Yes, around 650CT (325-0-325). Keep in mind that if you use the plexi-spec'd transformer (325-0-325 180 ma) in the champ, the B+ will run high due to being lightly loaded since you are only pulling 60-70 ma, and the supply won't sag nearly as much when overdriven. I imagine this will change the character of the amp a fair bit.
As already mentioned, these two designs are quite different and require different power & output iron. If it were me, I'd build the champ first, tweak/play with it for the next 6 months, then build the plexi if you are still think you need it. There are probably a few other vintage Fender designs that the champ iron will support.
Also keep in mind that you'll need a bias winding or a separate bias transformer for the plexi since it's fixed bias, another reason to start with a champ since the power supply is simpler.
correct me if im wrong, but it looks to me like the only stage that is fixed biased is the output stage. the preamp stages are all cathode biased from my understanding of it. Also, a little while ago i found an interesteing PS schematic for the plexi, which provides a negative DC volatge. i cant remember where, but I know essentially what it looks like..ill draw it up tonight to see what you guys think.
hey guys, ive noticed a lot of PP OP trafos have taps for screen voltages. are there any significant benefits/disadvantages to using these vs biasing the screens with the PS? I assume that biasing using the OP trafo taps would make it impossible to change the screen voltage without changing the anode DC voltage as well. it would probably be a little more efficient though right?
Last edited:
Furthering what MrCurwen mentioned above, the screen grid taps on the output transformer are for ultralinear operation, which provides less distortion a little less output power than pentode mode. It is widely used in Hi-Fi power amps. Think of it this way, you have triode and pentode operation, and ultralinear is in-between in terms of power and distortion. The screen grid UL OT taps are usually around 43% of the windings. If your output transformer has UL taps, that's fine, you don't have to use them.
this is what i meant by taking a bias supply from the HT winding, if your using a fullwave rectifier using a CT then basically just add some reverse biased diodes to take off the negative voltage, you can put a 2.2k probably 2 watt resistor before the reverse diode to lower the voltage, and then lower the voltage like you would a normal power supply. you can then make a potential divider at the end of it to provide variable bias.
This is what Mark Huss does on his design right? the one I found was very similar to that, only with a tube rectifier instead of the stacked UF4007 diodes...with this kind of biasing, the user can dial in warm vs cold biasing for different tone/distortion if they wanted to or adjust the bias to compensate for difference in tubes right? where as the cathode biasing will somewhat automatically compensate for the difference in tube types. am i on the right track here?
Yes, the Plexi design is getting the bias from the secondary transformer winding with a 220K dropping R and a half wave rectifier (SS diode) and a CRC (47uf-15K-47uf). A bias tap or separate winding is convenient, but not strictly req'd for fixed bias.
Fixed bias allows changing idle current with the pot adjust, with cathode bias you would need to change out the cathode R for a different value to change idle current. Keep in mind that increasing idle current usually drops the B+ and vice-versa....everything is related.
Last edited:
could cathode bias and a rheostat work the same way as a fixed bias with a pot adjust?
i assume turning the cathode R's value down would cause B+ to drop and give more or less clean headroom as desired. also, ive read (maybe at the valve wizard's page) that cathode bias encourages 2nd order harmonics, which is generally a good thing, and it seems a lot cheaper than trying to get a negative fixed bias in the power supply.
i assume turning the cathode R's value down would cause B+ to drop and give more or less clean headroom as desired. also, ive read (maybe at the valve wizard's page) that cathode bias encourages 2nd order harmonics, which is generally a good thing, and it seems a lot cheaper than trying to get a negative fixed bias in the power supply.
Cathode bias is a little easier and recommended for a first build.
Yes, you could use a variable R to change the idle current. If you are entertaining this, consider putting fixed R's either in parallel or in series (ohm's law is your friend here) to limit the min and max R the cathode sees. It prevents the tube from melting and keeps the smoke in. Too much R is OK it just puts the tube in or near cutoff, it's too low an R value that becomes problematic.
As a reference, look at the Tubelab Simple SE "tubes and transformers" link below. This is a basic SE hi-fi design that can use EL34, KT88, 6L6GC, and 6V6 output tubes (among others). A number of folks have built the Simple SE with a rotary selector that selects different R values for the cathode R. A simple switch can be used to switch between 2 values. They do this mostly to change the cathode R when changing from EL34s to KT88's, etc.
Look at the sim tables near the bottom of the page; they show power out, distortion, etc for various B+ voltages, cathode bias R's, output transformers, etc. The power dissipation values in red are exceeding the datasheet max power. Some tubes can take this happily for a long time and others can not. The last table is for 6V6's.
Tubes & Transformers
It shows some reasonable values for cathode R's with various B+ values. NOTE: The tables show operating conditions for triode wired outputs, so the power out may seem low.
Yes, you could use a variable R to change the idle current. If you are entertaining this, consider putting fixed R's either in parallel or in series (ohm's law is your friend here) to limit the min and max R the cathode sees. It prevents the tube from melting and keeps the smoke in. Too much R is OK it just puts the tube in or near cutoff, it's too low an R value that becomes problematic.
As a reference, look at the Tubelab Simple SE "tubes and transformers" link below. This is a basic SE hi-fi design that can use EL34, KT88, 6L6GC, and 6V6 output tubes (among others). A number of folks have built the Simple SE with a rotary selector that selects different R values for the cathode R. A simple switch can be used to switch between 2 values. They do this mostly to change the cathode R when changing from EL34s to KT88's, etc.
Look at the sim tables near the bottom of the page; they show power out, distortion, etc for various B+ voltages, cathode bias R's, output transformers, etc. The power dissipation values in red are exceeding the datasheet max power. Some tubes can take this happily for a long time and others can not. The last table is for 6V6's.
Tubes & Transformers
It shows some reasonable values for cathode R's with various B+ values. NOTE: The tables show operating conditions for triode wired outputs, so the power out may seem low.
Last edited:
ok, well i definitely want to avoid more power dissipation than the tube is rated for, since i dont really know them that well...acording to this with a 300v B+ the minimum cathode voltage is around 470 ohms does this mean that for a PP stage it would be 235 or can a PP stage not be modelled as 2 SE stages in parallel? if the resistors connected to the bias voltage in the plexi schem were wired to ground instead of a negative voltage, the DC bias at the grid would be pretty much 0 because of the caps after the phase inverter, so should i figure out the voltage (and therefore resistance) ranges that way using a load line?
also, is there anyway to find out how much the B+ will drop, it has something to do with the output impedance of the PS im sure
also, is there anyway to find out how much the B+ will drop, it has something to do with the output impedance of the PS im sure
For an excellent discussion about load lines and lots of other things:
education+diy
Miles Prower also has a great thread in a post here about calculating load lines, if I can dig it up, I'll post the link.
Yes, the B+ drop is a function of the PS output impedance. Your PSUDII model can easily calculate this (assuming the transformer secondary impedance that you used is reasonable); right click on the current tap representing the 6V6, and check the "stepped load box" then increase the current by say 10ma or so after 5 seconds. You want to let enough time pass to be on the steady state part of the B+ curve before the stepped load kicks in. The new lower B+ is shown for that current. The other thing that you want to look for is ringing/overshoot during the transition. In a hi-fi amp, you don't want any ringing/overshoot (underdamped behavior); with 8uf caps in the champ PS, you may have some, and it's probably fine, again a difference between MI and Hi-Fi amps.
Or you can just change the current draw for the 6V6 tap without using the stepped load feature, same thing, the new current will give a new B+, just won't show ringing/overshoot.
For power dissipation, it's the voltage drop across the tube times the current for triodes. For UL and pentode mode, the screen current comes into play a bit.
hey, so ive oredered most of the parts now, and i have a question for you guys regarding prototyping...do you guys use breadboards/perfboards at all? and if so how do you rig up the tubes so they arent moving around and whatnot?
also, the 6v6 has separate values for plate dissipation and screen dissipation, so that 14 W is for the plate only. thanks for helping me understand this stuff better...now just the long wait for the transformers...and hopefully by then everything should be in
also, the 6v6 has separate values for plate dissipation and screen dissipation, so that 14 W is for the plate only. thanks for helping me understand this stuff better...now just the long wait for the transformers...and hopefully by then everything should be in
George has a nice breadboard setup that could be easily built. Check out his "tubelab 3" breadboard
The Tubelab
thats a really good idea thanks paul! looking through the tonelab website, I noticed he built a lot of his final products of PC boards. I read somewhere that PC boards are not good for tubes particularly if the tubes are mounted directly on the pc board (as are the ones on tonelab) I understand its probably an issue with capacitance between the traces (most of which can be alleiviated with careful board layout), but are there any other reasons not to use a PC board. I have never really done point to point wiring before, so im not completely sure me doing point to point would lower capacitances much more than a well wired pc board. what do you guys use for your final products?
For breadboarding, a chunk of plywood works. You can get proto octal sockets from Steve at Apex Jr (as well as zillions of other parts cheap). They are the $2 Grey Potter & Brumfield parts. Apex Jr.Home Page
Tube Scokets
An aluminum baking pan is a cheap and easy breadboard chassis also, some people use them for finished amps.
Geek over in the vendor forum also sells some experimenter's PCBs, or you just drill a 7/8" dia hole in some scrap aluminum for the 9 pin socket. Search for "Geeks ye olde PC shoppe" or something like that. Panel mount octal sockets require a 1 1/8" dia hole. Greenlee punches are one way to make nice big holes in metal, esp the 1 1/8 inch dia ones. A unibit (stepped drill bit) will make a nice 7/8" hole for cheap. Greenlee punches are expensive, but if you only need to make a couple Harbor Freight sells knockoffs that are soft as butter and barely work, but hey they're cheap.
PCB's and tubes can work just fine together, and lots of projects use them, such as a few from Pete Millett and Tubelab et al. For a simple circuit like the champ, P-2-P wiring should be cake.
A few basic things: Use solid core wire for the heaters. Twist the heater wires tightly together and keep them away from the signal wires whenever possible. If the two must cross, try to do it at a right angle. Tuck the twisted heater wires into the edges and corners of your final chassis, only bringing them out to solder to the tube sockets. When continuing heater wires from on socket to the next, don't "hop over the socket with the wires, come up to the socket, then wire the continuing wires back to the edge/corners of the chassis, etc. This is for hum reduction. The solid wires keep the twist and stay in the edges corners well. Use 600V rated wire for the B+ stuff, since your B+ will be over 300V. You can use cat5 wire for the signal wire cheap/free and sometimes very good esp if teflon insulated. Small coax also works well for the signal wiring.
And very important, use a grounded mains cord, and 220K 3W or so bleeder resistors across your PS caps. If your choke will be easily disconnectable, put a bleeder R on the caps either side of it to assure that the caps bleed down even if you disconnect the choke. Use a star ground, and connect the safety ground there as well especially on the breadboard.
It's wise to get some clip leads for your meter, then you can be hands-off when testing a live amp. When using a probe, always keep one hand in your pocket or grab a belt loop around the back of your pants. A couple of DVMs are very handy, and harbor freight sells cheapies for like $6 or something. Typically you want one on B+ and another on what ever you are checking out minimum.
Good solder is worth it, as it melts like butter and is easy to work with. Kester 63/37, cardas quad eutectic, etc. You can get small quans of Cardas quad eut on ebay for a few bucks. 63/37 is easier to work with than 60/40.
I build my PS stuff on small perfboards, but P-2-P wiring works too.
Tube Scokets
An aluminum baking pan is a cheap and easy breadboard chassis also, some people use them for finished amps.
Geek over in the vendor forum also sells some experimenter's PCBs, or you just drill a 7/8" dia hole in some scrap aluminum for the 9 pin socket. Search for "Geeks ye olde PC shoppe" or something like that. Panel mount octal sockets require a 1 1/8" dia hole. Greenlee punches are one way to make nice big holes in metal, esp the 1 1/8 inch dia ones. A unibit (stepped drill bit) will make a nice 7/8" hole for cheap. Greenlee punches are expensive, but if you only need to make a couple Harbor Freight sells knockoffs that are soft as butter and barely work, but hey they're cheap.
PCB's and tubes can work just fine together, and lots of projects use them, such as a few from Pete Millett and Tubelab et al. For a simple circuit like the champ, P-2-P wiring should be cake.
A few basic things: Use solid core wire for the heaters. Twist the heater wires tightly together and keep them away from the signal wires whenever possible. If the two must cross, try to do it at a right angle. Tuck the twisted heater wires into the edges and corners of your final chassis, only bringing them out to solder to the tube sockets. When continuing heater wires from on socket to the next, don't "hop over the socket with the wires, come up to the socket, then wire the continuing wires back to the edge/corners of the chassis, etc. This is for hum reduction. The solid wires keep the twist and stay in the edges corners well. Use 600V rated wire for the B+ stuff, since your B+ will be over 300V. You can use cat5 wire for the signal wire cheap/free and sometimes very good esp if teflon insulated. Small coax also works well for the signal wiring.
And very important, use a grounded mains cord, and 220K 3W or so bleeder resistors across your PS caps. If your choke will be easily disconnectable, put a bleeder R on the caps either side of it to assure that the caps bleed down even if you disconnect the choke. Use a star ground, and connect the safety ground there as well especially on the breadboard.
It's wise to get some clip leads for your meter, then you can be hands-off when testing a live amp. When using a probe, always keep one hand in your pocket or grab a belt loop around the back of your pants. A couple of DVMs are very handy, and harbor freight sells cheapies for like $6 or something. Typically you want one on B+ and another on what ever you are checking out minimum.
Good solder is worth it, as it melts like butter and is easy to work with. Kester 63/37, cardas quad eutectic, etc. You can get small quans of Cardas quad eut on ebay for a few bucks. 63/37 is easier to work with than 60/40.
I build my PS stuff on small perfboards, but P-2-P wiring works too.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.