R/E/P Community

R/E/P => Klaus Heyne's Mic Lab => Topic started by: afterlifestudios on July 03, 2019, 06:02:00 PM

Title: NG Supply indicator lamps
Post by: afterlifestudios on July 03, 2019, 06:02:00 PM
Is there any negative (other than visual) impact of using incandescent or led indicator bulb in an NG supply?  (I’ve read that neon can act as voltage regulating in some applications.  But in the NG supply, line voltage just goes across the lamp, so I’m thinking this isn’t one of those applications?)

One of my NG’s bulb socket has been swapped for a miniature beyonet (t-3 1/4 I believe) and I’m having trouble locating 110v neon in that base type.

Title: Re: NG Supply indicator lamps
Post by: uwe ret on July 03, 2019, 11:35:57 PM
Any type of 110V ac-powered illumination (Neon, incandescent, LED, fluorescent, plasma...) will work in this application. The sole considerations will aesthetics and physical fit. It may be easiest to find an LED fixture (lamp and socket).
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 04, 2019, 01:20:58 AM
Excellent.  Thank you, Uwe.  There are lots of available incandescent T-3 1/4 (9mm) base bayonet bulbs available near me for less than $1.
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 06, 2019, 01:29:32 PM
One more question.  What about the heat?  This 110v incandescent bulb gets very hot to the touch...  The lamp holder has a “plastic” (?) screw on cover. I guess I’ll keep a close eye on it today...
Title: Re: NG Supply indicator lamps
Post by: klaus on July 06, 2019, 02:10:26 PM
Some people like their pilot lights bight, others like it barely visible. Install a 2W resistor in series with the bulb feed, after temporarily soldering in a trimmer, to determine your preferred illumination level for the bulb.
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 10, 2019, 05:45:37 PM
Well.  I just discovered that the indicator lamp sockets in BOTH of my NG PSU’s are run off the 105v DC.  Not the 110vAC like the schematics I’ve seen show.  So perhaps the neon IS part of the DC stabilization.

(Both u47 systems were bought at different times from different people in different places. One is Tele badged, the other is Neumann, so the chances that this was a mod done by the same person is extremely slim. And all the wire used corresponded with the rest of the PSU.)

I do not have a schematic that shows this arrangement.

Title: Re: NG Supply indicator lamps
Post by: klaus on July 10, 2019, 06:36:19 PM
I cannot see how the pilot light in NG supplies could act as voltage stabilization: it's positioned on the AC side, pre-input transformer, rather than the DC side with its five heavy-duty filter caps. 

All of my official Neumann schematics for the NG and NG 1/2 show only AC bulbs, all positioned before rectification.
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 10, 2019, 06:58:32 PM
I was surprised too.  But that’s what I’m saying: It is positioned after rectification and is being feed DC.  A wire comes from the last 40uf cap and brought back to the indicatotor bulb socket. Then the other side of the indicator socket is brought back via another wire along that same path to ground at the mic connector.  It doesn’t exactly look stock.

Photo attached.

What are the odds that two separate previous owners in 2 separate countries did the same mod using the same method and layout etc?

Should I rip that stuff out and wire it as per the schematic?
Title: Re: NG Supply indicator lamps
Post by: uwe ret on July 10, 2019, 08:33:15 PM
The heat shrink tubing and the two square 1kΩ/5W power resistors are clear evidence that the power supply shown in the photo has been serviced and modified.

None of the various implementations of NG-style power supplies ever utilized the neon indicator bulb (or anything else) for voltage regulation/stabilisation. The current to fire the small indicator bulb is far to small to be of any effectiveness for voltage regulation, that task was accomplished via a type 150B2 gas discharge tube in early regulated supplies.

If you cannot source a neon bulb, to avoid potential heat damage from incandescent bulbs I recommend a light emitting diode (LED) to illuminate the indicator. With a suitable current limiting resistor a wide range of desired brightness can be achieved. There even may be available LEDs operating of 110Vac, just like the original Neon bulb.
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 10, 2019, 09:00:14 PM
Thank you Uwe. I will remove the modified b+ lamp wiring (green twisted wires in photo) and connect the lamp socket to the AC110 as the schematic shows.   I have 110v neon bulbs but they were not firing, which is how I ended up discovering this.  They would only fire if the PSU was not under load (mic not connected), because they were getting a LOT more voltage.

Does anyone have simple layout diagram or photo of how the AC lamp sockets were originally wired?  As I mentioned, BOTH my NG’s have this b+ indicator lamp mod so I don’t have a normal one to copy.

Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 11, 2019, 01:01:32 AM
Is it possible that the mod is for over voltage protection?
Or is that the same thing I already asked above?
Just trying to figure out why two separate tech’s did this mod...
Title: Re: NG Supply indicator lamps
Post by: klaus on July 11, 2019, 12:12:57 PM
They probably did this because:

1. Original Neon bulbs of the kind used in Neumann NG supplies are hard to find and now quite expensive.

2. DC bulbs are the norm for powering pilot lights in power supplies, so techs are familiar with it, even without a schematic (before they became widely available on the internet).

As mentioned by several posters: power supply pilot lights just indicate that power is getting to the unit, nothing else.
That does not mean that it's always done intelligently: early N52 supplies used a pilot light feed in series with the heater voltage output: when the lamp's filament broke, the mic no longer received heater voltage.
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 11, 2019, 01:29:27 PM
Thanks Klaus.  Both these NG’s have the bulb socket in parallel (not series) with the b+, and someone (two separate people) deliberately made the effort to do so.
I just thought before I put it back to stock (where it truly does just indicate if there’s AC making it past the switch), I’d try to decipher the reason for the mod. It doubt it was bulb scarcity. The socket was not changed on one NG, and the other was changed to bayonet base and came with 4 spare ne51 (neon) bayonet base lamps.  Also, running an incandescent bulb in parallel off the b+ would pull too much, so these were meant to be neon I think. 

My thoughts are the possible overvoltage protection that neon provides, but I don’t have the math to back that up.

Or maybe they wanted an indicator that shows you’re getting at least (n)VDC.  n being whatever the firing point of the neon bulb is.  Maybe that’s just a more useful indicator?

But if no one has seen this before, and nobody chimes in with strong support and reasoning for this mod, I will revert to stock.

Update:  Dennis from Tab Funkenwerk informed me that Oliver had come across a similar mod a couple of times and always reverted it to stock, sighting the impedance and need for purity of the b+.

So it's unanimous.  I'm heating up the soldering iron...  Thanks Klaus and Uwe.
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 12, 2019, 01:09:33 AM
Does a stock NG connect the lamp socket leads to the two terminals I’ve indicated on the terminal block in the attached photo?  (Not a photo of my PSU)

The Ac wires from the lamp socket in the photo I found at not twisted.  Would you advise twisting?

I’m assuming that if switching from 110v to 220v main power, I would need to switch to 220v lamp as well?  Or when the jumpers are set to 220v does the lamp see half the voltage as well?

Sorry for all the questions. Thanks for helping me get these back to original design.
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 13, 2019, 01:13:57 AM
Ok. I removed the wiring to the lamp from the DC signal. I rewired the lamp socket to the AC with a 100kohm resistor in series with the ne-51 neon lamp as per schematic.  It glows!

Then I measured b+. (First I drained the caps, as I was doing the other work with the PSU unloaded.)

With my DMM set to measure the b+ VDC, (presently 0V) I connected the mic and powered up.  The first thing I noticed was that it got up to about 80VDC faster than I think it did before? About 2 seconds or less.  Is that normal? Then it climbed a little more slowly up to 105.  Then it kept climbing to 107 at which point I powered off.   Time to redo the dropping resistors.

I originally had all 1kohm dropping resistors. I substituted a pot for R2 and found that 1.3kohm brought the b+ down to 105VDC.  Is it ok to have that much of a mismatch on R2 (the first resistor after rectifier)?  Or do they all need to be the same?

Right now it’s:
R2 = 1.3kohm
R3 = 1kohm
R4 = 1kohm
R5 = 1kohm

Or would adding 300ohm 10W between rectifier and first cap be better? (heatsinking it to chassis as well.)


Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 13, 2019, 04:12:24 PM
So I ended up with this dropping resistor configuration

R2 1.2k ohm
R3 1k ohm
R4 1.2k ohm
R5 1k ohm

B+ is 105.4 after 10 mins.
Mains voltage at wall is 120 VAC today.

The B+ ramp up to about 80 VDC is only a couple of seconds but then it takes about 60 seconds to slowly climb from 80VDC up to 105 VDC.   Is that about right?

Is there any problem with this dropping resistor configuration? 

I rewired the lamp socket in the other NG supply to return it to stock as well.  Indicator is working. B+ is now at 107.3 on that supply.  Should I leave that alone?  Mains voltage is 120 VAC today and is usually lower. (115VAC) and I’ve never seen it above 120 VAC.  So it maybe ok to leave the dropping resistors alone in the 2nd supply?

Interestingly, I must have a very well matched pair of VF14’s in these u47’s.  They both measure almost exactly the same depending on which PSU they are plugged into.

Ie. The Neumann badged u47 is 105.4 on PSU #1 and 107.4 on PSU#2
And the Tele badged u47 is 105.3 on PSU #1 and 107.4 on PSU #2

Lucky me?

Title: Re: NG Supply indicator lamps
Post by: uwe ret on July 14, 2019, 04:14:12 PM
Try not to overthink the NG power supply, it is a rather simple affair with no need or intention of regulation. The power indicator is connected across one half of the power transformer primaries. Thus it sees nominal 110 Vac regardless whether the mains settings are for 110 Vac or 220 Vac. The slightly excessive 105 Vdc (under load of a microphone with properly functioning VF14 tube or 2625 Ω/5W dummy load) is properly due to your AC-mains at 117 V rather than the nominal 110 V. Typically the tolerance permitted iss +/- 5% from the nominal 105 Vdc. No harm can be expected from adjusting the dropping resistor values to get closer to this target. Congratulations!
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 14, 2019, 04:56:11 PM
Excellent.  Thanks Uwe.  Good points.  Could you describe which terminals to connect the lamp socket to so that is only connected across 1/2 of the primaries in the PT?  I think I did that correctly, but would like to know for sure. Working great at 110VAC.

The only other thing I’m still seeking confirmation of is the speed at which the b+ rises to about 80VDC before it more slowly climbs to 105. Is it normal for those first 80VDC to happen so quickly like that?  (2 seconds).  I don’t want to do anything to potentially shorten VF14 life.

Then I promise I’ll stop posting every day whenever my b+ changes by one tenth of a volt! 😉

Thanks again.

Title: Re: NG Supply indicator lamps
Post by: uwe ret on July 14, 2019, 05:39:21 PM
The power indicator neon lamp and its 100 kΩ series resistors originally connect parallel to the blue and red transformer primary winding wires. BTW. alternately a connection to the brown and white transformer leads would work as well.
Several factors are attributing to the time it takes the operating to settle to the final value. Initially the tube filament is cold and no current is drawn. The voltage rise is primarily determined by the staggered time constants of the series R/C combinations of the filter chain plus the time constant formed by the series impedance of the transformer secondary and the internal resistance of the selenium rectifier with the first filter capacitor. It takes around 4 time constants to reach 98% of the ultimate charge of each R/C combination, and since they are staggered, 2 seconds to around 80% seems right. Once the tube starts conducting and current is drawn, it does slow the voltage rise proportionally.
I hope this helps ...
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 14, 2019, 08:03:11 PM
Thanks Uwe.  Yes. Very helpful.  So I believe that would be where the arrows are in this picture?

(Transformer primary lead wire colours are not consistent with schematic.)

Title: Re: NG Supply indicator lamps
Post by: klaus on July 14, 2019, 08:20:40 PM
You are getting the light to illuminate, so your connections are correct. No need to fuss with it further. Uwe (!) explained the mathematical calculation of the slow rise to nominal voltage. I concur from my experience.

Let's not focus on B+ or the heater supply alone (by the way, 107VDC will give you almost exactly 36V on the heater in many cases of NG + U47 + VF14 I have observed): a good portion of the health of the tube can be deduced from the plate voltage generated. Anything above 32VDC is fine, anything much below this will be audible as tube noise. Observe* that voltage 5 minutes after start-up, then again after a few hours.  After a long time on the shelf, sometimes a lower than nominal plate voltage of  VF14 will come back up, once the cathode is freed of deposits.

* Don't do it if you are not familiar with the mic's circuit. The voltage inside the mic can be deadly. Even experienced technicians will keep one hand behind their back when measuring tube mic voltages!
Title: Re: NG Supply indicator lamps
Post by: afterlifestudios on July 14, 2019, 08:31:14 PM
Thank you Klaus and Uwe. (So sorry Uwe!)

I was getting 40VDC at the plate when b+ was 102 last week, so I think I’m all set.  Most importantly, they sound incredible.

Unless any problems arise, I will check B+ at the PSU every 2 months just to keep an eye on things.

Aside: While I was searching for photos of insides of NG supplies online, I found this one.  Seems like red and brown PT leads are reversed from schematic?  I think it’s ok because it is strapped for 220VAC so they are still in series.  But if this thing got strapped for 110VAC, it would go poof, right?