Need better #47 Lamps

Discussion in 'Technical' started by W5HRO, Dec 31, 2015.

  1. W5HRO

    W5HRO Administrator

    I've still been having trouble keeping the #47 lamps in my NC-303 receiver from burning out. I have them DC regulated and a 1-ohm series resistor to knock the voltage down to just under 6.0V. I've been tempted to retrofit it with LEDs instead. It took almost 1 year, but all 3 of the dial lamps burned out. One burned out last summer, then the next one burned out about 3 months ago then the last one early this week. Well, I replaced all 3 of them again on Tuesday and one of them has already burned out just today.

    Anyway, I think it's just the cheap ass lamps I bought online and they are the same kind you get from places like Antique Electronic Supply, etc. and they are probably made in China. I searched around and found the #755 lamps below. These are made by GE and they have a 20,000 hr average life span. They also make the #47's too, but those only have the 3000 hr life span. I'm hoping these will last, but I guess I will find out.

    Does anyone else know any that will last longer that are not made in China? If not and these #755 lamps don't last I might just retrofit the receiver with LEDs instead.

    6V.jpg

    Technical Specs
    • Item Miniature Incandescent Bulb
    • Overall Lamp Shape T
    • Trade Number 755
    • Lamp Shape T3 1/4
    • Base Type Miniature Bayonet (BA9s)
    • Voltage 6.3
    • Current Drawn 150mA
    • Watts 1
    • Average Life 20,000 hr.
    • Initial Lumens 1
    • Light Color Clear
    • Max. Overall Length1.19"
    • Lamp Type Miniature-T
    • Standards Meets 2012 Lighting Legislation Energy
    • Primary Application Indicator
    .
     
  2. W6MQI

    W6MQI Member

    Maybe try these guys http://www.marcospecialties.com/pinball-parts/05-44B-HO I've purchased some of their LED's retrofits thinking they would be as bright as an incandescent bulb not so at least the ones I picked out. Looks like now there's some more options available to pick from I might try again would like to change the 303 color to blue like the Drakes. I know these are #44 they may have #47 in white as well.
     
  3. W5HRO

    W5HRO Administrator

    Excellent..!!! Thank You...

    I just ordered 12 of the wide-angle ones below. I didn't even know they made these.

    #44_LED.jpg

    LED LAMP #44 BASE 6V WHITE WIDE
    New LED lamps provide upgraded lighting for your pinball machine or arcade game. They are state of the art, "green" technology devices that make your machine cooler, brighter, more efficient and more attractive with with a great selection of colors and diffusion patterns. LED's have very long life expectancies making them excellent replacements for difficult access playfield areas. They are vibration resistant and generate little or no heat -- ideal upgrades for pop bumper lamps.

    This white lamp is a replacement for #44 and #47 bayonet based miniature incandescent lamps. It operates in 6.3 volts AC or DC circuits and has excellent color rendition with a wide 120º dispersion pattern. Great for use in playfields, backboxes, coin doors in general or switched illumination applications.
    .
     
  4. W6MQI

    W6MQI Member

    Sure...Let me know how they work out like I said some that I purchased weren't very bright even though the description stated so.
     
  5. W5HRO

    W5HRO Administrator

    Even if they are not quite as bright and you can see the dial and they don't burn out then I will be happy.

    I have the other standard #755 lamps coming too, but it’s a PITA to keep taking that dial off. I know the trick on how to do it without scratching it or loosing those tiny 6 screws, but still, it's a PITA!

    What I was also thinking was to cut a long rectangle strip of perf or copper board and solder some (more than 3) of those big jumbo square surface mount LEDs on it and mount it on the back of the dial plate. You can remove those 3 bayonet lamp sockets on it.

    I'll try these wide-angle replacements first though and see what happens. I leave the receiver on when I get up in the morning until I go to bed at night when I'm home so they need to last.

    .
     
  6. W5HRO

    W5HRO Administrator

    This topic was moved here because it's better suited for technical discussions I think. The dial and/or meter lamp reliability issue is a common problem with most vintage equipment.

    I’ll post my results after I get and plugin the wide-angle LED ones to see if they are bright enough.
    .
     
  7. WD5JKO

    WD5JKO Member

    I made one prototype which is still quite crude. This was intended for 6.3 vac consisting of a 300 ohm resistor series feeding two LED's back to back in parallel. With one led, use a diode across it in reverse direction.

    Jim
    Wd5JKO

    led_47.jpg
    .
     
  8. W5HRO

    W5HRO Administrator

    It that something you did long ago or just recently?

    My #755 lamps will be here tomorrow and then the wide-angle LED's probably on Thursday. I think part of the problem is the quality of the lamps being made today really sucks compared to the ones made back in the old days. That first pic I posted is a #44 and has the brass base. You can tell by looking it was made very well. The new #755 lamps I ordered are probably silver in color and not as long lasting, but I won’t know that until I receive and look at them first.

    I searched around and you can get some really bright LED automotive lamps and they make 6V ones too, but they probably won’t last. My 2011 Honda Accord has those for the rear brake lights, but they burn out all of the time. That might be because they are 12V lights though and with the alternator going it’s more like 14V. Even LEDs won’t last if you run the voltage up too high.

    P.S. #44 lamps are 6.0V 2W lamps and #47 lamps are 6.3V 1W lamps so I wouldn’t run #44’s with the extra 0.3V on them. If using #44 LED’s retrofits then use a small series resistor to drop the voltage down to no more than the 6.0V max. Many of the old vintage receivers run the filaments way higher than 6.3V anyway so you need to watch out for that. I think it has to do with today’s AC line voltage being a bit higher than in the old days when the stuff was made so the filament transformers run hotter at their outputs now. Something originally designed for 110V to 115V on the primary is going to output a higher voltage with 130 to 140V on it instead. With my Fluke meter plugged into the wall outlet switched to the AC RMS scale I often see peak averages of 140V when the line voltage is up all of the way. Most of the time it's around 130V, but it does rise up to 140V during low load conditions which is 280V peak averages on the main running to my house from PG&E.

    .
    .
     
  9. W5HRO

    W5HRO Administrator

    Actually I spoke too soon, my #755 lamps were on my doorstep when I arrived home today. They have brass bases and look pretty good. The box says made in Hungary. Don't know whether that's a bad or a good thing.

    box_front.jpg
    box_back.jpg
    lamps.jpg
    .
     
  10. WD5JKO

    WD5JKO Member

    I did that experiment last summer. I bought 100 LED's on Ebay from China. These came 25 each in 4 colors, and came with 100 resistors at 300 ohms. All for $3.99 and free shipping from China. How the hell do they make money? The LED's seem to adhere to old style LED thresholds of 1.1v, and 1.5V (+/-) operating. Not related to this thread, but LED's of decades ago could be used as a low leakage voltage clamp. At 1/2v forward, the leakage is < a few picoamperes with an abrupt threshold around 1.1v. The low leakage was when dark since they do act as a crappy photo diode when in the presence of light. I remember some PCB's using LED's as clamps in high gain instrumentation amplifiers. The Led's were red, and had black heatshrink over them. I once played with LED's in a delayed AVC (DAVC) circuit in a RCA AR-88. The DAVC using low leakage LED's in series worked, but keeping the RF stages at full gain until some threshold was tricky since cross modulation was the result. I decided against this since it did more harm than good. The LED's though worked when zeners or silicon diodes would not have worked since the circuit impedance was too high.

    Jim
    Wd5JKO
     
  11. W5HRO

    W5HRO Administrator

    Ok, the LED's came early too and not too shabby. They are a little dimmer, but I can live with it. If I removed the 1-ohm series resistor on the output of the regulator they would be a bit brighter, but they are good enough with only 5.8V.

    I'm going to modify the S-meter lamp clip-bracket because that's always been an issue which makes the meter lighting never as bright. I will see how long these last. I hoping more than two years, but I will find out.

    The funny thing too is the white LED 's eliminate the 55+ years of dial and meter age yellowing and they look new again :mrgreen:

    dial.jpg
    rcvr.jpg
    leds.jpg
    .
     
  12. W6MQI

    W6MQI Member

    Right on! looks good I still may do this mod the 303 is my only rig that I haven't changed out the lamps to LED's. Still thinking about modding the Viking II meter to accept an LED somehow like the Ranger. Also thought about drilling a small hole in the brown plastic bezel in the plate dial window, and installing a small LED to shine down on the dial scale. Never liked the fact the Viking had no lighted dials, and meters.

    P.S. Been using the 303 with no issues since fixing it.
     
  13. W5HRO

    W5HRO Administrator

    If you want them brighter try the 6.3VAC and see what they look like. Just check the voltage first because with todays higher line voltage it’s going to be higher and also since you removed 4 of the tubes, which lightened the load on the transformer. The LED's might still last with 6.3V instead of 6.0V, but there's only one way to find out. Just add a series resistor to drop it down to no more than 6.3VRMS max and see what happens.
    .
     
  14. WD5JKO

    WD5JKO Member

    Are they rated for AC operation? I ask because DC operation in the most basic form (resistor-led in series) needs to get the polarity correct, and AC operation with one LED needs a diode to keep the LED from being reverse biased on the half swing where no light is emitted.

    Jim
    Wd5JKO
     
  15. W5HRO

    W5HRO Administrator

    This white lamp is a replacement for #44 and #47 bayonet based miniature incandescent lamps. It operates in 6.3 volts AC or DC circuits and has excellent color rendition with a wide 120º dispersion pattern.

    That's what the specs say
     
  16. W5HRO

    W5HRO Administrator

    I have an update on these LED lamps below that were discussed here, they are crap!

    They just do not seem to hold up with 6.0V regulated DC on them. They start out very bright and then get dimmer after a few days and start to flicker. Then they crap out shortly after.

    They are either Chinese made crap or they just don't like direct current applied and want alternating current instead. That could be possible, but beware...


    leds.jpg
    .
     
  17. W5HRO

    W5HRO Administrator

    I should have updated this topic long ago, but the lamps in my NC-303 receiver project turned out to be that the lamps were acting like a capacitive load on the regulator so it was NOT the LED lamps themselves. I was having the same problem with the stock lamps. What happened was there was a ton of AC ripple on the 6V DC line at the output of the 6V regulator so the average voltage level on the lamps was much higher. I remember when I put my voltmeter on the output of the regulator I saw a little over 8V and went WTF? So then I looked at it with the scope and saw the bad ripple. I think it's just a lot worse with the LED lamps. The next thing I did was check all the ground connection, but they were good. Anyway, that was the problem and it wasn’t the lamps themselves. They were somehow making the 6V regulated output oscillate. Using a non-inductive 1-ohm resistor in series with the lamps and a 1K load to ground resistor at the output of the regulator solved the problem. I also added the diode across the regulator from input to output, very important.

    Below is what I had to do to correct it. I think if you just use 6.3Vac on the LED lamps straight from the transformer they would be fine. I would still add a small non-inductive resistor though even if using AC to make sure the voltage stays just under 6.3Vac at all times.

    update_3.png
    .
     
  18. W5HRO

    W5HRO Administrator

    For the hell of it today I took out my ohm meter and put it across the good LED lamps I have left and they are open-circuit so that's why and there is no diode drop to ground in either direction either. They basically don't have any DC path to ground. I then measured the stock GE lamps and I see the typical 5-ohms or so and it varies from lamp to lamp due to the filament wire.

    So that all makes sense and is why I also needed to add that 1K resistor to ground at the output. It gives it a path to ground and it will always have a load with that resistor. At startup, the lamps appear as capacitors to ground blocking DC until a threshold is met and the output starts seeing resistance so that’s why adding the series non-inductive resistor and the 1K resistor to ground worked.

    I guess if you are going to use the lamps with DC instead of AC that’s what you must do otherwise the peak-to-peak ripple voltage on the DC will greatly exceed the 6.3V. I my case it was over 8V and is why the lamps burned out so quickly.

    The next time I have to work on the NC-303 or start on the NC-300 I will throw my capacitor meter across the 4 LEDs in parallel to ground and see what it is. My guess is the capacitance is pretty high.
    .
     
  19. W6MQI

    W6MQI Member

    Good find Brian I'll make note of this I never did change out the lights in my 303 with LED's not sure I ever will.
     
  20. W5HRO

    W5HRO Administrator

    The LED lamps being open-circuit is also the reason they work on AC. You basically have 2 pieces of metal inside with a gap just like a capacitor. That's all they are. It's just the design, size and shape of the metal on each side that makes them work. I'm not really sure they should be called LED's though.

    With the older lamps that have the filament wire the resistance when cold is very low, almost like a short and as they start getting bright the resistance goes way up. These other lamps are the exact opposite. If you use them with DC, but with no regulator you could just throw a resistor across each lamp to ground instead. You would divide by the number of lamps to make all of the resistors in parallel equal around 1K or whatever value that drops it down to 6V.

    The neat thing about using the white LEDs though is they turn the old yellowed plastic dials due to age into new looking white dials again. They are not clear lights, they are white lights.
    .