Electronics FAO Cpemma (Or other such Electronics Luminary) - RE Diode Bus

I'd like to build the Diode Bus (a couple actually - one for the intake fans and one for the exhausts). Problem is I need to put each one in a floppy bay blank on my LiLi as I've run out of 5 1/4" bays.

To save on space, what I'd like to do is ditch the rotary switch in favour of a push button. Is there anyway a circuit can easily be knocked together so that each push of a momentary switch would step the voltage up, in the same way as the rotary switch would? And when it got to max voltage, a further push would set it back to min power? (or the other way round if it was the only way to do it).

The eye candy LED's would be a must as well.

Can it be done?

Cheers,

Electronics Noob GOO

 

Ooooo Challenging

First off, you can do it, but it may end up bigger than you anticipate. (But the bulk can be hidden far from the switch)

You can do this with a counter and transistors.

Using a transistor in series at each diode will give you the switched effect. (And add about 0.2 volts drop)

Then you use a counter aranged so that when you reach the desired number of steps (5, 10) it will kick back to zero.
The counter would also have to keep the previous outputs on.
(Raising them one by one)

The output of the counter is connected to the base of each transistor turning each on in turn.

If made properly you could even have two buttons and have a up/down effect.
Also you could easily put in an automatic mode than read's the temp and set's the speed accordingly.

The LED's could simple tap off some current from the transistors for indicators.

(I am at work so no schematics right now)

 

I've seen the ics used in those domestic lights where you touch the switch and the brightness cycles up and down, which would be nice for a fan control, but IIRC they're geared up to run from a 50/60Hz mains, not 12v DC.

It can be done (anything can be done with the right chips ) but I can't think of a neat easy & cheapish solution.

 

A pushwheel switch would work.
Digikey and
Mouser carry them, no doubt someone in the UK as well.

 

That's what I'm doing for my fanbus. You'd need a 4017 IC. Connect the pushbutton switch to the clock input. Connect the output pin that corresponds to the number of steps to the reset pin. The outputs go high in sequence though, so you'll have to bus them and add diodes between each one.

 

Thanks guys - any chance of schematics?

 

...but viridius can

And if you adapted the back&forth KR circuit here it would go up and back down in speed, rather up to max then sudden drop to minimum.


Forget the 555 clock, the pulses come from your push-button. Instead of each led, you've the base of a NPN transistor that will take the fan current (2N2222A springs to mind or maybe a darlington). The tranny collecter to +12v, emitter to a point on a diode chain leading to the fan red, fan black to ground.

 

Quick and dirty little schematic here: http://www.geocities.com/alawanx/schematic.gif. You'll have to cut and paste the link, I'm afraid (anyone feel like hosting?). The switch on the left is the pushbutton switch. You may have to make a simple debouncing circuit to control stray pulses. The transistors are MOSFETs. Since this circuit uses CMOS, VDD can be anything from 5 to 18 volts. I wouldn't recommend using 12 V straight from the molex though. Use a 7805 or something. A capacitor may be needed across pins 8 and 16 also. This schematic is a design for a pushbutton-advanced four-state fanbus. It starts with fans off. Pin 2 is high, but since it's not connected to anything, nothing happens. Pushing the button makes pin 1 high. This turns on the MOSFET, allowing current to reach the fan after passing through two diodes. Another push of the button, and pin 3 is high. Current reaches the fan through one diode. One more push of the button, and pin 7 is high, letting current reach the fan directly. One last push, and pin 11 goes high. This gives a signal at pin 15, which resets the counter to zero, fans off. There are more pins to use, but for the sake of keeping the schematic small, they have been omitted.

 

Viridius: Can't see the schematic, even with cut and paste.

Cpemma: I'm guessing that all of the components around the 555 can be ditched as well? Just a momentary switch on the positive rail to the counter?

Once again - thanks all

Edit - I can see schematic now, after taking off the trailing 'dot'

 

With the viridius idea, no reason why you can't have 3 diodes in series bunched between the mosfets. Bunches of 3 will give about 11.5v, 9.25v or 7v to the fan.

Goo- yes, just a momentary switch from +12v, and it might also need a 10k-47k pull-down resistor for when the switch is off.

Contact bounce could be a problem, but WTH, it's no 3-Mile Island if you have to go round the cycle again occasionally.

 

Ok - am I on the right track with this, or am I mooing up the wrong tree? (56K nasty):

Diode Bus II

Cheers,

GOO

 

I'll have to print that out and look at it, but I thunked up a simpler system with the 4017 KR.

Where the leds are, goes one resistor (R2) of a potential divider, all commoned to a second resistor (R1) to ground.

From the junction, feed an adjustable regulator adj terminal. By varying R2, various output voltages will be set.

I also kept the 555, with a slow clock, so keeping your finger on the switch cycled the speed till it was as req'd.

This is now pat pending

 

That's exactly what my mate Rob at work started talking about! Stuff about regulators and holding buttons down until as the voltage increased. But as I'm a noob, most of it went right over my head (hell all of it did. I'm sure that diagram I've drawn will be a complete turkey....)

 

That was my idea for a fanbus. I scrapped it because it would require 19 ICs. I had a 555 driving a DS1804 digital potentiometer which controlled another 555 to provide PWM. I've still got the schematics if anyone's interested. The voltage regulator is a good idea, though.

 

nothing wrong with 19 IC fanbus...im working on a binary controlled fanbus atm(software, tempature, and manual control). Im using a typical rotary switch. But a slider switch could be substituted...check the themodfathers.com, it should be there when im done.

 

Sooo - has anybody been able to check out my schematic? Opinions? Will it blow up in my face?

 

Anybody......, Anybody......?

 

I think that should work, looks very good.

 

Thanks Zap- I guess I'd better buy some bits and protype it now.

I'm guessing that the values of the current limiting resistors will change depending on how many led's the current is flowing through?

And do I need those resistors between the bases of the transistors and the diode OR gates? weren't they there as current limiting resistors for the original LED circuit?

Cheers

GOO

 

Look up the specs on the IC
If it already limits the current or is TTL compatible then you can connect it straight up.
Otherwise just leave them.