Construction Diary -- Part 1, The LED matrix/display

[Andrew Davie]

Gents,

Klaas is quite right in what he says, but the 'protoboard' can only be used generally once, akin to stripboard (Veroboard). I tend to develop on the breadboard just a small section at a time, transfer this to a stripboard and move onto the next section.

Actually, I already have/use both. The protoboard was used to build my rectifying circuits. I settled on the exact-same no-holes-connected configuration. You can buy several with various configurations.

I think while designing I'd use the stripboard... I do so hate unsoldering, as I invariably degrade the copper traces.

[Klaas Robers]

Ok Andrew, and now for the maximum brightness....:

I assume that you still have the resistors in the gamma circuit as they are in my cicuit diagram. That gives you 40 mA of white current through your LED-cluster. You have still 8 strings of 3 LED's en series connected in parallel, so the peak current through each LED string is only 5 mA. You can have 4x more light........

Rewire your LED cluster into strings of 12 LEDs in series, give each serie one series resistor of 220 ohm, and connect both strings in parallel. Now you need a voltage much higher, about 50 to 60 volts.

Find a small transformer with 2 windings of 24 volts. The wattage might be as small as 2 watt, but more is not at all a problem. If not findable you can also use two transformers with one winding of 24 volts.

Now make two rectifiers for the 24 volts with each a capacitor of 470 uF minimal 40 volts. On each capacitor you will get about 30 to 35 volts. Then connect these capacitor voltages "on top of each other" to obtain a voltage of 60 to 70 volts. Connect the - to your ground and the + is going to be the voltage for your newly wired LED cluster.

Be a little bit carefull. 70 volts DC isn't too dangerous, but you will feel it well when you touch it. Connect a permanent resistor (22k) across the 70V to enable the capacitors to discharge (a bleeder resistor) when the 230V is switched off.

For your information: the safety limit for AC voltages (50 Hz) is 50 volts, for DC-voltages is 130 volts. Strangely 50 Hz is the optimal dangerous frequency for AC voltages........ Above these voltages they can be lethal.

[Andrew Davie]

Ok Andrew, and now for the maximum brightness....:

Thank you for the above information. I might go this way, but before I do, a question or two...

Am I correct in assuming that the 'bleeder resistor' is there to gradually (slowly) discharge the capacitor(s) solely to make the circuit safer after it has been switched off. That is, to reduce the likelyhood of the capacitors discharging an unpleasant shock if touched accidentally?

You are correct that I have left the gamma resistors untouched. I considered changing these, but I figured they were close enough to allow me to get a picture up and running first (my main goal) and they could be adjusted to improve the picture later.

I understand your calculation that there is only 5mA going through each string of LEDs (40mA / 8 strings). I also understand that the LEDs can be 4x brighter (or more, they handle 35mA, right? -- that's 7x brighter). What I don't quite "get" yet is why we increase the voltage/power supply. Couldn't we increase the current? If there were (say) 200mA of current then the current through the LED strings would be ( 200/8 = 25mA ). Wouldn't the LEDs, then, be 5x brighter?

I think I've forgotten the relationship between the "R" of the gamma circuit and the voltage/rating of the LEDs. I have to think about this some more, and go back through my earlier notes in this thread.

More to the point, though, I'm loathe to play with voltages that are approaching "dangerous". Remember this is my first ever electronics project and though I'm trying to be careful, I have already made a mistake or two. Sparks, yes... smoke, yes... but not any shocks (yet, fingers crossed, touch wood). Playing with 70V would make me very nervous indeed.

There's another thing, I'd find it quite elegant running all this (the sound circuit, the LEDs, the motor, and your boards... from a single transformer/rectifier. At the moment, the sound and motor are running from different supplies... but I plan to experiment and see how things look when driven from the same source.

[Klaas Robers]

Yes the bleeder resistor is just there to prevent the capacitors to hold their charge and bite you when you are thinking there is no voltage left.

Don't be too afraid of 70 volts. When I started this all in 1960 I worked with electron tubes running on 300V. Then I knew: switch off, wait a while, measure to be sure and then you can continue. Working, measuring etc in a powered circuit was absolutely not done. That bad habbit came with the introduction of transistors and IC's. With 70V this is not really needed, it is still half the 130V where voltages become dangerous. But first measure, then touch, still is a good idea.

Changing the resistors of the gamma network is not that easy. I just realised that after the simulations done by Pete. It works less simple than I hoped. In my wooden cabinet monitor (see the Convention report pages) I have one chain of 32 amber LED's (40mA peak white) on a voltage of about 100V. I started this with a Neon lamp on an even higher voltage, but this gave almost no light at all and burned dark very fast.

It is not silly to use different power supplies for different tasks. In CD-players it is usual to have different powers for the analogue part, the digital part and the servo part. Yes of course there is most of the time one transformer with more secundary coils and made especially for thier tasks. But this is just a matter of money for the company. In my monitor I have two transformers too. When the cabinet is closed nobody can see it..... And I am sure of no mutual influences.

If your LEDs can handle 35 mA continuously then you could think of making just one chain of LEDs and make an even higher voltage. I didn't know that and had the idea they shouldn't go higher than 20 mA. I would do that. During sync pulses the LED's are off, so 40 mA during (part of) the scan is no problem. Then make 120V DC, 4 times 24V rectified and see for good insulation.

[Andrew Davie]

Here's a link to the book...
http://www.artofelectronics.com/
Steve A.

I have now purchased a copy of this book through eBay (from India). Cost was US$29 including postage. Will advise if/when it arrives.

This book arrived safely from India today. Poor packaging, basically firm paper wrapped around the outside, a bit of string to tie it. Minor damage to the back lower corner -- overall, though, an excellent budget purchase. It's thick -- lots to read!

[Steve Anderson]

It's thick -- lots to read!

Yes there is a lot to read, but you don't have to read it all. Is it the first or second edition they have sent to you? And is it the hard-backed version or the soft-backed?

Steve A.

[Andrew Davie]

It's thick -- lots to read!

Yes there is a lot to read, but you don't have to read it all. Is it the first or second edition they have sent to you? And is it the hard-backed version or the soft-backed?

Steve A.

Soft-backed second edition. Special cheap version for 'backward' countries like Nepal, India, Bangladesh... and Australia. Actually, not supposed to be sold in Australia -- they cheated, and obscured the origin of the book. That's OK by me, I get a copy cheap.