Power supply

[TinMan]

The one thing im missing (amoung many) is a decent power supply.

As you would have seen in my video's,i have a very large 24 volt transformer.
I also have many cpu power supplies.

So dose anyone have a simple circuit for a variable power supply?

Thank's

[Chet K]

Go shopping for a variac, we'll figure a way to get it.
you build for all of our benefits ,one hand washes the other around here.

Thats "how it works"!!
thx
Chet
PS
Maybe when you post your next Graph you can place a toll at the Gate.
I'll collect

[ion]

The one thing im missing (amoung many) is a decent power supply.

As you would have seen in my video's,i have a very large 24 volt transformer.
I also have many cpu power supplies.

So dose anyone have a simple circuit for a variable power supply?

Thank's

Maybe a few more specs are needed, then we could have a go at it.

Output voltage and current requirement?

Regulated or unregulated?

What is the rating of your 24 Volt transformer (secondary amps). Is the input 220 or 110?

The simplest unregulated would be a Variac to your transformer to bridge rectifier to smoothing cap.

What is the intended end use?

[TinMan]

Maybe a few more specs are needed, then we could have a go at it.

Output voltage and current requirement?

Regulated or unregulated?

What is the rating of your 24 Volt transformer (secondary amps). Is the input 220 or 110?

The simplest unregulated would be a Variac to your transformer to bridge rectifier to smoothing cap.

What is the intended end use?

Hi ION

It needs to be just a simple bench top ower supply-DC.
1.5 to 24 volt's.
My transformer will do 20 amps at 24 volts each leg(AC)-it has 3 output leg's.
One i will be useing for a 24 volt AC power supply output.

I am drawing up a schematic at the moment,and would like you to ook it over when i have it posted.

Cheers

Brad

[TinMan]

Ok-i found this one by irondmax.
Sorry for the 1st grad school picture,but thats how it came lol.
But will this work?-something just dosnt seem right in reguards to the positive rail output going back through a diode and back into the base of the transistor?

[Allcanadian]

@Tinman

It needs to be just a simple bench top ower supply-DC.
1.5 to 24 volt's.
My transformer will do 20 amps at 24 volts each leg(AC)-it has 3 output leg's.
One i will be useing for a 24 volt AC power supply output.

You know in 30 years of working with electronics I never did get around to buying a good variable power supply,lol. For me it just didn't make sense because I generally always needed more than one and much of the left field stuff I was doing would have blown it up anyways. Whenever I wanted a DC power supply I would simply grab a 555 timer and an inductor and throw one together.
If you have built a Joule Thief then simply replace the tickler coil to the transistor base or mosfet gate with the output from the 555 and place a capacitor in series after the mosfet with a diode. You will need a blocking diode across the input leg, in parallel with it. When the pulse width is varied on the 555 the output voltage changes in proportion to it. If you want it regulated then use a voltage divider across the output capacitor which feeds back to the biased gate of a mosfet between the battery (+) and power pins 4 and 8. When the voltage on the capacitor rises this raises the voltage on the gate and at a perdetermined threshold turns off the power to the 555. What you want is called a buck converter which is a little different that the Joule thief.

Now I just grab a $14 Arduino microcontroller and use the onboard PWM to drive the gate of a mosfet/inductor then attach a voltage divider across the output capacitor and feed it back to an analog input pin. It takes about 5 minutes to rig the power supply anyway I want and costs under $20 a pop. I cannot stress how much time and money I have saved since getting back into microcontrollers and it's no big deal to throw together a power supply over 1Kw anymore. As well it's easy to throw in a couple 50 cent optoisolators and have voltage protection up to 7500v just because shit happens.
I can post a simple schematic and some code if you want as there is really nothing to it.


ac

[Allcanadian]

@tinman

Ok-i found this one by irondmax.
Sorry for the 1st grad school picture,but thats how it came lol.
But will this work?-something just dosnt seem right in reguards to the positive rail output going back through a diode and back into the base of the transistor?

To be honest I wouldn't touch that thing with a ten foot pole, the LM317 is a dog and overheats at the drop of a hat and nobody in their right mind would ever drive a mosfet with DC. Mosfets are switches and when they are used as a DC regulator instead of a switch they become very inefficient and get very hot very fast even at low power levels. This is why they use multiple mosfets which will require massive heat sinks because probably half the input energy will be dissipated as heat. Mosfets like to by driven hard and fast and a $1 mosfet driver is usually worth it's weight in gold.

AC

[Groundloop]

Ok-i found this one by irondmax.
Sorry for the 1st grad school picture,but thats how it came lol.
But will this work?-something just dosnt seem right in reguards to the positive rail output going back through a diode and back into the base of the transistor?

Tinman,

I did redraw your posted drawing in Eagle CAD to make it more readable.
The LM317 will need a medium heat sink and also the 6 PNP transistors will need a large heat sink.

GL.

[ion]

Nice job on the redraw. The Lm317 is used as a master with the 3055's slaving the high current to the load so it should be fine.

This is what's called a series pass regulated power supply. Properly built the output should be very low noise when compared to a switcher of the same power level and complexity. Since you are mostly doing brute force motor testing is this what you really need?

You never answered the original questions I posed so I will repeat:

Output voltage and current requirement? most important

What is the intended end use?

Regulated or unregulated? If regulated, percent regulation required.

and a few more:

Output noise requirement? Current Limited? Current source adjust?

There are many ways to design this once your specifications are firmed up.

For the record, as a serious researcher, I've got about 20 or more power supplies of all types including some very high voltage regulated types around the labs. When an idea hits, I need to test it, not spend time building a custom supply. As a past switchmode and linear power supply designer, I finalize needs and build a custom unit only when final specs are determined.

I'm not saying everyone should should take this approach as it can be expensive, but I get most of my quality equipment from surplus asset sales for pennies on the dollar. Or I'll buy a non working unit for cheap from ebay and repair it.

A lot of factors go into determining what type of power supply fits your needs. A general purpose bench supply of decent regulation can be built or purchased, the choice is yours depending on your time / budget.

This is Lab #2, in my garage. If you look closely you will see at least nine different power supplies of various types / voltages / current ratings.

[TinMan]

@Tinman
You know in 30 years of working with electronics I never did get around to buying a good variable power supply,lol. For me it just didn't make sense because I generally always needed more than one and much of the left field stuff I was doing would have blown it up anyways. Whenever I wanted a DC power supply I would simply grab a 555 timer and an inductor and throw one together.
If you have built a Joule Thief then simply replace the tickler coil to the transistor base or mosfet gate with the output from the 555 and place a capacitor in series after the mosfet with a diode. You will need a blocking diode across the input leg, in parallel with it. When the pulse width is varied on the 555 the output voltage changes in proportion to it. If you want it regulated then use a voltage divider across the output capacitor which feeds back to the biased gate of a mosfet between the battery (+) and power pins 4 and 8. When the voltage on the capacitor rises this raises the voltage on the gate and at a perdetermined threshold turns off the power to the 555. What you want is called a buck converter which is a little different that the Joule thief.

Now I just grab a $14 Arduino microcontroller and use the onboard PWM to drive the gate of a mosfet/inductor then attach a voltage divider across the output capacitor and feed it back to an analog input pin. It takes about 5 minutes to rig the power supply anyway I want and costs under $20 a pop. I cannot stress how much time and money I have saved since getting back into microcontrollers and it's no big deal to throw together a power supply over 1Kw anymore. As well it's easy to throw in a couple 50 cent optoisolators and have voltage protection up to 7500v just because shit happens.
I can post a simple schematic and some code if you want as there is really nothing to it.


ac

Hi AC
I most certainly would like that scematic,it would help me out a lot.
Also the code aswell ofcourse.

Cheers

Brad

[TinMan]

You never answered the original questions I posed so I will repeat:

Output voltage and current requirement? most important

What is the intended end use?

Regulated or unregulated? If regulated, percent regulation required.

and a few more:

Output noise requirement? Current Limited? Current source adjust?

There are many ways to design this once your specifications are firmed up.

Sorry ION

Just a plain old DC variable power supply
1.5v to 24 v out-smooth DC
No current limiting needed-although that would be nice if not to complex-time is short.
15 amps would be good enough for the time being.
Its just for general purpose use-nothing speacial.

Cheers

[TinMan]

Tinman,

I did redraw your posted drawing in Eagle CAD to make it more readable.
The LM317 will need a medium heat sink and also the 6 PNP transistors will need a large heat sink.

GL.

Thanks for the redraw GL
I was going to do it on autocad,but thought i would just post it as i downloaded it for some viewing of the circuit-and get your guys opinion

Brad

[TinMan]

In the schematic i posted above-that GL redrew,would it pay to have say a 2 ohm resistor on each base-so as all transistors get the same amount of current to the base?
Or will the .1 ohm resistors on the emmiter take care of that being that there pnp transistor's?

[Slider2732]

The PWM 555 isn't a bad idea, might build one of them myself.
The huge heatsink circuit looks to be very handy for the bench in winter !
As for Arduino, that whole thing is intriguing for short code running of many ideas for projects. Might I ask, when a 29F or whatever chip is burned, can it be simply socketed in a stand alone circuit ? (thinking EPROM). Have just never known.

Oh and ION, do you have a frequency analyser in that glorious man cave of equipment ? Asking because of Dr. Stiffler's inference of 13MHz anomalous energies. Just wondered if that is true about that, or if it's just a 1/4 wave type FM radio leak that people have talked about in that range.

Have you considered an old ATX PC power supply Brad ?
3.3V, 5V, 12V, -12V for a start.
Then you could use settable regulators for smaller projects to get the required voltage (up to around 2A). For larger amperage don't do what I did though and use several large diodes in series to set the voltage
Just thinking you probably already have one around.

[TinMan]

Have you considered an old ATX PC power supply Brad ?
3.3V, 5V, 12V, -12V for a start.
Then you could use settable regulators for smaller projects to get the required voltage (up to around 2A). For larger amperage don't do what I did though and use several large diodes in series to set the voltage
Just thinking you probably already have one around.

Yes i have many ATX power supplies,but would prefer a variable power supply.
I seen some where a guy who turned one of them into a variable power supply,and it was quite simple-but cant find it anywhere now.

[Allcanadian]

@Slider

As for Arduino, that whole thing is intriguing for short code running of many ideas for projects. Might I ask, when a 29F or whatever chip is burned, can it be simply socketed in a stand alone circuit ? (thinking EPROM). Have just never known.

I use the Arduino Uno and simply plug in an optocoupler or mosfet driver into the digital output(s) or the PWM and go. I like it because I can sequence multiple outputs for phase differentials or build complex waveforms to any spec I want simply by writing a few extra lines of code. It's so easy I could probably have this thing rigged to drive a boost converter or power supply before my damn soldering iron even warmed up.

Personally I have found that I spend way more time testing that building since making the switch to micro-controllers and I really do not miss making scratch built circuits at all. All that soldering tiny components onto proto boards was just a time consuming pain in the ass and I don't think anyone could pay me enough to go back. Basically we can take most all of the logic functions in the IC's were soldering onto our circuits and replace them with code, now that is cool.

I have also started using the I2C LCD's at $16 a pop which are brilliant, hook up 4 wires in about 30 seconds and we have a functional LCD display for whatever data we choose.
You can find the Uno on Ebay for $14 with free shipping .

AC

[TinMan]

@Slider

I use the Arduino Uno and simply plug in an optocoupler or mosfet driver into the digital output(s) or the PWM and go. I like it because I can sequence multiple outputs for phase differentials or build complex waveforms to any spec I want simply by writing a few extra lines of code. It's so easy I could probably have this thing rigged to drive a boost converter or power supply before my damn soldering iron even warmed up.




AC

Do you have a circuit of the mosfet driver?
Problem with PWM's hooked to a smoothing cap is,when a load is applied-the volotage drop's(not stable)
What i need is a supply voltage that will not drop with an applied load-voltage regulated.

Looking forward to the schematic.

Cheers AC

[mbrownn]

Do you have a circuit of the mosfet driver?
Problem with PWM's hooked to a smoothing cap is,when a load is applied-the volotage drop's(not stable)
What i need is a supply voltage that will not drop with an applied load-voltage regulated.

Looking forward to the schematic.

Cheers AC

If you charge an inductor with the power you need for the pulse, it will impedance match the load and supply the voltage. An unconventional approach I agree but it works, I don't know how semiconductors will hold up with it.

This is what I think the primary function of the trifilar coil on the lockridge is for, the secondary being core saturation.

[TinMan]

If you charge an inductor with the power you need for the pulse, it will impedance match the load and supply the voltage. An unconventional approach I agree but it works, I don't know how semiconductors will hold up with it.

This is what I think the primary function of the trifilar coil on the lockridge is for, the secondary being core saturation.

hi Mbrownn
thanks for the info.
I am looking forward to AC's info and setup on the arduino,as it would be good to have such a great range of control like that

[TinMan]

Well i tried 3 times with the circuit i posted,but this power supply circuit is shocking.The voltage is very unstable,and when a load is place on it-the voltage rises along with the current(ofcourse) by about 3 volt's-depending on load.The transistors that should handle 15 amps each,get hot with only a 2 amp load at 12 volt's-and that is all six of them on a very large heat sink. While running a 24 volt dc motor at 12 volts,when a load was applied to the motor the volts would climb.The more load you put on the motor,the more the volts would climb.Then when i hit 3 amps,one transistor blew out.

Second circuit(all new parts)-exactly the same as the first attempt.Then while hooked to a HHO cell,we hit 3.2 amps-and bang,another transistor gone and all the rest very hot

3rd attempt- I printed out the schematic on A4,and built the circuit right on top of it with all new components (just incase i was making some stupid mistake),but once again the results were exactly the same.This time i even added 2 ohm resistors to the bass of each transistor to make sure they were sharing the current evenly.Only thing different this time was it blew out the lm317 aswell as another transistor-all transistors hot at only a 3 amp draw @ 12 volts.The input caps(4x 10000uf 63 volt's high current) voltage remained very stable at 37.4 volts.

This schematic has now been thrown in the bin.

ION
Do you have a schematic for a simple variable power supply?
Voltage range-1.5v to 38v(or there abouts).Smooth DC-without current limiting.
AMP's-close to 20 amps would be good-give or take a few.

Something like i posted-but one that actualy work's.
Cheers

[ion]

From Tinman

ION
Do you have a schematic for a simple variable power supply?
Voltage range-1.5v to 38v(or there abouts).Smooth DC-without current limiting.
AMP's-close to 20 amps would be good-give or take a few.
Something like i posted-but one that actualy work's.
Cheers

Hi tinman;

I have many circuits developed and tested that are very robust for your use. Regulated supplies like the one you have built are fine for certain applications, but do not stand up under severe duty.

The circuits I can supply to you are variations of welder control whereby the primary of your transformer is controlled by a thyristor, in this case a triac. The beauty of this is that you can drive just about any transformer and thus use it for experiments where you need high current at low voltage or vice versa.

Driving inductive loads such as a transformer is difficult but not impossible. This yields extremely simple circuitry that is hard to beat. There is very little heat dissipated in the control element (triac).

I have developed a portfolio of these  circuits, built and tested. Some are proprietary and I will not post them on the web. Please contact me by mail or we can Skype and talk about it and I will send some to you.

Here, as an example is a welder control schematic, but I do not recommend it as it is unduly complex. I have much better variations that are simpler and some with regulated smooth DC output. Also attached is the app note from which it came.

If this method appeals to you, there are many variations that can suit your exact needs. Developing power supplies in the kilowatt range is quite easy using this technique and they are very robust.

[TinMan]

From Tinman
Hi tinman;

I have many circuits developed and tested that are very robust for your use. Regulated supplies like the one you have built are fine for certain applications, but do not stand up under severe duty.

The circuits I can supply to you are variations of welder control whereby the primary of your transformer is controlled by a thyristor, in this case a triac. The beauty of this is that you can drive just about any transformer and thus use it for experiments where you need high current at low voltage or vice versa.

Driving inductive loads such as a transformer is difficult but not impossible. This yields extremely simple circuitry that is hard to beat. There is very little heat dissipated in the control element (triac).

I have developed a portfolio of these  circuits, built and tested. Some are proprietary and I will not post them on the web. Please contact me by mail or we can Skype and talk about it and I will send some to you.

Here, as an example is a welder control schematic, but I do not recommend it as it is unduly complex. I have much better variations that are simpler and some with regulated smooth DC output. Also attached is the app note from which it came.

If this method appeals to you, there are many variations that can suit your exact needs. Developing power supplies in the kilowatt range is quite easy using this technique and they are very robust.

Hi ION
I have added you to my skype,as per Chets information given.

Cheers