70L7 PC speakers

[Opcom]

Not exactly AM, but heres what I've done this weekend. I have a case of 100 NOS 70L7's, and decided I could break it open and use two of them. The input transformer is a 48 ohms to 5K 3 watt unit, and the output transformer is a 120V to 6.3VCT 0.6A filament transformer.

I may have to go to a regular output transformer, because I have very little low frequency response. The response seems severely tilted to the highs, to the point of "hissy-ness" on music. The best size fil. xfmr to use for this is a 1A or 1.5A and I don't have any. I found (late) a pair of "50C5 output transformer" new in the packages.

I could use more gain as well, although the PC seems to be able to overdrive the amp if pushed..
48:5000 is only 10:1 voltage step up. Bias is 7V, so I would need 14V peak, meaning 1.4V peak from the PC - it should be able to do that.

The whole thing is transformerless for power like an All-American 5, with the isolation being the input and output transformers. Right now I am waiting to see if the heater dropping resistor is going to melt the speaker cabinet, and have only put one together.

The rules for tube PC speakers is that only the tube can show.. my rules anyway..

[W3RSW]

Very nice!

Why don'ch'a put two in push pull?  Man, if I had a 100... 

Hmm, then your power supply could be a full wave voltage doubler, the AF amp transformer input and output and 140 volts across the fils.   At 125 line voltage the 140 string ought to last forever.

Guess we're waiting for the 10 in pp/parallel, class C fully modulated AM rig. 

[Opcom]

I thought about the heater voltage. At Ef=62.5V, the emission is substantially down according to test. It was suggested in rec.audio.tubes to use a 4uF capacitor or somesuch to drop the heater voltage (If=150mA) and get rid of the 8.25 watts heat from the dropping resistor. Plate dissipation is 5 Watts and output is 1.8W in class A. Maybe a pair could do 15 watts in AB1. For RF, I've never tried that on a tube like this but I bet it would be OK to 7 and maybe 14MHz. I say maybe, because the ones i have are made with all the leads going through a button stem, then to fan out to the pins in the base. I do not think it is rated for grid dissipation, but it would not be much, maybe 1/2 Watt. Not sure if the drive would be sifficient before heating the grid. I have no experience with low power AM transmitters using unusual receiving tubes.

The plate voltage by the book is 110VDC, but if in push pull, it could go double on peaks and same for modulation, I see no harm in that. I'd concede the diode sections of the RF PA stage would be good for the "three diode negative peak clipper"

It would be a hoot to do am AC/DC 70L7 HF rig. I could use all 70L7's from speech amp and oscillator to PSU, mod, and PA. No voltage doubling though because the PIV is 350V on the diode and no bridge rectifying because the heater of two of the tubes would be at +/- 88V peak to B- reference. the H-K voltage is rated +/-90V on the beam power section, really close. Do I have the time to build it? I don't know. It took 7 hours to build the PC speaker ones and most of that was alot of fiddling in close quarters. I had already designed it some time ago.

so..
No issues with rectifiers if used in classic AA-5 configuration.
Rectifier is rated 70mA, which is 7.7 watts at 110VDC. This limits the available input power.
A 70L7 probably has alot of gain as a voltage amp.
Class AB1 output = 7.5W into 4K ct? per pair
Class C output = 4.2W carrier per tube (?)

60 tubes - 20 PA's and 40 modulators.. (reserving 40pcs for spares)
modulator = 150W audio into 200 Ohms CT. (power requirement is 110VDC@2.8A max.)
PA  = 84 watts carrier, modulator load is 78.6 ohms (110VDC@1.4 amps)
A 150VAC 2A power transformer having dual 120V primary windings in series could be used for modulation iron.

I can see this would be really, really sick, er.. slick! I think I better not, because it is becoming all to real.

The Arvin 40 "Mighty Mite" model 40 AM table radio from 1938 has only two tubes: 25B8 and 70L7, but maybe the best match up would be a 50L6 and 70L7. The 50L6 has a higher plate dissipation.

I'll make you a deal on some of them. You can even have all of them and the original bulk box that says "100 pcs of 70L7", for the right price and I'll go get two NOS ones off the internet for $4 each.

[Opcom]

Oh, what I was going to say was that I fixed the hissy sound.

It was the 48 Ohm to 5000 Ohm (1:10) transformer I had used as an input transformer.
I put a 600:600 transformer there and it's correct, but also very low volume.
I do not know why the high-ratio step-up input transformers I tried caused this hissy sound.
I also swapped the 6.3V to 120V transformer (1:19) to the input and got the same boosted highs/no lows thing.
Above tests were done both from a PC sound card output and from a ghetto blaster headphone output jack.

I elected to put the pair of identical NOS 50C5 output transformers in place as output transformers, the sound is as good, maybe a tiny bit better, and they are physically smaller. I might have to add a triode gain stage to the units. Maybe a 12AX7 or the like. Fortunately the heater droppings resistor is a dividohm type.

[W3RSW]

Sounds like your having a lot of fun with the possibilities.  Uh, thanks for the offer but half the magic is in using the 70L7's that are on hand. 

I have a similar situation with 6U8's, 5763's, 6CL6's etc. and have dreamed of all kinds of receivers, transmitters, etc. using a bunch.

I'd still try a voltage doubler, full wave power supply.    Hey, with a hundred, and the fact that tube specs can be somewhat exceeded in amateur use. might be worth a try.  So you bake a couple. The doubler can be tamed and better regulated by choke input too.  As you've noted, all sorts of possibilites.

And...  As far as the AF amp goes, 2 in PP is cleaner on second harmonic.  Quiescent current for 2 in PP is less than one in class A, etc.  The advantages of PP are awesome when you stack up the mirrored load lines.  Way cool.   I've always just loved that sort of stuff.  Say would an inductance boost circuit work on high current filament loads?  Don't see why not if your just looking at boosting 60 volts to 70. More components though. Anyway I'd at least try the 60 volts per fil. and see if the emission really does fall off as much.  For a first trial just put an additonal 70L7, filament only, across your existing single tube AF amp.

Too bad transformers are so expensive these days; you could sell a stereo pair version to Audio phools.  Let them supply their own speakers.  Another use for low power audio is for the 'rear' speakers or phantom mixed 'rear' sound in an afficionado's all tube setup.  Marketed properly, they'll not be able to stay away from your awesome  GLOWING PUBE REAR CHANNEL PHASORS

[Opcom]

Does anyone know where to find new or NOS 2K/8 Ohm OPT for a few bucks? Some looking around for new 50C5/70L6 and the like output transformers revealed only one that was well matched, 2.5K to 4/8 ohms. It is $60 from Hammond. This is not the optimum part, since it's designed for something more like a 6V6 and rated 5 watts, but it was the smallest one with 40mA+ current handling. I was a bit surprised at the pricing, but maybe it's a very low production volume. I think I will be saving every AA5 OPT I see now.

I'm afraid the 'Phools are on their own. I could always sell them the tube, a ceramic socket, the power components, and the instructions. Anyway it is great fun. Did not get to put in the driver stage yet, due to other work the lab today.