Cheap power supply for AC701 tube microphones

[RuudNL]

This week a Neumann M269 microphone came in.
The owner knew that the microphone worked, but he didn’t have a power supply for it.
So, if I could build a correct power supply for him…
Because it is a lot of work to create a single power supply, I investigated the possibilities to use a simple ‘Chinese’ power supply, like the ones used with the Apex 460 for example. Those power supplies can deliver +120V HT and ~6.7 V (7806 + diode) for the filament. The AC701 wants only 4 V, so this would need a modification. Since we don’t have something like a 7804 regulator, I used the well known LM317 in a pretty standard configuration. But with a little extra: a ‘slow start’ to hopefully extend the life of the filament in the (expensive!) AC701.
Because I wanted the circuit to be 100% safe, I also added a crowbar circuit.
In case the output voltage becomes higher than 4.1 V (adjustable), the thyristor is triggered and blows the fuse. I have build the circuit on a small PCB, that can easily be fitted inside the 460 power supply.
The original transformer delivers 9.5 Volts AC, this results in about 12 V after rectification and smoothing.
To reduce the dissipation in the LM317, I added a 47 ohm resistor where the 7806 regulator used to be. (Between input and output.)
This works fine. Filament current is only 100 mA and everything stays cool.
Here is the schematic:

 

[Vac11]

Very nice, thanks for sharing!

Some noobie questions:
1)where did you connect your PCB in psu circuit? On the output of the heather circuit?
2)you desoldered 7806 regulator and connect 47R(1W???) to its  in an out, ok?
3) What thyristor did you use?
4) how exactly did you adjust 4.1V in the crowbar?
5) Is it possible to use this circuit for other tube with diffrent heather voltage?

Best

M.

 

[RuudNL]

1+2: I removed the 7806 and connected a 47 (or 33) ohm/ 1 Watt resistor between in and out.
Both sides of the 7806 are connected to a 1000 uF capacitor.
I took the voltage from what was the original 'output' side of the 7806. (You could leave the diode in place, it is no longer connected.) So in fact you now have: rectifier-1000uF-extra resistor-1000uF-wire to extra PCB.
3: I used a TYN612 thyristor, only because I had them in stock. But I suppose any thyristor that can handle 2+ amperes will do.
4: I removed the fuse and put a 6 V lamp in place. I adjusted the voltage to 4.1 V and adjusted the crowbar trimpot in such a way that the lamp lit. After that I checked this again by turning the output voltage slowly up and made sure that the lamp lit again at a voltage higher than 4.1 V. Of course you should adjust the voltage while measuring at the end of the cable, that is: inside the microphone.
(To compensate for the voltage loss in the cable.)
5: Yes, the only thing you would have to change is the 470 ohm resistor in the regulator and the 4.7K in the crowbar circuit. (If you need 6.3 V for exemple)

 

[MagnetoSound]

Nice work!

 

[Gus]

I would add the RC filter after the 317 not before it.

 

[Khron]

In case you haven't read carefully enough (or figured it out), that's there to "burn off" some of the (high-ish) voltage coming from the transformer, to save the regulator from having to dissipate all the power on its own.

Some might argue, especially if there's some filter capacitance in the mic itself, that the cabling makes up some post-regulator RC filtering itself.

I would add the RC filter after the 317 not before it.

 

[Gus]

Khon why the attack?
I just added a  well known refinement to using a 3 terminal regulator when being used as a regulator into a known more or less constant load (fil of a tube heater)
The R in the RC would be after the reg not before REMEMBER it is a series circuit and the power dissipated in the regulator would be the same with the resistor pre or post regulator

Maybe you should read about noise in 3 terminal regulators and how to reduce it.

 

[Khron]

At least in Ruud's "donor" power supply, he gets about 12V DC after rectification. Do you mean to tell me that the LM317 would be dissipating the same power regardless of whether it had a series resistor with its input OR output?

I would add the RC filter after the 317 not before it.

The transformer would need to provide the same power, true, but the difference between that 12V and, in this case the 4V output at the quoted 100mA, would be split by the "upstream" series resistor AND the LM317 (100mA through 47 ohms drops 4.7V and burns 470mW, leaving 3.3V across the regulator which only needs to dissipate the resulting 330mW).

Sliiiightly different numbers if we move that resistor to the output, wouldn't you agree? The 8V difference would mean 800mW dissipation for the regulator, which approaches the "this thing needs a heatsink" territory.

The original transformer delivers 9.5 Volts AC, this results in about 12 V after rectification and smoothing.
To reduce the dissipation in the LM317, I added a 47 ohm resistor where the 7806 regulator used to be. (Between input and output.)
This works fine. Filament current is only 100 mA and everything stays cool.

ADDING another RC filter on the output is another story altogether, which would, of course, attenuate whatever noise comes out of the regulator. The resulting lowering of the voltage might need re-tweaking of the voltage-adjusting resistor divider, though.

 

[RuudNL]

The modified Apex 460 power supply, with extra regulator PCB.

 

[Matador]

Because I wanted the circuit to be 100% safe, I also added a crowbar circuit.
In case the output voltage becomes higher than 4.1 V (adjustable), the thyristor is triggered and blows the fuse. I have build the circuit on a small PCB, that can easily be fitted inside the 460 power supply.

Why the voltage clamp?  Is the AC701 that sensitive to slight over-voltage on the heater rail?

 

[RuudNL]

The main reason is that the AC701 is expensive! 
I didn't want to risk that the filament voltage would become significantly higher than 4 Volts, in case there is something wrong with the regulator circuit.
(This could happen if a resistor goes 'open', or a bad slider contact of the trimpot. The unregulated input voltage is 12V!)
Only few components are needed and it gives a feeling of security.

 

[Khron]

Would a current-limiter not have been enough? Like a series 4.7-5.6 ohm resistor on the output, with a BJT sensing the voltage drop across it, controlling the base of the "slow-start" BJT?

 

[RuudNL]

Maybe. As always: there is more than one way to kill a cat...
An other consideration is that the filament of a tube behaves like a PTC. So detecting current could be misleading.
Low voltage: low resistance, high voltage: higher resistance, very high voltage: open circuit!  ;D
Well, I will keep it as it is now. It works as it should, and that is the most important thing.

 

[ln76d]

Don't see anything what could be changed in this simply circuit or why it should be changed. It's really well designed
Thanks Rudolph for sharing!

 

[granger.frederic]

take a look at the TPS7A4700 or TPS7A4701

has soft start, built-in protections, uses few components and has better specs (ultra low noise and very good ripple rejection for fixed bias operation, amongst other things...)

http://www.ti.com/lit/ds/symlink/tps7a47.pdf

for example : https://www.ebay.fr/itm/TPS7A4700-Ultralow-noise-4-Vrms-Power-Supply-1-4-20-5V-1A-for-Audio-DAC/253306365034?hash=item3afa3c646a:g:VEIAAOSwSlBYts6j

above all,  i would put a soft start on the HT, if you want to extend the tube's life...

 

[Matador]

Maybe. As always: there is more than one way to kill a cat...
An other consideration is that the filament of a tube behaves like a PTC. So detecting current could be misleading.
Low voltage: low resistance, high voltage: higher resistance, very high voltage: open circuit!  ;D
Well, I will keep it as it is now. It works as it should, and that is the most important thing.

Makes sense.  Food for thought: on a few of my tube supplies, I add a 5W Zener clamp to prevent polarization voltages from exceeding 70V or so.  You only need two components: a Zener diode and a current limiting resistor (to limit the current through the Zener under fault conditions).  Perhaps 20 cents of parts.

It would save you a fuse, fuse holder, three of the resistors, and the BJT as well.

 

[Gus]

a link
http://repforums.prosoundweb.com/index.php?topic=37130.0

 

[ln76d]

take a look at the TPS7A4700 or TPS7A4701

has soft start, built-in protections, uses few components and has better specs (ultra low noise and very good ripple rejection for fixed bias operation, amongst other things...)

http://www.ti.com/lit/ds/symlink/tps7a47.pdf

for example : https://www.ebay.fr/itm/TPS7A4700-Ultralow-noise-4-Vrms-Power-Supply-1-4-20-5V-1A-for-Audio-DAC/253306365034?hash=item3afa3c646a:g:VEIAAOSwSlBYts6j

above all,  i would put a soft start on the HT, if you want to extend the tube's life...

Nice, thanks Fred!

a link
http://repforums.prosoundweb.com/index.php?topic=37130.0

Also nice, thanks Gus!

 

[Papa Tango Charly]

Bonjour à tous
J'espère que tu vas bien
Pensez-vous qu'il serait possible d'apporter une modification à ce bloc d'alimentation Tbone (qui a une très bonne réputation
the t.bone PSU SCT 2000). Pour l'adaptateur pour alimenter un micro M49.
Merci pour vos commentaires?

 

[Papa Tango Charly]

https://www.thomann.de/fr/the_tbone_psu_sct_2000.htm