Ah yes. That's the fun thing about doing digital electronics. You can't 
just *buy* a 7400. No, you have to decide whether you want a 74LS00 or a 
74HC00 or a 74HCT00 or...

Reading the datasheets is like walking into another world. (Especially 
since the datasheets are usually poorly-scanned grainy PDFs that don't 
match the part number you actually asked about!)

However, I just found the following pages:

http://www.kpsec.freeuk.com/components/ic.htm#logic
http://www.kpsec.freeuk.com/components/74series.htm
http://www.kpsec.freeuk.com/components/cmos.htm

Suddenly everything seems so much clearer. For example...

"74LS series: Inputs 'float' high to logic 1 if unconnected."

Oh crap. Well that would explain a thing or two! >_<

(I foolishly assumed that being unconnected would register as 0V - i.e., 
logic 0. Damn, this is going to make interactive testing *so* much harder!)

Also, it's nice to have some indication of the power requirements, 
switching frequencies, and what the hell all the letter codes mean!

Post a reply to this message

> "74LS series: Inputs 'float' high to logic 1 if unconnected."
>
> Oh crap. Well that would explain a thing or two! >_<

You should never be leaving any inputs as floating anyway, that's just 
asking for trouble (for example if you use a different IC series later). 
Best to always tie your inputs to whatever you want them to be.

Post a reply to this message

scott wrote:
>> "74LS series: Inputs 'float' high to logic 1 if unconnected."
>>
>> Oh crap. Well that would explain a thing or two! >_<
> 
> You should never be leaving any inputs as floating anyway, that's just 
> asking for trouble (for example if you use a different IC series later). 
> Best to always tie your inputs to whatever you want them to be.

Sure. But I'm sitting with the IC between my fingers, poking the pins 
with the end of a wire. Not so easy to tie it all down just for a quick 
test.

So it seems that every input pin is supposed to be connected to one rail 
or the other. (Or an output, which I guess is going to be the most 
common case...)

The page also claims that you can configure an output pin to drive an 
LED so that it lights when the pin goes *low*... Man, electronics is weird!

Post a reply to this message

> The page also claims that you can configure an output pin to drive an LED 
> so that it lights when the pin goes *low*... Man, electronics is weird!

Why is that weird?  The LED doesn't care which leg is connected to the logic 
output pin, it's just a voltage.  You could even connect the LED between two 
output pins*, then it would only turn on when one was high and the other was 
low.

* Using a series resistor if required...

Post a reply to this message

>> The page also claims that you can configure an output pin to drive an 
>> LED so that it lights when the pin goes *low*... Man, electronics is 
>> weird!
> 
> Why is that weird?  The LED doesn't care which leg is connected to the 
> logic output pin, it's just a voltage.  You could even connect the LED 
> between two output pins*, then it would only turn on when one was high 
> and the other was low.

I always thought of it as logic high meaning "connected to the +V rail" 
and logic low meaning "not connected to anything". From the looks of 
things, it seems that actually logic low means "connected to the -V 
rail". So it's like a bipolar setup...

> * Using a series resistor if required...

Yes, that's the other fun thing. Depending on which type of gates you're 
using, they supply about 20 mA. But some of the LEDs I'm looking it draw 
40 mA...

Post a reply to this message

Mmm, 74HC00, 79p. That's quite cheap...

Wait. If I need, say, 10 of them, that's £7.90. Suddenly that doesn't 
sound quite so cheap any more. o_O

Remind me, *why* am I embarking on this crazy project? ;-)



Oh, hey Mike, cheapest relay I can find is £1.91. Have a nice day. >:-D

Post a reply to this message

Am 19.05.2010 13:07, schrieb Invisible:
> Ah yes. That's the fun thing about doing digital electronics. You can't
> just *buy* a 7400. No, you have to decide whether you want a 74LS00 or a
> 74HC00 or a 74HCT00 or...
>
> Reading the datasheets is like walking into another world. (Especially
> since the datasheets are usually poorly-scanned grainy PDFs that don't
> match the part number you actually asked about!)

Texas Instruments has a nice collection of 74xx data sheets in a single 
proper PDF file at <http://focus.ti.com/lit/ug/scyd013b/scyd013b.pdf> 
("Digital Logic Pocket Data Book"), apparently covering all their 
current products (as of 2007) in the 74xx line, from single-gate 5-pin 
thingies like the 74x1G00 (single NAND gate) to 100-pin monsters like 
the 74x32543 (36-bit registered bus transceivers with 3-state outputs).

Post a reply to this message

Am 19.05.2010 14:51, schrieb Invisible:

> Yes, that's the other fun thing. Depending on which type of gates you're
> using, they supply about 20 mA. But some of the LEDs I'm looking it draw
> 40 mA...

That's why normally you'd use an additional transistor to drive the LED 
(if you have multiple LEDs to drive you may want to use a transistor 
array in an IC package).

Post a reply to this message

> I always thought of it as logic high meaning "connected to the +V rail" 
> and logic low meaning "not connected to anything".

Logic "high" and "low" are just names given to a certain agreed range of 
voltages.  The difference between "not connected to anything" (floating) and 
"logic low" is very important when working with data busses.  If you have 
several devices all with their outputs connected to a data bus, then only 
one device at a time must be driving the data lines to logic low/high, the 
rest must set their outputs to floating.

> From the looks of things, it seems that actually logic low means 
> "connected to the -V rail". So it's like a bipolar setup...

But don't assume a "logic low" output will be exactly -V, it is only 
guaranteed to be within the range of voltages given for "logic low" in the 
data sheet.

> Yes, that's the other fun thing. Depending on which type of gates you're 
> using, they supply about 20 mA. But some of the LEDs I'm looking it draw 
> 40 mA...

What you mean is, they draw *up to* 40 mA?  How much they actually draw 
depends on what value series resistor you put in there.  Using Ohm's law and 
the V-I curve of the LED from the datasheet you can work out exactly what 
resistor size you need for a given current.

Post a reply to this message

Am 19.05.2010 15:29, schrieb scott:
>> I always thought of it as logic high meaning "connected to the +V
>> rail" and logic low meaning "not connected to anything".
>
> Logic "high" and "low" are just names given to a certain agreed range of
> voltages. The difference between "not connected to anything" (floating)
> and "logic low" is very important when working with data busses. If you
> have several devices all with their outputs connected to a data bus,
> then only one device at a time must be driving the data lines to logic
> low/high, the rest must set their outputs to floating.

... except of course in data buses that operate with one dominant and 
one recessive symbol (e.g. CAN), in which sending the recessive symbol 
is perfectly equivalent to not sending anything - which allows for some 
neat tricks to deal with collisions (i.e. the situation where two 
devices start sending simultaneously).

Post a reply to this message

>> I always thought of it as logic high meaning "connected to the +V 
>> rail" and logic low meaning "not connected to anything".
> 
> Logic "high" and "low" are just names given to a certain agreed range of 
> voltages.

Indeed. You can use any two states you want, so long as it's easy to 
distinguish them, and all the components are using the same rules.

> The difference between "not connected to anything" (floating) 
> and "logic low" is very important when working with data busses.

There is that...

>> From the looks of things, it seems that actually logic low means 
>> "connected to the -V rail". So it's like a bipolar setup...
> 
> But don't assume a "logic low" output will be exactly -V, it is only 
> guaranteed to be within the range of voltages given for "logic low" in 
> the data sheet.

Heh. It gets weirder and weirder...

>> Yes, that's the other fun thing. Depending on which type of gates 
>> you're using, they supply about 20 mA. But some of the LEDs I'm 
>> looking it draw 40 mA...
> 
> What you mean is, they draw *up to* 40 mA?

No idea. The spec sheet just says "forward current".

> How much they actually draw 
> depends on what value series resistor you put in there.  Using Ohm's law 
> and the V-I curve of the LED from the datasheet you can work out exactly 
> what resistor size you need for a given current.

Heh. And here I was thinking it's just a case of "if the voltage is less 
than X, you don't need a resistor"...

Post a reply to this message

Hey, check this out:

http://www.maplin.co.uk/Media/PDFs/N24AZ.pdf

They only want £7 for it too. I have literally *no idea* what 98% of the 
content of this datasheet means! :-D

Post a reply to this message

On 5/19/2010 6:58 AM, Invisible wrote:
> scott wrote:
>>> "74LS series: Inputs 'float' high to logic 1 if unconnected."
>>>
>>> Oh crap. Well that would explain a thing or two! >_<
>>
>> You should never be leaving any inputs as floating anyway, that's just
>> asking for trouble (for example if you use a different IC series
>> later). Best to always tie your inputs to whatever you want them to be.
>
> Sure. But I'm sitting with the IC between my fingers, poking the pins
> with the end of a wire. Not so easy to tie it all down just for a quick
> test.

So, you put it on a breadboard with each pin tied to a pull-up or 
pull-down resistor, and probe with the opposite sense. If your resistor 
is pull down (i.e. connected to GND) then use VCC to probe (or better 
yet, just get a dip switch ...) If the resistor is connected to VCC, 
then use ground to set the signal to 0

> So it seems that every input pin is supposed to be connected to one rail
> or the other. (Or an output, which I guess is going to be the most
> common case...)

Yes. 5v is high, ground is low, not connected is neither of these, 
though TTL does float high, a non-connected input on a CMOS chip can 
lead to erratic behavior, or even destructive oscillation. It also 
leaves the chip wide open for ESD damage. FETs are interesting in that 
they don't sink or source any current on their inputs. without anywhere 
for the voltage to go, a charge can stay on that input. That charge can 
go dangerously (for the chip) high can cause a dielectric breakdown of 
the gate's insulating layer. Some CMOS gates will have a high-value 
resistor tied to ground on the input to bleed away some of the stray 
voltage as a safety measure for ESD. Don't rely on that.

> The page also claims that you can configure an output pin to drive an
> LED so that it lights when the pin goes *low*... Man, electronics is weird!

LED anode connected to VCC, cathode connected to out. ;) Not too strange 
when you think about it. A logic output is either connected to VCC or GND.

One other thing to be cautious of: If a TTL chip has an open collector 
output, make sure to use a pull-up resistor before connecting it to a 
CMOS chip. High is floating on an open collector output.

Oh! and by the way: with TTL chips VCC is *critical* due to the nature 
of the bipolar transistors used in the gates, they must have a regulated 
voltage. Look at Falstad's circuit simulator to figure out why. They 
have a resistor set to a particular value to bias the transistor. If the 
supply voltage goes out of range, the transistor is no longer properly 
biased, which will cause misinterpretation of signals.

CMOS doesn't require a bias, so can run on any voltage (within limits)
-- 
~Mike

Post a reply to this message

clipka wrote:

> Texas Instruments has a nice collection of 74xx data sheets in a single 
> proper PDF file at <http://focus.ti.com/lit/ug/scyd013b/scyd013b.pdf> 
> ("Digital Logic Pocket Data Book"), apparently covering all their 
> current products (as of 2007) in the 74xx line, from single-gate 5-pin 
> thingies like the 74x1G00 (single NAND gate) to 100-pin monsters like 
> the 74x32543 (36-bit registered bus transceivers with 3-state outputs).

Ah, neat!

Most of the books I've seen only cover the 7400, 7401, 7402, etc., and 
don't mention things like the 741G00. Could be useful to know. (Except 
that Maplin don't seem to sell this stuff... *sigh*)

Post a reply to this message

While we're here, does anybody comprehend what VCC, VSS, VDD and related 
symbols mean?

Post a reply to this message

On Wed, 19 May 2010 16:16:28 +0200, Invisible <voi### [at] devnull> wrote:
> While we're here, does anybody comprehend what VCC, VSS, VDD and related  
> symbols mean?

http://en.wikipedia.org/wiki/IC_power_supply_pin



-- 
FE

Post a reply to this message

>> What you mean is, they draw *up to* 40 mA?
>
> No idea. The spec sheet just says "forward current".

It's probably under a section titled "typical operating conditions" or 
"absolute maximum ratings".  There should also be a graph showing a curve of 
current against voltage, from zero up to the absolute maximum current 
rating.  Usually with LEDs the light output is fairly linear with respect to 
current, but certainly not with voltage.

>> How much they actually draw depends on what value series resistor you put 
>> in there.  Using Ohm's law and the V-I curve of the LED from the 
>> datasheet you can work out exactly what resistor size you need for a 
>> given current.
>
> Heh. And here I was thinking it's just a case of "if the voltage is less 
> than X, you don't need a resistor"...

If you want to run the LED at 10 mA you look up on the V-I curve in the LED 
datasheet what forward voltage corresponds to 10 mA.  Say for example 2.8 V. 
If your logic output is 5 V (and the other side of the LED is connected to 
0V), then you want 2.8 V across the LED and hence 2.2 V across the series 
resistor.  So your resistor needs to drop 2.2 V with a current of 10 mA, 
that makes it a 220 Ohm resistor.

Connecting the LED directly with no series resistor to a voltage source is 
usually a bad idea (even if you think the voltage is correct), because even 
slight changes in the voltage cause large differences to the current the LED 
draws which might damage your circuit.

Post a reply to this message

> They only want £7 for it too. I have literally *no idea* what 98% of the 
> content of this datasheet means! :-D

I can send you a longer one if you want :-)

Post a reply to this message

scott wrote:
>> They only want £7 for it too. I have literally *no idea* what 98% of 
>> the content of this datasheet means! :-D
> 
> I can send you a longer one if you want :-)

Why thank you, you're too kind. :-}

Post a reply to this message

>> Heh. And here I was thinking it's just a case of "if the voltage is 
>> less than X, you don't need a resistor"...
> 
> If you want to run the LED at 10 mA you look up on the V-I curve in the 
> LED datasheet what forward voltage corresponds to 10 mA.  Say for 
> example 2.8 V. If your logic output is 5 V (and the other side of the 
> LED is connected to 0V), then you want 2.8 V across the LED and hence 
> 2.2 V across the series resistor.  So your resistor needs to drop 2.2 V 
> with a current of 10 mA, that makes it a 220 Ohm resistor.
> 
> Connecting the LED directly with no series resistor to a voltage source 
> is usually a bad idea (even if you think the voltage is correct), 
> because even slight changes in the voltage cause large differences to 
> the current the LED draws which might damage your circuit.

With my old electronics kit, if I'm remembering this rightly, you 
connect through a resistor if you're using the 3V power source, and skip 
the resistor if you're using the 1.5V source.

Oh, and *don't* connect the LED to the 9V source at all. I did that once 
by mistake... that particular LED doesn't do anything now.

Ah, life seemed so much simpler back then. :-(

Post a reply to this message

Fredrik Eriksson wrote:
> On Wed, 19 May 2010 16:16:28 +0200, Invisible <voi### [at] devnull> wrote:
>> While we're here, does anybody comprehend what VCC, VSS, VDD and 
>> related symbols mean?
> 
> http://en.wikipedia.org/wiki/IC_power_supply_pin

OK, thanks for that...

Post a reply to this message

> Oh, and *don't* connect the LED to the 9V source at all. I did that once 
> by mistake... that particular LED doesn't do anything now.
>
> Ah, life seemed so much simpler back then. :-(

Reminds me of an interesting story.  We got a complaint from a customer that 
about 25% LCDs they have from us have a dark edge to the picture, either on 
the left or the right side.  When we checked the faulty ones we found that 
always the left or right LED was not working in the backlight (I think there 
were 8 in total).  This puzzled us for a long time, why would it always be 
the far left or far right LED that stopped working, and why were they 
breaking in the first place?

In the end we figured out what happened.  Actually *every* LCD we shipped 
had one broken LED, it's just they never noticed it when it was one of the 
central LEDs.  And it turned out the equipment at our outgoing inspection 
test was set to the wrong voltage, so the tester was just blowing one LED on 
every backlight that went out :-)

Post a reply to this message

Am 19.05.2010 16:04, schrieb Invisible:

> Most of the books I've seen only cover the 7400, 7401, 7402, etc., and
> don't mention things like the 741G00. Could be useful to know. (Except
> that Maplin don't seem to sell this stuff... *sigh*)

AFAIK they usually come in SMD packages only - which in turn means 
they'll usually be sold in bulk. And anyway - why would you buy, say, 
one 74x1G00 NAND and one 74x1G32 OR gate when you can get a 74x00 4xNAND 
IC for the same (or even lower) price, and wire up three of the gates to 
simulate the OR gate? (The 74x265 is also a neat beast - with its two 
(N)AND gates and two inverters/drivers it's an ideal candidate to 
simulate quite a wide variety of 2-gate combos you can imagine).

I guess the single-gate ICs are mainly targeted at applications where 
space, routing, speed or EMI constraints are serious enough to forbid 
sharing the gates of a single multi-gate IC among different sections of 
the circuitry.

Post a reply to this message

Am 19.05.2010 15:52, schrieb Invisible:

> http://www.maplin.co.uk/Media/PDFs/N24AZ.pdf
>
> They only want £7 for it too. I have literally *no idea* what 98% of the
> content of this datasheet means! :-D

Ah yes, the typical cheap alphanumeric LCD.

I guess most of the datasheet information says, "this thing does exactly 
the same as other manufacturers' devices of this type".

Post a reply to this message

On 5/19/2010 7:31 AM, scott wrote:

> Why is that weird? The LED doesn't care which leg is connected to the
> logic output pin, it's just a voltage. You could even connect the LED
> between two output pins*, then it would only turn on when one was high
> and the other was low.


There's even a name for that trick: Charlieplexing.

The 3 state had to be simulated by switched, but here's a circuit for 
the Falstad simulator (http://www.falstad.com/circuit/):

$ 1 5.0E-6 10.20027730826997 50 5.0 50
L 224 176 176 176 0 1 false 5.0 0.0
L 224 224 176 224 0 1 false 5.0 0.0
162 352 176 352 224 1 2.1024259 1.0 0.0 0.0
162 384 224 384 176 1 2.1024259 1.0 0.0 0.0
w 352 176 336 176 0
w 384 176 368 176 0
w 352 224 336 224 0
w 384 224 352 224 0
w 368 176 352 176 0
L 224 272 176 272 0 1 false 5.0 0.0
w 384 272 352 272 0
162 416 176 416 272 1 2.1024259 1.0 0.0 0.0
w 416 272 448 272 0
w 416 176 448 176 0
162 448 272 448 176 1 2.1024259 1.0 0.0 0.0
w 416 176 384 176 0
w 352 272 336 272 0
162 384 272 384 224 1 2.1024259 1.0 0.0 0.0
s 272 176 336 176 0 0 false
s 272 224 336 224 0 0 false
s 272 272 336 272 0 0 false
w 384 272 416 272 0
162 352 224 352 272 1 2.1024259 1.0 0.0 0.0
r 272 176 224 176 0 100.0
r 272 224 224 224 0 100.0
r 272 272 224 272 0 100.0


-- 
~Mike

Post a reply to this message

On 5/19/2010 8:13 AM, clipka wrote:

> Texas Instruments has a nice collection of 74xx data sheets in a single
> proper PDF file at <http://focus.ti.com/lit/ug/scyd013b/scyd013b.pdf>

Ah, sweet! Thanks for that link! <Bookmarked>


-- 
~Mike

Post a reply to this message

scott wrote:
> have several devices all with their outputs connected to a data bus, 
> then only one device at a time must be driving the data lines to logic 
> low/high, the rest must set their outputs to floating.

Yep. Indeed, there are a lot of busses that are organized such that you have 
an output to it and an input, and you start putting your address on the bus 
bit by bit until you see a bit different than your address. That means 
someone else put on a 0 while you put on a 1 (assuming floating high). Every 
packet gets delivered with no contention, with lower addresses getting 
priority.  It's a kind of cool way to work it.

-- 
Darren New, San Diego CA, USA (PST)
    Ada - the programming language trying to avoid
    you literally shooting yourself in the foot.

Post a reply to this message

Invisible wrote:
> that particular LED doesn't do anything now.

It's a DED component. It emits dark.

-- 
Darren New, San Diego CA, USA (PST)
    Ada - the programming language trying to avoid
    you literally shooting yourself in the foot.

Post a reply to this message

Mike Raiford wrote:
> There's even a name for that trick: Charlieplexing.

Cute. Is there a name for when you have a red LED in one direction and a 
green LED in the other and you use AC to make it yellow? I want to use that 
around my colleges and sound like I know what I'm talking about. ;-)

-- 
Darren New, San Diego CA, USA (PST)
    Ada - the programming language trying to avoid
    you literally shooting yourself in the foot.

Post a reply to this message

>> that particular LED doesn't do anything now.
> 
> It's a DED component. It emits dark.

Hey, so long as it doesn't start emitting negative light...

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

Darren New wrote:

> Cute. Is there a name for when you have a red LED in one direction and a 
> green LED in the other and you use AC to make it yellow? I want to use 
> that around my colleges and sound like I know what I'm talking about. ;-)

Most people call that a tri-colour LED - but sometimes that term is used 
to refer to an LED with three different-coloured dies inside it (and 
hence 3 pins). Which is obviously a different component.

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

scott wrote:
>> I always thought of it as logic high meaning "connected to the +V 
>> rail" and logic low meaning "not connected to anything".
> 
> Logic "high" and "low" are just names given to a certain agreed range of 
> voltages.  The difference between "not connected to anything" (floating) 
> and "logic low" is very important when working with data busses.  If you 
> have several devices all with their outputs connected to a data bus, 
> then only one device at a time must be driving the data lines to logic 
> low/high, the rest must set their outputs to floating.

OK, so if I'm understanding this right... The 74126 (go look it up) 
contains 4 "buffers" - gates who's output is logically equal to their 
input. But each gate also has an "enable" pin. When the enable pin is 
high, the gate works like normal. When the enable pin goes low, it's 
like the output pin isn't connected to anything any more.

And it seems that this allows you to connect several outputs together, 
forming a kind of wired-OR configuration, provided that at all times 
only one gate is "enabled".

Is that right?

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

Am 19.05.2010 16:16, schrieb Invisible:
> While we're here, does anybody comprehend what VCC, VSS, VDD and related
> symbols mean?

According to JEDEC,

VBB = substrate power voltage (bias voltage), a voltage < VSS/GND.

VCC = logic power voltage, serving as a reference for high signal levels 
(often sharing a common rail with VDD, in which case they may be used 
interchangeably).

VCCQ = output stage logic power voltage, separate (positive) power rail 
to drive output loads.

VDD = drain power voltage. separate (positive) power rail to drive the 
chip's internals.

VDDQ = output stage drain power voltage. Effectively the same as VCCQ.

VEE = emitter power voltage. Only used for ECL devices.

VHH = special function enable voltage. An extra-high voltage to enable 
certain special functions of a device.

VPP = programming voltage. An extra-high voltage to supply the power for 
the programming function of a device (e.g. EPROM)

VREF = reference power supply, serving as a special reference voltage, 
e.g. for ADC inputs.

VSS = GND = ground reference / source power voltage, serving as both 
(negative) power rail and low signal level reference.

VSSQ = GNDQ = output stage source power voltage / ground reference, 
serving as both a (negative) power rail and low signal level reference 
specifically for output signals (should always be at the same level as 
VSS=GND).

(This is just a subset of all the symbols used; for instance, 
digital/analog devices such as programmable amplifiers or ADCs/DACs 
often use separate voltage rails for the analog signals.)

Post a reply to this message

clipka wrote:

> According to JEDEC,
> 
> VCC = logic power voltage, serving as a reference for high signal levels 
> (often sharing a common rail with VDD, in which case they may be used 
> interchangeably).
> 
> VDD = drain power voltage. separate (positive) power rail to drive the 
> chip's internals.
> 
> VSS = GND = ground reference / source power voltage, serving as both 
> (negative) power rail and low signal level reference.
> 
> VSSQ = GNDQ = output stage source power voltage / ground reference, 
> serving as both a (negative) power rail and low signal level reference 
> specifically for output signals (should always be at the same level as 
> VSS=GND).

Suddenly the picture becomes even more complex than you feared...

> (This is just a subset of all the symbols used; for instance, 
> digital/analog devices such as programmable amplifiers or ADCs/DACs 
> often use separate voltage rails for the analog signals.)

Thank god I don't have to deal with anything analogue yet! ;-)

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

On 5/19/2010 12:22 PM, Orchid XP v8 wrote:

>
> Thank god I don't have to deal with anything analogue yet! ;-)
>

I spent hours trying to figure out why an op-amp wouldn't give me the 
output I was expecting. It finally clicked when I read the warning on 
the data sheet that pulling an input within 0.5v of V- would cause it to 
latch-up and force the output to near V+ no matter the state of the 
inputs. I went through a few chips thinking they were defective.... or 
that I toasted one (got V+ and V- reversed, which is apparently bad for 
op-amps...)


-- 
~Mike

Post a reply to this message

>> Yes, that's the other fun thing. Depending on which type of gates 
>> you're using, they supply about 20 mA. But some of the LEDs I'm 
>> looking it draw 40 mA...
> 
> What you mean is, they draw *up to* 40 mA?  How much they actually draw 
> depends on what value series resistor you put in there.  Using Ohm's law 
> and the V-I curve of the LED from the datasheet you can work out exactly 
> what resistor size you need for a given current.

See for yourself:

http://www.maplin.co.uk/Module.aspx?ModuleNo=35702

"Forward current: 25 mA"

If you can figure out what the hell that means, be my guest...

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

On 19-5-2010 13:07, Invisible wrote:
> Ah yes. That's the fun thing about doing digital electronics. You can't 
> just *buy* a 7400. No, you have to decide whether you want a 74LS00 or a 
> 74HC00 or a 74HCT00 or...
> 
> Reading the datasheets is like walking into another world. (Especially 
> since the datasheets are usually poorly-scanned grainy PDFs that don't 
> match the part number you actually asked about!)
> 
> However, I just found the following pages:
> 
> http://www.kpsec.freeuk.com/components/ic.htm#logic
> http://www.kpsec.freeuk.com/components/74series.htm
> http://www.kpsec.freeuk.com/components/cmos.htm
> 
> Suddenly everything seems so much clearer. For example...
> 
> "74LS series: Inputs 'float' high to logic 1 if unconnected."
> 
> Oh crap. Well that would explain a thing or two! >_<
> 
> (I foolishly assumed that being unconnected would register as 0V - i.e., 
> logic 0. Damn, this is going to make interactive testing *so* much harder!)
> 
> Also, it's nice to have some indication of the power requirements, 
> switching frequencies, and what the hell all the letter codes mean!

I think I resist any urge to comment on this. Other than "RTFM" and "buy 
any 'electronics for dummies book'" before embarking on such a project.

Post a reply to this message

andrel wrote:

> I think I resist any urge to comment on this. Other than "RTFM"

I don't think electricity comes with a manual. Neither do ICs, usually...

> and "buy 
> any 'electronics for dummies book'" before embarking on such a project.

Hey, what do you *think* I've got a dad for? ;-)

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

On 19-5-2010 22:59, Orchid XP v8 wrote:
> andrel wrote:
> 
>> I think I resist any urge to comment on this. Other than "RTFM"
> 
> I don't think electricity comes with a manual. Neither do ICs, usually...

They do, they are called datasheets.

>> and "buy any 'electronics for dummies book'" before embarking on such 
>> a project.
> 
> Hey, what do you *think* I've got a dad for? ;-)

Hard to say, other than the obvious task of a father ;) . But I guess 
you mean that he is the actual dummy that didn't know that a floating 
input is seen as a '1'?

Have you ever looked at e.g. 
http://en.wikipedia.org/wiki/File:TTL_npn_nand.svg ?

Post a reply to this message

On Wed, 19 May 2010 20:38:21 +0100, Orchid XP v8 wrote:

> See for yourself:
> 
> http://www.maplin.co.uk/Module.aspx?ModuleNo=35702
> 
> "Forward current: 25 mA"
> 
> If you can figure out what the hell that means, be my guest...

http://wiki.answers.com/Q/What_is_forward_current

Post a reply to this message

andrel <byt### [at] gmailcom> wrote:

> I think I resist any urge to comment on this. Other than "RTFM" and "buy
> any 'electronics for dummies book'" before embarking on such a project.

A good place to start would be Don Lancaster's two books, TTL COOKBOOK and CMOS
COOKBOOK. Wonderfully easy to read and comprehend. My copies are *old*,
dog-eared and marked up with my comments--I basically taught myself digital
electronics and logic from them, starting from scratch. The best books on these
subjects that I've ever run across.  (I see that they are still available at
Amazon.)

Ken

Post a reply to this message

> OK, so if I'm understanding this right... The 74126 (go look it up) 
> contains 4 "buffers" - gates who's output is logically equal to their 
> input. But each gate also has an "enable" pin. When the enable pin is 
> high, the gate works like normal. When the enable pin goes low, it's like 
> the output pin isn't connected to anything any more.

Yes exactly.

> And it seems that this allows you to connect several outputs together, 
> forming a kind of wired-OR configuration, provided that at all times only 
> one gate is "enabled".

Yep, you could make a simple address decoder logic circuit (ie have a 2 bit 
binary input to select which of the 4 buffers to enable) - that way you 
would guarantee only 1 would be active at a time.  If you were to connect a 
2-bit counter to the 2-bit address lines you would then have a 4-bit 
parallel to serial converter :-)

Post a reply to this message

> See for yourself:
>
> http://www.maplin.co.uk/Module.aspx?ModuleNo=35702
>
> "Forward current: 25 mA"
>
> If you can figure out what the hell that means, be my guest...

Load up the actual PDF spec that is linked from that page, it has more 
detail, just stating the forward current like that means nothing.

On page 3 there is the "Absolute max" table, the continuous current here is 
listed as 20/25/30 mA depending on the colour.  That is the maximum, above 
that they do not guarantee the LED will work.

On page 2 you see they've done all the spec values at 20 mA forward current.

Now look at the lower part of page 3, some interesting charts there 
(assuming the Bright Red colour, the others are on the next pages).

The one in the top right shows "brightness" against current.  You see it's 
linear up to 10 mA and then flattens out.  This is saying there isn't much 
point using more than 10 mA, it won't be very efficient.

The one in the top left tells you what current will be drawn for a 
particular voltage across the LED.  If you want 10 mA then you're going to 
need to give the LED 2.1 V.  You also see here that even adding just 10% to 
that voltage makes the LED current almost 3x higher!

Post a reply to this message

>> Why is that weird? The LED doesn't care which leg is connected to the
>> logic output pin, it's just a voltage. You could even connect the LED
>> between two output pins*, then it would only turn on when one was high
>> and the other was low.
>
>
> There's even a name for that trick: Charlieplexing.

Cool - is there also a name for when you connect a tiny speaker to two 
output pins and give one pin a square wave, and the other the inverse of the 
square wave - to make your speaker 2x louder than everyone elses? :-)

Post a reply to this message

Am 20.05.2010 09:11, schrieb scott:

> Cool - is there also a name for when you connect a tiny speaker to two
> output pins and give one pin a square wave, and the other the inverse of
> the square wave - to make your speaker 2x louder than everyone elses? :-)

That would be differential audio (also frequently but incorrectly 
referred to as balanced audio).

Post a reply to this message

>> OK, so if I'm understanding this right... The 74126 (go look it up) 
>> contains 4 "buffers" - gates who's output is logically equal to their 
>> input. But each gate also has an "enable" pin. When the enable pin is 
>> high, the gate works like normal. When the enable pin goes low, it's 
>> like the output pin isn't connected to anything any more.
> 
> Yes exactly.
> 
>> And it seems that this allows you to connect several outputs together, 
>> forming a kind of wired-OR configuration, provided that at all times 
>> only one gate is "enabled".
> 
> Yep, you could make a simple address decoder logic circuit (ie have a 2 
> bit binary input to select which of the 4 buffers to enable) - that way 
> you would guarantee only 1 would be active at a time.  If you were to 
> connect a 2-bit counter to the 2-bit address lines you would then have a 
> 4-bit parallel to serial converter :-)

I was thinking more about routing bundles of signals from place to 
place. The "obvious" thing to do is use logic gates to combine the 
signals, but I guess using these weird 3-state systems allows you to 
save a few logic gates (and hence ICs, wiring, and ripple time).

Post a reply to this message

scott wrote:

> Load up the actual PDF spec that is linked from that page, it has more 
> detail, just stating the forward current like that means nothing.

Oh, right. I hadn't even noticed that hidden away in a corner there! I 
noticed some of the other pages had this, but in *light grey* it's not 
especially prominent.

> On page 3 there is the "Absolute max" table, the continuous current here 
> is listed as 20/25/30 mA depending on the colour.  That is the maximum, 
> above that they do not guarantee the LED will work.

I see.

> Now look at the lower part of page 3, some interesting charts there 
> (assuming the Bright Red colour, the others are on the next pages).
> 
> The one in the top right shows "brightness" against current.  You see 
> it's linear up to 10 mA and then flattens out.  This is saying there 
> isn't much point using more than 10 mA, it won't be very efficient.

Right. So it will light up with just 10 mA, just not quite at full 
brightness. (On the chart I'm looking at, 15 mA looks like it's 
meaningfully brighter, but any higher than that has negligable effect.)

> The one in the top left tells you what current will be drawn for a 
> particular voltage across the LED.  If you want 10 mA then you're going 
> to need to give the LED 2.1 V.  You also see here that even adding just 
> 10% to that voltage makes the LED current almost 3x higher!

Right. So I'm looking at 2.1 V or 2.2 V or thereabouts.

Now looking at <http://focus.ti.com/lit/ug/scyd013b/scyd013b.pdf>, on 
page 231 (which is actually page 236 of the PDF file), we see that for 
the 74HC00 I'm looking at using, we have Icc <= 0.02 mA, Iol = -Ioh <= 4 
mA, and tPLH = tPHL <= 27 ns. Now, if I actually knew WTF that means...

Post a reply to this message

Orchid XP v8 wrote:

> Hey, what do you *think* I've got a dad for? ;-)

Wait - DON'T ANSWER THAT! >_<

Post a reply to this message

Am 20.05.2010 10:06, schrieb Invisible:

> I was thinking more about routing bundles of signals from place to
> place. The "obvious" thing to do is use logic gates to combine the
> signals, but I guess using these weird 3-state systems allows you to
> save a few logic gates (and hence ICs, wiring, and ripple time).

For unidirectional point-to-point connections, 3-state logic doesn't 
help you much, as there's no sane way to properly identify the 
"floating" state (unless you care to use some kind of ADCs at the 
receiving end).

Post a reply to this message

> Right. So it will light up with just 10 mA, just not quite at full 
> brightness. (On the chart I'm looking at, 15 mA looks like it's 
> meaningfully brighter, but any higher than that has negligable effect.)

Yes, every LED datasheet has a brightness against current curve.  Even for 1 
mA or 0.1 mA it will light up, but obviously it won't be very bright!  Note 
that even for 0.1 mA you probably still need a least a volt or so, which is 
why LEDs don't light up at all below a certain voltage.

> Right. So I'm looking at 2.1 V or 2.2 V or thereabouts.
>
> Now looking at <http://focus.ti.com/lit/ug/scyd013b/scyd013b.pdf>, on page 
> 231 (which is actually page 236 of the PDF file), we see that for the 
> 74HC00 I'm looking at using, we have Icc <= 0.02 mA, Iol = -Ioh <= 4 mA, 
> and tPLH = tPHL <= 27 ns. Now, if I actually knew WTF that means...

Dunno, what Icc means here, maybe the current consumption of the device 
itself?

From this page:

http://www.kpsec.freeuk.com/components/74series.htm

The HC series can sink and source up to 4 mA on the output pins if you want 
the signals to still be valid (eg for input to further logic gates).  If you 
are just using them to drive LEDs then apparently up to 20 mA is OK.

Post a reply to this message