Hi Beaky, I read with interest your posts on electronics stuff and find them interesting and informative, but to be frank, and I mean absolutely no offence and cheerfully add that you should feel free to ignore me, I reckon your led techniques and understanding are far from current and good practise.
The Vf of a led is a function of the junction material, they are not a linear device as you are presenting them. Most have a Vf of about 2.3 volts, the technique you've described works in limited cases with small numbers of leds (in each group) between the supply voltages and where the sum of the led Vfs (in each group) roughly matches the supply output voltage.
(This is also the reason the Cointaker and Ablaze leds can be used in parallel wired strings - as everyone who buys them uses them - they have an inbuilt resistor that acts as a current limiter, or a primitive current regulator if you like, in each led. This is also the poor feature that I alluded to in my earlier post)
Good practice drives leds in series and regulates current to vary brightness. A pretty fair summary can be found here http://www.ledsmagazine.com/features/4/8/1
A string of leds like the GI in a pin with about 40 to 50 driven all in parallel (with no current limit resister as described above) will suffer variations in brightness amongst individual leds due to variations in junction temperatures and the thermal runaway that will result. This would be made worse if you tried to use different coloured ones. I strongly suspect driving it this way will lead to failures of those that operate higher up on the Vf v If curve. This will lead to rapid increase in junction temp, lower junction resistance, higher current flow and ultimately failure.
One technique people use is to place several strings of series leds in parallel to even out these differences - this is how I would go about it if I were trying to control strings of GI leds. Take a feed of about 25V, loose a few volts headroom for the regulator and then arrange, say, 4 or 5 parallel string each with ten series leds in them.
Anyway. I'm happy for us to disagree and we can each continue to drive our leds in ways that we like, each of us secure in the knowledge that the other is wrong! Good luck to you and your leds.
no probs and please take no offence to my statements.
all of the leds that i have run on a string has their own resisitor on each pcb (if you look at the picture you will see 2 small resistors)
the reason why we ran them in paralllel is the machine that makes the strings was used originaly to make pixie light strings with incandecent lamps.
i agree that an led is not a linear device.
the led systems i made for the company i worked for where made between 2000 and 2005, the majority of the systems i made are still in service today.
some of the circuit boards i made for light fitings (pool lights, garden lights and various other low voltage light fittings) had up to 100 leds in them, in that instance the pcb had the leds wired in series and then wired in parallel, depending on the colour we would have 3 or 4 leds in series. (blue and white 3 in series and all other colours 4 in series).
so as far as using them in a pinball no one i know has rewired there G.I circuits in series.
and with my past experiences with leds they can be used in parallel with a resistors on each led and then be dimmed via dropping the voltage with out shortening the life of the led.
when you drop the voltage the current also will decrease.
If you have more than one set of leds running in series off a power supply and you dont use a resistor to current limit each set of leds and use a current limiting power supply you can run into some serious problems if 1 of the series go out.
I.E. if i had a power supply running 3 sets of leds that are running in series and each series of leds take 30ma, and you have your current limiting power supply set at 90ma (3 sets of leds @ 30ma = 90ma) and one of the series go out then you will be running 90ma into 2 sets of leds that only need 60ma, the 2 remaining sets of leds wont take long to die.
another problem you would have in the same scenario is 1 set of leds will have a lower foward voltage than the other 2 sets, it is very rare to have a identical foward drop in 2 leds.
in my opinion using a current limiting power supply to run more than 1 set of leds in series is not practical.
B.T.W. the power supply that i use to drive my leds at work and at home is a switch mode power supply. the way it regulates its output is via pulse width modulation so the power supply gives out very little heat.
again no offence
here is the desri p tion of the I.C. that is used in my power supplies:
The MAX724/MAX726 are monolithic, bipolar, pulsewidth
modulation (PWM), switch-mode DC-DC regulators
optimized for step-down applications. The
MAX724 is rated at 5A, and the MAX726 at 2A. Few
external components are needed for standard operation
because the power switch, oscillator, and control
circuitry are all on-chip. Employing a classic buck
topology, these regulators perform high-current stepdown
functions, but can also be configured as inverters,
negative boost converters, or flyback converters.
These regulators have excellent dynamic and transient
response characteristics, while featuring cycle-by-cycle
current limiting to protect against overcurrent faults and
short-circuit output faults. The MAX724/MAX726 also
have a wide 8V to 40V input range in the buck stepdown
configuration. In inverting and boost configurations,
the input can be as low as 5V.
The MAX724/MAX726 are available in a 5-pin TO-220
package. The devices have a preset 100kHz oscillator
frequency and a preset current limit of 6.5A (MAX724)
or 2.6A (MAX726).
And here is the link that has a basic power supply and the data sheet
http://pdf1.alldatasheet.com/datasheet-pdf/view/73748/MAXIM/MAX724CCK.html