triode + neon + volts = tremolo?

[duck_arse]

I have stolen researched this circuit from a 60's era australian (moody or vardis, probably a straight copy of a fender) amp schem and have it on my breadboard. it works, with the values shown, but has thrown up some questions and some observations.


and for the colour averse, as usual, this:.

I'm not smart enough to work out the voltages needed for the caps. C1 sees fullish B+ until warm, so 200V there, but C2 and C3 sit on top of the neon and R6, so what voltage is suitable for them? and C4 - I put the cro both ends of that and saw minimal voltage difference if any - what voltage suits there?

what sets the brightness of the neon, or is it a function of the particular part? does the discharge brightness ramp up/down, or does it discharge ON/discharge OFF?

observations - a 40 year old NE2 would not start with R6 less than 680k, but a newer one (pulled from a dead clothes iron) striked and ran with 470k. also, the discharge in the old neon wandered around the posts some, whereas the newer one just sat and covered one pole. the older neon showed some ?conduction spikes? (which happened to follow the discharge-wanders) on the cro trace at R6//R5, whereas the newer just turned on until it turned off.

why the odd caps? I look in my drawers and find very few usable values over 100V - hence the 165V. why an AX7? they are the only valves I have that will oscillate. why the circuit at all? it might end up being a tremolo, and just because.

any insights provided on theory of operation much appreciated.

[thermionix]

For C1-C3, just use 400v caps and for C4 50v should be more than enough.  A range of values will function for C1-C3, generally the larger the value, the slower the LFO.  Most common seems to be .02, .01, .01 (uF) respectively.  AKA the Fender values.

[duck_arse]

for those playing along, here are some waveforms, taken from the top of the neon (I think) in the circuit shown above.


this shows the difference between a new neon (the lower, flatter trace) and the old neon, with some spikes.


this shows a retrace of the older neon, to illustrate the dancing flame and the changing conduction spikes.

I have connected to an ldr and it does work as a tremolo, although I haven't listened to yet. the envelope seems very dependent on the ldr dwell characteristics, with a very sharp and deep pinch.


thanks, therm.

[thermionix]

the envelope seems very dependent on the ldr dwell characteristics

Ding!

[Kipper4]

"I have stolen researched this circuit from a 60's era ^australian (moody or vardis, probably a straight copy of a fender) amp schem"



I have little of worth to add except.

Hook up the trem and see if the Aussie distortion worky or no worky.

I have figuered out though through seeing your cro picture why mine wouldnt work well when i did my experiments. It was outside the range of my cro. So thanks for that.
Carry on.

[R.G.]

If you're using this as a neon light into an LDR for tremolo, you may want to read:
http://www.geofex.com/Article_Folders/sstremolo/sstremolo.htm
for an implementation of the Fender neon/LDR circuit. It can of course be done with a tube, but the MOSFETs are cheaper, don't need heaters, and don't wear out like tubes do.

Neons are insulators until the voltage stress on the gas causes it to ionize. When that happens, a cascade of ionization happens and the resistance of the gas drops suddenly with increases in current. There is a negative resistance region. When this bottoms out, the light emitted by the ionized gas depends on the current. More current, brighter light.

If you decrease the voltage on the ionized path below a certain "dropout voltage", the gas cannot continue to conduct, and reverts to non0-conducting.

[thermionix]

MOSFETs are cheaper, don't need heaters, and don't wear out like tubes do.

BUT require a power supply not already existent in an otherwise all-tube amp.  What else are you gonna do with all those noisy but still functional 12AX7s?

[duck_arse]

it's a funny thing you say, there, thermi - the 12AX7's I have here were given me by an amp repair man, from his box of shakes/rattles/rolls valves. they'll do me for the moment.

thanks RG, I had an idea you had something on neons, thought it was in mosfet follies.

an observation - the value of R6 varies the duty cycle of the neon flash, and I presume the light intensity. shifting that up to 2M2 gives from 35% - 46% on time, with the newer neon. added to a shunt trem circuit, it sounds pretty good, except for the tik ... tik .... tik ....

[thermionix]

The ticking can be dealt with.  Sometimes cured just by moving wires around.  Otherwise it can be shunted to ground with a cap.

[R.G.]

MOSFETs are cheaper, don't need heaters, and don't wear out like tubes do.

BUT require a power supply not already existent in an otherwise all-tube amp.  What else are you gonna do with all those noisy but still functional 12AX7s?

Actually, there are a number of MOSFETs, like the ones mentioned in the article and in MOSFET Follies, that can run quite happily on the 90-350Vdc environment of an otherwise all-tube amps. That was one of the points of MOSFET Follies.

No, given the proper choice of MOSFETs, they do not need a different power supply. They're a natural.

[thermionix]

Shows what I know about MOSFETs...jack snot!  Also realized later that I had assumed this trem circuit was going into an amp, where he actually never stated that.

[duck_arse]

ok - for those playing along, this is sitting on the breadboard, running on 9V feeding a 7555 based nixie supply, inductor wound on a PC psu drum ferrite, with 6V regulated heater supply. the 165V B+ is now feeding a B2 and a B3 supply, and after a pulldown/rebuild it still tikks.

it was never intended to go in an amp.

[thermionix]

Just for shiggles, try a 10nF cap across the neon.

Report back.

[duck_arse]

bad, very, very bad. well, I was trying a resistor in series w/ the neon, which helped, in a way, but messed up the light output. then I stuck a cap across the neon, and it rang.

so, the neon has left the oscillator, and now has it's own mosfet driver*. only 25% duty cycle at the moment, but NO sign (cro, not ear) of the spikes/tikks. messing with connection points and ground jackings is tomorrow, we'll see what is then.

* from RG's sstremolo linked article. except - when I tried 10M across the neon, it flashed, but the cro showed it was just a burst of damped ringing-like mess. 2M2 or 1M, the duty cycle the same, but the 1M has slightly slopier sides.

[thermionix]

Sorry, but what do you mean by "it rang"?

Also, what is (a) cro?

[Kipper4]

Cathode ray oscilloscope.

[thermionix]

Dig.  I guess "scope" is too long to type.

There's a tanning salon around here called Southern Rays Tanning.  I've always been tempted to call and ask to speak with Southern Ray.  Maybe he's related to Cathode Ray.

[duck_arse]

it rang. it took a single spike and multiplied it many times over, ringing like a bell. it is a bad thing, in this instance.

and cro, because it is a cro. scope is too long to ty

[duck_arse]

an update, a change of tack, and some bad news, everyone!

it seems to me that it isn't possible to have a pso flashing a neon and a standardish fender input stage in the same 12AX7 and not get massive ticking (600mV spike, input grounded). so I took the RG advice, and mosfetted the neon at a slight distance to the gain. the updated circuit is shown.



thanks also to transmogrifox, and his push for star ground, which is a bugger on the breadboard. I know this circuit works, and that there is not enough neon spike to trigger the cro, and that the trem depth is nice and deep (I saw it!), but didn't get to listen to the bloody thing, because all 3 of my (very second hand) valves have gone dead on the gain side. odd.

so, it isn't tuned, but it will work. the vibe version will now have to wait.

[edit :] corrected my paraphrase.

[thermionix]

all 3 of my (very second hand) valves have gone dead on the gain side. odd.

That is odd.  I don't see anything in your circuit that would be killing triodes.  And preamp tubes tend to go weak or too noisy to use sooner than going completely dead.  It does happen, but 3 in a row dying on the same side is suspicious.

These wouldn't be JJ brand tubes, would they?