Tweet
Originally posted by mooreamps
View Post
What is your point?
-
Wow... That's great. Oh, wait... My old Marshall doesn't have that... BUT IT DOES HAVE A STAND BY SWITCH!!!
What is your point?"Take two placebos, works twice as well." Enzo
"Now get off my lawn with your silicooties and boom-chucka speakers and computers masquerading as amplifiers" Justin Thomas
"If you're not interested in opinions and the experience of others, why even start a thread?
You can't just expect consent." Helmholtz -
Looked at this yesterday and had some comments but got distracted and never posted, but you've addressed those. I have two other comments now that I look at this again.
I would suggest pulling l the HT for the elevation circuit from the screen node, it will be more stable than the plate node and will keep things from wiggling around alot. I have seen people replace the lower resistor with a zener also to keep things locked in place.
I'm not sure how set you are on 4 100u caps on the plate node, but it's just as well to use two 200 to 220u caps.
Right now I'm having a brain fart, but I'm not sure R29 in the bias circuit is doing. Traditionally the wiper is tied to the negative rail so that if it fails open, bias goes full negative. I'm not sure you'll get that with your current setup, but again, I might be missing something.-MikeComment
-
Big bottles, big bottles,
Talk about red plates, my amp's gottle
... er, got 'em.
Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!
Oh, wait! That sounds familiar, somehow.Comment
-
Thanks for the comments.Originally posted by defaced View Post
Sure, I'll pull the elevation from the screens if thats more stable.
I'm not set either way. I have a big bag of 400V 100u caps but I also have 2 220u 400V snap ins that I could use too. Neither is particularly more space efficient I guess.
I'm using the universal bias supply from Merlins power supply book. There was an error in the original schematic and the correct one is here:
http://www.freewebs.com/valvewizard2/bias_adj1.jpg
From the book the equivalent resistor is there to apply the full raw negative voltage in the event of the wiper failing. The value I'm using may not be sensible though. I'll no doubt use a 19mm cement trimmer as I feel I have a better shot at mounting one of those on a turret board!
One thing that I'm slightly worried by is the grid leak I'm presenting to the KT88s. Should I scale this down? A 47k pot + 10k resistor going into 100k grid leaks is overshooting the 50k max somewhat! I've based this off other designs I've seen which do work but thats no reason to copy an error if its stupid.Comment
-
Big bottles drive me out of my mind. Don't stick them up your behi-hind.Originally posted by R.G. View Post"Take two placebos, works twice as well." Enzo
"Now get off my lawn with your silicooties and boom-chucka speakers and computers masquerading as amplifiers" Justin Thomas
"If you're not interested in opinions and the experience of others, why even start a thread?
You can't just expect consent." HelmholtzComment
-
100k max for KT-88s if dissipation is over 35w, per the JJ data sheet (not sure what brand of tubes you plan to use). When I get home I'll look at the bias circuit I have for my KT-88s as well as at least one other amp and see where I'm at on this. I have not had a problem and to say the least, our power amps are *very* similar.One thing that I'm slightly worried by is the grid leak I'm presenting to the KT88s. Should I scale this down? A 47k pot + 10k resistor going into 100k grid leaks is overshooting the 50k max somewhat! I've based this off other designs I've seen which do work but thats no reason to copy an error if its stupid.-MikeComment
-
Thanks. Much appreciated.
I was probably going to use sovteks but their data sheet stipulates 50k max grid leak so I'll use JJs instead. They'e not a great deal more expensive.Comment
-
The Sovteks should still work just fine, at least for a while- just keep in mind that over time as they get gassier any tube will benefit from lower value grid leak resistors. This is always made worse when a tube is running hotter, ie dissipating more heat.
jamieComment
-
I should also add- while you can still kill 6550's at these voltages I'd hardly call it full utilization of them. They'll happier run on much higher voltages. At the 450ish volts you're using an EL34 will probably make more power because it can pull the plates closer to ground.
Your grid leak resistance is technically a DC value and includes the value of the bias pot and any supply circuitry in series with it. This is why I advocate lower value grid leak resistors. Again, it's not crucial necessarily but it's a step on the path to "bulletproofing" the design if that's what you desire. This may be why "big bottle" amps like the Hiwatt tended to have a dedicated bias winding.
JamieComment
-
Ok, I have two amps, one commercial and one DIY which can be run in two modes, with and without MOSFET followers, here's what they have. I'm not advocating either, just kicking you some data to consider.Originally posted by Zozobra View Post
Commercial amp, Sovtek KT-88, maximum of 177k grid to ground resistance, B+ is in the 470-490v range (been a while since I measured)
DIY amp without MOSFET followers, JJ KT-88, maximum of 210K grid to ground resistance, B+ is around 530v.
DIY amp with MOSFET followers, JJ KT-88, maximum of 157K grid to ground resistance, B+ is around 530v.
Grid leak on both designs is 100k, any additional is from the bias circuit. Now that I know more (I designed that two years ago), I would probably reduce the 210k on the non-follower setup by about 50k.-MikeComment
-
When driving with a MOSFET, would it be better to use a CCS, rather than a grid leak resistor, for the connection to the bias supply? This would seem to give the best of both worlds: low resistance (for happy grids), and high impedance (for maximum signal).
What might be potential risks?Comment
-
Good question, it seems I misquoted the MOSFET numbers. The MOSFET is connected directly to the grid of the tube, so grid to ground in that case would be 47k (MOSFET source resistance). Good thing I posted alot of that work here.
Design thread: http://music-electronics-forum.com/t16854/
Here's a schem of the MOSFETs and the bias supply: http://music-electronics-forum.com/a...uffer-bias.png-MikeComment
-
I'll bear that in mind. I guess I could stand to drop the grid leaks.Originally posted by imaradiostar View Post
Thanks for taking the time to look at this. I could drop the grid leaks to something in the range 47-68k I guess if I'm using a 47k pot with 10k to ground. This would shoot a max grid to ground of:Originally posted by defaced View Post
2k2+(47/68k)+47k+10k = 106.2/127.2k, unless I've missed a trick here.
which doesn't seem too bad.Comment
-
Thanks. I'm using this B+ as thats the PT I already have and I want to design around KT88s as all my amps run EL34s
When I've built up my confidence from building this I'm getting a custom would toroidal PT wound which will shoot me into the 550V. I may well specify a bias winding on it while I'm there as its not going to add a great deal to the cost. More on this at the time though.Comment
-
Okay I've dropped the grid leaks to 47k. Schematic:
http://dl.dropbox.com/u/8802910/poweramp.png
I'm going to start working on the layout and then get this thing built in the next few weeks. Please let me know if you see any glaring omissions or oversights
Comment
Comment