Author Topic: 400V 4-8A DC supply  (Read 4620 times)

0 Members and 1 Guest are viewing this topic.

Offline Lord NicollTopic starter

  • Contributor
  • Posts: 17
  • Country: ie
400V 4-8A DC supply
« on: July 17, 2017, 12:35:35 am »
Hello guys, following on from a previous post about switching large capacitor banks, I have come to the part where I need a power supply that can charge the 400V capacitors. Initially I tried to charge them directly from the mains, but according to calculations, substantiated by the fact it tripped a 33 Amp breaker, a 55 Amp one and two 33 Amp ones when I tried in paralleled (yeah I know that's a very bad practice but I tried it anyway) I went and did the maths and realised it needed like 170 Amps ( |O ) in a single wave to charge them to the 322V DC that normal rectified mains provides, golly that was a pain. My solution was probably the worse way to do it, a large 1600 watt 36 ohm heating element used as a current limiting resistor. It worked and didn't even get that hot, but still used a lot of extra power that I would like to avoid wasting.

Now the question is: Do I used AC-AC and then RMS to get the 400v, or 240v and then DC-DC, or is there an even better way? I'm temped to get a big ass toroidal transformer and just put custom winding on it to make roughly the 290v (I thing my maths are off to get 400V rectified but it looks close) AC needed to be rectified to 400V DC, but idk if that's a good way, would it be better to use a step up switch mode power supply, There would be some what of a technical challenge to make it 400v 6 Amps (middle of the road for the current I want) and looses might be large, but probably on average with a transformer based solution.
 

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2222
  • Country: mx
Re: 400V 4-8A DC supply
« Reply #1 on: July 17, 2017, 01:14:45 am »
You don't specify how many microfarads your large capacitor bank is, to calculate the total energy you will be storing.

Anyways....you MUST limit the charging current.

A simple and time-proven concept is to use a power resistor in series, and then after a period of time, to bypass it with a power relay or thyristor.
Or even better, a two stage circuit: two power resistors in series, bypass the first after a period of time, then bypass the second after a few more moments.
 

Offline Lord NicollTopic starter

  • Contributor
  • Posts: 17
  • Country: ie
Re: 400V 4-8A DC supply
« Reply #2 on: July 17, 2017, 01:33:12 am »
You don't specify how many microfarads your large capacitor bank is, to calculate the total energy you will be storing.

Anyways....you MUST limit the charging current.

A simple and time-proven concept is to use a power resistor in series, and then after a period of time, to bypass it with a power relay or thyristor.
Or even better, a two stage circuit: two power resistors in series, bypass the first after a period of time, then bypass the second after a few more moments.

I have already started work on the triggering method, that has been designed and I'm just waiting in the parts. Really getting the 400V DC is the question of this post, the capacitor bank is 56,400 µF 400V, so not a trivial amount of power (just under 5000 kilojoules)
Current limiting will be done by the power supply on the mains main side, since AC is somewhat easier to limit before being rectified.
 

Online NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9014
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: 400V 4-8A DC supply
« Reply #3 on: July 17, 2017, 02:26:11 am »
Use a hair dryer as a current limiting resistor. Do realize that such a capacitor bank is at least as dangerous as the mains even when unplugged.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline station240

  • Supporter
  • ****
  • Posts: 967
  • Country: au
Re: 400V 4-8A DC supply
« Reply #4 on: July 17, 2017, 03:01:07 am »
8A x 400V = 3200W ouch.
Probably easier to get 4A or so.

There are some high power computer PSUs out there, I've got one that's 2450W and has PFC to charge the caps to 400V.
Without Power Factor Correction (PFC) it's going to draw a lot more current from the mains, as you found out.

I dunno if 3 phase is an option where you are, easier to get 400V DC from that due to higher average voltage.
 

Offline Lord NicollTopic starter

  • Contributor
  • Posts: 17
  • Country: ie
Re: 400V 4-8A DC supply
« Reply #5 on: July 17, 2017, 04:13:59 am »
Use a hair dryer as a current limiting resistor. Do realize that such a capacitor bank is at least as dangerous as the mains even when unplugged.
Hmm, people seem to be missing some points, I did use a large resistor, however mains voltage is only 330v DC when rectified, I want the full 400V, hence asking what you guys thought, I know how to do it, I just wanted a second opinion to test my own thoughts.
 

Offline Lord NicollTopic starter

  • Contributor
  • Posts: 17
  • Country: ie
Re: 400V 4-8A DC supply
« Reply #6 on: July 17, 2017, 04:16:25 am »
8A x 400V = 3200W ouch.
Probably easier to get 4A or so.

There are some high power computer PSUs out there, I've got one that's 2450W and has PFC to charge the caps to 400V.
Without Power Factor Correction (PFC) it's going to draw a lot more current from the mains, as you found out.

I dunno if 3 phase is an option where you are, easier to get 400V DC from that due to higher average voltage.

I'd rather not spend a massive amount more on buying it, plus this capacitor bank and charging supply are for use in a railgun so I want to design my own safety kill switches into it so make sure mains is never applied while the armature is in the rail assembly.

3 Phase is an option but it'd cost over €20,000 to get it installed, so while it's technically available, it's cost prohibitive.
 

Offline Lord NicollTopic starter

  • Contributor
  • Posts: 17
  • Country: ie
Re: 400V 4-8A DC supply
« Reply #7 on: July 17, 2017, 04:25:24 am »
They make "capcitor charging" PSUs.  Some old photographic flash / xenon strobe power supplies would put out such voltages, for instance, for charging capacitors to 300-400V range for the photo flash.

Anyway a common way as stated above is just to use a "large" current limiting resistor in series with the charger output and let the capacitors charge over XX seconds time acording to the 4.0*R*C time constant time or longer depending on the peak charge voltage needed and then the unit is ready for another flash.

If you wish to charge faster then you can use a smaller resistor or as given above a switched series of resistors so the charging current is higher and controller within limits.

If you want to provide a peak current limit and also have high efficiency (switched mode) charging then you could charge the capacitors by some means like connecting them to the output of a flyback type power supply that delivers bursts of energy somewhat more efficiently at controlled rates so as to limit the average current.

Or if you had a variac transformer or similar PSU you could manually or automatically adjust by the dial or control knob / means the PSU output voltage from 0VDC to 400VDC at a rate of change so that the output current limit is not violated.  That would be basically what a "capacitor charging PSU" would do.

It is not complicated but you just need the right equipment and setup to do the job properly and safely.

But anyway a capacitor is a "voltage source" and a voltage output PSU is also a voltage source.  Ideal voltage sources have zero output impedance.  So of course you cannot connect one voltage source to any other at a different voltage without too large of a surge current flowing.  So you must have some current limited "current source" type output for limiting the surge current or the voltage of the whole circuit must be changed all at once and gradually / progressively in a linear way so there isn't a surge of current.

If you don't know what you are doing to do the job properly in charging and discharging such sized capacitors, maybe you should not be working with this kind of apparatus before learning to do so appropriately and safely.

Charging the bank doesn't have to be a fast process, but charging 12 large capacitors from a flash unit might take a long time. I'd also like to avoid spending a massive mount more and I need to design in specific safety devices into the supply, so this is something I'll be building myself. I don't know if you read my original post, but I did mention I used a 1600 watt 36Ohm resistor to charge the bank, and would like something more elegant and that actually gets to 400V DC. While 8 Amps would be nice, I understand that is a lot of power, that is about as much as the 13 amp sockets can supply, the reason I choose that value actually. The flyback type SMPS might be the best bet for a reasonable current and voltage, I also have a few old welders I could use to make 290v AC and rectifies that as I previously mentioned, but that's not as efficient and lacks fine control.
 
The following users thanked this post: robocoyote

Offline abraxa

  • Frequent Contributor
  • **
  • Posts: 377
  • Country: de
  • Sigrok associate
Re: 400V 4-8A DC supply
« Reply #8 on: July 17, 2017, 05:25:15 am »
How about using a motor-generator to obtain 3-phase AC? That way, you can also limit the generator voltage by reducing the motor voltage, bypassing the need for current-limiting components on the generator side.
 

Offline mk_

  • Regular Contributor
  • *
  • Posts: 228
  • Country: at
Re: 400V 4-8A DC supply
« Reply #9 on: July 17, 2017, 05:35:21 am »
Hmm, people seem to be missing some points, I did use a large resistor, however mains voltage is only 330v DC when rectified, I want the full 400V, hence asking what you guys thought, I know how to do it, I just wanted a second opinion to test my own thoughts.

There are a lot of cheap 400>230V transformers aviable. Connect such a transformer reverse so that you can go 230 > 400V. Disconnect from mains if  VDC on your Caps has reached the voltage you need.
Don`t forget the currentlimmiter (Resistor or similar)

Michael



 
The following users thanked this post: jbb

Online NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9014
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: 400V 4-8A DC supply
« Reply #10 on: July 17, 2017, 06:08:34 am »
How about a voltage doubler with a bank of motor run capacitors as the input capacitor to limit the input current? You'll also need a cutout circuit to switch it off when 400V is reached.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline jbb

  • Super Contributor
  • ***
  • Posts: 1143
  • Country: nz
Re: 400V 4-8A DC supply
« Reply #11 on: July 17, 2017, 06:26:21 am »
As I recall, you're interested in a railgun or coil gun.  The massive magnetic fields these things kick up could be a problem for fancy power electronics, so I suggest you go old school and follow mk_'s solution: a step up transformer and diode rectifier.

I also recommend using an isolating transformer so that one rail of the cap bank (typically -ve side) can be connected to earth.  That way you'll know if you've got an accidental earth connection (and by 'know' I mean 'thank goodness we're wearing earmuffs and safety goggles).

To get 400V DC output, you need 280V RMS.  I suggest the following setup:
230V input -> big ass series resistor -> single phase variac -> 230 : 280V isolating transformer -> diode rectifier.

The variac isn't strictly necessary, but it will help you control the final charge voltage of the capacitor bank to achieve consistent results.  Please note that variacs are almost never isolated!

Now, an interesting trick you can use here is to deploy a non-linear resistor so that it doesn't take forever for the last bit of charging.  A couple of 500W halogen floodlights can be very effective in this role (just make sure they're not too close to flammable stuff as the beam out a lot of infra red).  When their filaments are hot, their resistance will be around 100 Ohms each, thus limiting peak current draw.  When their filaments are cold, their resistance will be along the lines of 5 - 10 Ohms each, which will help get the last few volts into the cap bank without waiting for ages.
 

Offline Lord NicollTopic starter

  • Contributor
  • Posts: 17
  • Country: ie
Re: 400V 4-8A DC supply
« Reply #12 on: July 17, 2017, 07:32:32 am »
As I recall, you're interested in a railgun or coil gun.  The massive magnetic fields these things kick up could be a problem for fancy power electronics, so I suggest you go old school and follow mk_'s solution: a step up transformer and diode rectifier.

I also recommend using an isolating transformer so that one rail of the cap bank (typically -ve side) can be connected to earth.  That way you'll know if you've got an accidental earth connection (and by 'know' I mean 'thank goodness we're wearing earmuffs and safety goggles).

To get 400V DC output, you need 280V RMS.  I suggest the following setup:
230V input -> big ass series resistor -> single phase variac -> 230 : 280V isolating transformer -> diode rectifier.

The variac isn't strictly necessary, but it will help you control the final charge voltage of the capacitor bank to achieve consistent results.  Please note that variacs are almost never isolated!

Now, an interesting trick you can use here is to deploy a non-linear resistor so that it doesn't take forever for the last bit of charging.  A couple of 500W halogen floodlights can be very effective in this role (just make sure they're not too close to flammable stuff as the beam out a lot of infra red).  When their filaments are hot, their resistance will be around 100 Ohms each, thus limiting peak current draw.  When their filaments are cold, their resistance will be along the lines of 5 - 10 Ohms each, which will help get the last few volts into the cap bank without waiting for ages.

Yes, that is correct, Railgun however I have tested both, Noted for the halogen light, the large element I used has the exact same effect (if it didn't I'd probably want to contact a physicist or write my own paper on it, as that would be breaking the laws of physics hehe), although not as much, it gets to a few hundred degrees, and then kinda stays there, even after charging is done (that poor bridge rectifier, I bed I'd fail if I left it on like that)

Note the gun will feature 4 stages, as in electrical safeties. Safe (no charge, mains off, all contacts bridged), Primed (charging, charge supply connected, or charged and charge supply connected) , armed (charged and charging supply disconnected, all contactors open except rail ones [Oh boy, I might make these ones out of copper myself for current]) and firing (SCR's activated, capacitors hopefully still alive, as would the SCRs and back EM protection). This is the main reason I'd want to construct my own supply, as to ensure this system won't randomly fire a hunk of steel into my chest or something else bad.

I will probably use the transformer method with 240 (Ireland still hasn't moved to the EU's 230v) to 280 (sweet, my maths was only about 10 volts off) and rectify that, I prefer simpler circuits, however there will be filtering even before the transformer, common mode chokes, MOV's and maybe not suppression capacitors, but if I feel they wouldn't hurt I have bags of them. I'll have a look to decide whether toriodal (I'm a big fan of them, I just like them) or a more standard will be used, however I might go with a standard one as they're easier to work with.

The gun's negative rail will indeed be tied to mains earth, and the positives rail by some 2KV capacitors, hopefully that also helps quell some magnetics. I might also use a plasma armature and not use SCR at all, instead fire a high voltage pulse to generate plasma and hope that dumps the power, but that would definitely destroy the rails faster, as it stands titanium was a massive sparkly failure (pretty to watch though). Also the miss alignment of the projectile becomes a factor, so it depends on small scale experiments.
 

Offline Lord NicollTopic starter

  • Contributor
  • Posts: 17
  • Country: ie
Re: 400V 4-8A DC supply
« Reply #13 on: July 17, 2017, 07:34:25 am »
How about a voltage doubler with a bank of motor run capacitors as the input capacitor to limit the input current? You'll also need a cutout circuit to switch it off when 400V is reached.
I though about the Cockcroft–Walton generator type deal, however the current is too high and 400VDC isn't that much higher than mains, so such a solution wouldn't be as pragmatic as the one I'm pretty sure I'll end up using.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19517
  • Country: gb
  • 0999
Re: 400V 4-8A DC supply
« Reply #14 on: July 17, 2017, 08:53:33 am »
Use a hair dryer as a current limiting resistor. Do realize that such a capacitor bank is at least as dangerous as the mains even when unplugged.
Hmm, people seem to be missing some points, I did use a large resistor, however mains voltage is only 330v DC when rectified, I want the full 400V, hence asking what you guys thought, I know how to do it, I just wanted a second opinion to test my own thoughts.
It's bad practice to run capacitors at their full voltage rating.

What about getting a 50V mains transformer, connecting the primary as normal and its secondary in series with the mains? Get the phasing right and you'll have 230V + 50V = 280VAC which will give you very close to rectified 400VDC.

I suggest adding a variac to the input because you might get a slightly higher voltage than expected, resulting in over 400VDC.
 

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2222
  • Country: mx
Re: 400V 4-8A DC supply
« Reply #15 on: July 17, 2017, 01:40:22 pm »
If you want to follow the transformer route, here in North America 230 to 277 volt transformers are common.
It will be expensive to ship them to Europe, though.
 

Offline C

  • Super Contributor
  • ***
  • Posts: 1346
  • Country: us
Re: 400V 4-8A DC supply
« Reply #16 on: July 17, 2017, 02:31:34 pm »
Many ways have been suggested.

One DC voltage with a variable resistance.
Variable DC voltage with voltage controlled by current.
The variac transformer is a version of this.

Why charge the 400V capacitors to 400v in one step?

Some have suggested a transformer to get extra needed to get to 400v.
Hero999 suggested a 50V mains transformer used as a boost.

You could use this transformer to charge from 0V to __V(<50v). Then switch connections so that this transformer reduces Mains to change higher. Then mains to charge higher yet. Then switch transformer to boost mains for final step.

With more steps you could also switch to higher step before you reach full charge for that step saving some time.
 
 

Offline Jeff1946

  • Regular Contributor
  • *
  • Posts: 104
Re: 400V 4-8A DC supply
« Reply #17 on: July 19, 2017, 06:20:09 pm »
Would the correct size light bulb or more in series work?   The resistance of tungsten increases rapidly with temperature so it would limit max current. Be interesting to do the math.
 

Offline jbb

  • Super Contributor
  • ***
  • Posts: 1143
  • Country: nz
Re: 400V 4-8A DC supply
« Reply #18 on: July 19, 2017, 08:26:15 pm »
Would the correct size light bulb or more in series work?   The resistance of tungsten increases rapidly with temperature so it would limit max current. Be interesting to do the math.

Yes, light bulbs work quite well for this sort of application.  Their hot resistance can be calculated from their rating (e.g. 100W / 230V = 0.44A, 230V / 0.44A = 522 Ohms approx.).  Their cold resistance will be 1/10th the hot resistance (warning - total guesstimate) or less for halogen types.  The best part is that they light up while working.

The downside is that they are only available in a few voltage ratings.  If you want 230V that's great.  But if you want 400V, you may have a problem sourcing a 400V bulb.

You unfortunately can't put them in series reliably.  This is because all light bulbs are not created equal; if bulb A has a slightly thicker filament than bulb B, bulb B will heat up faster.  Then bulb B will have more voltage across its filament (i.e. unequal voltage sharing), resulting in extra power dissipation and higher temperatures.  (This is a case of thermal runaway.)  The result will be that bulb A doesn't do much while bulb B gets hammered with over voltage and has a very short lifespan.

There's no problem putting bulbs in parallel, though.
 

Offline max_torque

  • Super Contributor
  • ***
  • Posts: 1280
  • Country: gb
    • bitdynamics
Re: 400V 4-8A DC supply
« Reply #19 on: July 20, 2017, 08:49:52 am »
You could probably find a cheap, second hand small petrol genset (ie honda unit etc), and remove the constant speed governor?  In fact, most genset motors will rev high enough to stick out quite a high voltage at low load.......
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf