4 Tube Amp From Scratch

[TubeNoob]

Steve Conner:

PS, my fanbase just doubled, I am delighted!

JMaf:

I watched every Nat Geo program there is...wanna know about the most poisonous jellyfish on Earth? Komodo Dragons? How houses are built on ice?

Thank you both for the only two belly-laughs I've had today Yeah, 7 days on my back and 30 more with the cast. But I'm cool, just using it as a research period!

I have seen the freewebs stuff and downloaded a lot of it as reference, excellent info - thanks for reminding me about it.

Steve - cool amps you've done. And that corvette sounded great. Dang you Scots have cool accents!

So, i guess the schematic is about ready for me to do the layout on. Tho i do want to double check my plan on the heater wiring...

As I mentioned, Fred Nachbaur said you can have the 12ax7 'sum' the currents allowing you to use 150ma and 300ma tubes together. Would something like the attached diagram do the trick then?

Fred said:

Note the slightly unusual series-parallel arrangment; each pair of 35C5 output tube filaments is in series with one half of the 12AU7 dual triode. The center-tap of the 12AU7 sums the two branch currents (150 mA each) into the preamp tube, giving 300 mA as required for this valve's heater supply.

While unusual, this geometry is eminently stable and safe. If any tube is removed (or its filament burns out), none of the others will suffer as a result, unlike some series-wired television sets where a single burnout could take out several other tubes at the same time.

Link is: "Spunky" Tube Amplifier - How It Works

[loudthud]

That won't work as drawn. You can't simply add the heater voltages, the heaters are not in series. To make that work you will need a 466 ohm resistor in series with the 35W4 and 366 ohm in series with the 50B5. The 466 ohm will dissipate 10.5W, use a 20W part. The 366 ohm will dissipate 8.3W, use a 15 or 20W part.

[TubeNoob]

Thanks for the input LoudThud. For some reason i seem to want to avoid having dissipating resistors (should I?). So, i'm wondering if i should approach it this way:

1:1 isolation transformer into the old radio's autoformer (which accepts 110-220v), then on to the heaters. The autoformer handled the different current tubes this way:

35w4 and 50b5 straight out of the autoformer first, then a 2nd line coming from the autoformer after those two tubes (which i assume added the other 150ma current), just before the rest of the tubes in the circuit which are all 300ma tubes (6at6, 6ba6, etc). Does this sound correct and does this sound better than dropping resistors? (Not really sure if i have any 20watt resistors around).

Thanks again...
-Steve

[jmaf]

IMO a 1:1 isolating transformer is the best solution there. Problem is they can be quite bulky....

[J M Fahey]

Dear tubenoob.
Iīve been following this thread getting somewhat nervous along the way, because *none* of the schematics showed a power transformer.
Hearing talk about an autoformer does not cut much ice with me.
Now you are mentioning an 1:1 isolation transformer, which, really, is the bare *minimum* safety you should add.
And you live in a 220V country, so 1:1 wonīt do either.
So it all turns out into a kludge to make a still kludgy solution work *somewhat*.
May I suggest that since you will need to invest in a transformer anyway, you buy the proper one, which can power +B, filaments, etc. the right way?
jm2c
PS: there *are* "safe" versions of those old Kay/Airline/Silvertone/whatever amps , youīll certainly find the one which is close to what you want to build.

[TubeNoob]

JMaf - thank you sir!

JmFahey - i've been following it nervously too!! Well, i figured since the autoformer takes 110-220v that a 1:1 would be cool, and that way i could use the same heater setup as is already installed on the radio (and could use it next time i'm in the States with it, too, as opposed to having a 2:1).

But if i do in fact have to buy specific iron for this, i reckon i could avoid the autoformer all together.

just trying to find the right solution (safe + parts on hand + cheapest parts i don't have).

i definitely appreciate the advice. i need this kinda input.

[TubeNoob]

Some questions, in case anyone has some ideas on 'em:

1) Could this tremolo circuit affect only amplitude (and not distortion) without too much of a change? i would LOVE to put it on a switch so i could select between the two options.

2) what should i expect in terms of output watts on this, around 4 watts on an 8ohm load?

3) Should R16 (the pot in the trem circuit) go to ground on that left lug? (not shown in the original Kalamazoo circuit)

4) Are there issues with dissipating that much power thru those dropping resistors in the heater string?

Updated schematic attached...

[jmaf]

1) Since that tremolo alters the cathode bias on V2A, it'll inevitably mess with the linearity in that stage. So some distortion will be inevitable, but I can't say how much or even if it'll even be noticeable for guitar standards of distortion.

2) Probably from 2 to 2.75 Watts, 2.75 being (close to a)maximum for that class A stage.

3) If S2 switch is open, that pot doesn't do anything. Is that intentional? If you want the pot to work when the switch is open as well, then you need to ground that lug.

4) There aren't issues besides the fact that you're throwing away power(and generating heat) in order to accommodate something unusual. But it's not wrong in any way IMO if the resistors are of the appropriate value to allow the needed current through, and rated at least twice the dissipation as you expect. In fact the resistors will protect the heaters from the initial surge current. Those older tubes can glow like a lamp at turn-on, that takes a little of their life away each time, so the surge protection is a bonus.

[TubeNoob]

Thanks JMaf and by the way, good luck on your open-source amp - super cool of you and that thing sounds amazing!

Attached is a MP3 of the Kalamazoo's tremolo modulating the distortion from the intro to one of my songs. I LUV this sound and would love to have it back in the arsenal (the Kalamazoo was on loan to me at the time).

Regarding the switch on the trem (on/off), i think it might be redundant since there's an intensity pot which i could simply turn down to zero, no?
In that instance, would i *not* ground that left lug of the frequency pot?

So, my transformer can put out 220v. The autoformer takes 220v and puts out a B+ as well as the 117v heater supply. It also has that line inserted into the heater string right after the 150ma tubes and just before the 300ma tubes (on the original radio). So, can it be done this way?:

Ditch the 6BA6 and 6BE6, and insert the 12AX7 right after the 50B5 (and before the extra line from the autoformer comes in.) Check out the attached graphic, top is the original, bottom is what i'm suggesting.

[J M Fahey]

Is that symbol below a transformer? Or an autoformer?
If I read it well: youīd apply 110V (if available) to the rightmost tap (the winding with most turns) and having successively higher line voltages, youīd move that connection towards the left, until if you have 220V they would go straight to the almost leftmost tap, the one that is connected directly to the series heater string: 35W4+50B5+12AX7+6AT6+lightbulb.
Because thatīs what I see.
Please confirm.

[TubeNoob]

hey JMFahey, thanks for the info. Yes, that's the autoformer depicted there from the radio's old schematic. (I would just put the transformer before it, hooked up to the 220v line-in on the autoformer).

As you can see, all i did was remove the two unused 6v tubes and put the 12ax7 to the right of that extra line coming in from the autoformer (it comes in between the 6v tubes and the 12v tubes).

Is this workable like that?

[J M Fahey]

Short answer: no.
The way itīs drawn, if you have 220V and set the selector switch accordingly, you apply 220V to a series filament string rated for around 100V: instant doom.
Going to the other extreme: if you set the selector to 110V with a 110V line, filaments get around 55V.
Since both options are impossible, the schematic is wrong , canīt be trusted at all.
I would leave that nice old radio as room furnishing, mood setting or whatever, UNPLUGGED, only as a decorative object and build some regular tube amp schematic, your choice, getting the suitable parts needed.
Forget that radio as basis for an amp.
The only part worth reclaiming might be the output transformer.
And a poor transformer, go figure.
Not even the tubes because they force you to use kludgy dangerous obsolete series filaments.
The kalamazoo was obsolete (and dangerous) already in the 40's , even in 110V USA.
Imagine in a 220V country.
Sorry.

[TubeNoob]

Hey JMFahey - thanks a lot for the no-nonsense reply.

Unfortunately, i'm far too stubborn to just 'build something that's already been done' and forget this. That's not my style. If i have to have 3 transformers and twice the parts, i'll still do it. I'm too anxious to hear what this amp will sound like and have something original.

So, given my insolence, and the fact that the only stopping block is the heater supply, what's the alternative?? This transformer has multiple secondaries - 3 14v outs, an 8v out, and a 185v, and i have a bunch of 6v regulators and diodes and a lot of time and solder, how can i make this work?!

and i know the OT is nothing special, but i can always upgrade that too later on, nay?

by the way, this radio is only the chassis, so it's not warming up the living room aesthetic in any way shape or form ;-)

Let me also add that even if the schematic is wrong, the heater is in fact wired this way - with a line to the heaters coming off the autoformer, and with a 2nd line coming from the autoformer right after the 150ma tubes and right before the 300ma tubes, so it MUST have worked this way before. So, since it did work that way, would there be a prob with ditching those two 6v tubes and putting the 12ax7 in with the 150ma tubes, despite what the schematic says?

Or, for example, what if i feed the 6AT6 with it's own 6v regulator and wire the other tubes in series like this amp:
http://1.bp.blogspot.com/_ZXngOXLq9f...matic+V1.1.JPG


Cheers,
Steve

[TubeNoob]

By the way, what would that matter if the filaments were getting 55v? The 50B5 needs close to that anyway and as i understood it, the tubes take what they need from the voltage - it's the current i have to focus on given they're in a string. I'm still learning so I'm looking for the education on these points.

[R.G.]

It doesn't work that way. heater filaments are like light bulb filaments. They're designed for a very specific voltage across them. They pull the current they need only at that specific voltage. Too much voltage and they overheat - FAST. The temp goes up as about the fourth power of the applied voltage, and the life goes down at about the 12th power of the voltage.

As to lower voltage, the temperature of the heater is what makes electrons come off the cathode and makes the tube work. The temperature goes down at about that same fourth power of applied voltage, so there is a sharp cutoff on enough heat/temperature to get electrons to flow. You really have to supply tubes their nominal heater voltage +/-10% for them to work as intended.