Homebrew concept: Chisel A20 (fo' shizzle)

[jecars]

Request for the moderators, experts and aficionados. 

Attached is schematic for a design that I am toying with.  Please take a look. I am seeking reference to any schematics with similar features....so that I might compare/contrast circuit details, component values, etc.  Of course, other comments or feedback welcome.

[2deaf]

NFB is DOA.

Why a 120K on the Volume?

Why a .002uF in front of the tone stack?

Why a voltage divider on the PI?

Where are you getting the 25.2V for the preamp heaters?

[John]

Your cathode bypass caps on the first 2 stages are larger than needed. In fact, they may muddy your tone. The 1uf caps coming off the PI I assume as typos; normally they're .01 - .022.
I also assume your lamp setup in the filament string is just your way of drawing an LED?
And yes, I wouldn't get too clever with the cathode biasing on those power tubes. Use a $1 10-watt 250 ohm resistor. Or at least, I would. :)

[tubenit]

This is how I would consider approaching it if I was attempting this build.  Not saying this is "better" but it reflects more of what I would like for a "transparent" or articulate tone.

With respect, Tubenit

[2deaf]

I still don't see how the preamp heaters are getting powered.  As drawn they need 25.2V and 150mA.  Are the 5881's supposed to idle at 125mA's apiece?

[PRR]

> comments or feedback welcome.

The hard-errors and the odd values suggest this was not thoroughly checked yet.

The NFB doesn't work.

V1-A has 1.5K+47u in cathode for <3Hz bass; but the 0.002u into a 170K load (pot full up) looks like _500Hz_ bass cut out of V1-A. If there is a logic to this, please explain.

> schematics with similar features....

The whole front-end can be stolen directly, numbers and all, from Fender __Face designs.

Likewise I would steal the driver V2-A V2-B from transitional Fenders or an Ampeg.
_____________________

> Why a voltage divider on the PI?

Not really a divider. The 1M (no part designator) is boot-strapped to >>1Meg. LF gain reduction is hardly-any. The 470K limits grid current in overdrive (and at cold-start), though I am not sure why it is so much bigger than any other grid-stopper.

> Where are you getting the 25.2V

Shows and says from power bottle cathode current. This will work. Heaters get clean DC. Power stage must run class A. Can it? He predicts 340V of B+, minus 20-24V at cathode, 320V across tubes. He can run 156mA at 25V Pdiss/each (50W total). 320V/0.15A is 2.1K equivalent load on B+. We would expect a 4.2Kpp load to be near-optimum. The 6.6K load shown does not extract "maximum" power on paper. I expect like 16 Watts, when 20-24W should be possible, on paper. In real world of loudspeakers, this is not enough different to warrant new iron.

[jecars]

^^^PRR, thanks for understanding my approach to the cathodyne PI.  Your comment seems to suggest my interpretation of Merlin's anti-blocking modification is roughly correct but could use a tweak. Also, thanks for understanding my approach to pure DC heater via PT cathodes.  I used the information per our discussion in a specific thread about this. I will take time to work through the values graciously provided to see if I get the same result.  If not, I'll post specifically about this.

Sounds like the only major flaw is my approach to NFB. Dang...I thought I understood the applicable Princeton schematic. Perhaps someone could suggest a schematic that shows a viable NFB for cathodyne.  There must be an iconic reference, right?   BTW, if the comments are mostly about polarity, then no worries (I already know how to swap OT leads)

Some suggest that I should "check" my circuit.  Tell me more, please. Should I analyze via SPICE simulation or the like?  If  "check" means to compare specific parts of my concept to their "favorite" schematic, then please point me to the reference.

Thanks for the other comments.  I have made note of the bypass cap suggestions. Just goes to show that there are many flavors of pre-amp choices.  For what it's worth, I borrowed all the pre-amp values from another amp of mine that I really like. Any suggestions on the power amp section?  I'm much more of a power tube distortion guy.



 

[sluckey]

Sounds like the only major flaw is my approach to NFB. Dang...I thought I understood the applicable Princeton schematic. Perhaps someone could suggest a schematic that shows a viable NFB for cathodyne.  There must be an iconic reference, right?   BTW, if the comments are mostly about polarity, then no worries (I already know how to swap OT leads)

Some suggest that I should "check" my circuit.  Tell me more, please.

Take a close look at where you actually connected the NFB. Stare at it as long as it takes... 

[kagliostro]

Also, thanks for understanding my approach to pure DC heater via PT cathodes.

Remember you must have that current disposable on your B+ winding

Franco

[2deaf]

V1-A has 1.5K+47u in cathode for <3Hz bass; but the 0.002u into a 170K load (pot full up) looks like _500Hz_ bass cut out of V1-A. If there is a logic to this, please explain.

I got 107K with a f_c of 743Hz, but we're splitting hairs here.

When I first started modifying Fenders in the '70's, I left the 25uF bypass capacitors in and reduced the coupling capacitors for the tone shaping.  It works fine with full bypass capacitors, but 47uF is a little overkill.

A resistor connected to the wiper (like the 120K) is a strategy that I have used many times to change the frequency response as the volume is adjusted.  I want less bass as the volume is increased and that is what happens with a resistor on the wiper.  The cutoff range here is from 80Hz to 743Hz, which may be a little extreme for an amp that isn't ultra gain.

Shows and says from power bottle cathode current. This will work.

I don't doubt that it will work, but it looks like the small heaters are going to be running cold.  It also looks like the 5881's will be idling in excess of the 23W P_diss.   

[PRR]

> I got 107K with a fc of 743Hz, but we're splitting

This is low enough that the preceding stage plate resistance is non-negligible.

[jecars]

Thanks, Sluckey.  I missed a resistor, clearly.  You know, you could have just said so 

And, I get it now. A "checked" schematic means all careless errors have been fixed. Thanks to all for "checking" my work. BTW, I have updated the schematic in original post (v1.1 now), having fixed the details that were spotted by "fresh eyes".  I also bow to peer pressure and switched all bypass caps to 22 uf.  I have noted all of Tubenit's suggestions, thanks for taking the time to share.

Regarding using the pre-amp heater filaments as bias resistors, I really like this aspect and wouldn't dream of changing it.

[2deaf]

This is low enough that the preceding stage plate resistance is non-negligible.

OK.  For a 12AX7 with an E_bb of 250V and an anode voltage of around 167V and a cathode voltage of around 1.3V, I get an r_a of about 68K using my handy-dandy RCA chart of average characteristics.  An R_a of 100K in parallel with r_a is about 40.5K.  40.5K in series with the parallel combination of 1M and 120K is about 147.6K.

Giving the same haircut to a 12AY7, I get an r_a of about 29K for an output resistance of about 22.5K.  22.5K in series with the parallel combination of 1M and 120K is 129.6K.

For a 12AX7, I get an f_c of 540Hz and 614Hz for a 12AY7.

[2deaf]

. . . having fixed the details that were spotted by "fresh eyes".

I didn't mean to discourage you from putting a .002uF capacitor in front of the tone stack.  I just wanted to hear why you were doing it.  I've seen people do that over the years and I even tried it myself early on.

Regarding using the pre-amp heater filaments as bias resistors, I really like this aspect and wouldn't dream of changing it.

Good.  I want to hear how it works out.

[PRR]

> 540Hz

So I was close.

[jecars]

I didn't mean to discourage you from putting a .002uF capacitor in front of the tone stack.  I just wanted to hear why you were doing it.  I've seen people do that over the years and I even tried it myself early on.

That .002uF coupling cap would keep DC off of the 100K resistor just downstream. Since voltage + resistance = hiss, this approach could theoretically reduce hiss.  In practice however, I think it is likely that this cap does very little. 

My reasoning is that since V1 is upstream, hiss is already present to some extent...mostly from the 10K resistor input to V1A. Going into the tone stack, I would guess that hiss is already >20dB above whatever the tone stack impact could be, so any additional hiss won't be audible.

The EL34 moderators and other experts have more practical experience than I do, so let's see if they chime in.

[2deaf]

The EL34 moderators and other experts have more practical experience than I do, so let's see if they chime in.

OK.  I have some points to make, but if you only want to hear from moderators and other experts, that's cool.

[jecars]

OK.  I have some points to make, but if you only want to hear from moderators and other experts, that's cool.

Fire away, I'm open to all. I have learned more by listening than I ever have by speaking 

[2deaf]

There is a hiss that is proportional to the direct current through a resistor, but there is no direct current through the 100K in the tone stack with or without the .002uF cap.

Random hiss from two or more sources doesn't directly add to give the total hiss.  Instead the total is the square root of the sum of the squares of each hiss voltage.  Because of this equation, if one of the voltages is significantly larger than the others, the total hiss voltage is essentially just the larger voltage.  As you already know, the amplified hiss from the first grid stopper resistor is significantly larger than any other hiss.  So even if the tone stack did generate some hiss, the equation would render it negligible.

The .002uF capacitor in front of the tone stack changes the tone stack from a mid-scoop type to a mid-and-low-cut type.  This could be desirable in a high gain amp where excess bass is a problem.  When I did it, I was trying to get a tone stack that was more usable for high gain.  With a standard stack, I always wound up with the bass clear off, the mid at max., and the treble being adjusted in a small range near max.

[jecars]

The .002uF capacitor in front of the tone stack changes the tone stack from a mid-scoop type to a mid-and-low-cut type.  This could be desirable in a high gain amp where excess bass is a problem.  When I did it, I was trying to get a tone stack that was more usable for high gain.  With a standard stack, I always wound up with the bass clear off, the mid at max., and the treble being adjusted in a small range near max.

I believe you.  Learn something new every day.  To prove it to myself, I created a SPICE model of a Fender-ish tone stack, with or without a .002uF coupling cap.  The model and frequency sweeps with and without the cap are attached below. The plot with the coupling cap is called "Fender-X." It sure looks to me that the cap cuts bass (low-cut), just as you said it would.

For Chisel A20, I started with 47 uF V1A bypass cap and .002uF coupling cap to tone stack. Based on forum input, I changed to 22uF V1A bypass cap and deleted the coupling cap.  I wonder if both choices are actually roughly equivalent?