6ck4 pp amp

[audiousername]

But with, at best, a third the plate dissipation.

Ah, but the 6EM7 and friends are rated at 10W Pa for section 2 - two-thirds of the 2A3's rating!

There is a simple 6EM7 PP amp here - it includes the use of the Bottlehead C4S, but with 2.6V and 3mA for the CCS, you'd be better off using an LM334Z IC or something similar and doing away with the negative power supply rail. A CCS for the output stage would be nice too.

 

[Sheldon]

Dangit, stop talking about that 6EM7, you're gonna get me sidetracked.* That is a nice schematic, and very similar to the 6CK4 one.

Anyway, regarding DC coupling and balance, I ran across this:

http://www.wdehaan.demon.nl/el84dc.html

The last schematic on that page shows a nice simple way to plug in a balance pot. I was thinking about something like that anyway. It would be a minimal deviation, without requiring change of any circuit values. If I found that I was constantly adjusting the pot, I could then try cap coupling of the stages for the final build. He also shows some negative feedback. I'll try that too, after I've tried without, just to educate myself on its affects - again no other changes required to plug it in. This'll be fun.

Sheldon

*You know. this whole thing started a year ago because I wanted to build some speakers with higher efficiency than my present ones (they are about 88db) and I wanted to learn to make crossovers. So I started reading up and gathering information. Still haven't built the speakers (though I did make a little set of backloaded single driver horns for a second system), yet I've built two chip amps and a remote shunt resistor attenuator, and now I've got two tube projects on the dock. And the speaker parts are on the way (I think, but that's another story) This forum is dangerous. Or maybe I just need Ritalin.

 

[Sch3mat1c]

http://www.wdehaan.demon.nl/el84dc.html

The last schematic on that page shows a nice simple way to plug in a balance pot. I was thinking about something like that anyway. It would be a minimal deviation, without requiring change of any circuit values. If I found that I was constantly adjusting the pot, I could then try cap coupling of the stages for the final build.

Yeah, that'll do it. What you really need is an NFB loop that adjusts it pretty much the same way, automatically, but like I said, it's a hassle. [Well, now that I think about it, simply returning the inverting input to the respective output cathode resistor, adding another grid leak to the negative rail to balance it to 0V, would do it.]

He also shows some negative feedback. I'll try that too, after I've tried without, just to educate myself on its affects - again no other changes required to plug it in. This'll be fun.

Yes, that's one big advantage of differential amplifiers.

*You know. this whole thing started a year ago because I ... the speaker parts are on the way (I think, but that's another story) This forum is dangerous. Or maybe I just need Ritalin.

Nah, sounds just normal to me

Tim

 

[Sheldon]

Yeah, that'll do it. What you really need is an NFB loop that adjusts it pretty much the same way, automatically, but like I said, it's a hassle. [Well, now that I think about it, simply returning the inverting input to the respective output cathode resistor, adding another grid leak to the negative rail to balance it to 0V, would do it.]
Tim

I think I get that. It does mean that the negative rail can't wander relative to B+ though?

Nah, sounds just normal to me

Now I'm really worried

Sheldon

 

[Sheldon]

No-no number three: your negative voltage supply isn't. It shows a coupling cap (from a filtered DC rail) to a filtered resistor divider. If you had included a diode, it *might* develop some voltage from the residual ripple, but nowhere near 30V. This needs to be a diode in series with a resistor to one 5U4 plate, to a filter cap and so forth, for instance as I used in Frankenhouse.

Tim

Well, I'm starting to breadboard this thing up to get an idea of where to go from the basic schematic (duly noting suggestions herein for improvements as I move along). I was going to do the negative supply as you had suggested. But, when buying parts, an old radio guy said (paraphrased) "what you have there will work but the center tap connection to ground is wrong, it should be connected to the negative terminal of the 68uf cap on the negative supply". Sure enough, if I look at it that way, it makes sense, and plugging in a total current draw of about 160ma (40ma per 6ck4) across the 170ohm resistor, puts me at -27v. Close enough to start, I think.

the schematic: http://tinpan.fortunecity.com/saints/668/primer/6ck4-pp.gif

Sheldon

 

[Sch3mat1c]

Ah, yes! Hadn't noticed the magnitude of the resistors there. Self bias sounds like it was the missing trick...

Tim

 

[noisenyc]

interesting thread...

...just noticed it.

i built this amp about 8 yrs ago.

it was designed by my friend jc morrison who is a member of this board, tho not very active.

here is a link to his comments reprinted from the original "fi primer":

http://tinpan.fortunecity.com/saints/668/primer/pp.html

also see:

http://tinpan.fortunecity.com/saints/668/primer/

i vaguely remember doing some minimal tweaking due to my transformer choices. i am pretty sure i could find the docs if you want them.

i built it bare bones (cheap parts choices). it sounded really beautiful.

a well dressed italian heard it and gave me $1200 cash on the spot.

 

[Sheldon]

Re: interesting thread...

...just noticed it.

i built this amp about 8 yrs ago.

it was designed by my friend jc morrison who is a member of this board, tho not very active.

here is a link to his comments reprinted from the original "fi primer":

http://tinpan.fortunecity.com/saints/668/primer/pp.html

also see:

http://tinpan.fortunecity.com/saints/668/primer/

i vaguely remember doing some minimal tweaking due to my transformer choices. i am pretty sure i could find the docs if you want them.

i built it bare bones (cheap parts choices). it sounded really beautiful.

a well dressed italian heard it and gave me $1200 cash on the spot.

Thanks,

If you run across your notes, I'd be interested in any significant issues you encountered.

I tried to contact the owner of the fortunecity site to note the error in the schematic, but apparently the address is blocked. Perhaps you can pass on the corrections.

Sheldon

 

[noisenyc]

Re: Re: interesting thread...

Thanks,

If you run across your notes, I'd be interested in any significant issues you encountered.

I tried to contact the owner of the fortunecity site to note the error in the schematic, but apparently the address is blocked. Perhaps you can pass on the corrections.

Sheldon

i'm out of town for a week but when i return i'll see if i can find my final schematic and scan it for ya...

at any rate this amp is well worth building and tweaking. it's a fun build and the results are very nice, even with cheapo hammond iron, etc.

btw jc's (original designer) username is nanana on this board. he's a great guy, maybe a bit tired of having his aesthetic choices challenged on boards by folks with different opinions...which as we know, there is NO end to

anyway i suggest you contact him, he was once extremely active in the DIY and i'd love to draw him back into the tumult hehe

 

[Sheldon]

Re: Re: Re: interesting thread...

i'm out of town for a week but when i return i'll see if i can find my final schematic and scan it for ya...

at any rate this amp is well worth building and tweaking. it's a fun build and the results are very nice, even with cheapo hammond iron, etc.

btw jc's (original designer) username is nanana on this board. he's a great guy, maybe a bit tired of having his aesthetic choices challenged on boards by folks with different opinions...which as we know, there is NO end to

anyway i suggest you contact him, he was once extremely active in the DIY and i'd love to draw him back into the tumult hehe

Thanks. And I'll try to contact JC after I've got a working breadboard and before I make the final design.

Sheldon

 

[Sheldon]

Well, I bread boarded up one channel. I followed the schematic, except all the pos. power supply caps are 47uf. I also used 47uf for the first neg. supply cap, and the indicated 220uf for the neg. output cap. I didn't have a Siliconix current reg. diode, so I made up a ccs from two J310's. I used a voltage divider to get about a 10v drop across the ccs, as those Jfets are only rated to 25V. I also put a 100 ohm pot between the driver cathodes, with the wiper to the ccs, so I can balance the output. I used a cheapie Hammond 125c pp output transformer for testing.

As far as results; The B+ and filament voltages are a little high but in the ballpark, as expected, since the power supply is for two channels and I'm only using one. Otherwise, my first tube project works. Even makes music and sounds fine, considering the limited transformer. No gross distortion. Not much hum either, considering I didn't worry much about proper star grounding for this round. I'm encouraged enough to build a proper chassis and assemble a stereo amp for real testing/tweaking, and to fit real opt transformers.

One question for the experienced among us: How much DC imbalance would be considered acceptable? I only started it up an hour ago, and haven't done much testing, but with the current portable cd source, I get about 7 volts across the output cathodes (measured from one tbe to the other). That's afer balancing the cathodes with the input shorted and a bias voltage of about 100V.

Sheldon

 

[Sch3mat1c]

Oh, 5 or 10mA tops. Within 1mA is usually easy enough.

Sounds like you're within 10%, which may be okay. Depends on how the OPT feels.

Tim

 

[Sheldon]

Thanks,

Good enough to start then. As I said, the grounding was nowhere near optimum. Minor fiddling got the difference down to 3 volts or so with the amp hooked up, vs. the input shorted. With the 2k cathode resistors, that's around 1.5 ma. I'm sure I can get it tighter. I'm guessing that a resistor on the inverting input to ground would help. But I'm not worried about static trim, rather, concerned about drift over time. If it's fairly low (adjust once a month or so) I'll just do it that way. If I have to adjust often, it would give me an excuse to make a DC servo, which I may do anyway just for the experience.

Sheldon

 

[Sheldon]

I'll be away from the bench for a couple weeks, so won't be able to get the full amp up and running for a bit. But I have a question about the correct cathode bypass cap for the output tubes. The schematic specifies 470uf across a 2k resistor. My calc's give a lf pole of 0.16hz. Any reason that cap needs to be that large?

Sheldon

 

[Sch3mat1c]

On a cathode, the "cathode follower" impedance (i.e., 1/Gm) is usually more significant than the resistor. Current flowing determines the time constant. 470uF is an all-around pretty good value, though a bit much for small stages like 12AU7s and a bit low for very low impedance tubes.

Tim

 

[Sheldon]

Thanks,

sheldon

 

[Sheldon]

Built

Well, I built this thing:

http://tinpan.fortunecity.com/saints/668/primer/6ck4-pp.gif

A major correction and a few minor changes:

Correction: The center tap for the transformer secondaries goes to the junction of the 170ohm/800ohm resistors for the neg. ps., not to ground.

Changes: 47uf, instead of 68uf for the neg. ps.. Added a meter and pots across the driver cathodes, so I can balance the output tubes.

The power transformer is a Hammond 272HX, 300-0-300, rated at 186VA, spec'd at 200ma DC.

Measurements:

Transformer primaries - 310v.
B+ - 310V.

Now the questions, for those experienced in these arts.

Question 1: Why is the B+ so low? The scematic indicates 365v with a 600v ct, which might be a bit optimistic, but PSUD indicates about 340V with 300v on the primaries. I measured the cathode voltage of the output tubes at around 70v (2k cathode resistors), so the total current draw should be less than 150ma (only 4ma total for the driver tubes). The other voltages are in proper proportion to B+, so I don't think that there is any current loss elswhere.

Question 2: When I short the inputs, I can balance the cathode voltages of the output tubes (about 70v). But if I leave the inputs open, or connected to my pre-amp, I get a large imbalance (80 on the side connected to the imput, about 50 on the side connected to the ground side of the 12BZ7). Do I need a resistor to ground? I set up the grounds as star grounds, with the power grounds to one point, the signal to another, then signal connected to power with a single wire.

Question 3: Maybe related to 1. The power tranny gets quite warm (40-50 degrees C). Is that normal. Remember, this is my first foray into the thermionic side.

Thanks,
Sheldon

 

[Sch3mat1c]

Voltage might seem low because PSUD can't include for instance the 170 ohm bias resistor. You'd be simulating voltage across the CT to wherever on the B+ rail, whereas you might be measuring to circuit ground instead of CT. That'd account for the 30V or so. Unless you already thought of that, in which case... eh... only 10% error. Close enough for me!

As for the inbalance, that's what you get for going DC coupled. Simple as that. Either live with it... or add a grid leak to the input (you skipped it? For shame!) and coupling capacitor (in case your signal source has a DC offset, which it shouldn't, but you never know).

Is that absolute temp, or temp rise? 40C temp rise is a typical maximum spec, a bit much for a Hammond but acceptable. If you *measured* 40-50C above a room temp of 10-20C, that's just fine.
In general, you can tell a transformer is operating within limits by the ten second test... hold your hand tightly against the unit. If you can withstand the heat for ten seconds (or longer), it passes the test.

Of my amps on now, Hept'AU7 currently only passes a four second test, while Frankenhouse passes with time to spare.

Tim

 

[Sheldon]

Thanks for the reply Tim,

Voltage might seem low because PSUD can't include for instance the 170 ohm bias resistor. You'd be simulating voltage across the CT to wherever on the B+ rail, whereas you might be measuring to circuit ground instead of CT. That'd account for the 30V or so. Unless you already thought of that, in which case... eh... only 10% error. Close enough for me!

As for the inbalance, that's what you get for going DC coupled. Simple as that. Either live with it... or add a grid leak to the input (you skipped it? For shame!) and coupling capacitor (in case your signal source has a DC offset, which it shouldn't, but you never know).


Tim

Yeah, I thought of that one too. The neg. supply resistor is a 120 and a 50 ohm in series. If I short the 120ohm, I get a B+ of 320 volts and a negative voltage of about -6 volts (actually, wouldn't that be enough for the ccs?). If I measure from B+ to the CT, that gives about 325v total, so that adds up, but still seems low given that my primary has 320v AC on each input leg to the rect. tube, under load. The data sheet for the 5u4gb (if I read it correctly) seems to show an output voltage of 290 (at 300ma output)with a plate to plate input RMS of 600 (without load). Well I have a no load plate to plate of about 650 (about 620 under load) and with 200ma load I get about 325volts DC. So maybe the schematic was optimistic, and my tranny has fairly poor regulation?

Across the 170ohm, I get about -21v. When I channeled Mr. Ohm, he said that my total current draw is about 200ma. Where did those extra 50 or so ma come from? (remember, I have about 35ma for each of the output tubes and 4ma for the input tubes)

By grid leak, I assume you mean the 100k resistor from input to ground. Yup, that's there. I did measure voltage on the input grid with the nothing on the input and get about -400mv, so something's sucking electrons off that grid. The plate is at about 45v and the cathode is about 250mv, so how can the grid be negative with nothing connected to it except a 100k resistor to ground?

Is that absolute temp, or temp rise? 40C temp rise is a typical maximum spec, a bit much for a Hammond but acceptable. If you *measured* 40-50C above a room temp of 10-20C, that's just fine.
In general, you can tell a transformer is operating within limits by the ten second test... hold your hand tightly against the unit. If you can withstand the heat for ten seconds (or longer), it passes the test.

Of my amps on now, Hept'AU7 currently only passes a four second test, while Frankenhouse passes with time to spare.

Tim

I didn't actually measure the temp, but I'm guessing that it's about 70 degrees absolute. Might pass the 10 second test. But that leads to another question. Say I wanted to make a hybrid bridge, or even just ss. Can I open the tranny and will I find the individual ends of the CT so that I can parallel the sections, or will I have to dig deep?


Sheldon

 

[nanana]

hey sheldon,
congratulations on building your first tube amp...! (did i understand that right?) i hope you have enjoyed yrself? i built that amp in 1986 and i don't remember much about it other than it worked fine and i was into coltrane and ornette back then. oh yes, it was fisher 7591A outputs from a 400 series amp that came out of the garbage. you are absolutely correct at pointing out the mistake with the power trans center tap... it goes to the 170 ohm res. that is a back bias circuit which is just a cheapo neg. supply. the CRD2.0 needed the -30 to really nail the AC balance. i checked that on a scope until it worked right. it would be much less with a J310. i won't argue the direct coupling comments other than to say that i just don't agree... this is a very simple amp that works better than it has any right to. i will say the one caveat with that is that this is a DC coupled differential amp with a current sourced first stage. if you get that, then you understand why it works as well as it does. it is also a class A amp, with some headroom... there's not really enough swing or current in the driver to do much AB1. also, there were never enough 6CK4s to get a good match and i didn't want any fancy bias crap so i arranged the amp so that it wouldn't matter. i never had a matched pair of tubes and "DC balance", it isn't necessary except if you are desperate to maximize power output. it might mean a watt more or less in class A. not even 3dB more sound pressure in a 96dB per watt speaker at nominal listening level. the seperate cathode resistors on the outputs allow for different tube currents. all it means is that the clipping point is determined by the tube that goes first... the forced AC balance from the output tranny straightens things out. the 12BZ7 is OK. a 12AX7 is just as good. i have no preference. i did it first with a 12AX7 and got bored... they both worked. i probably had to adjust the dropping res a bit... you can change it if you like, obviously! it sounds to me that you have really taken this project to heart and have learned something from it?! i have never used a modeling program in my life and it would seem that you have used yours to get a better understanding of what is doing what... i really take my hat off to you. as for your questions, i don't remember where or what the power trans i used came from or was... the line voltage here in NYC is generally on the high side (120+) and it also doesn't matter that much unless you need more watts. i don't know anything about the hammond you have... is the heater winding rated enough for the 6CK4s + the driver tube. i seem to remember they draw a lot of current? did you make a DC supply for that? that would increase the current draw by 1 and half times and perhaps pull down the mains B+ by heating the tranny. i don't think i can help much without more info... but perhaps you should just listen to some music for awhile? you can optimize it later...
jc