Heater voltage filtering.

[industrialarts]

And the first switched mode power supplies (using vibrators) were invented.

When I was a young technician, a friend asked me to fix the tube AM radio in his 1960 Cadillac. Sad to say, it was one of the first things I encountered that I couldn't fix. It was something in the multivibrator circuit but I was not yet familiar with the ways of vintage technology, so I was lost.

 

[industrialarts]

And the first switched mode power supplies (using vibrators) were invented.

Cheers

Ian

When I was a young technician, a friend asked me to fix the tube AM radio in his 1960 Cadillac. Sad to say, it was one of the first things I encountered that I couldn't fix. It was something in the multivibrator circuit but I was not yet familiar with the ways of vintage technology, so I was lost.

 

[vmanj]

Yes, Id only go with the rechargeable types for cathode bias as its has to safely withstand the anode current of the tube under long term usage , I proportion the cell in relation to the anode current of the tube , so it just keeps topped up at around 1/10th nominal charge current .
coin cell is fine for grid though , also in 1.5v format .

Thank.
Is it possible to use "AAA" format - they are smaller, or preferably "AA" - large ones, so will Li-Ion type be good for this?

 

[abbey road d enfer]

Thank.
Is it possible to use "AAA" format - they are smaller, or preferably "AA" - large ones, so will Li-Ion type be good for this?

You could do much better. You could use the battery to provide negative grid bias. The current draw is nil, so the batteries would last very long (typically as much as the shelf life). You woudn't have to disconnect them when the mic is not in use.
Alternatively, a supercap would also work well.

 

[Tubetec]

Ive used Ni-cad or Nimh , an AAA cell is more than enough in terms of capacity even for a high current preamp . Normally they are charged for around 10 hours at 100ma .
A small Nimh memory backup battery is probably more appropriate to the kinds of anode currents in a tube mic .

A 40Mah Nimh cell like below is fine for the job . You may need to access the cell to measure the voltage once in a while or maybe even provide external charge or discharge to a pre determined voltage if it changes much over time .
A regular lithium button cell is perfectly usable as a grid battery with the minus end pointed up , the coin cell carrier makes maintenance easy enough , it doesnt need to charge like the cathode arrangement , and should give out a very stable voltage for years .

Id avoid Li-ion battery as a special charge regime is required and overcharging could be catastrophic . In any case a Li-ion cell puts out around 4.2 volts fully charged .
A Nimh cell on trickle charge will top out at around 1.45 volts eventually , but it might take weeks or months for an initial charge voltage of 1.25v to gradually rise to the final float charge voltage ,especially at the fraction of a mA anode current found in tube mics .
Any battery requires maintenance which is a major downside , lets say for some reason the battery voltage drops away considerably since the last time you used the mic and you power it up it may take a long time before the voltage stabilises where you want it to be .
A slotted coin cell holder of the kind below could easily be incorporated in a tube mic .



I was wondering about supercaps also , but do they have a habbit of leaking corrosive electrolyte all over the place when they go bad ?
It can happen with batteries to of course ,

A question to Abbey,
suppose we need 1V grid bias from a 1.5v cell can we use the traditional resistances of a few hundred MOhms to proportion the voltage without killing battery life too quickly ?

 

[vmanj]

You could do much better. You could use the battery to provide negative grid bias. The current draw is nil, so the batteries would last very long (typically as much as the shelf life). You woudn't have to disconnect them when the mic is not in use.

I already did a grid bias (like U67), but I liked the sound more when the offset was through the cathode.

 

[vmanj]

Alternatively, a supercap would also work well.

supercap is a high capacity capacitor shunting a cathode resistor?

 

[vmanj]

Ive used Ni-cad or Nimh , an AAA cell is more than enough in terms of capacity even for a high current preamp . Normally they are charged for around 10 hours at 100ma .
A small Nimh memory backup battery is probably more appropriate to the kinds of anode currents in a tube mic .

A 40Mah Nimh cell like below is fine for the job . You may need to access the cell to measure the voltage once in a while or maybe even provide external charge or discharge to a pre determined voltage if it changes much over time .
A regular lithium button cell is perfectly usable as a grid battery with the minus end pointed up , the coin cell carrier makes maintenance easy enough , it doesnt need to charge like the cathode arrangement , and should give out a very stable voltage for years .

Id avoid Li-ion battery as a special charge regime is required and overcharging could be catastrophic . In any case a Li-ion cell puts out around 4.2 volts fully charged .
A Nimh cell on trickle charge will top out at around 1.45 volts eventually , but it might take weeks or months for an initial charge voltage of 1.25v to gradually rise to the final float charge voltage ,especially at the fraction of a mA anode current found in tube mics .
Any battery requires maintenance which is a major downside , lets say for some reason the battery voltage drops away considerably since the last time you used the mic and you power it up it may take a long time before the voltage stabilises where you want it to be .
A slotted coin cell holder of the kind below could easily be incorporated in a tube mic .



I was wondering about supercaps also , but do they have a habbit of leaking corrosive electrolyte all over the place when they go bad ?
It can happen with batteries to of course ,

A question to Abbey,
suppose we need 1V grid bias from a 1.5v cell can we use the traditional resistances of a few hundred MOhms to proportion the voltage without killing battery life too quickly ?

Yes, it's not easy to maintain...

I tried connecting a regular, non-rechargeable battery (1.5V) to the cathode.
The microphone worked for 30 minutes, everything seems to be in order ..
Can anything serious happen to the battery (explosion, etc.) for a longer time (2-3 hours)?

 

[Lawrocker]

This thread has renewed my curiosity about the Neumann N52A supply. It used two sets of 'rechargeable' batteries and a choke for filtering the Filament voltage. I had marginal success replacing the fifty year old cells with custom voltage NiCD packs from an RC car battery supply outfit in the Bay area. Other than creating a new regulator, as was done with my other units, has anyone done a rebuild like that? Not having the specs of the original cells, there is a little guesswork there.

 

[Tubetec]

The small current involved probably makes little difference even to a non rechargeable cell , monitor on charge cell voltage over time and see what happens .
An alkaline cell is likely to have much better shelf life than a Nimh .

 

[Gus]

interesting paper on cell noise
https://tf.nist.gov/general/pdf/1133.pdf

 

[Tubetec]

I thought this was worth a look relating to bias battery ,
https://antiquewireless.org/wp-content/uploads/09-the_mallory_bias_cell.pdf

 

[Tubetec]

It mentions in the Nt52a spec that it can withstand temporary mains outages , the batteries continue to supply the heater voltage obviously , Ht voltage probably lasts for minutes after the mains supply has been removed due to the large filter caps , even by olden day standards (40uF).

Modern electrolytic caps with x10 times or more the capacity of old you may find HT voltage can be sustained for many minutes after power ceases to be applied , especially in something like a tube mic where a fraction of a mA is the usual current involved . The battery charger/heater arrangement with extra large HT caps might allow an extra quiet mode where the mains is disconnected via relay from the mic psu while in record.

I found this link with schematics with extra service voltage measurements shown , It also mentions the batteries being able power the unit during short term mains dropouts .
https://www.filmsoundsweden.se/backspegel/nat-n52a.html

 

[Tubetec]

I found this hammond Dc choke , 320mH 600ma
comes out around the same DC resistance as the one used in the N52a , 7 ohms by my calculation .
Its around 10 dollars to buy .

Thanks again for the docs on battery noise Gus ,
Its in nice clear language anyone could understand .

Seems like Ni-Cad is the winner in terms of noise ,
AA size apparently came out best in the testing ,
Nicads arent as common as they once were , but there are still plenty of cell packs available for emergency lighting systems and medical equipment , AA nicads arent that common any more , as in domestic use their likely to end up in the trash .

As far as I can tell the batteries in the N52a are 300mah , so maybe an hour or twos worth of heater supply max .
A modern Nicad AA cell has around 900mah ,so it could last several hours without the need to recharge . Larger 3000-4500mah cell packs are also available .

https://jandkaudiodesign.blogspot.com/2015/04/2-coil-choke.html
I was wondering how common mode filtering might be applied to a battery heater supply , there is a 0.1H common mode inductor easily available , dc resistance and current capacity are about right at 3.2ohms/450mA
https://uk.farnell.com/triad-magnetics/cmf16-104450-b
I was thinking of two three cell battery packs center tapped and connected to cathode/ground , then the grid bias tapped off one cell down on the minus side . Charge is applied to either end of the chain of cells .

 

[Tubetec]

I found this other Neumann supply , destined for MSC2(2 volt), Ac701(4 volt) depending on how the 2.25 ohm series resistor is connected .
It has a battery regulated heater supply similar to what you find in the N52a .

 

[Tubetec]

I wouldnt be at all surprised if the original Neumann Stablyt cells were mercury based , there isnt much more in the way of pictorial evidence of their existance because its all probably melted down into a pile of toxic waste by now .

The N52A claims 6uV of noise , obviously with mains voltage applied , thats very good compared to your average modern regulator supply on load .
Question is on battery heater and residual HT voltage what was the noise figure? better, a lot better I bet ,

 

[Gus]

https://repforums.prosoundweb.com/index.php?topic=37491.0

 

[Tubetec]

Thanks for the link Gus ,
I did discover one more stablyt based supply which seems destined for the U67 called the NUK , but I cant seem to find the schematic though , maybe someone here has it ?

 

[moamps]

.. As far as I can tell the batteries in the N52a are 300mah , so maybe an hour or twos worth of heater supply max .
A modern Nicad AA cell has around 900mah ,so it could last several hours without the need to recharge . Larger 3000-4500mah cell packs are also available ...

I’ve always wondered if the user manuals for Neumann power suplies that contain NiCad stabilizers suggested that the microphone be disconnected from the power supply when the power supply is disconnected from the mains. If this is not done, the batteries continue to power the heather of the vacuum tube until it is completely discharged, which is certainly not healthy for the batteries. Nor for the AC701. And even when the power supply is turned on again, if the batteries are empty, it probably takes some long time for that heating voltage to stabilize at critical 4V. I only had the opportunity to service that type of power supply once, and I changed the battery assembly with the LM317 and additional protection circuits.

 

[Tubetec]

Good point Mo ,
deeply discharging Ni-cads is problematic and needs to be avoided . Larger capacity cells would help a bit.
I was thinking about a means of cutting the mains power to the mic psu only while the DAW/recorder is in record mode , that way as soon as the take is over power is restored , keeping both heater batteries and HT supply well charged . With several extra large HT caps like 680uF it probably takes many hours for for the 0.5-1mA HT current of the mic to discharge the HT . Im not sure how to calculate the time frame of the HT discharge ,maybe someone can help ?

Yeah its fiddly and has all kinds of potential pitfalls using batteries , upside would be if even for a limited amount of time ,no mains noise at all in your mic circuit . Its definately not something you would want or need on every microphone , but for a favourite vocal or instrument mic all the extra hassle might be worth the effort .
Its seems Neumann had a patent on the stablyt cell arrangement , be nice to get the details and translate the original text .

The VF14 , Msc2 and Ac701k are probably too rare and precious to risk any kind of battery supply , but with an EF86 or 6AU6 based mic , it might be worth a try .
The 6uV of the N52a corresponds to a noise level of -104 dbv , it seems like Nicads might beat that by at least 50 db , of course were still left with a much higher noise level due to the circuit itself , but it means any contribution from the PSU would be insignifigant .

Thanks again to Lawrocker for reviving the battery heater supply idea .