Small Line Amp (Tube)

Here's just an idea, untested for now, for a small line amp... for when ya need something a little beefier than a regular preamp, but don't need a big thundering +30dBM line amp.

Dave, that looks like another winner. You probably could drop the input tube if you didn't need that much gain -- you should get some 33dB without it.

Peace,
Paul

> some 33dB without it.

20dB?

On thumbs: 100K/2K is 50, then 4:1 in the output iron, is 12. Doing the math right on the calculator gave 11.59 and 21.some dB. The real answer discounts for finite open-loop gain, and may be a hair less than 20dB.

[quote author="PRR"]> some 33dB without it.

20dB?

On thumbs: 100K/2K is 50, then 4:1 in the output iron, is 12. Doing the math right on the calculator gave 11.59 and 21.some dB. The real answer discounts for finite open-loop gain, and may be a hair less than 20dB.[/quote]

Duh. I was ignoring the output tranny. Duh.

Peace,
Paul

Paul S: if that's the biggest mistake you've made today, you're doing better than most of us :wink:

OLG of the feedback pair is 50dB, CLG is 32dB. So the net gain of the feedback pair through the output xfmr is 20dB, maybe more like 19dB if you allow for xfmr parasitics.

The input stage gain is 26dB, or 20dB if you leave the cathode unbypassed. In the latter case, you'd probably want DC heaters--any hum coupled into the cathode at that point would be amplified way up by the remainder of the circuit.

Add a 1:10 or 1:15 transformer at the input, use a selected AV7 for V1 and metal film for the plate and grid resistors, and it should make a nice high-gain preamp (66dB or more).

This circuit came about because I was thinking of a combined "program amp" (booster and line amp in one) for that tube mixer that I never seem to get around to building :wink:. I was going to go for the classic topology of a booster amp followed by a powerful +27dBM or +30dBM line amp, then it occurred to me that I'm not going to need to drive peaks like that in my home studio. It's not like a broadcast plant of yore, where one might be feeding though a 6dB line isolation pad to a "bridging buss" or a 10-mile Telco line to a transmitter with a lossy equalizer at the far end. So, I spent part of the evening with pencil, paper and a tube manual, and this amp was the result.

For what it's worth, it performs in the simulator as predicted--but I never trust the simulator, so I'm going to try to find some time to lash together a breadboard next week.

dave-

i've built your with-feedback 12av7/bh7 line amp, and it works great. i'd like to be able to cram multiple "one bottle" 12av7 line amps (20db gain into 600ohms) in a compact chassis for another project; the second stage of this design seems like a possible candidate for that.

one thing i'd like to try, especially now that edcor is willing to wind output transformers with B+ taps for cheap (XPP & XSE series), is a 1 tube lineamp that can run off the lower B+ available from a 115VAC winding. this would allow me to use ordinary dual 115V primary transformers for the power transformer. does this seem feasible? is there another topology that would be better suited to my purposes?

thanks,

ed

> a 1 tube lineamp that can run off the lower B+ available from a 115VAC winding. this would allow me to use ordinary dual 115V primary transformers for the power transformer.

You get ~150V DC. Cut the voltage in half, on the same tubes, you get one-quarter power output, maybe less. If you must drive 600 ohms at nominal +4dBm, NYD's designs are already near their limit with 250V-300V, will not be happy at 150V.

But given a 120V winding, and cheap electrolytics, it is easy to wire-up a voltage doubler to give 300V. Since we have existing plans for that voltage, this is the obvious route.

The other is to redesign around fatter tubes, the ones used in US TV sets, which were designed for adequate performance at lower B+. You get less Mu per tube, but you will use a lower-ratio output transformer, so it isn't radically different gain. You may end up with one more stage mike-to-recorder than a 300V design. You will need more heater power to feed the fatter cathodes, and in a compact chassis the heat can be troublesome (TV set size was ruled by CRT and cabinet size: they ran HOT but not as hot as just a bunch of bottles in a minimum case). TV tuner tubes can be fine mike-amps and +12dBm line stages. V-sweep tubes seem ideal for Line Amp work and strong headphone drivers (even, push-pull, little speaker amps). Tuner tubes are often fairly "bent" for RF AGC action, but if you want tube-sound that may not be bad. The V-Sweep tubes are often very linear, so the TV raster was not distorted.

I'm not sure that, after allowing for heat and passive components, Miniature tubes are any smaller than Octal tubes. Octals are IMHO more robust and easier to work on. Also octal TV tubes approached the zenith of routine high-quality tube production. But if you gotta go mini, 6DR7 ($8) is interesting as a high gain driver with a grunty power side.

There is also the 6HB6/6HA6 which is a mini 10-watt V-sweep pentode with 20,000uMho at 40mA and Mu(G2) of 33. This computes out to triode Rp= 1,650 ohms, workable in a 2:1 2K4:600 transformer. If we can flow that much DC in the winding, that's an astonishing Gv=20 to the plate and Gv=10 to the 600 ohms. If we have to resistor-load it, figure half the voltage gain, still good. At 100V on the plate it really wants to be worked above 15mA for good gain: logical trade-off for lower voltage.

The TV V-Sweep tubes are often available in 13V or 15V-heater versions, and sell for a buck less that way. Tuner tubes sometimes come in 4V versions.

TV V-Sweep tubes will work AT their ratings for a couple hard years: if a TV is on, it is V-sweeping as hard as it can (and they usually pushed the ratings). Tuner tubes usually won't take full rated power for years: if a TV is getting a decent signal, the AGC will reduce tuner current, the rating only ensures no quick-death if the signal goes off the air.

[quote author="PRR"]
But given a 120V winding, and cheap electrolytics, it is easy to wire-up a voltage doubler to give 300V. Since we have existing plans for that voltage, this is the obvious route. [/quote]

You already know this, but for anyone who doesn't...

If using a voltage doubler (full wave), the current rating of the winding must be 4X the DC output current.

Ed: I know that dual primaries have been used with one of the primaries acting as a secondary, but it's not something I would recommend without knowing the particulars of a given transformer. Some may not have adequate isolation between the two primaries to allow them to be used safely in this fashion. Also, if the two primaries would normally be used in parallel for a given input voltage, you would have to derate the transformer appropriately if only using one as the input.

True isolation transformers are available at reasonable prices. I bought a 115:115 30VA Stancor or Triad from Mouser a few years ago, and I recall that it was pretty inexpensive. And I'm sure you're familar with the method of using two filament xfmrs back-to-back. If the second xfmr had a 230V winding, you wouldn't even have to mess with a voltage doubler. Back-to-back is a lossy arrangement due to all the copper in the path, but has been used with success in some lower-current setups.

And don't forget, Edcor is doing power transformers now, too. They could probably wind the exact transformer you need (e.g., 250V plate and 6.3V filament) at a low cost, and you wouldn't have to bother with any kludges.

Hey Ed: Just for fun, next time ya talk to Edcor, ask if they can make something to this basic spec. Something like this would work well for a number of projects, including the one you have in mind.

Ed,

Has the mic input transformer thing worked out with Edcor?
I have been thinking about asking them for a plate transformer Dave has posted. The only thing holding me back is the lack of time and money. :thumb:

Cheers,
Tamas

[quote author="PRR"]
I'm not sure that, after allowing for heat and passive components, Miniature tubes are any smaller than Octal tubes. Octals are IMHO more robust and easier to work on. Also octal TV tubes approached the zenith of routine high-quality tube production. But if you gotta go mini, 6DR7 ($8) is interesting as a high gain driver with a grunty power side. [/quote]

http://www.2tough.com/~kiira/rec/6DR7.pdf

That seems like a really interesting tube. Is there an octal version or something similar PRR? I like octal tubes. I've been using sweep tubes for rectifiers in parallel, like 6by5,6cj3,6au4,6ax5, but I think building something with a tube like this would be a good learn for me... maybe even get a nice preamp out of it. Maybe a one tube mixer mic /line pre requiring lower plate voltage?

64mu/6mu ha ha. cool.

Kiira

6EM7

first of all, thank you both, nydave and prr, for the wealth of ideas.

[quote author="PRR"]given a 120V winding, and cheap electrolytics, it is easy to wire-up a voltage doubler to give 300V. Since we have existing plans for that voltage, this is the obvious route. [/quote]

this will probably be my answer. i tend to resist using doublers, but under the circumstances, i'll try this first. i am a bit leery of the saftey issue pointed out by dave; does the lack of interwinding isolation bother you also?

[quote author="PRR"]The other is to redesign around fatter tubes, the ones used in US TV sets, which were designed for adequate performance at lower B+[/quote]

[quote author="PRR"]At 100V on the plate it really wants to be worked above 15mA for good gain: logical trade-off for lower voltage.[/quote]

the higher current demand would have negative repercussions for the low frequency performance of the output transformer, wouldn't it? this seems like an argument for going with the higher voltage plan.

dave, your power transformer concerns are well taken. my reason for going with this "primary as secondary" method is cost. the 50VA toroids from digikey are $19ea, which is quite appealing. i doubt there is much insulation between the two primary windings beyond the enamel on the wires themselves. i tried looking through the UL and international saftey regs cited for those transformers, but couldn't find any specific construction details.

mouser does carry the triad N68X which is a 50VA EI construction isolation transformer, reasonably priced at about $12. it has a multi-chamber bobbin, so it would be plenty safe, and the primary is split for 230v hookup if desired. a good option (glad you pointed it out), but it still requires another transformer for the filament supply. i do like it better than the back to back option for about the same amount of money, though i miss the pleasing doughnut shape.

i wish there were a simple AND inexpensive power transformer solution available from the usual sources, but unfortunately it isn't a mass market item.

as for a custom power transformer, i should talk to edcor, but even their smallest stock power transformer is $35. money well spent, but higher than what i was hoping for.

thanks for the basic output transformer spec. i'm sure edcor can wind something along those lines, and i bet (with a bit of extra attention) it could turn out great. if i manage to wrangle something from them i will post it here. i have a few other transformer projects ahead of that one, so if anyone else wants to have a go, feel free.

ed

[quote author="Tamas"]Has the mic input transformer thing worked out with Edcor?
I have been thinking about asking them for a plate transformer Dave has posted. The only thing holding me back is the lack of time and money.[/quote]

because edcor only uses M6 laminations, i was never able to get the sort of mic input transformer i was looking for from them. i have found other off the shelf sources (the pikatrons for $20 from buerklin, oeps from rs & farnell) that are cheap and suit my purposes better. i was surprised by how interesting the edcor mic input sounded, so maybe i'll go back to it someday.

you should talk to them about that plate transformer. they really are great; i just hope all of our custom orders don't run them into the ground.

ed

[quote author="PRR"]6EM7[/quote]

the first "tube" project i ever built was fred nachbaur's miniblok amp, which uses the 13EM7. i haven't listened through it in a while, but i remember liking it, especially for the money - the tubes are dirt cheap. if you have any ideas for a *EM7 (or other dual dissimilar) lineamp, at any plate voltage, i'd be interested to hear them.

thanks,

ed

> ask if they can make something to this basic spec

The DCR numbers are probably unrealistic for power-iron technology. I'd rather have 2X, even 4X, that much DCR if it bought better bass.

I'd start design on a "25VA" core. More-bigger iron is better than a marginal design on a small core. We are just to the point that capacitance becomes a problem; still, with triode drive, we should try the bigger (higher capacitance) winding.

It must be insulated 500V between windings. It should not be split-bobbin. It should be interleaved. For a power-iron winder, that suggests half primary, fishpaper, secondary, fishpaper, half primary, with the outside end of the winding identified (put that to the Plate).

[quote author="edanderson"][quote author="PRR"]6EM7[/quote] if you have any ideas for a *EM7 (or other dual dissimilar) lineamp, at any plate voltage, i'd be interested to hear them. [/quote]

We started discussing a two-triode mike amp using 6EM7 in this thread:
http://www.groupdiy.com/index.php?topic=12637&start=30

I intend to draw up something based on those recommendations. I was originally looking at the 12DW7 for a dissimilar twin triode, but the 6EM7 offers more power and lower output impedance at a lower cost. It didn't take much convincing to get me onboard.

Meanwhile, I'm stocking up on AV7s for Dave's designs. :thumb:

I have messed with the 12DW7 to make a one bottle similar to Dave's design. The 12AV7 or even the lowly 5965A works better for some reason, especially when it came to driving capability. No time to explore it further.

Edcor just agreed to make a sample of the plate transformer to the specs Dave published. It will be around $60 a pop. I can't wait to test it!!!

Cheers,
Tamas

That's expensive by Edcor's price standards. Are they using a nickel core or somethin'?

Yes, it will have a Ni core. It is not cheap although you know how much more competitors ask for similar things. It could be used with the One-Bottle micpre also. Thanks for the specs Dave!