tube current sources

Working on a HV signal generator and need a load? ;-)

I've drawn the left one before, but it takes three tubes (well, I suppose the pentode and 'error amp' triode could be a pair, but the wide difference in H-K voltage rather stinks).

The right one has rather huge dropout (70V for the neon, plus "Rds(on)" for the triode, which will be pretty poor as it's not far into positive grid bias, given safe values of current-limiting resistor).

It actually works quite well with a pentode (G1 to K, no cathode R, neon from K to G2), having a typical variation of say 10% in current, over the linear range. Looks just like Early effect, though the mechanism is actually electric field leakage, not depletion width. The variation can be tweaked by applying a little positive feedback (think foldback current limiting). I forget if this can be done with the help of a small cathode resistor, or if an 'error amp' is required.

Obviously, the dropout isn't any better for pentode, but the neon can be supplied from a higher voltage, if one is available, which solves that issue.

For the 100V, 100mA operating conditions I posted about originally, a

6CW5/EL86 is probably the least heater power candidate. For triodes, 12B4 is about right, but its plate resistance is terrible. (Phil can tell you all about the poor performance of followers using low-plate-resistance-equivalent devices...)

The little R+C compensation is interesting, but probably unnecessary. I'd be surprised if the bandwidth on these was beyond a few MHz. The loop gain is a whole lot less than doing it with transistors!

To do more bandwidth, you'll have to do your favorite: add a big whopping choke. It better be low capacitance. Can you get pi wound chokes in SMT?

We should add Jeorg to this. Maybe we can replace his high voltage PNPs with something better stocked. LOL!

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com 


"John Larkin"  wrote in message  
news:imml7bdehk6cm7v9hbikuv2ci1pfmv0j1f@4ax.com... 
> 
> 
> https://dl.dropboxusercontent.com/u/53724080/Tubes/Tube_CC_Sources.JPG 
> 
> One of those is sort of a current mirror. 
> 
> 
> https://dl.dropboxusercontent.com/u/53724080/Tubes/Pentode_CCsource.JPG 
> 
> Neon lamps are tubes, no? 
> 
> 
> 
> --  
> 
> John Larkin         Highland Technology, Inc 
> 
> lunatic fringe electronics 
>

The cool thing about playing tube circuits is that I'll probably never actually build another one.

How about triode, resistor, lithium battery?

Bootstrap current sources are cool, too. I use them in high-speed stuff (with *silicon* parts) because they can be very wideband without hassle.

Something like that could be done with toobs.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

You can drop the cathode follower from the screen source. Replace the neon lamp with a voltage reference tube (OA2, VR150 ...).

A pentode with stable screen voltage is a good constant current source in the plate circuit.

--

-TV

I saw a patent from the early 1930s where a tube was used as a virtual choke (big inductors are like current sources, right?).

It was basically version of this one transistor gyrator circuit using a big power triode, like a 2A3 or whatever it was back then, strapped across the B+ supply:

--


----Android NewsGroup Reader---- 
http://usenet.sinaapp.com/

If you place it on top of a triode, and take the output from the top cathode, the audiophiles call it a 'mu follower'.

It's hard to make it work at DC, though...

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com

If the need is simply for a HV current sink, simply put a NPN current sink under the cathode of any old tube you have in the junk box.

I seem to recall that someone named Philbrick...

made an OpAmp out of toobs ;-) ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

I don't remember how much screen current a pentode needs. I wanted to keep the right-most resistor big, to keep Zout high.

In the cathode circuit, too!

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

I once did a second-order version of that, with just an emitter follower worth of gain, and it oscillated at low frequency. Seems I accidentally stumbeled onto the RCRC passive circuit that has gain >

1.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Depends on screen area and alignment with the grid. And obviously, operation, since if you turn away all the electrons headed to the plate (by dropping the plate voltage, or screening it off with another grid as in dual-control pentodes and heptodes), they have to go somewhere.

Most small signal pentodes are around Ig2 = 16-33% of Ik, at normal operating conditions.

(Because of the current steering effect of a suitable g3, you can make an LTP with a single pentode -- take output from g2 and a. Miller capacitances are all kinds of screwed up, of course.)

The alpha, if you will, is usefully close to 1. :)

Power tubes with aligned grids typically fare in the 20:1 or better range. A sweep might be capable of sinking 1A to the plate with only 50mA screen, and that with only 50V on the plate.

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com

I guess the cathode follower should follow the final output, with the neon after that. So there is no additional load on the output.

It's really hard to design stuff with just tubes. No PNPs, no zeners, no opamps, rotten diodes.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Now now -- the diodes are actually pretty good, if you don't mind the stupendous heater power. Here's a 6AL5 (single half) model:

And the graph as measured,

Note that it has about 4 decades of exp response, and the slope implies a Vth temperature several times room temp (though the cathode isn't actually that hot; I think the difference is the emission coefficient). It also generates power, a few microwatts peak (for a thermal efficiency around

0.0001%). This is different from a light shining on a silicon diode, because the reverse leakage is small.

Better made thermionic generators use cesium vapor (think phototubes with heat), which apparently can organize into Rydberg matter (i.e., molecules of cesium atoms, stuck together in high excitation states, therefore very large radii and very low work functions), though I don't think I've heard of them being better than single digit percent. (NASA's favorite RTGs of course are fully solid state, and have reasonable efficiency as TEGs go.)

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com

The neon lamp tube will work but the other tubes will not because you let the vacuum out.

All my designs are space-qualified.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Here's my entry. I'm sure we're allowed to use a cheap semi voltage regulator IC. It'll still have the special tube "sound".

Precision (1%) HV tube current source.

| +HV supply | P tube G -------, K | | | +-----, | \_|__ | | /_\ R2 | | +--' Iz = Vz - Vg/R2 | _|_ | | | IC +---|FB | | |___| ** R1 | +-----' Ics = Vfb / R1 + Iz | CS out

** The regulating IC can be an LM334 (62mV), or an LMV431 (1.24V, >55uA) or even a plain jane TLV431 (2.5V, >400uA for Iz).

--
 Thanks, 
    - Win

See...

A triode (12AU7) doesn't do as well until ~150V. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Ah, nice, you drive the cathode. In a current sink, this trades a zener diode for a +5 supply.

. | +HV supply . | . P tube . ,-- G . | K ,-----+5 . | | | . | | R2 . | \|___| . | ,| | . | | _|_ . | | | | IC . | +---|FB | . | | |___| ** . | R1 | . | +-----' Ics = Vfb / R1 . '---+ . | . CS out

But for a current source, not so simple.

Doesn't the feedback loop get it under control?

--
 Thanks, 
    - Win

The cathode-to-ground voltage gets too low... not enough headroom under the cathode for the control circuits... maybe I should try a triode and tie the gate to +5V ??

That might work. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

That fixes it rather nicely...

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.