Help - U47 Regulated power for VF14

[maxwall]

I wanted to help a friend who wanted to see if it were possible to design a regulated power supply for a U47
with a VF14 tube . The power supply is currently outputting a unregulated 105.1 (+/-) 1-3 VDC to the microphone.
I thought about it and the TL783C regulator came to mind. Below is a schematic of how to implement the
regulator after the last filter capacitor or RC capacitor in the chain. Generally to get the voltage to a unregulated 105 VDC I would adjust the value of the last dropping resistor in the unregulated RC network (stock power supply). The value of that resistor is 1K, but found a resistor of around 850 to 950 Ohms would dial in the correct voltage for this tube microphone. I then would check voltages inside the mic to determine that the mic was distributing voltages correctly to the tube heater (VF14 - 36VDC ) and the plate ( 34 VDC ) and the capsule ( 63 VDC ) according to schematic.  And the mic sounds fine. However, can the 105 VDC be regulated or is it better to send a regulated 36 VDC to the heater removing the large 1780 dropping resistor inside the mic and leave the 105 side unregulated for plate voltages etc.

Schematic 1(above) , Also what is best for R1 and R2 to get 105VDC ?  R1=82 ohm R2= 6800 ohm (104.9 VDC regulated )

Schematic 2(below), U47 unmodified power supply ( unregulated )

Also should I use a 500ma or higher  120VACin to 125VDC out power transformer with a rc network that brings down the voltage to the TL783C regulator to around 106 VDC unregulated ?

Also , I noticed that Gyraf uses an array of 39V zener diodes to protect the regulator from unbalanced voltage swings. they sit in series above the protection diode. In my case I would probably use some zener diodes that were rated for a closer match to 105 V, yes ?

Schematic 3, Gyraf TL783 used for plate voltage(taken from G9 PS). How does R34,R35,R36 when plugged into the formula make for the regulated voltage needed to get to +245. Confused ? does R34 and R35 below combine as the equivilent to R2 and R36 as R1 in above data sheet ( schematic 1) Above ?

 

[bockaudio]

You'll get the best results from the stock supply. I't very well documented that your regulation "improvement" will not really improve anything.
I'm sure there are other DIY projects that will be much more worth your time.
regards,
David Bock
www.bockaudio.com

 

[Martin B. Kantola]

If you measure 105V going in to the microphone, you're fine. If your AC power changes a lot, you might want to to something about that anyway, some kind of power conditioner in your studio.

Your suggested modifications as such will not improve the microphone as David says, the original RC filters work really well. There also very little danger of damaging the expensive VF14 from some temporary overvoltage, remember that it's seriously underheated. But the original PSU provides a smooth, ramped voltage at startup, you might not want to mess with that.

To me, the only problem with the original PSU is that AC voltage changes can affect the sound from day to day.

Martin

 

[maxwall]

David,
thanks for the advice, it seemed like a worthy DIY experiment, but I realize its probably been tried many times
since 1949 or so.

To me, the only problem with the original PSU is that AC voltage changes can affect the sound from day to day.

Martin

Martin ,
I suppose the AC mains would be more effected in a commercial or undustrial zoned area rather than residential , but not in every case. I have monitored my mains AC before with less than +/- 1% fluctuation, provided my test equipment was'nt fibbing on me.

It would still be interesting to implement the idea as a voltage regulator experiment , sounds like it could be useful on other tube gear , not just mics.

 

[Martin B. Kantola]

If you want to experiment, using your friend's U47 and VF14 has its risks...

Anyway, since this is the drawing board, why not put the voltage regulator first instead, before the RC filters? Another idea is to replace R2 in the original with a current regulator.

Martin

 

[maxwall]

Current regulator is listed on the datasheet will give this a try. I think the mic draws low current ~40ma

RC after the regulator , for further filtering and voltage smoothing , effects on mic frequency response due to quality of the dc voltage ?

Yes, might be risky, but I think I would put a dummy load on the PS first and take current and voltage measurements first. The heater would be an easy dummy load to mock up on the bench, not sure about the plate loading.

 

[Martin B. Kantola]

Current regulator is listed on the datasheet will give this a try. I think the mic draws low current ~40ma

To clarify, the idea is that with the correct current set you will also have the correct voltage entering the microphone. Be aware of the voltage ratings of your regulator and that there are lethal voltages here!

RC after the regulator , for further filtering and voltage smoothing ?

Yes, ramping and filtering. Microphones need low noise power, regulators can be surprisingly noisy. Hum rejection is far from everything.

Yes, might be risky, but I think I would put a dummy load on the PS first and take current and voltage measurements first. The heater would be an easy dummy load, not sure about the plate loading.

Good idea to try with a 2k62 or so dummy load (resistor) first. 10W maybe. Still feel you're fixing something that isn't broke

Martin

 

[Gus]

This so easy just use OHMs law.  105VDC / .040amps gives the load resistor for the TOTAL u47 microphone circuit. Next use DC power math for the wattage.  Like Martin posted.

Also in the past stuff about power supplies for microphone were posted about here.  B posted a very good link for power supplies and noise.

Why do you want a three term reg? other stuff works "better" but you need to read up on power supplies.

Part of the beauty of the stock supply is if the caps fail open the series resistors still current and voltage limit and you will hear them start to fail with noise in the DC. 

When an active device fails, open what happens? shorted what can happen? whats a vf14 u47 go for?

Just because a design is old does not mean the designers did not do a very good job at the design.

The plate current part is low I believe it is even marked on a schematic you can find on the web so you don't even have to measure or do math.

You can use a variac and adjust your AC in.  Yes the variac costs money but what is the value of a vf14 or u47?

I think it is important to understand a design before changing it.

 

[maxwall]

Martin,

I suppose this is considered 'not broken'.   Notwithstanding,  I set up the u47 mic with a voltage test and watched where the voltage settled after warm up (1/2 hour). Due to the powersupply being unregulated naturally you'll see voltage fluctuations, which is what I have been observing lately. And may be attributed to the slight day to day inconsistency in sound you've described. Although the mic should run at 105 vdc it rarely stays put at that desired voltage in a consistant manner. It might rise to 105vdc on power up then level off at between 101.5 to 104.3 VDC constantly moving up or down slightly. I have tried adjusting the last resistor (1k) down to a lower value to get the volage to settle closer to 105 VDC, but its tricky working without tight regulation. I thought of putting a 2k resistor in parallel with a 2k 2W reostat so I could make more precise adjustment to the final output voltage. Its nearly impossible to dial in the correct voltage with standard over the counter 5% and 1% fixed resistor values. Ultimately, I was thinking that a regulated supply would tighten up this fluctuation and keep the mic more constant sounding. I think this was my objective in the first place, only your suggestions help confirm it. The only issue is making a regulated circuit that will not harm the filament of the VF14 in case of a power supply short/ open condition as stated earlier by Gus. Where runaway current in a PS failure could ruin something valuable.

Also just tried a power regulator (Sola),  It does improve the sag by about 2% from just plugging direct to the wall outlet. It now sits closer to 105 vdc, with about +/- .8 v fluctuation. Keep in mind this is not like one of those rack mount power conditioners , this one actually regulates and conditions the voltage from a wall outlet. there is a big difference. After trying two different length cables ( 10ft vs. 25ft ) for the mic there was a no additional difference in vdc drop.

so my wall outlet is flucuating AC much more than I realized, like you suggested.

I can say with confidence the original power supply,  as good as it is designed, is very sensitive to line voltage as Martin said, so if regulation can be introduced then maybe more consistency in they way the mic sounds, maybe , maybe not. I listened to the mic at different voltage levels 101 -105vdc and did not hear any appreciable difference in the overall sound.  So David Bock's point is noteworthy.

sometimes you set out to do something noble and find yourself no further along than where you started. At least , I can say it with experience from the test bench , rather than not.

In my opinion , I think I would install a small volt meter display in the power supply box and a rheostat in parallel over the last dropping resistor so I could make adjustments in voltage at any time to keep the 105VDC as close as possible to Neumann specs. Much like biasing a power output tube.

 

[o3misha]

Regulated power supply can give better results. I suggest you refer to A.Grosser. His PSU is made with the regulation transistors. PSU is incredibly quiet. Since the chain filter is resin film capacitors  - the microphone sound is much cleaner (in the original NG used electrolytic capacitors, which are much worse than film caps for sound ). If you have 101VDC voltage, -than you have  increased level of K3 distortion . Not much. But it is noticeable on loud sound sources. If you increase the feeding(say, up to 107-110 volts) then the level distortion k3 decreases slightly.  May also decrease the noise, since the increase of the heater voltage better effect on emissions. Do not worry about the rapid deterioration of the lamp, as the intensity increases by only 2-2.5 volts. So, you may just to decrease value of last R3 in NG . Also, I tried PSU with tube rectifier instead diod rectifier . This conversion method used in tube technology frequently. For example, in guitar amps. Tube  rectifier slightly softens the sound. Obviously, this needs a transformer auxiliary winding to supply heating kenotron.

 

[emrr]

Although the mic should run at 105 vdc it rarely stays put at that desired voltage in a consistant manner. It might rise to 105vdc on power up then level off at between 101.5 to 104.3 VDC constantly moving up or down slightly.

I listened to the mic at different voltage levels 101 -105vdc and did not hear any appreciable difference in the overall sound.  So David Bock's point is noteworthy.

Ah, a revived older post.  Love it.  Anyway, I'd be extremely surprised to hear a sonic difference with 4% B+ voltage change on any class A tube device.  That's what they call 'negligible'.