New Circuit, New Build, New Questions...

[flood]

EDIT: oh dear... did i post this in the wrong section? would someone be so kind and move it to the right one?

Hello to all - it's your favourite n00b from Mumbai, India (and possibly the only one) writing in with a few new questions. As with every good student, i suppose i'm just trying to gain a blessing from the gurus on this forum (which is undoubtedly the nicest one i frequent - all the others have a lot of unnecessary fighting and bickering and flaming going on) before i embark on my next quest.

The success of my Class A build, no small thanks to all the inputs i got on it here, inspired me to go ahead and try out a new circuit idea for somebody (forum member on an indian rock forum) who commissioned an amp from me. he was looking for a class A amp, single ended that is versatile within the space occupied by fender and vox. the amp should be clean, going from smoky jazz tones, to shimmering cleans, to fat but not overdistorted lead tones. so essentially a very ballsy champ.

my idea sheet:

a. two discrete channels, pentode dark (6SJ7) and triode bright (ECC83) with a switch to link the inputs in parallel or cascade the triode into the pentode.the triode channel has a boost switch, and the pentode has a 6 way brightness selector. a display LED glows brighter or darker depending on the position.

b. a cathode follower with a 59 bassman EQ section, with a switchable mid-shift.

c. the fun (or not) part: a totally flexible power section - allowing the use of 6L6, 6V6 and variants, EL34 and a noval socket in parallel for EL84. the output transformer is custom wound for 5k, 4k and 2.5k taps (overkill?). a 4-way rotary switch allows the selection of 4 cathode resistors, as shown on the schematic. a second switch allows the user to use a solid-state CCS (idea stolen from projet G5) to dial in the required bias current. i've foreseen 5W resistors for the cathode bias, if this doesn't suffice i can pick up larger ones.

i've scoured the market looking for axial capacitors, but it's impossible to get current production ones here,. since the customer is on a budget, i'm forced to stick to radials. is this a problem, or am i just worrying for no reason?

i'd be really grateful, as i always am, to get your inputs on my circuit. you're more than welcome to use it for whatever purpose - if it's commercial (which i doubt, because i'm sure that i've totally overestimated my abilities), it'd be nice if you gave me a little mention, but i'm not into building amps for the money anyway.... this is a hobby that finally needs a bit of external financing for it to sustain itself, although i have no qualms if i can earn a bit of money from it too

since there were readability problems with the last circuit i uploaded, i've upped this one as a (virus free) PDF attachment instead.

thanks to all in advance, and best regards from bombay.

credits for stolen ideas: allen, fender, vox, trainwreck, doug hammond (octal fatness), projet G5, THD, this forum and a lot more i can't even remember now.

[flood]

some changes were made:

-RF cap on pentode input is now 470p. will increase it to make it darker if i need to.
- channel mixing resistors are 220k. R22 is now 1M.
- power tube grid resistor is now 2k2, parallel cap deleted
- solid state CCS may be changed to an LM317 or discarded altogether, since the 1500uF/35V cathode bias cap on the power tube will probably approximate the effect of the CCS.

would 5W cathode resistors suffice, or should i up some/all of them to ten watts or more?

am still trying to figure out a use for the unused triode... was thinking along the lines of a opto-trem using a power mosfet for the second stage, like RG keen describes on geofex. or i could parallel the triodes for more current, but i believe that would be tonally too close to the pentode sound?

finally, here is a draft of an instruction chart i plan to give the man. is my insistence on turning off the power correct? would it suffice to simply put the amp on standby? it IS equipped with a bleeder resistor and an LED charge indicator...

again, thanks in advance for any help.

best regards, aditya

[flood]

...nobody? does the design suck that bad?

here's a picture of the chassis that came in day before yesterday from powder coating - it's BLUE!

[IMG:640:480]http://i39.tinypic.com/2wqwwzr.jpg[/img]

P.S. thanks for moving the thread!

[Cygnus X1]

5W on the cathodes should be good.

Use the extra for a tube driven effects loop perhaps?

And the powder coating is good!
Sharp looking build.

[flood]

hey cygnus, thanks for the reply. i don't know how to use a single triode for an effects loop - is it possible to do this? i suppose one side would have to remain unbuffered, or a solid state buffer would need to be used. if you do know of a way to incorporate it without having to put in one more 12ax7, please do let me know! i (and the recipient) would be very grateful!

i was thinking of using it as a light footswitchable boost in front.. perhaps using a 220k/2k2 and no cathode bypass cap. it would probably still result in OTT gain, especially in cascade mode.

is the instructional chart i prepared in order?

thanks, aditya

[M Fowler]

Flood,

Hats off to you for trying out new designs. I don't think your design sucks actually it will a great platform with all the switching you have going on to see what combination works. I look at it as a good testing amp and fun to play through.

I think the reason I didn't reply earlier was that there are so many switches and things to consider that I passed, sorry.

Mark

[flood]

Mark, thanks for the encouragement. i realise that the design (not the design per se, but the switching) is quite complex and could put people off quite a bit.

for some other considerations... has anyone used a 6SJ7 going into a cathode follower? schematics i see, particularly of the octal fatness, have no cathode resistor on the 6SJ7, and only 2.2M to the screen grid. will this suffice? r should i consider putting a cathode resistor + bypass cap in there?

in the 6SJ7 channel, i'm aiming for a darker but grainier sound.

[Andy Le Blanc]

Double check resistance coupled amplifier tables for the components vs. your
plate supply for the 6sj7. Pentode's can make a lot of gain, it might be a surprise
when your first power the circuit. You might want to try mixing the pentode
signal after the follower. Your using "zero-bias", and the grid return resistor
should be on the order of 5M to 15M. Your screen grid resistor may be too large.
But track down a table for the tube type and you should be fine.
When you select component values for the pentode gain, consider trying to make it roughly the same
as the triode if your looking to mix ahead of the follower.
It looks like it'll be a fun little amp.

[flood]

Hey Andy, thanks for stopping by and the tips... wow, i'm almost scared of the black metal can thingy now i've heard about pentodes having a lot of gain, but didnt expect transconductance to be in the order of thousands... i've never used a pentode as a preamp or line stage before.

i suppose that makes the cathode follower practically unnecessary as a current amplifier for the pentode preamp, except to add a touch of dynamics... also, as far as i know, the pentode output impedance is a bit high for driving the graphic EQ and the output tube. i read that somewhere on the intertubes, am not sure of it's accuracy.

the reason i want to mix it in before the CF is because i plan on cascading the triode gain stage into it for a high-gain sound that has it''s own character.

i suppose i should get cracking first with parts placement et al though... will write back on mon/tue with more pictures. happy long weekend to you all and don't party TOO hard. i'll be doing that for the entire forum

[flood]

this little amp has made quite a bit of progress, and i'm coming to the end stages (yeah, yeah, i've been slacking a bit...). a quick question:

the output stage has an octal and a noval socket wired in parallel for output tube flexibility. however, as advised, i've wired a 2k2 resistor to g1 on each socket for the input from the master volume knob. i want to add a switch that lets me go between pentode and triode mode (through a 100 ohm resistor).

is it necessary that each socket gets it's own 100 ohm resistor at g2? or can i simply wire one to the octal socket, and then the noval in parallel to it?

it's probably obvious, but only one of the two sockets will have a tube in it at any given time.

thanks again... pics coming soon!

[flood]

ok, this amp has gone a long way since inception...

1. EQ has bright and deep switches that add caps in parallel or series
2. the unused triode is sused as an additional gain boost stage for both channels. i am now officially scared of the gain potential here - 3 stages and a cathode follower. well into bogner territory!
3. there are a total of 5 selectable values for the output tube bias- 150R, 220R, 270R, 390R, 470R. all resistors are 5W

a couple of things i'd like to ask here.

1. as you know, there is an octal power soocket and a noval power socket. the PSU is assembled on a separate PCB. the screen grid resistor for both sockets is 1k, on the octal it is 5W and mounteed directly on the socket, for the noval it is part of the PSU PCB. i am wondering - is the 1k value ok for ALLn tubes that i intend to use? i know that 1k is good for EL34s, and acceptable for the 6*6 series too. is it ok for EL84?

2. assuming the buyer rebiases his power tubes using the aforementioned swwitch (see 3 above), is there any problem posed by the screen grid resistor?

3. i want to use the filament winding to power a number of 6V DC circutis using stabilizers and the like. these would include LEDs and relays, stabilising would be done by large caps and the 7806.... the problem that i see is that the filament wwiindings are 3.15-0-3.15, with the center tap being referenced to the power tube cathode. taking all precautions in mind to seperate this ground from the amps ground... would it be possible to get 6V (and more ) from a 3.15-0-3.15 wwinding? could i use a bridge rectifier instad of a FW, and if so how?

thanks to all in advance, i really am desperate to get this infformation as i need to close this build soon.

[Andy Le Blanc]

hey there its been a while.....

With the screen grid resistor I'd make sure that you can swap out values for
other tube types and to taste, and an accommodation for triode hook up.
If thats outside your plans and you plan to not change out your screen grid
resistor, I'd go with a higher value, say 2.2k, It'll be safer in general terms
for more tube types and operating points.

Stick 1 ohm resistor on the ground side of the cathode circuit, give your self
the opportunity to measure for plate dissipation.

The issue with your filament supply is going to be its PT current rating,
Rectification has a trade off, you'll get D.C. but the current after rectification
might not be enough. A bridge will get you: I D.C. = .62 x Sec. I A.C.

[flood]

hey andy,

thanks a lot for writing in. your insights have been incredibly helpful with my other build too!

hey there its been a while.....

With the screen grid resistor I'd make sure that you can swap out values for
other tube types and to taste, and an accommodation for triode hook up.
If thats outside your plans and you plan to not change out your screen grid
resistor, I'd go with a higher value, say 2.2k, It'll be safer in general terms
for more tube types and operating points.

noted. will do accordingly.... i would not hesistate swapping out the R myself, but the buyer is not too keen on fiddling around inside the amp (although there are safeguards - 220k 2W cap bleeder resistor with charge indicator LED). i am thinking of using a 5W screen grid resistor, but have heard that it's actually better to use lower wattage so in case something happens, the R will burn out and protect the tranny. what's your call - failsafe or overspec?

Stick 1 ohm resistor on the ground side of the cathode circuit, give your self
the opportunity to measure for plate dissipation.

1 ohm 1W 1% in place with multimeter test points - no fumbling around innards required! is it a problem though if the resistor is placed BETWEEN cathode and the cathode resistor//bypass cap duo?

The issue with your filament supply is going to be its PT current rating,
Rectification has a trade off, you'll get D.C. but the current after rectification
might not be enough. A bridge will get you: I D.C. = .62 x Sec. I A.C.

my tranny is rated - 250V-0-250V 100 mA, 3.15-0-3.15 3A. i was thinking about DC heating the preamp too, but i think it should be fine even otherwise. so even with an EL34, there's about 600 mA surplus... the relays will consume a total of 60mA, LEDs about the same, and the speaker sim circuit shouldn't be more than a few mA...

thanks again and let me know what you think about screen grid resistor wattage... regards, aditya

[Andy Le Blanc]

I've seen more fenders with the screen resistors replaced with higher wattage
parts, The tube fails and the fuse pops, a fuse in HT supply is a good safety
to add, the failure isn't allowed to cascade thru other parts, better than
constantly having to service the amp with lower wattage parts.

you want the ma. thru the tube, it doesn't matter where in the circuit you
put the 1 ohm, but the ground side of the cathode is the safest.

Its sounds like you have enough juice, always double check, if the PT hums
or gets hot you'll have to recheck..... but sounds good so far.

[flood]

thanks andy. i'm actually more worried about the PT's HV supply in case the man decides to use 6CA7s, not sure if 100mA will cut it (ok, the other tubes are using about 10 mA in total.... ). will have to recheck EL34 data. for everything else, it should be fine....

i've decided to overspec just about everything on anything i build.

there are fuses for both the primary and secondary of the PT.

i should really post the current schematic and photos soon. hope i can get it up and running (to some extent) tomorrow!