high power supply

R!f@@

Joined Apr 2, 2009
9,682
is it linear or sm ? ive found a couple of linear circuits on the web, but absolutely nothing on cc + cv pwm. except for driving LEDS. there also seems to be confusion with regards to CC and current limit on any psu designs ive found, calling CL cc etc. :(

i have now rewound the seconday to 24v 3kva.
so now i have 24vac and looking at the possibility of scr switching as ive done a bit of design on this before.
thanks for the help so far
It's a linear design...with CV and CC.
Output current can be any value you want.
Quite feasible u know.
 

R!f@@

Joined Apr 2, 2009
9,682
Don't go around and make the exact one.
For ur needs we need to make changes.

I have this build on vero boards fir testing and it ended up in my bench :D.
But now I have taken it apart to make a chassis of reasonable size and beauty.

I am going to make some PCB's for my design. You are welcome if you like to make one for u too. We have one member who can do this and ship the PCB if you need.

What say you that we discuss the changes.

Side note.
This circuit shows the CV part only. The CC circuit is completely separate. Once done, only 4 connections are required to complete the circuit. The CC circuit simply monitors the current drawn and activates the CV current limit integral transistor accordingly.
 

Attachments

Thread Starter

spankey666

Joined Nov 30, 2011
91
excellent, thanks :)

where do we start ?
i can quite easily re-wind the secondary again to 35v so that we are on the same playing field,
first off, i would think the soft start and r13 be removed.
next would be to up the o/p stage. a couple of questions on this, ive seen very similar circuits . on your output mj15024's have a current handling of 16a each so why 3 for just 10A? is it to make heat dispersal easier ? or is there something i'm missing ? and what is the purpose of R2,3,4 ?
R20 i would replace with a 75mv shunt i have ordered so that i have correct calibration for an digital ammeter
is this enough to start a discussion ? LOL
 

R!f@@

Joined Apr 2, 2009
9,682
i can quite easily re-wind the secondary again to 35v so that we are on the same playing field,
If you don't want higher Voltage then go ahead.

first off, i would think the soft start and r13 be removed.
I was thinking of doing that too. :D

next would be to up the o/p stage. a couple of questions on this, ive seen very similar circuits . on your output mj15024's have a current handling of 16a each so why 3 for just 10A? is it to make heat dispersal easier ? or is there something i'm missing ?
Here the thing is I went with what I have in my junk. I had a heat sink with three of 'em in it. :D

Best part is the heat sink is quite cool at 10 amps. So I do not need active cooling. And the transistors are well below SAO, I think. hehe, but I did not check that. I might change the transistor to make the sink smaller and include active cooling so I can use a smaller chassis.

Here you have to find good transistors, cause you are going for 30 Amps. In this area you need to be very careful in dividing the current across parallel devices so that the transistors remain quite happy when you loading it to 30 Amps. :)

You need to find the transistor and proper heat sink before doing anything else. Calculations are needed here.

and what is the purpose of R2,3,4 ?
They are emitter resistors, and is a MUST
They are necessary for proper current sharing amongst the transistors. Remove it and one will definitely blow before u know WTF happened. :D

R20 i would replace with a 75mv shunt i have ordered so that i have correct calibration for an digital ammeter
Yup...that would definitely work.. :D

Now I think I have to go and start the CV area since u are at it...bummer...damn busy these days.
 
Last edited:

pilko

Joined Dec 8, 2008
213
Why do you need to rewind the transformer?---check post 19
"As I mentioned in post 13, the max rated secondary current is 30 Amps.
If you apply 240V to the first primary you will get 60V at the second secondary.
If you apply 120V to the first primary you will get 30V at the second secondary.
If you apply 110V to the first primary you will get 27.5V at the second secondary."

After rectification your voltage will be about 38V.

pilko
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
i think both of have forgotten , i have already re-wound and have 24v at 3kva !!
so if i rewind again to 35v (as an example) i will have 90 odd amps to control :eek::D
 

joejoenikk

Joined Nov 23, 2011
9
A transformer is a power supply. If your trying to build a well regulated one the Ham radio guys seem to be the place to look.

Just on the chance that that is your area, the 12- 15 volt RF transistors are much more cost effective than the 24-28 volt ones.

My current work is on using a PNP Driven Darlington. That method of regulation gets the cost down really quick.

I don't know what your calling "high power", but knowing the price and quality of this series of regulators, I'd say that you wouldn't be wasting a lot of time in learning how to use it, if the link or brand doesnt agree, you can find Motorola's equivalent (2n5883). It is also, and intended to be an AUDIO AMPLIFIER Transistor!!!

http://www.onsemi.com/pub_link/Collateral/2N5883-D.PDF

One thing you will need to do in your build, is get the cost of the regulators down, real fast. Don't be suprised if you find out that it is often, more practical to build a battery charger and buy a big lead cell, in fact it has become difficult to buy a decent used ham radio unless it is a 12 volt model.
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
i electroplate images onto aluminium, but i want to have a crack at titanium colouring (hence the high voltage) and chrome . realistically i only want 70amps in the 8v-24v range, but a higher ceiling would be good (just in case ) so the plan is to make it as powerful as possible , even if i only use half of its potential most of the time.
stuff like this :)
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
i realistically i only want 70amps in the 8v-24v range, but a higher ceiling would be good (just in case )
as i said previously adjustable, voltage 8-24v cv, with cc control as well for as many amps as i can get out of my transformer. worst case just cc comtrol as aluminium is done by current, the volts take care of themselves. but if i look at other plating practices, some are voltage driven

the link i posted for the hydrogen producer is in production , so obviously works and in principle is the same as what i am doing. my only concerns are with supplying the circuit with a transformer as to how it would behave. i'm sure if i put enough caps on the rectifier there wont be an issue, but this is outside my knowledge area :(
 
Last edited:

R!f@@

Joined Apr 2, 2009
9,682
If you can get a 100 amp at say 24VAC secondary winding...and with a bunch load of Power transistors like the 250W types ( TO-3 ), it won't be a problem for me.

Plus you might need some power diodes in the range of 200 Amps to make the bridge and ummmm.. say 10,000μF 50V electrolytics, like ummm 10 of em to get the ripple down ..yup that should do it.


Ahh! yes...power diodes tend to drop higher Vf so better make the secondary to say like 30VAC. So you can regulate without loss to 24VDC.

For this u definitely need surge current limiter for safety reasons.
And yes for 90 amps u will need around 10 of those 16Amp 250W transistors to keep them in SAO. I think you should go for 15 transistors with 0.33Ω 10W emitter resistors.

How big is your wallet ? :D
 

Kermit2

Joined Feb 5, 2010
4,163
a LARGE set of batteries with a series resistor composed to two copper plates in a fish tank full of water. Resistance is set by increasing or decreasing the distance between electrodes. Pure water has high resistance and water with an ionic solid dissolved(salt etc.) has a much lower resistance. Experimentation would determine your needs quickly.

Very crude but useable if you add a volt meter and amp meter to it.
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
If you can get a 100 amp at say 24VAC secondary winding...and with a bunch load of Power transistors like the 250W types ( TO-3 ), it won't be a problem for me.

Plus you might need some power diodes in the range of 200 Amps to make the bridge and ummmm.. say 10,000μF 50V electrolytics, like ummm 10 of em to get the ripple down ..yup that should do it.


Ahh! yes...power diodes tend to drop higher Vf so better make the secondary to say like 30VAC. So you can regulate without loss to 24VDC.

For this u definitely need surge current limiter for safety reasons.
And yes for 90 amps u will need around 10 of those 16Amp 250W transistors to keep them in SAO. I think you should go for 15 transistors with 0.33Ω 10W emitter resistors.

How big is your wallet ? :D
i already have a 180A bridge,
power transistors maybe 4 of these ? http://cpc.farnell.com/stmicroelectronics/bux98a/transistor-npn-to-3/dp/SC06777?in_merch=Products From This Range would these be a suitable replacement in your circuit ??
and one of these for smoothing ? http://cpc.farnell.com/vishay-bc-components/mal210128104e3/capacitor-63v-100000uf/dp/CA05330
how accurate does the emmiter resistor be ? cant find .3 ohm but .47 ohm 50w are cheap enough http://cpc.farnell.com/bi-technologies-tt-electronics/mhp500r470f/resistor-0-47r-1-50w-to-220/dp/RE04815
If so, it brings it all down to £100.00 or so . quite do-able :) i also have various heatsinks and 1 in particular lenght of extrusion i could even possibly water cool (rather not though if i can help it )
 
Last edited:

R!f@@

Joined Apr 2, 2009
9,682
Holy ****. !!

U have nailed it down eh!.
SO let's see. A bit of advice or more of a better way.

If you are going to rewind the Txx. then go for center tap. Like 24V-0-24V.
This would increase current handling capacity and better in a away I presume. Don't ask for math.
At 24-0-24 VAC, Full wave with two diode will give a max Raw DC of
24V X 2 X 0.707 = 33VDC at the capacitor.
Current will be the secondary AC current times 1.414, so by rewinding the current secondary gauge size wire doubled will handle around 84 amps..

As for the transistor quantity, I need to know your minimum output voltage u will ever use with the highest current output... Then I can find out how much current the tr can handle continuously and safely.

When using filter caps, it is advisable to use more low value caps in parallel than one larger one. Gives better filtering..

And for active cooling means just use a fan at the sink and u will do fine but you need to find the thermal conductivity of the sink The specs of a TO-3 style heatsink.

I never used that sorta resistors but if it can be mounted on sink then I guess it's pretty good. Any value between 0.2 to 0.4 will suffice. Power rating need not be that high. 10W is typical.
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
Hurrah !!! it feels like i might actually be getting somewhere :)
Ive done some more reading up on some of the processes i would like to have a crack at, and it would seem that i need to go down to 5v . so that means i would now require full current capabilities from 5v-24v + a little headroom

so to sum it up 5v-24v 0-70A (min)
Edit Added
just found this http://www.qsl.net/vu2upx/Projects/40apsu.htm has the current limit already sorted. so is this workable ? this is where my lack of knowledge fails me, by replacing the output 2n3055's with some thing bigger like posted above, what changes would need to be made ? it looks like its supposed to be fixed voltage, but has a V control pot ??and i'de loose the over volt and relay etc
 
Last edited:

R!f@@

Joined Apr 2, 2009
9,682
That Circuit is do-able but has a flaw in current limit accuracy. The current limit is determined by the TR3 and R21, due to component variation across all the outputs, Ie won't be exact That is why the emitter R's like R21 via R25 are used. Meaning the current thru all of them won't be exact has the voltage drop across them will not be equal.

At high currents this is a problem. What I have designed lately is more accurate in current sensing as the current is measured at the out put path only. Besides my version use more accurate methods for exact current limit. I can do without a shunt measuring device I am using nichrome wires to measure the current or as the shunt regulator. Or I can use the heat sink mountable 50W resistor u showed before. I think I will use that.

My design can go as low as 0V u know.

so for ur needs the worst case measurement for the output element is 5V at 70A.
Assuming the transformer has a secondary at 24-0-24 VAC the RawDc will be ≈ 34VDC.
The voltage drop across the transistor is 34-5 = 29V

So at 70Amps the power dissipate by a single BUX98A will be 29VDC X 70A = 1890 Watts. Which is hell lot more it can handle.

But this figure is wrong since the Ic can only be 30amps , so the Max Ic 30Ams but at 29V across it the power is 870W.

Which still is far more that it can handle.

Now we have divide the 70A and power of 1890W.

Assuming Maximum power of output which is 250W we have 1890/250 ≈ 7.56 which rounds up to 8 transistors.
Hence output current divided by 8 tr's ≈ 70Amps /8 ≈ 8.75 Amps per transistor.

So at a load of 5V @ 70 amps, (assuming raw DC of 34VDC) each transistor will handle the power of ≈ 29VDC X Ic (8.75A ) ≈ 253.75Watts. Which still is more than it can dissipate.

So we have to use more than 8 devices.
What I would use is 12 BUX98A's ...

so again, 70A / 12 = 5.83A of Ic
P = 29 X 5.8 = 171W..which is quite a lot of heat but with a fan and big fin heat sink the transistor will be happy.

Bit this is at 25°C junction temperature. I have to do a lot more math to figure out the cooling levels. but for now this figure will do.

But Since u can rewind the Transformer I know a way of cutting down the no. of transistors. Minimum is 3 ouputs BUX98A's

So tell me...are u willing to rewind the output to the way I want.
My way is a multi-tap secondary winding.

The the Taps can be switched into the circuit automatically when u advance the Voltage level. This ways the Raw DC will be always the minimum headroom required to regulate the output. This way no excess heat or energy is wasted and the heat sink can be way smaller than before
 
Top