transformerless tube mic

[Larrchild]

Seems we are just a microcosm of a world gone mad these days.

I wouldn't leave, analag, cause yer over the hump here. More people "get" you than not.

For me it was quite recently, tho, :wink:

 

[Gus]

I just looked at the schematic. I see a phase splitter driven by a gain stage, yes I see the direct coupling.

We don't need any overall gain more than 1 with microphones. We don't get the "free" capsule voltage with the first stage of the Williamson.

What am I not seeing in the Williamson?

check the 4060 thread, look close at neumann schematics (KM5x etc)there is a german word "adjust" I believe for the adjusting of the final overall gain.

To be clear I don't have an interest in building an all tube transformerless microphone it is a silly idea to me. A hybrid might be fun for a different sound maybe good maybe not so good.

 

[Larrchild]

I built a totem-pole CF U-47 once. But just so I could put the output xfmr in the power supply, hehe.

 

[sismofyt]

[quote author="analag"]I was just about to post it...only to see this ********. F U too.
You think I need to **** with this or any other forum...watch this.

analag[/quote]

damn, you're a crybaby!

btw, you agreed to these terms when you signed up

"Leave the Flame-war mentality at another forum. Personal attacks as well as general hateful comments (regarding race, religion, gender, sex, etc...) will not be tolerated. "

so no F U

love ya :grin:

 

[sismofyt]

[quote author="sismofyt"] love ya :grin: [/quote]

sorry if i offended you man, no bad intentions :thumb: :sam: :wink:

come back and talk about something ... more interesting than that lame mic. let's talk about linelevel output mics instead. that's cool i think

jumping over that whole chain of gain, iron, lowimpedance conversion and whatnot do have it's own charm. i've often used the sennmann M147 and M149 straight into the tape / converters. those mics don't seem to care about the load anyway. it can sound damn nice. sometimes

what do you think, gus?

well, one could use a capsule, a tube and jam it straight into the converter chip. i wonder how that would sound .. ? (yes, i know the company not named gefell has such a similar thing on the market)

ok, no more (alfred e) neumann jokes :green:

 

[analag]

Here is a quickie and like Gus says, it's kind of a silly idea to begin with.
My prefence is a triode strapped pentode going into a good quality OPT.
Nice and simple is simply nice.

Anyway let's break this schemo down and see if it makes sense. In this case we are discussing configuration not the actual parts values, hence just Rs and Cs.

analag

 

[ioaudio]

i often heard the argument against line level mic is that its better to have mic and pre seperated, cause than you can choose and combine different ones.
but i´m not sure about that - one can even run a line level mic into a preamp to add some colour.
apparently i´m thinking about making a pair of small diaphragm mics, and i also thought about making them line level out.
how long could the cable between mic & psu/preamp be when i take a normal grounded cathode circuit like u67 without out transformer? just stay high impedance without cathode follower and such direct into the micpre, where it goes directly to grid.

sismofyt, what about a srpp mic? :grin:

 

[Svart]

What's with all this Tube **** anyway?

FETs or die *******!

 

[Gus]

I have thought about a two gain stage line level microphone with a vol control between stages.

Now the cool thing is the older model car, plane remote control servos(I have not checked what the new stuff does) have about 270 degrees of rotation and used about a 50 hz PWM 1.5ms 1/2 .5ma each way 1 to 2ms IIRC. One could build a 555, PIC circuit whatever and have a RC servo turn a real pot, only turning on the sevro circuit when when changing volume. One could remote pres etc.

I do think I might build a phase splitter tube front to BJT output hybrid just for fun not using a 12ax7.

The current is not balanced in the ouitput stage in the schematic because it is direct coupled one to the plate and the other to the cathode?

Change the 12au7 to something like a cap coupled IRF840 SF and have a hybrid high level microphone

 

[sismofyt]

[quote author="Gus"]I have thought about a two gain stage line level microphone with a vol control between stages.

Now the cool thing is the older model car, plane remote control servos(I have not checked what the new stuff does) have about 270 degrees of rotation and used about a 50 hz PWM 1.5ms 1/2 .5ma each way 1 to 2ms IIRC. One could build a 555, PIC circuit whatever and have a RC servo turn a real pot, only turning on the sevro circuit when when changing volume. One could remote pres etc.[/quote]

oh, fancy. i'm just experiemtning with an ordinary motor pot

 

[sismofyt]

[quote author="ioaudio"]i often heard the argument against line level mic is that its better to have mic and pre seperated, cause than you can choose and combine different ones.
but i´m not sure about that - one can even run a line level mic into a preamp to add some colour.
apparently i´m thinking about making a pair of small diaphragm mics, and i also thought about making them line level out.
how long could the cable between mic & psu/preamp be when i take a normal grounded cathode circuit like u67 without out transformer? just stay high impedance without cathode follower and such direct into the micpre, where it goes directly to grid.

sismofyt, what about a srpp mic? :grin:[/quote]

your hear a lot of things then. i haven't heard that, and even i had, i'd say wtf

the zout depends on the tube of course, but something like the 12B4A or the WE471A / 5842 can drive very long cables without problems. most problems arise when you use a multicore and / or are in a nasty (electrically) enviroment

i dunno what the zout of the 67 circuit is, that's an AC701, right?

i like to run my transmissions around 1k into 100k for all my devices, generally unbalanced (but often with iron). yes, that is very illigal :roll:

i'm tired of srpp

 

[Svart]

I'm not sure what you are doing with the servos and pots but I have some suggestions for you if you are interested.

digital pots are great things, I use them all the time. Most servos don't use pwm, they used to use pulses but they were not pwm. there is a fancy PIC program that inputs servo pulses directly from the receiver and outputs a PWM train that can be used pretty easily. I don't have the link but they gave you the source and the binary for the pic on some website.

 

[sismofyt]

why complicate matters so much? i use a motor driven ALPS with a couple of switches (on the remote box) to alter the +/-3VDC .. up it goes .. or down. i have no problem feeding that lot, B+ and heater through a cable to the mic

digital pot? i wouldn't wanna smoke that! :green:

 

[Larrchild]

digital pot? i wouldn't wanna smoke that!

No..you'll just start ordering digital pizzas.

I like the motor-pot idea. My line level mic had an internal gain trim between stage 1 and 2 but thats mucho better.=)

 

[Gus]

some stuff from googling

http://www.seattlerobotics.org/encoder/200009/S3003C.html

http://www.epanorama.net/circuits/servo10v.html

http://www.epanorama.net/documents/motor/rcservos.html

http://www.geocities.com/BourbonStreet/3220/servobasics.html

http://www.rc-cam.com/servotst.htm

http://www.pitt.edu/AFShome/s/o/sorc/public/html/robotics/handbook/ServoControl.htm

 

[PRR]

>> a tube microphone without a transformer output

> What would the output Z be on something like that?

The output Z can be anything we want. Put a hi-Mu tube in front and wrap feedback around the whole mess.

The question is: what is the LOAD? Are there no engineers here? This isn't rocket science. No degree needed.

On the one hand, "most" inputs are 2K. But some are lower.

On the other hand, sometimes we have 300 feet of cable. That can be 10,000pFd, 0.01uFd, or less than 1K at the top of the audio band.

What is the output level? Big diaphragms will often top 1 Volt on orchestral, and 2V or 3V or who-knows on close-up rock cymbals.

Being worst-case, we could need 3V across 500 ohms, or 6mA! Twice, because there isn't a real elegant way to do it in one string. When I say 6mA, that's idle, the tubes have to go down to zero and up to 12mA. With 12AU7-class tubes, we need 100V across the tube.

Analag's plan seems rational. Use a high supply voltage and take a gain of 20. Cathodyne. CF. Then use a resistor divider in each CF to give a loss of 40 with 150 ohms output impedance. Say 4K7 and 150 in series. The output will be about unity gain and about 300 ohms Zout, which is acceptable. The 12AU7 CFs will be "colored" by the heavy 5K load; they and the volt-amp will be colored by the ~40V signal swings. Heat in the can is 7 Watts, which is a hot can.

This could be refined, but with common tubes the heat load is hard to reduce.

You could consider 6DJ8 long-tail pair. The longtail would be a fine input for a condenser capsule. Run at 6mA per side. Put 150 ohms in each plate. The overall gain would be somewhat less than unity, Zout is again 300 ohms. There's no internal feedback for a triode working with such a low load, so you are up against the tube's raw nonlinearity. However there is even-order cancellation, so numeric THD isn't so bad, up to a point where it goes grossly clipped.

You could just admit that tubes and transformers go together like men and women, and any transformerless tube amp is probably some kind of lame or else over-built.

 

[Gus]

I was working in the constrants of the Gemini description I could find.

I don't like the extra gain of A's circuit. We only need 1 an overall gain of about 1.

If one draw's out the circuit below there is a way to get the heat down a little 1/2 of a 12ax7 and charge the capsule and be a phase inverter.

Some china microphones have 200V supplies that where 200V came from. Would I use a 12ax7 or au7 NO. I would use a different first tube(not a 6dj8) and semiconductors for the follower output If I used a follower out.


I don't like 6dj8s for tube microphones that grid "touches" the capsule I think it has to do with the element and grid winding spacing and possable Islanding.

This is kind of like semiconductor Schoeps circuit change to use tubes.


Lot of first guesses for china tubes I looked at curves of NOS ones

200V B+, 5.6 to 6.3 fils raised above ground would need to do some tests.

front end 1/2 12ax7 47K plate R
1.2K from cathode to another 47K to ground, 100uf or what ever bypass cap across the 1.2K (Look at a fender deluxe) Match the 47K if you want, the cap should drop the 1.2K out for AC balance

200meg to a 1gig, grid to 1.2K 47K node cathode leg

one side of the capsule to the grid other side ground.

film caps from plate and cathode nodes can be smaller because the 1 meg appears bigger than 1 meg to the first stage adjust to taste 1/2piRC

one to a cathode follower set about 3ma 1/2 12au7 470ohm cathode to a 33K to ground 1 meg from the 33k 470 ohm node to grid
cap couple out from node with maybe a 1meg pull down at the output of the cap, could be electro. Should be formed even with 48V phantom on.

same CF circuit using the other 1/2 of the tube hooked up to the cathode leg.

The front is a phase splitter like a shoeps fet but the cool thing it charges the capsule cardiod or omni. Look at AKG circuits.

200V seems to be common with china tube power supplies
looked at the 12ax7 12au7 curves and tossed some standard values at it .

I don't care to much for the DC points are because the signal might only be 1 or so volts

1st Phase splitter and capsule charge voltage 2nd CFs with some drive current. Maybe I am (whatever) but I don't think in terms of small signal output Z more in terms of advailbe current and voltage.

I am looking for at least 1V of input headroom and some drive from the push pull CF outputs.

BJT EF outputs would make more sense to me or even MOSFET SFs

 

[PRR]

One objection to your scheme: the cathode impedance of 12AU7 at 3mA is about 700 ohms each (Gm =~1,400uMho). So the mike's output impedance is around 1,426 ohms, rather than the 100-300 ohms we normally use.

On short line, no problem. I run medium long lines at 470 ohms. But in general use on long lines, lower would have less trouble with capacitance. And lower would give better bass response on transformer inputs.

We don't need voltage gain to get acceptable levels, but it is an interesting trick to use voltage-dividers to reduce output impedance. If we have 1V signals under a 200V supply rail, we can take some gain without pain.

Another issue is turn-on/off thump. (Don't tell me users will ALWAYS unplug the mike before the power supply.) The dividers limit the thump. Yours will have 100V transients, mine more like 1V.

Either plan can work. I'm not ready to rush either plan to production.

> BJT EF outputs would make more sense

I agree heartily. But this isn't, really, about technical issues. It is about how many tubes can be claimed on the Sales Brochure. WOW! Two tubes! It must sound twice as good as one tube!

And consider the lowly 1D8GT. Triode-PowerPentode in a bottle with a grid-cap. The common cathode (actually filament) is awkward, but something could be done, power is low, cost is very low.

 

[Gus]

I agree with the output of the circuit idea I posted being weak.

I do wonder what the Gemini circuit is?

I agree the output should run at higher current I GUESSED lower because of heat and the size of the microphone.

The point about driving a cable at 300 feet is a good one I wonder how the real Gemini circuit drives something like that.

 

[penquista]

The only thing i could find about the circuit used in the gemini was this published in the Music Tech Magazine review:

METHOD SPOT
Valve design
Though the distributors weren?t able to confirm it, it appears that the first half of the ECC83 triode provides the gain while the second half works in ?cathode follower? mode to lower impedance. The ECC82 is used as a phase splitter to provide the balanced output, making transformer and solid-state output circuitry unnecessary.


But maybe somebody could try the free 7-day loan service they offer and find out more.