[VicPiMakers Projects] LED and Flourescent tube voltages

[George Bowden]

Well, I survived.  But these LED tubes need a ballast to produce a higher
voltage than 110v.  The internal capacitors limit the current way too
much.  Replacing the capacitors with a single 0.1 uF capacitor, for current
limiting, allowed the dim output I desired.  But you have to get inside the
bulb to do that, and the tube is made of glass...



On Sun, May 8, 2022 at 11:30 AM George Bowden <gtbowdeng at gmail.com> wrote:

> If anyone has some ideas on LED and fluorescent tubes, pile in.
> I got some LED T8 tubes that are for direct replacement with
> fluorescents connected to instant on ballasts.
> I carefully looked up ballasts that are compatible to this LED tube
> <https://www.gelighting.com/tubes-ballast-compatibility> and chose to
> look at wiring diagrams for a B132IUNVHE-N ballast[image: image.png]
> Clearly, the two pins at each end are joined together.
>
> I then took apart one LED tube and drew its schematic. The electronics are
> repeated at each end of the tube (drawn one above the other), which makes
> sense because tubes can be inserted in either direction. I then drew AC
> power connections per the diagram above. Pins 1 and 2 are at one end of the
> tube, and pins 3 and 4 are at the other.  Given the shorting of pin1 to 2
> and pin 3 to 4, capacitors 1,2,6,&7 are shorted out, as are resistors
> 1,2,3&4. Not shorted are the capacitors and diodes next to each other.
> Tubes also could be inserted into ballast wiring where pin1 and pin2 are
> not shorted, and it is interesting to have a tube design that must work
> with 4 different power connections!!
>
> [image: image.png]
>
> Simulating this design in LTspice shows 70v peak across the bank of
> leds, with current through each bank at 3.6 milliamps.
> There are actually 3 banks connected in parallel of 22 leds connected in
> series. I've only shown one bank.
> I've guessed at the surface mount capacitor values, and the simulation
> does not include the loads of LED bank2 and bank 3, so it is not quite
> realistic.  Also, I am guessing at the LED type.
> Given that ballasts try to maintain a certain current, using a 110v AC
> input is probably wrong, but that is what I will use to power the tube.
> I might include a 1.5uF 250v cap on the input to limit current, but the
> simulated current through the LEDs is very low already...
> My application is for a very weak light over my keyboard, so 3.6ma might
> be fine.
>
> [image: image.png]
>
> Now to plug it in....AAGGGHHH
>
> --
> George Bowden  250-893-7423
> Victoria BC
>


-- 
George Bowden  250-893-7423
Victoria BC
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://vicpimakers.ca/pipermail/projects_vicpimakers.ca/attachments/20220509/e4316842/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 33751 bytes
Desc: not available
URL: <http://vicpimakers.ca/pipermail/projects_vicpimakers.ca/attachments/20220509/e4316842/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 42437 bytes
Desc: not available
URL: <http://vicpimakers.ca/pipermail/projects_vicpimakers.ca/attachments/20220509/e4316842/attachment-0001.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 52876 bytes
Desc: not available
URL: <http://vicpimakers.ca/pipermail/projects_vicpimakers.ca/attachments/20220509/e4316842/attachment-0002.png>