HE-104 AC p/s
kevincw01
pc/104 board as well as the he-104 12v vehicle power supply
(http://engineering.tri-m.com/products/power/he104.htm). The system
will enventually reside in a car with 12VDC power but during
development
I need to provide the 12vDC to the he-104 in my house. I'm thinking I
can just go pick up a 12v AC/DC adapter from radio shack and cut the
connector off but I can't figure out how many amps the adapter needs.
I went through both PDFs on the he-104 site and didnt find anything
relating to the input amp requirements. They constantly repeat that it
will accept 6-40VDC input, but thats it.
Maybe it's obvious but I do not have an EE background. Please excuse
my ignorance in advance :)
Thanks for the help.
-Kevin
Palindr☻me
The spec says 50W at "up to" 95%.
I would allow 100W - so that is basically a 12v 8A supply.
As it takes such a wide range of input voltages, a regulated supply
isn't needed. It is evidently a switch mode power supply - so will have
its own local energy storage capacitors.
I would just get a continuously rated 8A 12v battery charger and hook
it up to that.
--
Sue
kevincw01
-Kevin
Pop
"kevincw01" <kevi...@gmail.com> wrote in message
news:1138125099.3...@f14g2000cwb.googlegroups.com...
: Hi, I'm building a pc/104 PC as a side project for fun. I
:
kevincw01
13.8v/15A unit for $80. Are there cheaper alternatives? I know that a
PC power supply has a +12vdc but I'm not sure what modifications I'll
need to do to get it to work off the motherboard.
-Kevin
Palindr☻me
Good Idea!
A PC power supply *should* work. Look at the rating on the unit and see
what its 12v output current rating is - it needs to be rated at 8 A or
more.
The problems you may encounter are:
1) You may need a load on the 5v output as well - particularly if it is
a cheap power supply. Something like a one ohm, 25W resistor should do.
This is because some of these power supplies use the current in the 5v
rail, as load is applied, to provide feedback to tell the power supply
to work harder. Without the load on the 5v rail, the power supply thinks
it has nothing to do - and the output to all the rails, including the 12
v rail, can be low or very unstable.
2)An AT supply may be easier for you to use than an ATX. The former just
uses a mains on/off switch - whereas the latter has to connect a switch
to the right pins of the output connector.
3)The supply was intended to be put in another case. The output wires
are short and there isn't much strain relief to stop them being pulled
too hard. Make sure they are secured properly.
4)The power supply needs the fan to work to keep the electronics cool.
So it will be noisy. You can get fanless models but they are very expensive.
Basically, you could buy a new, or second hand, AT power supply for very
little. Just tie-wrap up all the cables except one 4 wire cable that
normally goes to something like a disk or CD drive. Stick a 50 watt car
headlamp between the 12v (yellow) wire and a return (black) wire. If it
looks to be on full power (or you can measure that it has 12v across it
with a meter) - then you don't need a 5 v load. If it is dim, much below
12v or flashing, get the 1 ohm resistor and stick that between the red
and black wires. That should get the lamp nice and bright and stable.
Then you can replace the lamp with your unit. Mount the resistor, if you
need one, with care - you may need to buy a heatsink to mount it on or
stick it in the airstream coming out of the power supply.
--
Sue
VWWall
Here are instructions for modifying a PC supply to get 13.8V/ 15A
http://www.qrp4u.de/docs/en/powersupply/
This modification re-works the supply to obtain a higher voltage and
current than the original design.
This assumes an old AT PC supply. The ATX supplies have a "soft start"
circuit that requires pin #8 on the main 20 pin connector to be grounded
to turn it on after the input 120/240 AC is applied. You might find an
ATX supply whose 12V output is sufficient without modification.
--
Virg Wall, P.E.
Palindr☻me
little experience of such things.
Not only is there a good chance of electrocution and of burning the
house down, there is a good chance that the modified unit will not work.
I had to depart from the circuit mods shown in published designs in
several areas. Easy for an electronics engineer but less so for someone
not familiar with switch mode power supply design.
This particular design uses the original 12v winding of T1 - at a
current well in excess of what it was designed for. Even if the wire
thickness is the same as for the 5 v output (and I have yet to strip a
supply where this is the case), it has far more turns and may still be
unsuitable. There are alternate designs which require re-winding of T1
which are better and safer.
--
Sue
VWWall
> VWWall wrote:
>> kevincw01 wrote:
>>
>>> I took a trip to radio shack and the closest thing I could find was a
>>> 13.8v/15A unit for $80. Are there cheaper alternatives? I know that a
>>> PC power supply has a +12vdc but I'm not sure what modifications I'll
>>> need to do to get it to work off the motherboard.
>>
>>
>> Here are instructions for modifying a PC supply to get 13.8V/ 15A
>>
>> http://www.qrp4u.de/docs/en/powersupply/
>>
>> This modification re-works the supply to obtain a higher voltage and
>> current than the original design.
>>
>> This assumes an old AT PC supply. The ATX supplies have a "soft
>> start" circuit that requires pin #8 on the main 20 pin connector to be
>> grounded to turn it on after the input 120/240 AC is applied. You
>> might find an ATX supply whose 12V output is sufficient without
>> modification.
This should be pin 14, (not 8). This is usually yellow, and can be
grounded to any black wire to start the PS.
>>
> I have done such conversions but wouldn't recommend them to someone with
> little experience of such things.
I thought there might be some interest, but I agree using the +12 V
output of a standard ATX supply would be easier. A load on the +12 V
alone is sufficient for most units. I use your suggestion of a 12 V
automobile bulb. I still have a few 6 V lamps from my old auto days.
They work well on the 5 V output for testing.
>
> Not only is there a good chance of electrocution and of burning the
> house down, there is a good chance that the modified unit will not work.
Note the first four, (red), lines in the reference.
> I had to depart from the circuit mods shown in published designs in
> several areas. Easy for an electronics engineer but less so for someone
> not familiar with switch mode power supply design.
It's also probably hard to find an AT PS anymore. :-( I've seen some
for about $10, but the un-modified 12 V is only 6 A. It's a 225 W unit,
so rewinding the output transformer would work.
One would think a battery charger would provide the easiest answer, but
those having an output above a "trickle" charge are expensive.
--
Virg Wall, P.E.
Palindr☻me
LOL. What it doesn't mention is the chance the thing won't work - even
after risking life and limb...
>
>> I had to depart from the circuit mods shown in published designs in
>> several areas. Easy for an electronics engineer but less so for
>> someone not familiar with switch mode power supply design.
>
>
> It's also probably hard to find an AT PS anymore. :-( I've seen some
> for about $10, but the un-modified 12 V is only 6 A. It's a 225 W unit,
> so rewinding the output transformer would work.
>
> One would think a battery charger would provide the easiest answer, but
> those having an output above a "trickle" charge are expensive.
>
Second-hand AT supplies can be found for next to nothing, if not
nothing. I have some "Intel" 500W ones that would do the job nicely,
with more than enough amp rating on the 12v rail.
I had hesitated to suggest two or three identical low cost chargers with
their inputs and outputs paralleled. I can't think, at 01:15 in the
morning, why that shouldn't be fine. I have done that to charge a car
battery, more than once. In fact more often than I care to remember -
as I tend to try to keep resuscitating old car batteries when most
people would have pulled the plug and let them go to that big battery
heaven in the sky. But a PC supply will probably be cheaper and better.
--
Sue
ehsjr
MPJA has a 24v CT 10 amp transformer for $17.99 stock# 7846 TR
http://www.mpja.com/directview.asp?product=7846+TR
Connect stock # 4248 ($2.70) - a 25 amp bridge rectifier - to one
12V leg of the transformer lead and the center tap. Feed the output
of the bridge to your he-104
Ed
kevincw01
bridge. Thanks Ed. In the mean time, I canabalized an atx power
supply I had on hand. It's only 6amps but I'm hoping it works anyways.
I just need to get my hands on a car lamp.
-Kevin
ehsjr
I thought the he-104 needed DC input?
If so, you need the bridge.
Ed
kevincw01
auto bulb to the yellow/black on one of the 4 pin connectors. Then I
connected the yellow and black wires on the 20 pin harness. This did
nothing so I did some further reading which said it should be green.
Connecting these two did the job...sort of. The light got really
bright and the p/s fan spun up but then everything turned off. Then I
went over to radio shack and found the largest watt-rated 1ohm resistor
they had which was 10 watts. It came with two. So I wired up one of
the resistors across the black and red of the same 4 pin connector and
tried again. This time, the fan and the bulb stayed lit. Soon, I
smelled a strange odor :). I touched the resistor and it was extremely
hot so I turned everything off. I wired up 2 resistors in series,
thinking I would distribute the load and the resistors seemed to be
hotter, faster. At this point, I figured I should wait for what the
experts think.
For shock value, here is a picture of my setup. I'm sure it's not the
safest aperatus at this point.
http://netkev.com/drupal/gallery2/main.php?g2_view=core.DownloadItem&g2_itemId=4617
Palindr☻me
> OK, I went ahead with my 6am/200watt atx p/s. I connected a 12v/50watt
> auto bulb to the yellow/black on one of the 4 pin connectors. Then I
> connected the yellow and black wires on the 20 pin harness. This did
> nothing so I did some further reading which said it should be green.
> Connecting these two did the job...sort of. The light got really
> bright and the p/s fan spun up but then everything turned off.
Sounds like it did need the 5 volt load then..
>Then I
> went over to radio shack and found the largest watt-rated 1ohm resistor
> they had which was 10 watts. It came with two. So I wired up one of
> the resistors across the black and red of the same 4 pin connector and
> tried again. This time, the fan and the bulb stayed lit. Soon, I
> smelled a strange odor :). I touched the resistor and it was extremely
> hot so I turned everything off.
It is only rated at 10W and you are shoving 25W into it - yes it will
get extremely hot. The resistor will start to "cook" and probably will
smell not nice at all.
I wired up 2 resistors in series,
> thinking I would distribute the load and the resistors seemed to be
> hotter, faster. At this point, I figured I should wait for what the
> experts think.
This should only be 12.5W total - about 6W per resistor. They will still
get pretty hot getting rid of that much power, even though they are
rated for 10W. It shouldn't get hotter, faster. But maybe it was still
pretty warm from last time..
If, however, you put the two in parallel, not series, that would be 25W
each. Yes they would then each get extremely hot.
>
> For shock value, here is a picture of my setup. I'm sure it's not the
> safest aperatus at this point.
> http://netkev.com/drupal/gallery2/main.php?g2_view=core.DownloadItem&g2_itemId=4617
>
Looks like you are almost finished.
You just need to mount the two resistors (wired in series) where they
can get lots of airflow and get rid of their heat safely. I would
suggest a couple of bits of thick aluminium sheet, bent into L shapes.
Drill suitable holes in the sheets and bolt the two sheets together,
with the resistors sandwiched betweeen them. You can use the "feet" of
the assembly to screw it to somewhere suitable. The aluminium sheets
will help the resistors get rid of the heat - but if you can mount it
somewhere that it gets moving air from the/a fan - even better.
Crimping connecting wires onto the resistor leads is better than soldering.
--
Sue
VWWall
> OK, I went ahead with my 6am/200watt atx p/s. I connected a 12v/50watt
> auto bulb to the yellow/black on one of the 4 pin connectors. Then I
> connected the yellow and black wires on the 20 pin harness. This did
> nothing so I did some further reading which said it should be green.
The power on is a green wire on pin #14, which when grounded, should
turn on the PS.
> Connecting these two did the job...sort of. The light got really
> bright and the p/s fan spun up but then everything turned off. Then I
> went over to radio shack and found the largest watt-rated 1ohm resistor
> they had which was 10 watts. It came with two. So I wired up one of
> the resistors across the black and red of the same 4 pin connector and
> tried again. This time, the fan and the bulb stayed lit.
Some supplies require a minimum load on the +5 V rail to stay on.
> Soon, I
> smelled a strange odor :). I touched the resistor and it was extremely
> hot so I turned everything off. I wired up 2 resistors in series,
> thinking I would distribute the load and the resistors seemed to be
> hotter, faster. At this point, I figured I should wait for what the
> experts think.
A one ohm resistor will draw 5 Amperes at 5 V. This is 25 Watts. The
series resistors equal 2 ohms and draw 2.5 A. This is still 12.5 W, or
6.25 W/resistor. They will get hot, even though they're operating
within their power rating.
You might find that the +5 V needs less than the 2.5 A you're providing.
If you have a 6V electric torch lamp, you might try that. A low power
6V automobile lamp would be worth trying, but they're almost extinct!
Even a low power 12 V auto bulb might work. If you can find a lamp that
works you can avoid the need for a heat sink for the resistor(s). A
lamp also makes a nice power on indicator.
>
> For shock value, here is a picture of my setup. I'm sure it's not the
> safest aperatus at this point.
> http://netkev.com/drupal/gallery2/main.php?g2_view=core.DownloadItem&g2_itemId=4617
Looks like you can handle the hardware part OK. Your bench looks a lot
neater than mine!
--
Virg Wall, P.E.
kevincw01
idea. Maybe there's an LED out there at 6v.
VWWall
> Thanks, I'll try and go the 5v light route since I like your "on lamp"
> idea. Maybe there's an LED out there at 6v.
>
for the regulator circuits to work. A LED draws only about 20/1000
ampere, and their voltage is usually 2-3 V, requiring a series resistor
to limit the current. You've shown that two 1 ohm resistors in series,
(2 ohms drawing 2.5 A @ 5 V), will work OK. It probably doesn't need
2.5 A, and the goal is to find a smaller load that will work. A lamp
has the advantage that it may be easier to find, doesn't need a
heatsink, and provides a "pilot lamp". It won't hurt to try various low
power lamps rated at least 5 V.
Good luck!
--
Virg Wall, P.E.
Palindr☻me
I mostly agree, FWIW.
I suggested a 25W load as a starting point - a little less than what
could be assumed to be 10% of the rated output. I was pretty certain
that would be enough load to satisfy whatever unit the OP used. Below
about 5% then things start becoming less certain.
I, personally, would not rely on a single lamp as the minimum load. If
the lamp does fail, then the resultant instability can cause the other
rails to exceed nominal values. These rails seldom have crowbar
protection. In the OP's case - it would be unlikely that the rails could
go to voltages that could damage his particular load, however.
A standard suitably rated resistor will last the lifetime of the system.
--
Sue
Roy L. Fuchs
<sb38...@hotmail.com.invalid> Gave us:
>I suggested a 25W load as a starting point - a little less than what
>could be assumed to be 10% of the rated output.
It only needs to be a few watts of loading to correct the regulation
circuits.
kevincw01
resistor then I calculated that a 2.5ohm resistor would give me 25
watts at 5vdc. I'm guessing, I'll have to move up to 3 ohms...
Palindr☻me
I have one such supply here that very certainly will not stabilise
without a minimum of 15W loading on the 5v rail.
However, you are perfectly correct in that it *usually* only requires a
few watts. Many need no external minimum load at all.
I haven't come across a single ATX supply that needs a minimum load.
IIRC, some are even marketed on that basis.
I also have some 5v 250A power supplies - with 4 terminal outputs such
that the actual supply can rise to 8v on full load.
I suppose it was easier to design as they are only single output but
they need no minimum load at all. Still, I wish I had been given that
project.
I keep hoping one will go wrong so I have an excuse to open it up and
work out what the design is....
I don't really know why I bought them - they were sitting there in a
scrap box and I just wanted to have them. Hundreds of amps has always
appealed. My ex-husband didn't mind me buying such things - I suppose
that I was a lot cheaper than someone who does the same thing with cars
or clothes ;)
--
Sue
Roy L. Fuchs
<sb38...@hotmail.com.invalid> Gave us:
>I don't really know why I bought them - they were sitting there in a
>scrap box and I just wanted to have them. Hundreds of amps has always
>appealed. My ex-husband didn't mind me buying such things - I suppose
>that I was a lot cheaper than someone who does the same thing with cars
>or clothes ;)
Make me your current husband. I won't ever object, and I even
participate in "junk" purchases. I have six "old" "screamer"
computers to prove that! :-] They were new though at one time.
Some of my ancestors ran junk yards! I think a modern junk yard would
be cool!
I do dishes and windows too! ;-]
Palindr☻me
You realise, of course, that a proposal on usenet in front of two or
more Chartered Engineers is legally binding in 43 countries?
--
Sue
kevincw01
Roy L. Fuchs
<sb38...@hotmail.com.invalid> Gave us:
Well, unless you are just more ugly than Minnie Pearl...
Michael A. Terrell
>
> I have one such supply here that very certainly will not stabilise
> without a minimum of 15W loading on the 5v rail.
>
> However, you are perfectly correct in that it *usually* only requires a
> few watts. Many need no external minimum load at all.
>
> I haven't come across a single ATX supply that needs a minimum load.
> IIRC, some are even marketed on that basis.
>
> I also have some 5v 250A power supplies - with 4 terminal outputs such
> that the actual supply can rise to 8v on full load.
>
> I suppose it was easier to design as they are only single output but
> they need no minimum load at all. Still, I wish I had been given that
> project.
>
> I keep hoping one will go wrong so I have an excuse to open it up and
> work out what the design is....
>
> I don't really know why I bought them - they were sitting there in a
> scrap box and I just wanted to have them. Hundreds of amps has always
> appealed. My ex-husband didn't mind me buying such things - I suppose
> that I was a lot cheaper than someone who does the same thing with cars
> or clothes ;)
>
> --
> Sue
I'll bet my collection of junk is bigger than yours! I can't park
inside my 30' * 40' garage, the 18' * 12' machine shop is so full that I
can barely get to my drill press and the 12' * 12' tool shed is full to
the door. Good thing that door opens out, not in. I have enough
projects lined up to last me the rest of my life. ;-)
--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.
Michael A. Terrell
Central Florida
Palindr☻me
A tad tricky as, when I looked up her biography she appears to be dead.
However, borrow any of my tools without asking ...
--
Sue
Roy L. Fuchs
<sb38...@hotmail.com.invalid> Gave us:
That's the first time in my life that remark has ever made my mouth
water... (not the minnie pearl remark) :-]
Michael A. Terrell
>
> Well, unless you are just more ugly than Minnie Pearl...
What a guy. :( For your information, Minnie Pearl was a country
comedy stage act, not a real person. Sarah Cannon who was known as
"Minnie Pearl" on the "grand Ole Opry" and at other country venues spent
most of her life raising millions of dollars to help children around
Nashville, Tennessee.
Michael A. Terrell
>
> However, borrow any of my tools without asking ...
You tell him! There are no second chances. ;-)
Roy L. Fuchs
<mike.t...@earthlink.net> Gave us:
>"Roy L. Fuchs" wrote:
>>
>> Well, unless you are just more ugly than Minnie Pearl...
>
>
> What a guy. :( For your information, Minnie Pearl was a country
>comedy stage act, not a real person.
You're an idiot (what a dolt?). Minnie pearl was a regular on "The
Beverly Hillbillies", as well as "Hee Haw". That was years AFTER she
already had made her own fame doing country comedy. Granted it was
her stage name, but so too was "Marilyn Monroe" a stage name. So
what?
> Sarah Cannon who was known as
>"Minnie Pearl" on the "grand Ole Opry" and at other country venues spent
>most of her life raising millions of dollars to help children around
>Nashville, Tennessee.
Yeah... so... she also wasn't the cutest peach in the bushel
basket. Jeez... it was a joke... Get over it.
snip
You link twinks are funny. I knew exactly who she was when I posted
what I said, dufus.
Roy L. Fuchs
<mike.t...@earthlink.net> Gave us:
>Palindr?me wrote:
>>
>> However, borrow any of my tools without asking ...
>
>
> You tell him! There are no second chances. ;-)
I saw it as an offer, not a warning.
Michael A. Terrell
No, you're the joke, get over it.
Roy L. Fuchs
<mike.t...@earthlink.net> Gave us:
> No, you're the joke, get over it.
Fuck you, retard. Fix your filter list so I don't see your crap
anymore. I know there is a 90% chance that will go right over your
head, but...
kevincw01
fighting.
Palindr☻me
> oh gosh, guess I have to start a new thread to get away from the
> fighting.
>
That isn't fighting btw. You must be new to usenet.
--
Sue
kevincw01
build a new work bench. I'll post a new topic once I get re-setup :)
Michael A. Terrell
You barely hit .003% on the "Troll-O-Meter." Grow up, little girl.
You sound more like Phil every day.
Art Deco
Numby! Nice to see you're still working on those hard bits of usenet
posting, Genius!
--
Official Associate AFA-B Vote Rustler
Official Overseer of Kooks and Saucerheads in alt.astronomy
Co-Winner, alt.(f)lame Worst Flame War, December 2005
"Causation of gravity is missing frame field always attempting
renormalization back to base memory of equalized uniform momentum."
-- nightbat the saucerhead-in-chief
"Have patience. First I shall deal with the State of Oregon
and County of Josephine, Then the AFAB, government/media
disinformation Agents with whom you conspire to libel me and my
family. Your time will come."
-- Raymond Ronald KarczewskiŠ, usenet "christ"
"Classic unsubstantiated and erroneous claim, and rather ironic, coming from
someone who obviously has no understanding of what a signature is. Tell me,
Haslam, do you sign your checks as 'Can't you show a little restraint?'"
-- David Tholen, Clueless Newbie of the Month, February 2003