12V DC to 9V AC/50 Hz
Paul Förster
... I found the clock generator in section 12.7 (step 8) of the
comp.sys.cbm FAQ. It generates 60 Hz out of 12 V DC.
How would I have to modify the circuit if I wanted 9V AC @ 50 Hz
output? Yes, I live in PAL country and need 50 Hz here and also need 9V
for CIAs and user port...
I thought about using an 7809 to convert the 12V DC input to 9V DC.
Then generate the 50 Hz AC pulse using a variation of the circuit
mentioned in the FAQ.
Any help is appreciated. Thanks very much in advance.
--
cul8er
Clocky
Do you have a link to the circuit?
Paul Förster
> Do you have a link to the circuit?
... yes, it's right here in the c.s.c FAQ:
http://www.faqs.org/faqs/commodore/main-faq/part7/
Here's the circuit:
------------------------------------------------------------------------
8. You could generate this clock using a variety of methods. Here's one:
+12V
__________________________|
| |
| C1 R1 ______|_______
|-)|-+-/\/\/-+-----|6 8 |
| | \ | |
| XTAL1 / R2 | MM5369 |
| | \ | |
|-)|-+-------+-----|5 1|---> 60 Hz (to R37 pad)
C2 | 2 |
---------------
|
GND
C1 = 30 pF
C2 = 3-15 pF (variable)
XTAL1 = 3.57 MHz crystal
R1 = 1k
R2 = 10M
It is suggested that you use a short length of shielded coax cable to
connect the 60 Hz output to the R37/R100 pad.
Also, it is suggested that you use a frequency counter to tune the above
circuit to exactly 60 Hz.
This conversion would work well if you wanted to use the C64 in a car
or powered
by a solar set-up.
------------------------------------------------------------------------
What I need is 9V AC @ 50 Hz output. Input would ideally be 12V DC, but
I could do with putting a 7809 after 12V DC to make it 9V DC. That
wouldn't be the problem. The problem is the 50 Hz there.
--
cul8er
Andreas Beermann
depending on the accuracy you require you could use a 3MHz crystal
rather than the 3.57MHz one. This would result in 50.3 Hz or a
deviation of 0.6% - when running a clock from this frequency if would
run ~8.5 minutes too fast per day.
Don't know if you can accept this.
Andi
Clocky
Replace the crystal with a 2.982950 MHz crystal for 50Hz output.
See here for a possible source for everything you need
http://www.angelfire.com/electronic2/index1/mm5369.html
Here is another design to suit a PAL C64, for your consideration (no idea if
it works, just had it bookmarked for future reference)
http://www.ide64.org/stuff.html
Paul Förster
On 2009-09-28 10:59:55 +0200, "Clocky" <nic...@migo.com> said:
> Replace the crystal with a 2.982950 MHz crystal for 50Hz output.
> See here for a possible source for everything you need
> http://www.angelfire.com/electronic2/index1/mm5369.html
>
> Here is another design to suit a PAL C64, for your consideration (no idea if
> it works, just had it bookmarked for future reference)
> http://www.ide64.org/stuff.html
... sorry for getting back so late. Well, in the meantime I checked my
options. Seems like I have three of them:
a) Go with angelfire.com
b) Build the ide64.org converter
c) Take a 12V AC power supply and do the following:
12V AC ----+---- rectify -----+----------- 12V DC
| |
| +-- 7805 --- 5V DC
reduce to 9V AC
Something like that. I don't know which one to give the first shot,
tho. Solution a), I don't know. It may work or it may not work. And if
it works, how accurate is it? I can generate 50 Hz using a simple
HTC555 without any chrystal at all. But that'd run at 5V and is
anything but a sine wave...
For solution b) I need it to be way smaller, half of it's size at best...
Solution c), well, how do I reduce 12V AC to 9V AC without a real
transfomer? Using a resistor requires knowing U and I. I know U
obviously but not I. So how can I know R in an equation with three
variables and two unknown?
I could, however, ad the HTC555 solution to solution c) so that I could
run the C64 on both 12V AC and 12V DC. The HTC555 would drive the TODs
on the 5V DC basis. That'd mean, I'd have a perfect 50 Hz on DC on the
TOD pins but no 9V AC on the user port in case I plug it into 12V DC
(e.g. a car).
What do you think?
--
cul8er
bogax
Paul Förster wrote:
> Hi Clocky, hi Andy,
9 VAC is 12.7 volts peak which is why the original clock circuit
uses 12 VDC it's close enough. (I'm guessing, but it seems obvious)
If you're trying to build a replacement power supply that runs from eg
a
battery that you can just plugin in in place of the original, you
really need
the 9VAC to be floating relative to the rest which pretty much
requires
a transformer or perhaps a completely seperate supply with it's own
battery.
If you're willing to hack into the computer, it'd probably be simpler
to
make you're own 12 VDC source and wire it in, but it sounds like you
don't realy know enough to attempt that.
The reason the original clock circut uses a 3.58 MHz cristal (and the
reason the MM5369 uses that frequency) is that 3.58 MHz is the NTSC
color frequency so there's lots of those crystals around, they're
cheap
and easy to get ahold of.
Building your own equivalent to the MM5369 wouldn't be all that much
harder than the original clock circuit (just an oscillator and a
divider
basically) and you could build it to use what ever you can get ahold
of.
In some ways it might be simpler. You could use a canned osillator
for example (a crystal oscillator in a can that you just solder in
like
any other component)
You can get CMOS 555s and I think they go to 18V, but a 555 isn't
going to be very accurate.
If you have an (approximately) 12 VDC source (a car battery) then
you can try and get your 9VAC from that which can be a relatively
simple
square wave oscillator or a slightly more complex modified sine wave
supply. but they'll probably be (electrically) noisy
A modified sine wave is (usually) a square-ish wave with a peak
voltage
similar to a sine wave and the duty cycle messed with to give it an
RMS
closer to a sine wave.
Since 12 VDC to 120 VAC modified sine wave inverters are fairly
common, a little Googling should turn up several circuits varying
from purely awful to passably good that could be adapted to your
purpose (probably just use a different transformer)
Paul Förster
On 2009-10-09 00:27:19 +0200, bogax <bog...@gmail.com> said:
> If you're willing to hack into the computer, it'd probably be simpler
> to make you're own 12 VDC source and wire it in, but it sounds like you
> don't realy know enough to attempt that.
... yes, I don't know that much. But I'm doing exactly that and I'm
learning. Up to now, I didn't destroy anything but fry a CIA, which I
replaced ok. :-P
Fact is, I have 12V and 5V DC available on the board but I'd need 9V AC
as well.
> The reason the original clock circut uses a 3.58 MHz cristal (and the
> reason the MM5369 uses that frequency) is that 3.58 MHz is the NTSC
> color frequency so there's lots of those crystals around, they're
> cheap and easy to get ahold of.
... ah, ok. Didn't know that.
> You can get CMOS 555s and I think they go to 18V, but a 555 isn't
> going to be very accurate.
... the 555 has a deviation of some 5% or so, depending on the
temperature, which is way too much in my eyes...
> If you have an (approximately) 12 VDC source (a car battery) then
> you can try and get your 9VAC from that which can be a relatively
> simple square wave oscillator or a slightly more complex modified sine wave
> supply. but they'll probably be (electrically) noisy
... more noisy than the normal power line coming out of the wall? If
not, then that would be an option if it can be done *small*. I've been
looking into converting 12V DC to 9V AC (sine or at least similar)
quite some time now and all I get is "this is impossible". So, if you
have a solution, I'm open to it.
> Since 12 VDC to 120 VAC modified sine wave inverters are fairly
> common, a little Googling should turn up several circuits varying
> from purely awful to passably good that could be adapted to your
> purpose (probably just use a different transformer)
... yes, and they're the size of a brick... Can't that be done very small?
--
cul8er
Paul Förster
... in the meantime since my original post, a friend came up with the
following:
+---------- GND
|
/-\ +----+
12V AC / 50Hz ---+----< >---+ +---|7805|--- 5V DC
| \+/ | | +----+
/ | | +----+ |
R=6 Ohm / 7 W \ +----------|7812|---+------------ 12V DC
/ | +----+
| |
+------------------ 9V AC / 50Hz
| |
/ |
R=18 Ohm / 7 W \ |
/ |
| |
12V AC / 50Hz ---+--------------+--- 9V AC / 50Hz
He said the resistors should get only slightly warm. If this would work
then I would have 5V DC, 12V DC and 9V AC directly from the power
supply.
That would be a solution to my problem because it not only provides 9V
AC, but it also provides the 50Hz without having to explicitly generate
it.
With this, I could revive User Port pins 10 and 11 with real 9V AC sine
wave as well as the CIAs' TOD clock pins 19.
And if I for example plugged it into a DC source like a car, I would
not have 9V AC at the User Port and also not have the CIAs' TOD
working. But the C64 should work otherwise, right?
What do you think of this? Is this a viable option? I'd like to hear
your opinions.
--
cul8er
bogax
Three problems come to mind immediately.
1 if that's a block diagram sort of thing, fine
if it's supposed to be a schematic you need some filter capacitors
at least.
2 it's a simple resistive divider for the 9VAC so if you load the 9VAC
at all, the voltage will droop and it will vary with the load.
3 iirc The 9VAC is tied to ground through diodes in a charge pump
voltage doubler in the C64 so I think you would at least have to
get the poarity right if it worked at all.
if you have a source of AC it would (IMO) be a lot easier to just
use a transformer, even if you had to rewind one for 12VAC to 9VAC
Also you'd be wasting some power in the resistive voltage divider
much more than you would be drawing on the 9VAC if you don't
want the voltage to vary much.
It would probably make more sense to use some power rectifiers
in series to drop the voltage.
Assuming .7 V per rectifier gives you 2.8 V for a string of four
with two strings of four in inverse parallel so your output is
12VAC-2.8 = 9.2 VAC, which should be close enough.
There would be a dead spot at the zero crossings that would
add some noise, but probably not enough to cause problems.
As to a 12VDC to 9VAC inverter the biggest bit (I would guess)
will be the transformer. For something simple I would expect it
to be approximately the same size as an equivalent wall wart,
Or maybe a little more allowing for heat sinks.
Roughly speaking, making an inverter efficient (and simple)
is what makes it (electricaly) noisy if you don't mind it not
being efficient then it can be relatively quiet and simple.
If you're not drawing much power on the 9VAC maybe
efficiency won't matter.
I can't remember just off hand what the user port is rated for,
but say it's 500ma then you might be talking 6W to supply
that relatively efficiently (75%) but noisy and 12W to do it
not so efficiently (37%) not so noisy.
(I'm thinking cheap audio amplifiers or maybe a couple as
a bridge amp)
Paul Förster
... ok, it's been two weeks and I have much to do and little time. :-(
A friend suggested the following in the meantime:
+----------+------+------+-------+------+---o GND
| | | | | |
| === C1 | === C2 | === C3
/-\ | | | +--+--+ |
Vin o---+---< >---+ | | +----| VR2 |---+---o Vdc1
| \+/ | | | | +-----+
/ | | | +--+--+ |
R1 \ +----------+---| VR1 |---+------------------o Vdc2
/ | +-----+
| |
+-----------------o Vac
| |
/ |
R2 \ |
/ |
| |
Vin o---+-------------+---o Vac
Vin = 12V AC, 50 Hz
Vac = 9V AC, 50 Hz
Vdc1 = 5V DC
Vdc2 = 12V DC
VR1 = LT1083CP-12
VR2 = 7805
R1 = 6 Ohm, 7W
R2 = 18 Ohm, 7W
C1 = 3000 �F (Elko, proof voltage min. 25V)
C2 = 3000 �F (Elko, proof voltage min. 15V)
C3 = 10 �F (Tantalum capacitor, proof voltage min. 7V)
Sorry if the terms might be wrong, but I don't know the exact English
terms. That'd be for a 3A supply. He said C1 and C2 should be 1000 �F
for each amp.
--
cul8er
bogax
> Hi Bogax,
>
> On 2009-10-19 02:11:20 +0200, bogax <bog...@gmail.com> said:
> > 3 iirc The 9VAC is tied to ground through diodes in a charge pump
> > voltage doubler in the C64 so I think you would at least have to
> > get the poarity right if it worked at all.
I was wrong it's not the charge pump, it's another bridge rectifier.
here's a (not very readable) schematic:
http://www.zimmers.net/anonftp/pub/cbm/schematics/computers/c64/manual-html/Page_04.html
> A friend suggested the following in the meantime:
All I can suggest is imagine your circuit with both sides of your
9VAC clamped to ground with diodes on the negative excursions
and then make it work.
ie for both sides of your 9VAC attach the cathode of a diode and
ground
the anodes (that's the diodes in the internal bridge rectifier).
Even if you pull the C64's internal brdge, anything you plug into the
user port
might have the same thing.
And don't forget there's (probably) a good reason for those seperate
internal regulators..
.
saab...@gmail.com
Would you be able to help me design the circuit needed to create 9VAC from 12 VDC? If I had a schematic, I'm pretty sure I could go out and get the parts and solder it together.
This is the first time I've tried messing with building an AC circuit, so I am not sure where to start
Thanks in advance if you can help.
On Friday, November 6, 2009 3:19:31 PM UTC-6, bogax wrote:
> On Nov 5, 3:58 am, Paul Förster