DIY Tube Mic project

[Emmathom]

Hi !
After having built several microphone preamps (AOP / BJT / Tube) and tweaked a few commercial microphones (for example a pair of t.bone SC1100 with the help of @micolas > thanks to him for his valuable advice) today I want to build a tube mic from scratch.
I would like to point out that I am not trying to imitate or specifically copy an existing model but simply build myself a “good microphone”. A "do-it-all" microphone which would have a frequency response without exaggerated peaks, good sensitivity, a good signal/noise ratio...

I know that the manufacture of microphones has been the subject of many topics but because of this proliferation of scattered information I find complicated and risky to take information here and there and put it end to end. I also understood that the main component, the one which gives all its personality and all its qualities to a microphone is the capsule. On the condition of course of being supported by an efficient amplification circuit.

I therefore inform that I do not intend to acquire an overpriced capsule (for my budget in any case) such as a Thierst, ADK, Studio939... But perhaps an Arienne K87? a JLI TSC-2? a 3uaudio M7? The M7 Fs response posted by 3U seems quiet flat...
It's my first DIY tube mic built that's why I want to begin with something quiet simple (I do appreciate simple things as they often reveals quality)
I should point out that in terms of large capsule microphones I already own a pair of modified t.bone SC1100s (3 patterns), a Stam Audio U47Fet (very nice sounding), a pair of sE Electronics sE2300s (3 patterns). The sE2300 are on sale to finance this tube mic project. The SC1100 pair will soon receive Arienne K47 flat capsules which I recently got.

So as a base I thought of using a t.bone SC1100 for:
- it is very solid and its body is wide and high enough to accommodate a lamp + a transformer
- it has a good quality capsule (type U67) which can be kept initially (at least it is not bad at all)
- it can be found second-hand for less than €100
On the other hand, I don't know if its head basket is appropriate... I also notice that the capsule is offset towards the top of the head basket and not verticaly centered (?) but this can be restored if necessary by using another capsule support > although when listening this decentering does not cause any problem to my opinion

If this is possible (but it's a detail) I would like to use 12AX7 / 12AY7 because I have a few in stock and from different manufacturers (Sovtek / EH Gold / EI long-plate / JJ / Harma Cryo).
But this is not an obligation if another tube ref. is more suitable and since it's not a "magic tube" from the 50's or 60's that costs an arm or a leg.

I am posting here some diagrams that I have gleaned here and there and asking for your opinions and advice to help me with this project. I must say that all the schematics I encoutered are much closed to each other and that, as always with tube, there's no "mistery" in these schematics : a few resistors, 2 or 3 very good capacitors, a good Tx and a good capsules are "the ingredients" for a good result I guess.
I think the right biasing of the circuit & a proper wiring are the key to success...

Thank you in advance

 

[thor.zmt]

After having built several microphone preamps (AOP / BJT / Tube) and tweaked a few commercial microphones (for example a pair of t.bone SC1100 with the help of @micolas > thanks to him for his valuable advice) today I want to build a tube mic from scratch.

Good luck. You will probably need it. And you will need to build not just a Mic, because tube mic's pretty much do not run on Phantom, power, but you need to build a power supply and possibly control box.

And you need to run fairly high current (heaters) and lethal voltage connections to the Microphone, safely without electrocuting recording artists.

IF I may be so bold, you are trying to start a project I feel may leave you disappointed after a LOT of effort.

I would like to point out that I am not trying to imitate or specifically copy an existing model but simply build myself a “good microphone”. A "do-it-all" microphone which would have a frequency response without exaggerated peaks, good sensitivity, a good signal/noise ratio...

If I just wanted a Microphone that was a good "do it all" I would use one of those dedicated Microphone J-Fet's in a modified Shoeps circuit. I find these J-Fet's very "tube like" sounding, in fact more "tubey" than a 5670.

I build a few Mic's like that and people who used them at my place often preferred them to the stuff they dragged in.

It would also give you an exercise in building Mic's, an experimental platform and more experience.

Apparently China has now a LDC Electret Capsule that looks nice enough. This one I think. It has been discussed here and I remember they view was not altogether negative.

https://www.jlielectronics.com/microphone-capsules/jli-3412/



I'm sure the same capsule is on Taobao / Aliexpress for under a Pavarotti.

Get some really cheapo chinese junk mikes as "body donor" and put in your own gut's with this capsule. If you pay attention and avoid most of the mistakes chinese OEM's for many of relatively expensive Midmarket Microphones, you will probably have great all around Mic's at a "silly price".

You could even use multiple capsules back to back or vertically stacked for spherical or figure 8.

Once you get a flat measurement Mike (they come in USB these days) and a decent single driver wideband speaker you can even measure frequency response and add EQ circuits in the Mic to tame HF spittiness etc.

In the process you will learn a lot that you will be able to use building your dream tube mic and you end up with a bunch of dirt cheap mike's that will take some beating.

Lacquering the bodies in green hammertone will provide instant street cred.

I know that the manufacture of microphones has been the subject of many topics but because of this proliferation of scattered information I find complicated and risky to take information here and there and put it end to end. I also understood that the main component, the one which gives all its personality and all its qualities to a microphone is the capsule. On the condition of course of being supported by an efficient amplification circuit.

Correct. But actually it is the combination of Capsule, Head Basket assembly, mounting assembly and all that stuff that is responsible for the acoustic side. Once you make (say) 4 identical cheapo 32mm Electrets as described above and start experimenting, you will learn how this shapes sound.

I therefore inform that I do not intend to acquire an overpriced capsule (for my budget in any case) such as a Thierst, ADK, Studio939... But perhaps an Arienne K87? a JLI TSC-2? a 3uaudio M7? The M7 Fs response posted by 3U seems quiet flat...

Still a fair bit of bread plus you need capsule Bias. Another brick in the wall.

So as a base I thought of using a t.bone SC1100 for:
- it is very solid and its body is wide and high enough to accommodate a lamp + a transformer
- it has a good quality capsule (type U67) which can be kept initially (at least it is not bad at all)
- it can be found second-hand for less than €100

Maybe, maybe not. Next to the old Tube Mikes they seem small.

Look at the Mike-parts kit here and the size of that body:

https://microphone-parts.com/collections/microphone-kits/products/v47-tube-microphone-kit

Plus, why miss out on the opportunity to make something serious cool and vintage retro looking?

You could make a Mic in the old "suspended capsule" style with a tube attached to it, so it's in view. At least go for maximum koolness. It might also minimise head basket issues.

If this is possible (but it's a detail) I would like to use 12AX7 / 12AY7 because I have a few in stock and from different manufacturers (Sovtek / EH Gold / EI long-plate / JJ / Harma Cryo).

They are for many reasons uncommon in tube Mic's, despite they ubiquity, that alone should give reason for pause. These tubes are indirectly heated and fairly "hot".

In 2024 if I absolutely needed a tube in a Mic, I would probably use "subminiature" or pencil tubes, pick a nice 6AK5 parallel type pentode.

https://www.effectrode.com/knowledge-base/subminiature-tubes/

Or even use a 6AK5 (like here: Real simple tube mic w/ 6AK5 (Royerish)) ) they are just a bit chonky, like my Katz. I just got a pair of GE 6AK5 special quality US Version (JAN5654W) for a less than a Pavarotti for a Transformer plus Tube DAC output for a ESS DAC Chip.

The most simple and kool looking baby bottle with 6AK5 Mike gotta be the Swedish SELA T12. Runs on 60V & 5V (underpowered heater, trickey so-and-so's). Safe(ish) voltages, bias and Tube HT same, super simple.

https://www.filmsoundsweden.se/backspegel/sela.html

Well, the Transformer is going to be a problem. Good, really small and high ratio (10:1 is) are not excessively common.

A possible trick is to keep the transformer and most other parts out of the Mic-Body and put them into the power supply. If at that point you wire the tube as follower from the Mic, or perhaps as concertina splitter, you get low impedances in the multicore wire and potentially a very simple circuit.

Capsule voltage and anode voltage 60V, one wire.

Capsule backplate to grid, gridleak bias (add a diode if you must)

Cathode second wire.

Ground (case) and Heater two more wires.

So a 4/5 Pin Standard XLR type plug/socket will work, there are minimal parts in the microphone.

At the PSU you simply do +60V / -6V and use JFET CRD as Cathode load (say E562) plus coupling cap (Nichicon ES Muse?) and a 1:1 600 Ohm transformer (OEP?) if you insist on a transformer out all inside the (smallish) PSU Box.

Well, you wanted my thoughts. Sorry for the long shaggy dog who's owner I met in Saigon in 1973 story.

Thor

 

[Emmathom]

Well...
About 2 months ago I acquired two "Schoeps-like" circuits + two DC/DC circuits from a very nice & cool Englishman discovered during research on the net: Thet
https://www.halfshavedyaks.xyz/pcb/BM700-schoeps.php
I built and installed these circuits in a pair of sE Electronics X1S and they work very well.
With the original capsules of my t.bone SC1100 pair (that will be replaced by Arienne K47flat) I just miss 2 donor bodies to get two more 3 patterns good mics...

About what @thor.zmt said, I think I know how to design & wire safely HV in the mic, same for the PSU, that's not the problem... The problem is which "good-simple-efficient" schematic choose ?
I already thought about using a subminiature tube (6N17B / 5719 / 6H7B / etc) because they're small, just need a hundred volts or so and don't generate too much heat > qualities for a tube mic
I don't know 6AK5 and I'm gonna search for some infos...

 

[thor.zmt]

About 2 months ago I acquired two "Schoeps-like" circuits

Far from how I would do it. The problem is it's mostly a copy without questioning the circuit.

This circuit will work (with reduced max SPL) with P12. Which we do not need.

This circuit uses a high Idss and high Transconductance Fet, so it will not sound "tube like".

Using bipolar emitter followers there, where it should use small signal P-MOS.... Tsk, tsk, tsk.

This will sound like a cheap SS Mike.

I already thought about using a subminiature tube (6N17B / 5719 / 6H7B / etc) because they're small, just need a hundred volts or so and don't generate too much heat > qualities for a tube mic

Watch out in the datasheet. Some have very short lifespan, because they were designed for proximity fuses for AA Shells and similar application. You need the "computer grade"types.

I don't know 6AK5 and I'm gonna search for some infos...

It's a common Pro-Audio tube used in Microphone Pre's and Microphones. Also known as EF95.

It is 7-pin (so a bit smaller than a 12AX7 physically), 6V/175mA heater120V +B in triode mode.

The SinoSov 6J1 (6Ж1П-EB) version of it is currently big in small cheap tube preamp's for Hai Enten Fai.

Thor

 

[Emmathom]

Far from how I would do it. The problem is it's mostly a copy without questioning the circuit.
This circuit will work (with reduced max SPL) with P12. Which we do not need.
This circuit uses a high Idss and high Transconductance Fet, so it will not sound "tube like".
Using bipolar emitter followers there, where it should use small signal P-MOS.... Tsk, tsk, tsk.

So the Jfet is not well chosen and the output by bjt (a classic in many mics) is not a good choice...
Output by an P-Mos like IRF5210 for example ? source follower... ?

It's a common Pro-Audio tube used in Microphone Pre's and Microphones. Also known as EF95.
It is 7-pin (so a bit smaller than a 12AX7 physically), 6V/175mA heater120V +B in triode mode.
The SinoSov 6J1 (6Ж1П-EB) version of it is currently big in small cheap tube preamp's for Hai Enten Fai.

Have you please got a mic schematic using 6AK5 ?

>>> https://groupdiy.com/threads/s16-tube-microphone-6028-408-and-6ak5-5654.62423/

 

[thor.zmt]

So the Jfet is not well chosen

It is well chosen for a relatively low distortion, "not tube sounding" design, but it still gets killed by any Op-Alice for that sort of thing.

Using a 2SK660 or 2SK596 would be a better choice for "tube sounding" mic.

Output options, a few exist, BSS84 is low noise.

We now can use MOhm bising resistors and thus use coupling capacitors as small in value as 10nF without loss of bass response.

Next, we can maximise the output levels. Because tube mikes can take high spls without clipping, they just progressively distort more, we need to handle more than common.

We don't need P12/P24 compatibility. Let's set the output at 36V DC and the MOSFET Drain's at 24V.

This runs the Mic on 3.5mA which is more than enough for the front end.

With extra filtering, let's say, we have 22V for the J-Fet frontend, 11V for the J-Fet and 5.5V for each resistor maximise the available swing.

We operate the J-Fet at 0V bias, so it's gate biases up to ~ 5.5V. so Capsule bias is needed at ~65V.

We can now handle >3V rms without hard clipping. If we have a capsule with -34dB Sensitivity we get 20mV @ 94dB; 200mV @ 114dB and 2,000mV @ 134dB, SE we can handle 134dB without pad on the input.

A pad on the Mic Pre will be needed to handle the resulting 4V RMS @134dB SPL, but the mic will handle it, a few percent HD mainly H2 probably.

This gives the classic tube mic's that extra warmth, slight compression and great tone on close miked loud vocals. The classic sound of Lady Bird, Lady Ella, old Blue Eyes and Harry Belafonte.

Output by an P-Mos like IRF5210 for example ?

Nah, bad choice.

source follower... ?

Yes, instead of emitter follower.

The Shoeps design is super clever.

The 6.8k (nominal) phantom power resistors become the emitter (source) load resistors.

The output devices are connected (if done right) to drop a certain fixed amount of voltage, kind of like zener diode.

The current entering emitter/source exits from the collector/drain and there commonly finds a zener diode used as "shunt voltage regulator".

An alternative circuit can be made using a transistor (Capacitor follower) that filters and supplies the J-Fet. Using a small N-channel MOSFET again allows us high value resistors and relatively low value film capacitors.

As the J-Fet draws typically a few 100uA max a load resistors is added to draw the extra current. Or you can use the bias generator (plus resistor if needed).

Now we have 4 active parts, no zenners, no electrolytic capacitors, very stripped down circuit that will behave more "tubey" in all the good senses of meaning.

Otherwise an OPA1642 Op-Amp will be hard to beat for "clean" output.

Have you please got a mic schematic using 6AK5 ?

There were links in the thread, but I'd probably design from scratch, for a maximally stripped down and simplified design.

The biggest issue I see is the output transformer.

Thor

 

[Emmathom]

Hi Thor
Thanx for your inputs.
As I understand you're now speaking about improving a transistor circuit and no more about tubes...

1) So you mean that I use 2SK660 or 2SK596 as front Jfet and go output with a BSS84 (only available in SOT-23 > a bit tuff to solder but I think I can)
- 2SK660 = 24v at drain (B+ = 36v so 12v to loose by a divider ?)
- BSS84 = 11v at drain (another divider ?)
- Capsule voltage 65v
I don't know how to "design" such a circuit from scratch you know...

2) Yes I've read about using an AOP instead of a Jfet in front. OPA1642 is chosen for Alice project. Some also use OPA1611/1612... It's sure that it offers high Z input, very low Z output, "silence" when supply is well filtered

3) About 6AK5... I did some research and found interesting things...
- a mic schematic using this tube (poor quality pict but enough to read values)
- a thread talking about a "Royer modification" with a 6AK5 :
https://groupdiy.com/threads/real-simple-tube-mic-w-6ak5-royerish.57234/
- In76d thraed : https://groupdiy.com/threads/st12-6...-tube-microphone-in-the-world-full-pro.62601/
... and (very simple) schematic (attached) without any grid resistor (a hot debate ! with other members)
- a swedish brand (SELA) which first used 6AK5 and then moved to AC701 (far out of my budget!) :
https://www.filmsoundsweden.se/backspegel/sela.html

If the 6AK5 schematic attached is ok (with R4 1G instead of 2G > the guy put 2G because he lacks 1G but should have corrected his schematic) I could go on this way. I may use a Neurik NTE10-3 for 10:1 Tx output. I think it offers cardio & omni (by switching the back diaphragm or not) wich is sufficient (for me anyway) in most cases...

 

[thor.zmt]

As I understand you're now speaking about improving a transistor circuit and no more about tubes...

Not so much "improving" but "tubifying" the sound.

1) So you mean that I use 2SK660 or 2SK596 as front Jfet and go output with a BSS84 (only available in SOT-23 > a bit tuff to solder but I think I can)
- 2SK660 = 24v at drain (B+ = 36v so 12v to loose by a divider ?)
- BSS84 = 11v at drain (another divider ?)
- Capsule voltage 65v
I don't know how to "design" such a circuit from scratch you know...

Well, yes, re-design from scratch to get these results.

2) Yes I've read about using an AOP instead of a Jfet in front. OPA1642 is chosen for Alice project. Some also use OPA1611/1612... It's sure that it offers high Z input, very low Z output, "silence" when supply is well filtered

OPA1611/12 is a terrible choice.

As modern Op-Amp's have good supply rejection and low current, you can make a microphone with large dynamic range and very low distortion. The opposite in most ways of a tube microphone.

3) About 6AK5... I did some research and found interesting things...
- a mic schematic using this tube (poor quality pict but enough to read values)
- a thread talking about a "Royer modification" with a 6AK5 :
https://groupdiy.com/threads/real-simple-tube-mic-w-6ak5-royerish.57234/
- In76d thraed : https://groupdiy.com/threads/st12-6...-tube-microphone-in-the-world-full-pro.62601/
... and (very simple) schematic (attached) without any grid resistor (a hot debate ! with other members)
- a swedish brand (SELA) which first used 6AK5 and then moved to AC701 (far out of my budget!) :
https://www.filmsoundsweden.se/backspegel/sela.html

I linked Sela in my post. The Royer mod as well.

And yes, omitting the grid resistor is valid. This is called "grid leak biasing.



Note also that the heater voltage is reduced, which reduces the tube's transconductance, current capability and noise.

Extra trick, connect the transformer primary to the cathode, not ground.

Further trick, use the heater voltage to bias the Cathode, perhaps via a divider.

Further trick, replace the 100k with a "CRD" a j-fet connected as constant current source. There are some rated 100V max. More than 60V +B may be worthwhile.

Again, the transformer is going to be really hard to get and not cheap. At least not something decent. Of course that applies to any tube microphone.

You also need to handle the Capsule connection differently for 3-Patterns.

Something I might try instead might be this:

Cathode/Source follower, no transformer.

Thor

 

[LevinGuitar]

Using a grid leak bias, why not to ground the cathode? Since the tube will bias itself, what is the R1 function in the Sela ST12?

 

[Emmathom]

I wish I could design as fast and as easy as you Thor...

Thank you for this "smart" schematic using just 2 coupling caps (only) and no need for a Tx output.
And minimising the number of components inside the mic.
A very "pure" & light design. Great !

* As I understand, the phase is inverted at 6AK5 cathode output and then restored at BSS drain output
* C1 65pF is the capsule own capacitance (not a capa to place here)
* E562 > all I found is a kind of diode and so a dipole. I didn't find any fet under this ref...
* For the caps. polarisation 60v I can use the 70v B+ with a divider (large R values I guess like 470K + 82K ?) or use 70v straight (with good filtering)
* I have Muse BP 100uF 50v in stock for coupling output
* For 70v & 6v : 6AK5 datasheet says heater needs 175mA • 6v (so let's say I need more-or-less 350mA) and for the plate ? which current will it draw ? datasheet says max. 7,5mA but under 120v...
For 6v I could use a Meanwell SMPS RS15-12 (1,3A) (got it in stock) and lower to 6v with a 33Ω 2W or 3W (to be adjusted)
For 70v I could use a 2*18v Tr (got it in stock) + a voltage doubler (72*1,414 = 102v) and add a regulator LM317HV (TL783?) or lower the voltage by 4,3KΩ R serie (7,5mA^2 * 4300 = 250mW >> 1W)

And last but not least, which capsule use ??? Maybe start with the U67 style from one of my t.bone SC1100...

 

[thor.zmt]

Using a grid leak bias, why not to ground the cathode? Since the tube will bias itself, what is the R1 function in the Sela ST12?

That is a good question, it may be that without resistor the anode current was a bit high (and voltage low). I find circuits like this often are designed empirically. It may of course also just have been tradition and we can short it out.

Just ran a Sim, the 510R makes very little difference in the Sim.

Something that looks great though is bumping up the +B to 100V and using a E562 CRD as anode load. Still clips early for my taste (~ 1.4V RMS or +6dBu) on the output.

Thor

 

[thor.zmt]

And minimising the number of components inside the mic.

This was standard with many Capacitor Mike's from the 60's/70's I encountered....

* As I understand, the phase is inverted at 6AK5 cathode output and then restored at BSS drain output

No, no phase inversion at all.

* C1 65pF is the capsule own capacitance (not a capa to place here)

Correct, plus there is "generator" to get signal into the simulation.

* E562 > all I found is a kind of diode and so a dipole. I didn't find any fet under this ref...

It is a "Current Regulator Diode", inside it actually is a J-Fet with source and gate linked together. So I use a J-Fet as model. It is like a CCS in parallel with a few 100....1,000kOhm.

* For the caps. polarisation 60v I can use the 70v B+ with a divider (large R values I guess like 470K + 82K ?) or use 70v straight (with good filtering)

The tube grid will be a little positive, I'd go direct.

* I have Muse BP 100uF 50v in stock for coupling output

50V is not needed, 16V should be fine.

* For 70v & 6v : 6AK5 datasheet says heater needs 175mA • 6v (so let's say I need more-or-less 350mA)

You need 175mA +/-20% and around 5.6mA (E562) for the cathode follower tail and mosfet tail, so say 200mA max.

and for the plate ? which current will it draw ? datasheet says max. 7,5mA but under 120v...

Determined by E562 regardless of tube parameters.

For 6v I could use a Meanwell SMPS RS15-12 (1,3A) (got it in stock) and lower to 6v with a 33Ω 2W or 3W (to be adjusted)

For 70v I could use a 2*18v Tr (got it in stock) + a voltage doubler (72*1,414 = 102v) and add a regulator LM317HV (TL783?) or lower the voltage by 4,3KΩ R serie (7,5mA^2 * 4300 = 250mW >> 1W)

Or you could make a boost converter that uses the 6V to make 70V/11mA. Schematics exist, e.g. here:

https://threeneurons.wordpress.com/nixie-power-supply/hv-supply-kit/

And last but not least, which capsule use ??? Maybe start with the U67 style from one of my t.bone SC1100...

I would not expect a final product on the first try. So use "fungible" parts, where you don't mind burning the diaphragm with a soldering iron etc. - I'd use the cheap 32mm Electret for starters.

Thor

 

[rock soderstrom]

Something that looks great though is bumping up the +B to 100V and using a E562 CRD as anode load. Still clips early for my taste (~ 1.4V RMS or +6dBu) on the output.

SELA T12 clips relatively early with the 5654/6AK5/E95F (this type of tube produces nice harmonic distortions), that's part of the sound. I like it very much for music styles from the 50s and 60s.

I've built 6 of them in the last few months, some are now being used by newcomer bands who love them...from my point of view, the 10:1 mini transformer from Neutrik is a good match for the T12. Of course there are better OPTs out there, but the sound and price of this one fits the whole package very well.

Inexpensive Chinese K67 clones are a very good start and sound less treble emphasized with this circuit. Super cheap KK12s electrets are also very good for this kind of circuit.

The power supply can be realized conventionally with something like this project and a small tube transformer or you can use this PCB module, which also works well. Do not forget the PE!

SELA T12 is a very good project for beginners, as you can simply try out everything on it. Variable B+ and polarization voltage are fun, the rest is so straightforward that almost every (component) change is audible.

Everyone should have built a SELA T12, it's worth it.

 

[Emmathom]

That is a good question, it may be that without resistor the anode current was a bit high (and voltage low). I find circuits like this often are designed empirically. It may of course also just have been tradition and we can short it out.

Just ran a Sim, the 510R makes very little difference in the Sim.

Something that looks great though is bumping up the +B to 100V and using a E562 CRD as anode load. Still clips early for my taste (~ 1.4V RMS or +6dBu) on the output.

Thor

Should we go to 100v ? or stay with 70v as in your schematic ?

I've done some shopping yet.
BSS84 / 6AK5 (never used) on ebay / E562 / 22MΩ + 12MΩ (for 33MΩ > didn't get 33MΩ)
I went on 5 pin XLR (to have one more wire in case of) : 1 XLR 5 femelle chassis (PSU) + 1 XLR 5 male > I want to fix a cable straight out of the mic (no connector) since it's a real DIY concept it has no need to be "universaly compatible" and the less connection, the better. I'll put a fixed 3-5 meters cable (4 pairs shielded > Sommer Quantum Highflex Multipair 4 from Thomann) I could choose a 2 pairs but I want to have at least one more wire left (in case of) and the next step beyond 2 pairs is 4 pairs...

I forgot I've got 2 cheap Ali Express one diaphragm LDC caps and also a BM700-like empty body > these will do to begin with...

For PSU, if you don't mind, I'll use what I've got in stock.
1) HV : Transformer (2*18v)*2 for 70v [or (2*24v)*2 if 100v is chosen for B+] > I have them both in stock in 12VA (12w/72v = 160mA / 12w/96v = 125mA)
2) Heater + E562 + BSS84 : SMPS 12v 1,3A reduced to 6v (by R serie or regulator or zéner).

Let's go !

 

[Emmathom]

An option for PSU's... Of course different solutions are available.

Here HV is set by zéner and heater 6v (6,2v) set by level up the ground ref. by 1,2v on a 7805 fixed regulator.

But I also need -6v for E562 & BSS84 : I can't get +6v & -6v from the same PSU... can I feed the heater with -6v and ground ? I don't think so...

I'm not sure that the 10KΩ R is well calculated to provide 7,2mA (should I calculate 24v/10mA for each zener instead of the sum of them ?) and if 7,2mA is enough current for these 24v 1,3w zener (IZT 10mA / IZMax. 47mA)

Edit : Since the heater will draw around 175mA I don't need any zener polarisation resistance I guess... A polarisation R should be necessary for a voltage ref. (to a power transistor for ex.)

 

[thor.zmt]

SELA T12 clips relatively early with the 5654/6AK5/E95F (this type of tube produces nice harmonic distortions), that's part of the sound.

Also a result of low current/voltage in the tube

from my point of view, the 10:1 mini transformer from Neutrik is a good match for the T12.

I looked and have never seen a "Transformer Datasheet" so utterly devoid of ANY AND ALL information. Did you measure the frequency response and distortion levels?

SELA T12 is a very good project for beginners, as you can simply try out everything on it. Variable B+ and polarization voltage are fun, the rest is so straightforward that almost every (component) change is audible.

Everyone should have built a SELA T12, it's worth it.

Seconded and 100% agreed with.

Thor

 

[thor.zmt]

Should we go to 100v ? or stay with 70v as in your schematic ?

Just try it. At 100V we are pushing the E562 ratings

BSS84 / 6AK5 (never used) on ebay

Make sure to have a component tester available, the one's I bought were useless fakes.

If in doubt, a 2SA970 (where even fakes are decent) might be used instead, though it is a BJT.

I forgot I've got 2 cheap Ali Express one diaphragm LDC caps and also a BM700-like empty body > these will do to begin with...

Agreed.

For PSU, if you don't mind, I'll use what I've got in stock.

I do not mind at all. It's your project.

1) HV : Transformer (2*18v)*2 for 70v [or (2*24v)*2 if 100v is chosen for B+] > I have them both in stock in 12VA (12w/72v = 160mA / 12w/96v = 125mA)
2) Heater + E562 + BSS84 : SMPS 12v 1,3A reduced to 6v (by R serie or regulator or zéner).

7912 or so...

Here HV is set by zéner and heater 6v (6,2v) set by level up the ground ref. by 1,2v on a 7805 fixed regulator.

Note, heater is NEGATIVE and is meant to power the -6V!

So 7905. And perhaps make the diode(s) in the ground pin switchable and try how it runs at 5V, 5.6V and 6.2V. Together with the adjustable HT.

Thor

 

[Emmathom]

Make sure to have a component tester available, the one's I bought were useless fakes.

If in doubt, a 2SA970 (where even fakes are decent) might be used instead, though it is a BJT.

about bought on ebay I was meaning the 6AK5 tube, not the BSS84

7912 or so...

???

Note, heater is NEGATIVE and is meant to power the -6V!

So 7905.

ok for negative regulator and -6v for the heater as well as for BSS84 & E562 tube source

And perhaps make the diode(s) in the ground pin switchable and try how it runs at 5V, 5.6V and 6.2V. Together with the adjustable HT.
Thor

Together witg the adjustable HT ? don't understand... And yes I will test the output voltage with a test-charge to adjust it as closed as possible to -6v (but I think +5% is acceptable here)

May I ask again :
- do you agree with the HT schematic ? because there's some other ways to do else than zeners... does my 10KΩ for zener 24v polar. ok ? to set the min. current for zeners polar. do I must calculate the whole volatge (3*24v/7,2mA = 10KΩ) or one zener voltage (24v/7,2mA = 3,3KΩ) ? I would say the whole voltage is more logical...
- is the 6v schematic ok (except that I need a 7905) ? SMPS 12v followed by a R serie to loose some volts and filter and then a neg. regulator ? It avoids to buy a 230v-6v Tr (I don't even have a 230v-12v now in stock)
BR

 

[ran]

CIAO !
Dopo aver costruito diversi preamplificatori microfonici (AOP / BJT / Tube) e messo a punto alcuni microfoni commerciali (ad esempio una coppia di t.bone SC1100 con l'aiuto di @micolas > grazie a lui per il suo preziosi consigli) oggi voglio costruire da zero un microfono a valvole.
Ci tengo a precisare che non sto cercando di imitare o copiare appositamente un modello esistente ma semplicemente di costruirmi un “buon microfono”. Un microfono "tuttofare" che avrebbe una risposta in frequenza senza picchi esagerati, una buona sensibilità, un buon rapporto segnale/rumore...

So che la fabbricazione dei microfoni è stata oggetto di molti argomenti ma a causa di questa proliferazione di informazioni sparse trovo complicato e rischioso prendere informazioni qua e là e metterle fine per fine. Ho capito anche che il componente principale, quello che dà tutta la sua personalità e tutte le sue qualità ad un microfono è la capsula. A patto ovviamente di essere supportati da un efficiente circuito di amplificazione.

Informo quindi che non intendo acquistare una capsula troppo cara (per il mio budget in ogni caso) come una Thierst, ADK, Studio939... Ma forse una Arienne K87? un JLI TSC-2? un 3uaudio M7? La risposta dell'M7 Fs pubblicata da 3U sembra piuttosto piatta...
È il mio primo microfono valvolare fai da te, ecco perché voglio iniziare con qualcosa di abbastanza semplice (apprezzo le cose semplici perché spesso rivelano qualità)
Tengo a precisare che per quanto riguarda i microfoni a capsula larga possiedo già una coppia di t.bone SC1100 modificati (3 pattern), uno Stam Audio U47Fet (suono molto bello), una coppia di sE Electronics sE2300 (3 pattern). Gli sE2300 sono in vendita per finanziare questo progetto di microfono valvolare. La coppia SC1100 riceverà presto le capsule piatte Arienne K47 che ho ricevuto di recente.

Quindi come base ho pensato di utilizzare un t.bone SC1100 per:
- è molto solido e il suo corpo è largo e abbastanza alto per ospitare una lampada + un trasformatore
- ha una capsula di buona qualità (tipo U67) che può essere tenuta inizialmente (almeno non è affatto male)
- si trova usato a meno di 100€
D'altra parte non so se il cestello della testa sia adeguato... noto anche che la capsula è spostata verso l'alto del cestello della testa e non centrata verticalmente (?), ma questo può essere ripristinato se necessario utilizzando un altro supporto per la capsula > anche se durante l'ascolto questo decentramento non crea alcun problema secondo me

Se questo è possibile (ma è un dettaglio) vorrei usare 12AX7 / 12AY7 perché ne ho alcuni in stock e di diversi produttori (Sovtek / EH Gold / EI a piastra lunga / JJ / Harma Cryo).
Ma questo non è un obbligo se un altro tubo rif. è più adatto e poiché non è un "tubo magico" degli anni '50 o '60 che costa un braccio o una gamba.

Pubblico qui alcuni diagrammi che ho raccolto qua e là e chiedo le vostre opinioni e consigli per aiutarmi con questo progetto. Devo dire che tutti gli schemi che ho incontrato sono molto vicini tra loro e che, come sempre con le valvole, non c'è alcun "mistero" in questi schemi: qualche resistore, 2 o 3 ottimi condensatori, un buon Tx e una buona capsula sono "gli ingredienti" per un buon risultato immagino.
Penso che la giusta polarizzazione del circuito e un cablaggio adeguato siano la chiave del successo...

grazie in anticipo

 

[rock soderstrom]

I looked and have never seen a "Transformer Datasheet" so utterly devoid of ANY AND ALL information. Did you measure the frequency response and distortion levels?

This "datasheet" is a bad joke!

The transformer was recommended to me here on groupdiy years ago and I only wanted to use it to set up the mic and then replace it with something better. But then I liked it so much for this mic that I ordered more. Unfortunately, Thomann has recently discontinued these transformers, the price was really good.

I have not measured this transformer, the frequency response seems to be sufficiently broadband, I think the weakness of this transformer lies in the distortion due to its size. Although for many applications such as acoustic guitar, lead and backing vocal many other sources these transformers work astonishingly well for its size and money.

I wouldn't put them in a bass drum, although I'm sure there is someone out there who likes exactly that.

pinout for NTE 10.3 transformer attached