high power supply

R!f@@

Joined Apr 2, 2009
9,682
This my basic idea in multi tap secondary windings

U can rewind the secondary and tap 9 volt taps of 9. Meaning 3 9V taps but the coil should handle 100amps if you want 70Amps DC
 

Attachments

Thread Starter

spankey666

Joined Nov 30, 2011
91
i'm happy to go along with what ever you recommend. i need to get this sorted.
i can re-wind to what ever you want. i should have room on the core for double windings, 4awg guage . if double winding, then could i go down on size which might make it less of a squeeze ?
 

R!f@@

Joined Apr 2, 2009
9,682
So what exactly do you have right now, can u show me.

{ED}
Wind to this spec and let me know ( attached file )
 

Attachments

Last edited:

R!f@@

Joined Apr 2, 2009
9,682
The winding will switch into the circuit automatically when needed.

For this U will need to find Relays with a contact current rating of 100Amps.

coil voltage can be anything.

There are ups and downs u know... ask ur self this.

Do I want 100amp heavy duty contactor's with 3 BUX98A's in a small heatsink without much heat loss

or

Do I want 12 BUX98A's in insanely big heat sink with lotta heat emitting at lower voltages.

Which is cheaper for u?

by the way ...u will need 3 separate 100 amp relays
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
my concern is this, when anodising, the layer is insulating as it grows, thats why constant current is required to "punch" through the layer to keep a constant at the base metal. so i would expect to see roughly 0.1-0.3v increase per micron of layer over a 2 hour period (to get 25 microns) how would the relays react at the cross over point with such a small rise in voltage ? as the voltage doesnt rise smoothly, it fluctuates +/- 0.2v as it goes up. this is because of temperature rise on the surface, cooler cutting in etc.

and having a quick google, 100a (at 12v) relays are not cheap LOL so lets see you suggestion :)
 
Last edited:

R!f@@

Joined Apr 2, 2009
9,682
U do not understand what the relay does do you. Here is how it works.

When u switch on the raw DC will be at 12V at the caps. Headroom will be 12-7 ≈ 5V..
Here u have 0-5V Variation at any given amps. The maximum power Tr can handle will be 12X70 = 840W. but surely you would not draw 70 amps at 1V would you ?

Let's say u are drawing 70A @ 3VDC out which is 210W. With three transistors current will be shared and each will dissipate around 12-3 x 70/3 ≈ 207W. Which is with in limits. They will get hot but a good heatsink with a fan will cool them indefinitely. Get my point.?

So here now the out put is adjustable between 0 to 5V.
Next, if you want 10V, you will turn the pot. When you turn the put the detector circuit will switch off the 1st relay and switch on the second one. Thus the Raw DC will rise from 12V to 25V. increasing the headroom for the transistors. And it will give a max out of 25-7 = 18V.

And it will go on if u turn the pot.

This way at lower voltages you are giving the transistors just enough headroom to regulate and not dissipate unwanted heat.
 
Last edited:

R!f@@

Joined Apr 2, 2009
9,682


See the blue color contactors...they have a 50A contact rating and with 4 contacts in parallel u can switch 200A.

We can use these too u know...we have to change the circuit a little bit. :D

I believe these will be cheaper for u..
and they are more effective u know...
See how much it cost for 3 of them.........

See these
 
Last edited:

Thread Starter

spankey666

Joined Nov 30, 2011
91
i fully understand what relays and contacters are, what i am trying to say is , with plating the volts control is turned to max, and the current is set by the control knob. volts required to maintain the current is relative to surface area. so at the beginning of a run, the part may pull 8v and by the end may pull 19v. at very small increments and fluctuations at the crossover point wont the relays cut in and out with these small fluctuations ? contactors make a dreadful noise when in buzzer mode LOL
 

R!f@@

Joined Apr 2, 2009
9,682
Aah ! Damn..not good..
U need strict current control and the let the voltage fly..
Darn it....... f i knew plating then I wouldn't have recommended so many things.

U have one way.
Rewind the secondary to 24-0-24 VAC..
Wire gauge should be able to withstand around 70 Amps. This way you can load the transformer 24/7

get 12 of em Bux's and a suitable heatsink..so we can get this going already.
Get 10,000uf caps. Parallel around 7 of em minimum, figure is taken like a 1000uf / amp.

{ed}
I have attached a schema of ur Power input stage. It will be something like that. The bridge is what u have, we can use two diodes and leave the other two.
Surge limiter should be there or u might blow the bridge eventually when powering up. You'll never know. :eek:.
The Relay should handle 100Amps easily or it will melt u know.

By the way...you should include a 100A fuse too between the bridge and the caps.
 

Attachments

Last edited:

Thread Starter

spankey666

Joined Nov 30, 2011
91
i guessed it :(
ok, thinking laterally again. i cant get out of my head some sort of switching to create an "averaging" output. we know dimmers dont work well on inductive loads, but as i mentioned at the beginning of this thread, whats the possibility of the same sort of idea on the secondary pre rectifier. a standard dimmer works with a diac on the gate which switches in at around the cross zero point. what if the diac was replaced with an optoisolator diac that is supplied with a square wave oscillator ? this would cause the diac to switch on with frequency but only once the trigger point voltage has been crossed so giving an averaged voltage out put and keeping it in phase. pic below shows roughly what i mean.my software doesnt have diacs or triacs so it shows a thyristor and opto-transistor. R1 and C1 create a snubber for better switching. the diac Live is taken to before the circuit so that the voltage will always be above its switching limit.
i just need to be told this wont work for my head to be cleared and accept that a full linear circuit is the only way LOL
this is a link to the opto diac http://kubuntu.free.fr/wiki/data/fp/moc3041.pdf
 

R!f@@

Joined Apr 2, 2009
9,682
Full linear is not the only way, in fact the SCR method is possible but too complicated.

I don't think I have any diagrams of the sort.

Another method is to use switch mode. Problem here lies with ferrite core transformers. Do you have any knowledge on winding SMPS ferrite transformers ?

As far as I know, linear is the easiest way and far easy to fix too
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
nope. over the period of this tread i have been reading and reading (i have chronic insomnia :( )trying to understand various methods and ways around working with what ive got. the dimmer type thing works on arc welding machines on the primary. so in theory i could put the above on the primary which would keep components to a minimum and only have to deal with a max of 13amps or so. but i dont understand what the output current would be if the primary input sine wave is chopped up with a dimmer. if the idea is workable, a simple ne555 could supply the square wave with a fixed frquency and an op-amp current / voltage sensing circuit would change the duty cycle to regulate the primary to keep output constant.
 

R!f@@

Joined Apr 2, 2009
9,682
Yup that might work...

Arc welding does not require precise current control. They are just low voltage high current and dimmers as such works quite well with them

Is your work precise or anything goes..u decide and come back.
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
i'm only bouncing ideas around lol.
lets have a look at your relay idea, maybe we could use scr's instead of relays and some sort of voltage sense switching ??
 

R!f@@

Joined Apr 2, 2009
9,682
Nope...it's not that easy.

Triggering SCR is not my line of work. Meaning I am not familiar with them quadrant switching and all that.

Can't help you in tht area.

you cannot use relays if ur load fluctuates.
I was thinking once a voltage is set it will stay at that level for that job.

Only option for u is a SMPS (pretty powerful one at that) or Linear.

I don't do SMPS although I can very well repair them..

I can only help you with a linear design which I have done.
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
hello again,
after much reading etc, i'm throwing in the towel with scr type switching, too little info on the net (and i'm too old to soak up what there is ), and i really need to get something going :(
reading through this forum, i believe you have knowledge on the circuit below thats lifted from the lm317 data sheet. ive made a few adaptions, and would just like for you to take a look at it and advise me of any alterations that need to be made before i order the few bits i dont already have ie transistors. the ones chosen are 300w . i have decided that in order to get something done, i will limit this project for the moment to 40Amps.
maybe if its deemed possible, at a later date build another identical unit (double wind the transformer )and paralel the outputs so doubling the capabilities ???.
thank you ever so much for all of your help, this has been one hell of a learning curve :)

 

R!f@@

Joined Apr 2, 2009
9,682
There is a catch in that one.

The adding of pass transistors varies the base drive, so U have to change the R1 value accordingly and the shunt too.

U have to make it and do some or measure what u need.
Inshort experiment.

I wonder why u are throwing away something that is already been done.
 

Thread Starter

spankey666

Joined Nov 30, 2011
91
oppps, should have been 75mv :mad:

i am still quite happy to go with your circuit Rif@@, but i was under the impression it was un finished ,
why the change of heart ? well i re-wound the transformer to your spec , stuck it under load (in my acid tank) to test voltages etc and for some reason that i cannot figure out the transformer started pushing out 90v across each set of windings fully loaded .i have spoken to someone in the know who seems to think the core is saturating and giving the strange results. previously it only gave 30 ish v. and going down on wire size has fixed this but ive lost the higher current as a trade off.
 
Top