On Friday, 1 March 2019 19:48:36 UTC, Andrew Gabriel wrote:
On 01/03/2019 16:30, whisky-dave wrote:
On Friday, 1 March 2019 16:02:52 UTC, NY wrote:
"whisky-dave" wrote in message
...

Dimmable fluorescent fittings have been around since the early 1960s..
Nothing new there.

How are fluorescents dimmed?

I waqsnt; sure so looked it up.

https://www.etcconnect.com/Support/A...ures-Work.aspx

I'd always thought that it was impossible,

well I knew it was possible but a bit more difficult than just replacing a light switch with a dimmer switch.

It is basically replacing the light switch with a dimmer (suitable for
inductive loads), but you also need to add a separate pair of isolated
supplies to keep the filaments hot enough to continue operating in
thermionic emission mode when the tube current is too low to do it just
by itself. This will also enable the tube to strike without a starter.

Indeed, I made just such a light as a teenager.

I'm not allowed to just replace the lighting in a teaching lab with something I did at home. I also have about 45 double tubes in the lab.



Commercially, Transtar was a well known manufacturer of dimming magnetic
ballasts (and very high quality non-dimming ballasts).

Good for them but I leave that to those designing the spaces in the lab.

Just has a academic asking for a 96V or higher PSU I politely laughed.

I've been asked to suggest the purchase of some new bench supplies for the lab but I doubt I'll get permission to order ones that got up 96V.
I'm not looking at anything about 30V at the moment.





and
was surprised when the fluorescents in my school lecture theatre could be
dimmed. They were probably installed in the early to mid 70s, which gives an
idea of what technology was available then.

One issue with banks of dimming fluorescents is all the tubes had to be
replaced together, otherwise newer and older tubes are significantly
different light output, and different makes can be different too.

Yes we have that problem witout dimmering, 3 faulty in the lab at the moment with 1 or 2 on their last legs.


Also, you get colour shift and a significant drop in CRI when dimming
fluorescents, as the ratio of light output from the different mercury
line changes, which upsets the colour balance from the arc and the
phosphor (tend to lose more green and end up with purple from the over
balance of red and blue).

Makes me think that perhaps LED tubes are better. We did ask for them rather than florries but the cost would have be over £3k so was rejected.



Nowadays with microcontroller based electronic ballasts, adding dimming
is trivial, although it's not done by using triac dimmers. 0-10V, DALI,
switch-dim, or photocell to generate constant lux level are common, and
many electronic ballasts support multiple of these.

why not use LEDs ?


Indeed, many electronic ballasts support multiple different tubes that
run at different currents, and having identified the tube in use from
the electrical characteristics, effectively use the internal dimmer
logic to limit the current to the max allowed for the identified tube.

I think that's what our new LED tubes in the new lab do, on walking over to them after a few seconds they come on whether wanted on or not.

WHich is going to be fun when they want the light out when using photocells in a lab experiment can't wait to have to explain that one the the 'designers'