I've just started with some very basic digital electronics articles, and in one of those, I'm supposed to test the behavior of some electronic gates (TTL 7400 series). In order to do that, I need a power supply (they recommend something between 7-12 V, minimum 100 miliamps). They use one 5-volt voltage regulator and two electrolytic capacitors to supply clean regulated 5V to the breadboard.
The thing is that I already have a PC power supply spare, and I think I could use it, and if I'm not wrong, the voltage coming out of a PC power supply is already regulated (at least there are lots of electonic components inside my power source).
My problem: the damn thing is an ATX power supply. There's no ON/OFF button. The PC motherboard is responsible to send this signal to the power supply. How do I simulate this signal to make the power supply working constantly?
Thanks
PS: I don't know if this is the most appropriate newsgroup to post this question, if not, please advice.
"Padu" <p...@merlotti.com> wrote: >I've just started with some very basic digital electronics articles, and in >one of those, I'm supposed to test the behavior of some electronic gates >(TTL 7400 series). >In order to do that, I need a power supply (they recommend something between >7-12 V, minimum 100 miliamps). They use one 5-volt voltage regulator and two >electrolytic capacitors to supply clean regulated 5V to the breadboard.
>The thing is that I already have a PC power supply spare, and I think I >could use it, and if I'm not wrong, the voltage coming out of a PC power >supply is already regulated (at least there are lots of electonic components >inside my power source).
>My problem: the damn thing is an ATX power supply. There's no ON/OFF button. >The PC motherboard is responsible to send this signal to the power supply. >How do I simulate this signal to make the power supply working constantly?
>Thanks
>PS: I don't know if this is the most appropriate newsgroup to post this >question, if not, please advice.
An ATX power supply is *perfect* for your needs. There are just a few odd things you need to know about how it works... Below is an article that I posted to a different newsgroup yesterday, responding to a question of "how do I test an ATX psu". You'll find all the information you need!
There are two requirements for an ATX power supply to "act normally". One is the PS-ON lead (green wire) must be grounded (to a Black wire). The other is there must be a minimum load on the 5 volt line. I've looked around, and cannot handily find any specification on what the "minimum load" must be though. It seems that it is extremely small, and an old disk drive or cdrom will be enough. Another guess would be a 25 ohm 5 or 10 watt resistor, which would draw about 200ma of current across 5 vdc.
If you really want to test it, here's a proceedure, which assumes a 20-pin plug (the colors would be the same, but see below for a 24-pin plug pinout).
1) Plug AC in.
2) Measure pin 9 (Purple wire). Should show about 5.0 vdc.
3) Measure pin 14 (Green wire). Should show some voltage between 3.5 and 5 vdc. The exact voltage is not critical.
4) Unplug AC.
5) Put a load across the 5 vdc line. You can do this by using an old disk drive, cdrom, or a resistor.
6) Jumper pin 14 (Green wire) to ground (any Black wire).
7) Plug AC in.
8) A. The fan should run. B. All voltages should measure close to their nominal values (Note that -5v and -12v may be very poorly regulated).
Note that pin 14 is the PS-ON signal line, which is normally wired to the on/off switching circuit on the motherboard. It has a 1000 ohm pull-up resistor connected (internal to the PSU) to the +5v Standby line, and if there is no other connection it should probably read close to 5 volts. That voltage will cause the power supply to be "off". The nominal switching point for PS-ON is 0.8vdc, and standard operating voltages are less than 0.4 volts for "on" and more than 3.5 volts for "off".
Here is a list of leads on a 20 pin connector,
PIN COLOR NAME DESCRIPTION ------------------------------------ 1 Orange +3.3V Power, +3.3V 2 Orange +3.3V Power, +3.3V 3 Black GND Ground 4 Red +5V Power, +5V 5 Black GND Ground 6 Red +5V Power, +5V 7 Black GND Ground 8 Gray PWR-OK Power OK 9 Purple +5V VSB +5V VSB 10 Yellow +12V Power, +12V 11 Orange +3.3V Power, +3.3V 12 Blue -12V Power, -12V 13 Black GND Ground 14 Green PS-ON PS Remote on/off 15 Black GND Ground 16 Black GND Ground 17 Black GND Ground 18 White -5V Power, -5V 19 Red +5V Power, +5V 20 Red +5V Power, +5V
And this is a list of leads on a 24 pin connector,
PIN COLOR NAME DESCRIPTION ------------------------------------ 1 Orange +3.3V Power, +3.3V 2 Orange +3.3V Power, +3.3V 3 Black GND Ground 4 Red +5V Power, +5V 5 Black GND Ground 6 Red +5V Power, +5V 7 Black GND Ground 8 Gray PWR-OK Power OK 9 Purple +5V VSB +5V VSB 10 Yellow +12V Power, +12V 11 Yellow +12V Power, +12V 12 Orange +3.3V Power, +3.3V 13 Orange +3.3V Power, +3.3V 14 Blue -12V Power, -12V 15 Black GND Ground 16 Green PS-ON PS Remote on/off 17 Black GND Ground 18 Black GND Ground 19 Black GND Ground 20 White -5V Power, -5V 21 Red +5V Power, +5V 22 Red +5V Power, +5V 23 Red +5V Power, +5V 24 Black GND Ground
The -5 VDC line may not exist.
The +5 VSB supply is Standby Power, which supplys parts of the motherboard which are always powered up (to allow options like "wake-on ..." to work).
The PWR-OK line is at ~5 VDC if the AC input and the +5V and +12V lines are within specifications. If either the +5V or the +12V line falls to below the specified voltage tolerance or if AC is lost for more than one power cycle interval, then PWR-OK, will drop to ~0 VDC. (If it helps any, if AC is lost, PWR-OK is supposed to drop at least 1 ms before the +5V and +12V lines go below specified voltage tolerances!)
> "Padu" <p...@merlotti.com> wrote: > >I've just started with some very basic digital electronics articles, and in > >one of those, I'm supposed to test the behavior of some electronic gates > >(TTL 7400 series). > >In order to do that, I need a power supply (they recommend something between > >7-12 V, minimum 100 miliamps). They use one 5-volt voltage regulator and two > >electrolytic capacitors to supply clean regulated 5V to the breadboard.
> >The thing is that I already have a PC power supply spare, and I think I > >could use it, and if I'm not wrong, the voltage coming out of a PC power > >supply is already regulated (at least there are lots of electonic components > >inside my power source).
> >My problem: the damn thing is an ATX power supply. There's no ON/OFF button. > >The PC motherboard is responsible to send this signal to the power supply. > >How do I simulate this signal to make the power supply working constantly?
> >Thanks
> >PS: I don't know if this is the most appropriate newsgroup to post this > >question, if not, please advice.
> An ATX power supply is *perfect* for your needs. There are just > a few odd things you need to know about how it works... Below > is an article that I posted to a different newsgroup yesterday, > responding to a question of "how do I test an ATX psu". You'll > find all the information you need!
> There are two requirements for an ATX power supply to "act > normally". One is the PS-ON lead (green wire) must be grounded > (to a Black wire). The other is there must be a minimum load on > the 5 volt line. I've looked around, and cannot handily find > any specification on what the "minimum load" must be though. It > seems that it is extremely small, and an old disk drive or cdrom > will be enough. Another guess would be a 25 ohm 5 or 10 watt > resistor, which would draw about 200ma of current across 5 vdc.
> If you really want to test it, here's a proceedure, which assumes > a 20-pin plug (the colors would be the same, but see below for a > 24-pin plug pinout).
> 1) Plug AC in.
> 2) Measure pin 9 (Purple wire). Should show about 5.0 vdc.
> 3) Measure pin 14 (Green wire). Should show some voltage > between 3.5 and 5 vdc. The exact voltage is not critical.
> 4) Unplug AC.
> 5) Put a load across the 5 vdc line. You can do this > by using an old disk drive, cdrom, or a resistor.
> 6) Jumper pin 14 (Green wire) to ground (any Black wire).
> 7) Plug AC in.
> 8) A. The fan should run. > B. All voltages should measure close to their > nominal values (Note that -5v and -12v may be > very poorly regulated).
> Note that pin 14 is the PS-ON signal line, which is normally > wired to the on/off switching circuit on the motherboard. It > has a 1000 ohm pull-up resistor connected (internal to the PSU) > to the +5v Standby line, and if there is no other connection it > should probably read close to 5 volts. That voltage will cause > the power supply to be "off". The nominal switching point for > PS-ON is 0.8vdc, and standard operating voltages are less than > 0.4 volts for "on" and more than 3.5 volts for "off".
> Here is a list of leads on a 20 pin connector,
> PIN COLOR NAME DESCRIPTION > ------------------------------------ > 1 Orange +3.3V Power, +3.3V > 2 Orange +3.3V Power, +3.3V > 3 Black GND Ground > 4 Red +5V Power, +5V > 5 Black GND Ground > 6 Red +5V Power, +5V > 7 Black GND Ground > 8 Gray PWR-OK Power OK > 9 Purple +5V VSB +5V VSB > 10 Yellow +12V Power, +12V > 11 Orange +3.3V Power, +3.3V > 12 Blue -12V Power, -12V > 13 Black GND Ground > 14 Green PS-ON PS Remote on/off > 15 Black GND Ground > 16 Black GND Ground > 17 Black GND Ground > 18 White -5V Power, -5V > 19 Red +5V Power, +5V > 20 Red +5V Power, +5V
> And this is a list of leads on a 24 pin connector,
> PIN COLOR NAME DESCRIPTION > ------------------------------------ > 1 Orange +3.3V Power, +3.3V > 2 Orange +3.3V Power, +3.3V > 3 Black GND Ground > 4 Red +5V Power, +5V > 5 Black GND Ground > 6 Red +5V Power, +5V > 7 Black GND Ground > 8 Gray PWR-OK Power OK > 9 Purple +5V VSB +5V VSB > 10 Yellow +12V Power, +12V > 11 Yellow +12V Power, +12V > 12 Orange +3.3V Power, +3.3V > 13 Orange +3.3V Power, +3.3V > 14 Blue -12V Power, -12V > 15 Black GND Ground > 16 Green PS-ON PS Remote on/off > 17 Black GND Ground > 18 Black GND Ground > 19 Black GND Ground > 20 White -5V Power, -5V > 21 Red +5V Power, +5V > 22 Red +5V Power, +5V > 23 Red +5V Power, +5V > 24 Black GND Ground
> The -5 VDC line may not exist.
> The +5 VSB supply is Standby Power, which supplys parts of the > motherboard which are always powered up (to allow options like > "wake-on ..." to work).
> The PWR-OK line is at ~5 VDC if the AC input and the +5V and > +12V lines are within specifications. If either the +5V or the > +12V line falls to below the specified voltage tolerance or if > AC is lost for more than one power cycle interval, then PWR-OK, > will drop to ~0 VDC. (If it helps any, if AC is lost, PWR-OK is > supposed to drop at least 1 ms before the +5V and +12V lines go > below specified voltage tolerances!)
To make sure you know this, PC power supplies are not supposed to operate without any load. So if you plan to use this kind of power supply make sure you have always a load attached (for example, a low voltage light bulb, a power resistor etc.) -- VortexOne ************************************************************************ vort...@hotmail.com (Messages trashed without the word KEYWAY in subject line) ************************************************************************
Minimum PSU load is a function of PSU design. Just another reason why all PSUs require specifications.
I believe Radio Shack sells something like a 10 ohm 10 watt resistor. That would be more than sufficient load for an ATX PSU.
This load requirement varies from supply to supply. Some that required a minimum load often would still power up - almost every time. Therefore they required maybe a 100 milliamps load. But again, that is why specifications are always provided by a responsible supply manufacturer.
Welcome to TTL. Where supply meets test circuit, put at least one 10 uf capacitor, +5 to ground. Also a 0.01 uf ceramic capacitor is installed in the area of TTL power and ground pins - maybe one capacitor for every three or five TTL chips.
Superior than TTL is CMOS Logic. One basic difference. CMOS inputs must always have a pullup or pulldown connection - something that does not adversely affect TTL.
> To make sure you know this, PC power supplies are not supposed to > operate without any load. So if you plan to use this kind of power > supply make sure you have always a load attached (for example, a low > voltage light bulb, a power resistor etc.)
In many supplies, only a watt is more than sufficient. Until one has manufacturer's specs, then one cannot say how much or if any load is required. It varies for every design.
> "w_tom" <w_t...@hotmail.com> wrote in message > news:3E540B36.F83183C@hotmail.com... > > Minimum PSU load is a function of PSU design. Just another > > reason why all PSUs require specifications.
> > I believe Radio Shack sells something like a 10 ohm 10 watt > > resistor. That would be more than sufficient load for an ATX > > PSU.
> Not at 12V. 12V/10Ohm = 1.2A 1.2A * 12V = 14.4W Put two in series. > ;-)
"Anthony Fremont" <spam...@houston.rr.com> wrote: >"w_tom" <w_t...@hotmail.com> wrote in message >news:3E540B36.F83183C@hotmail.com... >> Minimum PSU load is a function of PSU design. Just another >> reason why all PSUs require specifications.
>> I believe Radio Shack sells something like a 10 ohm 10 watt >> resistor. That would be more than sufficient load for an ATX >> PSU.
>Not at 12V. 12V/10Ohm = 1.2A 1.2A * 12V = 14.4W Put two in series. >;-)
>michael
If that were what one wanted to do, three in series would be suitable.
However, it isn't the 12v bus that needs a load. Either the 5v or the 3.3v bus must have a load on it. 10 Ohms isn't a good idea there either. Something like 25 ohms on the 5 volt bus would be just fine. That will dissipate 1 watt, so a 10 Watt resistor would be just fine... and two of these 10 Ohm 10 Watt units in series would be OK too.
> > "w_tom" <w_t...@hotmail.com> wrote in message > > news:3E540B36.F83183C@hotmail.com... > > > Minimum PSU load is a function of PSU design. Just another > > > reason why all PSUs require specifications.
> > > I believe Radio Shack sells something like a 10 ohm 10 watt > > > resistor. That would be more than sufficient load for an ATX > > > PSU.
> > Not at 12V. 12V/10Ohm = 1.2A 1.2A * 12V = 14.4W Put two in series. > > ;-)
> > michael "w_tom" <w_t...@hotmail.com> wrote in message
> In many supplies, only a watt is more than sufficient. Until > one has manufacturer's specs, then one cannot say how much or > if any load is required. It varies for every design.
No qualms with that, I was only pointing out the potential for overheating the resistor if he attached it to the 12V supply. I have no idea what a "suitable" load would be for operating the supply like this. As you said, I'm sure that it would vary by supply. ;-)
I solved the problem. I had an old HD around and then I've plugged it in one of the available wires. Works perfectly.
This weekend I've tested the behavior of a NAND port (TTL 7400). Being a software engineer, I know what's a NAND port and what it should behave like, but for my surprise, I cought myself amazed when leds turned on and off with my experiments.
I think I'm too used with 0's and 1's instead of 0V and 5V.
> Minimum PSU load is a function of PSU design. Just another > reason why all PSUs require specifications.
> I believe Radio Shack sells something like a 10 ohm 10 watt > resistor. That would be more than sufficient load for an ATX > PSU.
> This load requirement varies from supply to supply. Some > that required a minimum load often would still power up - > almost every time. Therefore they required maybe a 100 > milliamps load. But again, that is why specifications are > always provided by a responsible supply manufacturer.
> Welcome to TTL. Where supply meets test circuit, put at > least one 10 uf capacitor, +5 to ground. Also a 0.01 uf > ceramic capacitor is installed in the area of TTL power and > ground pins - maybe one capacitor for every three or five TTL > chips.
> Superior than TTL is CMOS Logic. One basic difference. > CMOS inputs must always have a pullup or pulldown connection - > something that does not adversely affect TTL.
> VortexOne wrote:
> > To make sure you know this, PC power supplies are not supposed to > > operate without any load. So if you plan to use this kind of power > > supply make sure you have always a load attached (for example, a low > > voltage light bulb, a power resistor etc.)
You are now ready to get your hands dirty. Simplest solution is to find a book entitled "TTL Cookbook" or "CMOS Cookbook". Concepts associate with digital logic are explained. You now have a good platform to demonstrate what you read.
Also learn about a famous chips called the 555. It is a 'classic' cousin to digital logic.
> I solved the problem. I had an old HD around and then I've plugged it in one > of the available wires. Works perfectly.
> This weekend I've tested the behavior of a NAND port (TTL 7400). Being a > software engineer, I know what's a NAND port and what it should behave like, > but for my surprise, I cought myself amazed when leds turned on and off with > my experiments.
> I think I'm too used with 0's and 1's instead of 0V and 5V.
