On Fri, 21 Sep 2018 22:11:47 -0700 (PDT), whit3rd
wrote:

On Friday, September 21, 2018 at 5:36:05 PM UTC-7, Ned Simmons wrote:
On Fri, 21 Sep 2018 12:59:12 -0700 (PDT), whit3rd
wrote:

... The 'require ballast' lamps get about 24V of
excitation, which means seven or eight LED diodes in a series string, with
other strings in parallel. The 'full voltage' get 240V, so that's
seventy or eighty in series.

Where do you get 24V? A standard ballast for 8 foot F96T12 lamps
outputs 750V for starting the lamp and 425mA operating.

The LEDs are more efficient, but the ballast still chokes current at circa 425 mA.
So, the voltage is lower for LEDs than for tubes. My '24V' number is only order-of-magnitude.

Which means the LED driver in the tube has to deal with a constant
current source that can deliver up to 750 volts instead of a
well-behaved 120V constant voltage source. Doesn't seem advantageous
to me.



...Now, in addition to
the internal LED driver(s) in the lamp, you have a point of failure,
the ballast, that's not doing anything useful.

But, it IS buffering the LEDs from voltage spikes,

The ballast is intentionally designed to develop a 750V spike.

without the power-wasting resistors
used in some lamp-replace series strings.

I've not seen dropping resistors in the replacement tubes. All I've
seen have switching LED drivers. Look at the specs for the LED
replacement tubes I pointed to. They run on 100-265VAC or 85-265VAC.

The starter-boost is a complex part
of the ballast, but isn't required for the LEDs; if it fails, you'll never notice.

Perhaps for an electronic ballast, not so for a magnetic ballast.

The 'internal LED drivers' in the one lamp I dissected, were... only rectifiers to prevent
reverse voltage.

I don't understand. Why would you need to protect an LED from reverse
voltage?


--
Ned Simmons