistance. This time pot only went to 70k, voltage =214.
BTW, I only have 0.22uF as first cap, so I can tailor voltage upwards when needed.
Dont forget that the 56k resistor needs to be able to dissipate 3 watts. I use a pair of Panasonic 3 watt resistors.
Sorry, I wasn't clear in my last post. I set aside the 120/240 TXR and put in the big guy (360-0-360). It has extra volts, so I am using quasi-choke input with 0.22uF, then choke, then 100uF cap, then the regulator board.
So, no change for the 56k resistor. Again, dummy load is 25k resistor (about 12 mA.)
So, no change for the 56k resistor. Again, dummy load is 25k resistor (about 12 mA.)
Update:
With current setup, and 25k dummy load, no reg, V = 333
Left channel, 72kR reading on pot, V = 294 V; with 0k reading on pot, V = 269 V
Right channel, 68k reading on pot, V = 292 V; with 0k reading on pot V=294 V
I found a missing connection on right channel reg; when re-wiring I had neglected to tie 56k resistor with 200/180k junction. I think Woodturner Fran did the same thing.
So, am I OK?
With current setup, and 25k dummy load, no reg, V = 333
Left channel, 72kR reading on pot, V = 294 V; with 0k reading on pot, V = 269 V
Right channel, 68k reading on pot, V = 292 V; with 0k reading on pot V=294 V
I found a missing connection on right channel reg; when re-wiring I had neglected to tie 56k resistor with 200/180k junction. I think Woodturner Fran did the same thing.
So, am I OK?
Looking good! Depending on your EL84 (there's a lot of variation between "classic" ones and modern ones), you may have to adjust the voltage range; if the idle current is too high at the minimum setting, reduce the set resistor a bit, the opposite if you can't get the idle current high enough. But the range you have is fine for the JJ tubes I've been using.
I was reading through the article on SY's website and noticed this statement:
"For example, if a 47k resistor is chosen, any grid current will be limited to a milliamp or two. This will limit any tendency of the output stage to block via the grid. The downside of this approach is the high frequency roll-off set between the grid resistor and the input capacitance. For a triode configuration of the output stage, the capacitance is high enough that the pole will be dangerously close to the top of the audio band, and certainly close enough to the output transformers high frequency limit to make stability problematic."
Speaking about triodes I get the part that the pole is dangerously close to the top of the audio band but can someone explain the second part about this pole being problematic in regard to stability? My guess is it would only be problematic when feedback is used which loop includes the output transformer? Or is it a problem no matter what?
For example you have an EL34 amp triode strapped with zero global feedback, there is some Schade feedback. Gain of the EL34 stage is 7.5 and a 1k5 grid stop resistor is used putting the pole very close to the top of the audio band, is this an issue?
I apologize for quoting you SY, you didn't make it easy for me to copy and paste from your site......I had to view page source to do so. I will delete the quote if you wish.
"For example, if a 47k resistor is chosen, any grid current will be limited to a milliamp or two. This will limit any tendency of the output stage to block via the grid. The downside of this approach is the high frequency roll-off set between the grid resistor and the input capacitance. For a triode configuration of the output stage, the capacitance is high enough that the pole will be dangerously close to the top of the audio band, and certainly close enough to the output transformers high frequency limit to make stability problematic."
Speaking about triodes I get the part that the pole is dangerously close to the top of the audio band but can someone explain the second part about this pole being problematic in regard to stability? My guess is it would only be problematic when feedback is used which loop includes the output transformer? Or is it a problem no matter what?
For example you have an EL34 amp triode strapped with zero global feedback, there is some Schade feedback. Gain of the EL34 stage is 7.5 and a 1k5 grid stop resistor is used putting the pole very close to the top of the audio band, is this an issue?
I apologize for quoting you SY, you didn't make it easy for me to copy and paste from your site......I had to view page source to do so. I will delete the quote if you wish.
I'll talk to my webmaster (aka my wife) about making copy and paste easier. Sorry if it was a problem- I'm pretty clueless about that stuff.😀
In any case, you're absolutely correct, the issue is closed loop stability. Not a problem for pentodes, but very much one for triodes.
In any case, you're absolutely correct, the issue is closed loop stability. Not a problem for pentodes, but very much one for triodes.
Hello SY, I wouldn't bug your webmaster, she is doing a good job!
Thank you for confirming my suspicion. Sometimes I shock myself......literally 😀
Thank you for confirming my suspicion. Sometimes I shock myself......literally 😀
I have another theory question about the coupling capacitors chosen for your amp.
Plugging the numbers into the equation 1/(2*3.14*270000*(.1*10^-6))=5.89Hz
Is this the -3db or -6db breakpoint? My simulations and everything else I read say it's -3db
But when I calculate the reactance of a .1uF cap at 5.89Hz I get;
1/(2*3.14*5.89*(.1*10^-6))=270349
When viewing that reactance value and the 270k load as a potential divider I get;
Vin*(270000/(270000+270349))=-6db
Plugging the numbers into the equation 1/(2*3.14*270000*(.1*10^-6))=5.89Hz
Is this the -3db or -6db breakpoint? My simulations and everything else I read say it's -3db
But when I calculate the reactance of a .1uF cap at 5.89Hz I get;
1/(2*3.14*5.89*(.1*10^-6))=270349
When viewing that reactance value and the 270k load as a potential divider I get;
Vin*(270000/(270000+270349))=-6db
So... I finally dropped the hammer on some output James output iron (a pair o' 8k JH-6228HS) especially for this project. It's apparent I have a lot of reading to do while I wait in greedy anticipation, what with SY's web article and 1055 forum posts. While I promise to try and my homework before asking any stupid questions, nothing is proof against a resolute fool.
I've been tossing around various physical configurations in my mind for some time now, and I think I've settled on a "Musical Machine" (aka Eric Kingsbury) - type deal with the not-so-separate separate power supply. Any thoughtful input/advice would be greatly appreciated...
BTW: I'm probably one of the few people on this forum who thinks calculus is fun (I didn't say I was good at it, though)! 😀
I've been tossing around various physical configurations in my mind for some time now, and I think I've settled on a "Musical Machine" (aka Eric Kingsbury) - type deal with the not-so-separate separate power supply. Any thoughtful input/advice would be greatly appreciated...
BTW: I'm probably one of the few people on this forum who thinks calculus is fun (I didn't say I was good at it, though)! 😀
The RLD amp that SY developed puts out about 18 watts per channel. The musical machine as designed (6v6) is about 4 or 5 watts(triode strapped). Is that enough for your present system? My speakers are 95 db at 1 watt so i'm fine with 4 watts, but I also use SYs amp in my system and it sounds great!
Just my .02 ¢
Just my .02 ¢
Last edited:
Hi rock4016.
My first reaction on reading your post was "huh?" - then I got it. Yes, my current system (RH84 into a pair of modded Klipsch RF-20's) is more than up to the task. The problem is I've got the itch to build!
Thanks for your input though. I just didn't do a good job of explaining myself. What I'm intending to do is adopt (read: shamelessly steal) Poinz's idea of building one chassis with separate PS and signal sections. I sorta' did that in the RH84 mentioned above (one chassis, PS shoved to the right, but one top plate), with excellent results.
That said, I admire both designs (RLD and Musical Machine) and am in fact collecting parts for both!😀
(And parenthetically - welcome to the Mr_Zenith Parenthesis Festival! 😉)
My first reaction on reading your post was "huh?" - then I got it. Yes, my current system (RH84 into a pair of modded Klipsch RF-20's) is more than up to the task. The problem is I've got the itch to build!
Thanks for your input though. I just didn't do a good job of explaining myself. What I'm intending to do is adopt (read: shamelessly steal) Poinz's idea of building one chassis with separate PS and signal sections. I sorta' did that in the RH84 mentioned above (one chassis, PS shoved to the right, but one top plate), with excellent results.
That said, I admire both designs (RLD and Musical Machine) and am in fact collecting parts for both!😀
(And parenthetically - welcome to the Mr_Zenith Parenthesis Festival! 😉)
Last edited:
