dropping 3 volts off power supply

Sawmiller

hi all,
i need to drop about 3 volts off a 35 volt dc· supply
my cnc needs max 32 volts
i have some scavenged parts that i have piled together
120 v -> transformer -> 24vac -> full bridge -> 70 Mfd cap -> 35 volts..
is there a easy way to drop 3 volts ? i thought about 1n4001 diodes· in series, but that s a handful of diodes
btw this 35 volts is with a 32,000 ohm·resistor as load , i'm tring to build a cheap unregulated power supply
as they say on the cnczone that a unregulated is the way to go.( plus i have all these neat scavenged parts)
thanks for suggestions,
dan

kjennejohn

Hello Sawmiller.

A few ideas come to mind:

1. Use a Variac to drive the transformer down to an output voltage that gives you 32V at the output. Keep in mind that the mains don't stay steady at 110V AC, they go up and down, so your DC output WILL CHANGE in an unregulated supply. How far off from 32V can you go safely?

2. If·this set up·requires·a constant current, you can simply introduce a series resistor to drop the extra three volts. A large, well-cooled, probably expensive, resistor. Or a clutch of power resistors in various configurations to arrive at the desired resistance value. Remember, resistor wattage adds up, so the more resistors you add to this setup the smaller the wattage per resistor is needed.

3. Run this through a light socket (sure, a typical 110V lamp socket) and keep screwing in various bulbs 'til you get the required voltage drop. This might even work with a string of Christmas bulbs in a parallel cable, where removing bulbs doesn't kill the rest of the bulbs.

Etc. Since I don't know what you have in the way of scavenged parts, I can't comment any further.

'Luck.

kenjj

PS How can I do a simple new line without jumping an extra line every time for·an Enter/Return press?

Phil Pilgrim (PhiPi)

I'm not sure of a good way to drop that much voltage at the current you require. (A Variac on the primary side would work, but they're pretty expensive.) But I do know that you will need a much bigger capacitor to quiet the 120Hz ripple at full load. Depending on your current requirements, 35,000µF wouldn't be too much.

-Phil

bennettdan

Sawmiller,
Your powersupply is unregulated and you say you have a 110vac to 24vac transformer and when your wire it up you get 35 volts out of it. Is that wil no load you should test it with a load on it like a DC motor and then see what voltage you are getting. I bet it drops below the 32vdc you need with a load on it.

Sawmiller

thanks for the input guys, looking thru the scavenged parts today for a bigger cap
i can order some diodes and a big cap, just wanted the instant gratification thing.
i do know i dont have a variac, so that is out.
later all, take care
dan

Dennis Ferron

A bigger capacitor will only get you so far - the problem is not the size of your filter capacitor; the problem is that your transformer is nominally 24 volts. What that means is that if you draw a line through the AC wave form, so that some of the area of the humps is outside the line, and some is inside the line, that average is at 24. The peaks go outside the line at no load and build up to 35 volts, but as soon as you draw off any current it drops quickly.

How much current do you need? If you don't need very much current, then if I were you I would just use a higher voltage transformer and regulate it down with a 317T linear regulator IC. You'll get a guaranteed output voltage and your filter capacitor will not need to be so big because the linear regulator will take care of some of the ripple.

If you need larger amounts of current, you can make your own linear regulators out of 2N3055 transistors, which are monster power transistors that can dissipate 115 watts.

In any case you need a transformer with a higher nominal voltage than 24.

Phil Pilgrim (PhiPi)

I inferred from the fact it's a CNC app that the power was for driving motors, so the current requirements would be in the neighborhood of 12 amps (4 amps per axis * 3 axes). In that case a cap on the order of 35,000uF will be needed. Typically, you can calculate the output voltage from this kind of power supply to be 1.414 * RMS input voltage or, in this case, about 34V. Undervaluing the cap to reduce the average output voltage is not a good idea, since at no load (i.e. motors idling) the output will still jump to the peak voltage and possibly fry the motor driver.

At these currents, a linear regulator will get HOT, so I'd be reluctant to go that route. The beauty of modern motor drivers is that they don't need a voltage-regulated input to regulate the current in the motor windings.

If it were me, I'd be looking at the following options:

····1. Buy a different transformer.
····2. If they're accessible without disturbing the primary, remove some windings from the transformer's secondary.
····3. Buy a different motor controller that will tolerate a higher voltage.
····4. Use a Variac on the input. (A new transformer would be cheaper, though.)

When you have mismatched components like this, I'm always reluctant to take a bandage approach to mate them. Whenever I've done so, I've ended up paying for it later.

-Phil

Sawmiller

yup, just about perfect phil.
running 3 axis , max 2.5 amps ea, so i've put that transformer back and am looking around for a different one, or buying one.
not in a big hurry , since i have those daisy chained computer power supplys, and i know they work, just bulky .
i might end up going and winding my own secondary to get the right voltage, i have 4 of those transformers and several others , so theres a possiblity, cnc zone has a good transformer winding article.
thanks again
dan.
ps 24volts ac when run thru a full wave rectifier, yields 34-35 volts dc, due to the fact that the ac value is actually rms , the dc includes those peaks.
·so 32 volts dc ( desired) * .707 = 22.6 vac ....the fun of DIY, you learn so much

Post Edited (Sawmiller) : 6/26/2007 10:10:56 PM GMT

Dennis Ferron

When I was in high school I modified a computer power supply so that I could regulate the output voltage. By moving the regulator input from the 5 volts line to the 12 volts line, I gained direct control over the 12 volt output. Then I placed a potentiometer on the reference input to the regulator chip so that I could choose what fraction of the 12 volt line the reference input measured. It yielded a benchtop supply that can be controlled from 2.5 to 15 volts and can handle 10+ amps of current. I still use it today many years later.

It could be possible to mod a computer power supply to even produce as high as 32 volts on its 12 volt line, if you replace all of the output caps with higher voltage rated units and disable the "crowbar" over-voltage protection. In fact, when I first built the power supply, I mistakenly mis-wired it to produce 36 volts on the output, and blew up all my filter caps & had to replace them. But if you could get the supply to run steady at that voltage, it would provide plenty of amps.

Sawmiller

cool, another idea.
thanks dennis
just got done rebuilding my table larger, now i have a 20 x16 inch cutting area , still need to put on my limit switches, then i'm going to sit down and look at the rewiring transformer vs rewiring computer power supply.
and dust collection, cant forget that cough,cough