Maker Pro
Maker Pro

36V >6A power supply?

D

Dan

Jan 1, 1970
0
Hi, I'm looking to build the aforementioned supply for a 6A 36V motor.
Since I can't find anyone with 10:3 transformers, I guess I'm stuck with
building a power supply that takes the ~60V half wave (after
rectification) that the wall gives me and outputing 36V from that.
Unfortunately, a standard zener/transistor setup would require such
incredibly high wattage that building one would probably cost about as
much as buying a small country.

So instead I thought about building a switched supply based on a SCR.
The schematic is at http://www.nmia.com/~overhaul/regulator.gif Not
being an electronics engineer, I don't know if this circuit will even
work. But here's the idea of operation:

Start with capacitor discharged and no input voltage applied. Then
"turn on the power switch." The voltage at 'b' is 0V, and at 'a' is
about 3.1V right? This should trigger the SCR into conduction, charging
the capacitor. When the capacitor reaches 39V-Vgt, the SCR will not
turn on again, and after a slightly higher voltage will even turn itself
off. This should hold the voltage close to 36V across the capacitor.

So the obvious questions are: Will this circuit work? Is it practical?
Is there a better way to accomplish what I want to do?

Thanks
 
A

Activ8

Jan 1, 1970
0
Hi, I'm looking to build the aforementioned supply for a 6A 36V motor.
Since I can't find anyone with 10:3 transformers,
?!

I guess I'm stuck with
building a power supply that takes the ~60V half wave (after
rectification)

Where'd you get the idea that 120 Vrms rectifies into a 60 V half
wave?
that the wall gives me and outputing 36V from that.
Unfortunately, a standard zener/transistor setup would require such
incredibly high wattage that building one would probably cost about as
much as buying a small country.

So instead I thought about building a switched supply based on a SCR.
The schematic is at http://www.nmia.com/~overhaul/regulator.gif Not
being an electronics engineer, I don't know if this circuit will even
work. But here's the idea of operation:

Start with capacitor discharged and no input voltage applied. Then
"turn on the power switch." The voltage at 'b' is 0V, and at 'a' is
about 3.1V right?

That's a schottky?
This should trigger the SCR into conduction, charging
the capacitor. When the capacitor reaches 39V-Vgt, the SCR will not
turn on again,

You're thinking it will turn off. If it did, it would turn back on
on the next half cycle provided the cap discharges into a load.

Problem is, the AC on the SCR anode will keep rising to about 170
Vp. The SCR won't shut off until the current through it drops below
its holding current. The 2 diode branches are going to fight each
other.
and after a slightly higher voltage will even turn itself
off. This should hold the voltage close to 36V across the capacitor.

So the obvious questions are: Will this circuit work?
No.

Is it practical?

It's insane. Where'd it come from?
Is there a better way to accomplish what I want to do?

Yeah, find some SCR app notes and try a light dimmer. But check it
out here first. Otherwise I figure you'll fry.

Or buy a motor controller.
 
J

Jan Panteltje

Jan 1, 1970
0
Hi, I'm looking to build the aforementioned supply for a 6A 36V motor.
Since I can't find anyone with 10:3 transformers, I guess I'm stuck with
building a power supply that takes the ~60V half wave (after
rectification) that the wall gives me and outputing 36V from that.
Unfortunately, a standard zener/transistor setup would require such
incredibly high wattage that building one would probably cost about as
much as buying a small country.

So instead I thought about building a switched supply based on a SCR.
The schematic is at http://www.nmia.com/~overhaul/regulator.gif Not
being an electronics engineer, I don't know if this circuit will even
work. But here's the idea of operation:

Start with capacitor discharged and no input voltage applied. Then
"turn on the power switch." The voltage at 'b' is 0V, and at 'a' is
about 3.1V right? This should trigger the SCR into conduction, charging
the capacitor. When the capacitor reaches 39V-Vgt, the SCR will not
turn on again, and after a slightly higher voltage will even turn itself
off. This should hold the voltage close to 36V across the capacitor.

So the obvious questions are:
Will this circuit work? No
Is it practical? No
Is there a better way to accomplish what I want to do? Yes
Thanks
SRC normally only turs off if the current through it drops to a very low value.
The 'wall' gives you about 170V peak if rectified, but if single phase rectified
only half of the time.
JP
 
D

Doug Goncz

Jan 1, 1970
0
Dear Dan,
Hi, I'm looking to build the aforementioned supply for a 6A 36V motor.

216 W.

Jameco has #194837Cl, 12 VDC @ 0- 6.5A, 75 W, and ##123449CL, 24 VDC @ 0-6.5 A,
150 W, which in series provide 36 V, 6A, 225W, for $140 together.

Anybody here in sci.electronics.design, can you see if putting them in series
like this is as easy as adding a 12 V battery to a 24 V battery, and as
reliable?

They are described as "Switching DC Power Supplies" and the 0-6.5 A rating, I
think, is important, as some switchers require a load to remain stable sources.


On the theory side, 120 VAC is 120 VAC rms sine wave, and is 120 * 1.414 peak
to peak.
Unfortunately, a standard zener/transistor setup would require such
incredibly high wattage that building one would probably cost about as
much as buying a small country.

You're paying for efficiency when you buy a switching supply, and I don't know
of a square foot of dirt in Northern Virginia that can be had by itself for
$140.

Hm. Looks like it was drawn with a burnt paper match on a brown paper bag, then
faxed from a small country selling for less than $140.

You're always better off buying than building if you can take it off the bottom
line.

By the way Dan, if you get a chance, wander over to 1303 13th Loop and see
where I used to live, back when it was a base. Those were good times, and I
recently joined the Atomic Museum and got the Einstein equation coffee mug.



Yours,

Doug Goncz ( ftp://users.aol.com/DGoncz/ )

Read about my physics project at NVCC:
http://groups.google.com/groups?q=dgoncz&scoring=d plus
"bicycle", "fluorescent", "inverter", "flywheel", "ultracapacitor", etc.
in the search box
 
T

Tim Wescott

Jan 1, 1970
0
Dan said:
Hi, I'm looking to build the aforementioned supply for a 6A 36V motor.
Since I can't find anyone with 10:3 transformers, I guess I'm stuck with
building a power supply that takes the ~60V half wave (after
rectification) that the wall gives me and outputing 36V from that.
Unfortunately, a standard zener/transistor setup would require such
incredibly high wattage that building one would probably cost about as
much as buying a small country.

So instead I thought about building a switched supply based on a SCR.
The schematic is at http://www.nmia.com/~overhaul/regulator.gif Not
being an electronics engineer, I don't know if this circuit will even
work. But here's the idea of operation:

Start with capacitor discharged and no input voltage applied. Then
"turn on the power switch." The voltage at 'b' is 0V, and at 'a' is
about 3.1V right? This should trigger the SCR into conduction, charging
the capacitor. When the capacitor reaches 39V-Vgt, the SCR will not
turn on again, and after a slightly higher voltage will even turn itself
off. This should hold the voltage close to 36V across the capacitor.

So the obvious questions are: Will this circuit work? Is it practical?
Is there a better way to accomplish what I want to do?

Thanks

If you're going to do an off-line switcher you're probably better off
using "standard" switching technology -- but a 220W power supply is
pretty heavy metal to build yourself.

Why don't you try looking in the surplus catalogs (MPJA associates, All
Electronics, Herbach and Rademan & Surplus Sales of Nebraska are
probably all good for this).

You either want to get three 12V, 6A supplies with _isolated_ outputs,
or you want a 36V, 6A transformer (which you may find). If you can't
find one 36V transformer, you can use three 12V, 6A transformers with
the primaries in parallel and the secondaries in series. For driving a
motor you could probably just rectify the voltage and leave it at that,
but you could throw a cap in there to keep your diodes honest.

Frankly if you're not a circuits guy and don't want to be one I'd
recommend the three-isolated-supply approach. You'll learn a lot less,
but you'll burn a lot less, too.
 
A

Activ8

Jan 1, 1970
0
Dear Dan,


216 W.

Jameco has #194837Cl, 12 VDC @ 0- 6.5A, 75 W, and ##123449CL, 24 VDC @ 0-6.5 A,
150 W, which in series provide 36 V, 6A, 225W, for $140 together.

Anybody here in sci.electronics.design, can you see if putting them in series
like this is as easy as adding a 12 V battery to a 24 V battery, and as
reliable?

They are described as "Switching DC Power Supplies" and the 0-6.5 A rating, I
think, is important, as some switchers require a load to remain stable sources.

On the theory side, 120 VAC is 120 VAC rms sine wave, and is 120 * 1.414 peak
to peak.

Putting transformer secondaries in series might be easier.
 
A

Activ8

Jan 1, 1970
0
You either want to get three 12V, 6A supplies with _isolated_ outputs,
or you want a 36V, 6A transformer (which you may find). If you can't
find one 36V transformer, you can use three 12V, 6A transformers with
the primaries in parallel and the secondaries in series. For driving a
motor you could probably just rectify the voltage and leave it at that,
but you could throw a cap in there to keep your diodes honest.

But what's the startup current of the motor going to do to a 6A
xfmr? The thing could draw 4X that.
 
T

Tim Wescott

Jan 1, 1970
0
Doug said:
Dear Dan,

you wrote:

--snip--
Jameco has #194837Cl, 12 VDC @ 0- 6.5A, 75 W, and ##123449CL, 24 VDC @ 0-6.5 A,
150 W, which in series provide 36 V, 6A, 225W, for $140 together.
--more snip--

Look for the word "isolated". If the output is isolated from the case
and the ground conductor of the line cord then yes, you can safely
connect them in series. I'd be inclined to use three matching supplies,
but that's mostly paranoia.
 
T

Tim Wescott

Jan 1, 1970
0
Activ8 said:
But what's the startup current of the motor going to do to a 6A
xfmr? The thing could draw 4X that.

Oh, details, details.

The transformer will probably be fine, since it'll be wound with pretty
much the same design rules as the motor.

The rectifier may have problems...

It would probably be a good idea to limit the starting current to the
motor -- if you use switchers their behavior will depend on the model
that you buy.

Come to think of it, if all you want to do is run a small motor at
constant speed off of the AC mains, why aren't you just using an
induction motor?
 
A

Activ8

Jan 1, 1970
0
Oh, details, details.

The transformer will probably be fine, since it'll be wound with pretty
much the same design rules as the motor.

At high Duty Cycle?
The rectifier may have problems...

It would probably be a good idea to limit the starting current to the
motor -- if you use switchers their behavior will depend on the model
that you buy.

Come to think of it, if all you want to do is run a small motor at
constant speed off of the AC mains, why aren't you just using an
induction motor?

I'm not using any motor. This isn't my project, but that's one
question I thought the OP should anyswer - I thought I'd best stop
him from using that goofy circuit first. OTOH, there are quite a few
apps for 36 VDC motors, so I figure he's stuck with an old Az-El
motor or something.
 
T

Tim Wescott

Jan 1, 1970
0
Activ8 said:
At high Duty Cycle?

Then the motor will burn up along with the transformer. The motor and
transformer will both have similar thermal time constants, while the
rectifier will have one short enough to really fry on start up.

Limiting the current will be much kinder to the power supply and to the
commutator (which will also have a much shorter thermal time constant
than the motor windings).
I'm not using any motor. This isn't my project, but that's one
question I thought the OP should anyswer - I thought I'd best stop
him from using that goofy circuit first. OTOH, there are quite a few
apps for 36 VDC motors, so I figure he's stuck with an old Az-El
motor or something.

Sorry, that question was for the OP.
 
D

Dan

Jan 1, 1970
0
I guess I'm stuck with
Where'd you get the idea that 120 Vrms rectifies into a 60 V half
wave?

Average voltage measurement between the rectifier's positive output
(it's only a half wave rectifier) and the return wire with a DC
voltmeter. Makes sense--the voltage peaks between roughly +85v and
-85v, but the diode cuts out the lower half.
That's a schottky?

No, a zener. The intent is to make sure Vgt stays < the breakdown
voltage. The other diode prevents flow the other way either through the
39V zener or if the voltage across the load increases above 39V (which
unfortunately it will with this design).
You're thinking it will turn off. If it did, it would turn back on
on the next half cycle provided the cap discharges into a load.

Yeah that's the one fatal flaw I thought of after I posted this message.
I was thinking of a GTO. And unfortunately the local parts stores
don't carry any GTO's that handle more than 5A.
 
A

Activ8

Jan 1, 1970
0
Average voltage measurement between the rectifier's positive output
(it's only a half wave rectifier) and the return wire with a DC
voltmeter. Makes sense--the voltage peaks between roughly +85v and
-85v, but the diode cuts out the lower half.

The peak voltage of a half wave of AC is 120 * sqrt(2) about 170 Vp.
Again, restated. Where'd you get the idea that a 120 Vrms AC wave
has a Vp of 60 V or Vpp of 120 V? rms = 1.414 * Vp ...


170*1.414=120
No, a zener.

I can see the zener. look at the numbers. If a were at 3.1 volts,
3.1 - 2.7 = .4 which is *not* the forward drop of a normal
rectifier, but more like that of a schottky.

<snip>
 
K

KR Williams

Jan 1, 1970
0
The peak voltage of a half wave of AC is 120 * sqrt(2) about 170 Vp.
Again, restated. Where'd you get the idea that a 120 Vrms AC wave
has a Vp of 60 V or Vpp of 120 V? rms = 1.414 * Vp ...
Umm...

170*1.414=120

....want to try again? ;-)
 
A

Activ8

Jan 1, 1970
0
^^^^^^ 170/sqrt(2) = 120
...want to try again? ;-)

LOL! Oops! I think the most important number I can give the OP is

911 assuming he's in the US and his town is on the system.
 
D

Dan

Jan 1, 1970
0
Activ8 said:
The peak voltage of a half wave of AC is 120 * sqrt(2) about 170 Vp.
Again, restated. Where'd you get the idea that a 120 Vrms AC wave
has a Vp of 60 V or Vpp of 120 V? rms = 1.414 * Vp ...


170*1.414=120

Agreed (after fixing the math). But the AVERAGE DC is 170/pi, or
54.112... That's where the 60V estimate came from.

Anyway, I may have found a better solution: A used General Instruments
satellite actuator 36V power supply. If I can get it for $50 or less,
that will be cheaper than any power supply I can build.
 
T

Tam/WB2TT

Jan 1, 1970
0
KR Williams said:
...want to try again? ;-)

I thought he was half wave rectifying. The DC value would be
(.63/2)x120xSQRT(2)=53.5 VDC; assuming no capacitors. If I was doing this, I
would get a 28V 10 or 15 amp transformer, and run it hrough a 30 amp bridge
rectifier. The peak voltage at the rectifier output would be 39V. Now add an
electrolytic cap to that, to get about 5V ripple, around 10,000 ufd,

Tam
 
A

Activ8

Jan 1, 1970
0
I thought he was half wave rectifying. The DC value would be
(.63/2)x120xSQRT(2)=53.5 VDC; assuming no capacitors. If I was doing this, I
would get a 28V 10 or 15 amp transformer, and run it hrough a 30 amp bridge
rectifier. The peak voltage at the rectifier output would be 39V. Now add an
electrolytic cap to that, to get about 5V ripple, around 10,000 ufd,

Tam

Look at the origianal post:

"I guess I'm stuck with
building a power supply that takes the ~60V half wave (after
rectification)"

nothing about average or rms there. I though he figured that 120
VAC, half wave rectified, would give him 60 Vp.

Then in his next post he said,

"Makes sense--the voltage peaks between roughly +85v and
-85v, but the diode cuts out the lower half."

That indicated to me that he didn't know wtf because

60 Vrms * 1.414 = 85 Vp

which would mean he thought that Vpp = 2 * Vp = 170 Vpp as if 120
VAC swung +/- 85 V
 
K

KR Williams

Jan 1, 1970
0
ONE ^^^^^^^^^^^^^^^^
TWO ^^^^^^^^^^^^^
^^^^^^ 170/sqrt(2) = 120

LOL! Oops! I think the most important number I can give the OP is

Well, if you hadn't made the same mistake twice in the same post,
I wouldn't have corrected you. Mistooks happen, but when you're
correcting others...
911 assuming he's in the US and his town is on the system.

There is a huge misunderstanding here (you're not wrong, just
mistooken;). ...see above. ;-)
 
K

KR Williams

Jan 1, 1970
0
Agreed (after fixing the math). But the AVERAGE DC is 170/pi, or
54.112... That's where the 60V estimate came from.

Huh?! WHere the hell did 170/pi come from? The "AVERAGE DC" is
*ZERO* (as is the average). The Peak is 170. The peak-to-peak
is 340. The RMS is 120. There ain't no '60' anywhere in there.
Anyway, I may have found a better solution: A used General Instruments
satellite actuator 36V power supply. If I can get it for $50 or less,
that will be cheaper than any power supply I can build.

I fear for your family's life.
 
Top