From: Tim Wescott on
lerameur wrote:
> On Mar 3, 9:20 pm, Hammy <s...(a)spam.com> wrote:
>> On Thu, 04 Mar 2010 03:03:26 +0100, Sjouke Burry
>>
>>
>>
>> <burrynulnulf...(a)ppllaanneett.nnll> wrote:
>>> lerameur wrote:
>>>> hello,
>>>> I have a very simple circuit here, I am charging a 900uF capacitor to
>>>> 600v. (900v max capacity)
>>>> I want to discharge this at 0.01seconds interval. I have 30 ohm
>>>> resistance, without so from my calculation I should have 20 amp max.
>>>> All this element are in series with each other, I want to add a
>>>> transistor (or mosfet) to make and break the intervals I need.
>>>> DO I need to get a 600v and a 20 amp rated transistor ?
>>>> Thanks
>>> Always apply a safety factor, I would use 800 and 30.
>>> Whether the 30 amp is needed depends on your duty cycle,
>>> if the dutycycle is low, 20 and even 10 might be enough.
>>> Just check the allowed peak current.
>> If he is saying
>>
>>> " (900v max capacity)"
>> Shouldnt he be useing a 1kV-1.2kV fet or IGBT with maybe a drain
>> clamp?
>
> The capacitor can take up to 900v, but I am only charging it to 600v..
> I thought maybe if I can get some transistors in parallel to drain
> the current, but using transistor rated below 600v. But I do not think
> it will work.
>
Certainly not in parallel -- a parallel connection sees the full voltage
across each device.

In theory you could connect transistors in series, with appropriate gate
or base drivers. But it'd be a lot cheaper to just get a 900V transistor.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
From: Tim Wescott on
lerameur wrote:
> hello,
>
> I have a very simple circuit here, I am charging a 900uF capacitor to
> 600v. (900v max capacity)
> I want to discharge this at 0.01seconds interval. I have 30 ohm
> resistance, without so from my calculation I should have 20 amp max.
> All this element are in series with each other, I want to add a
> transistor (or mosfet) to make and break the intervals I need.
> DO I need to get a 600v and a 20 amp rated transistor ?

Were you discharging down to nothing I'd suggest an SCR (or stack
thereof). But you're not...

You need to get a transistor that's rated for _more_ than 600V -- the
voltage rating on any device is "don't ever exceed this, even once, or
it'll die, you'll die, men, women and children will die, aughhhhhhhh!".
Different companies and industries treat this absolute maximum rating
with different attitudes -- electrolytic capacitors are generally rated
much closer to the edge than semiconductors, for example.

I'd use a transistor that's rated for at least 800 or 900V, and if that
900uF cap is electrolytic I'd consider finding one that's good to 1200V.

The current ratings that they put in big print on FET data sheets is
based on a well heat-sunk transistor that's operating continuously at
room temperature. You really need to do the thermal analysis for your
particular situation to make sure that your power devices aren't going
to go 'poof!'. When you're done you may find that you can get by with
one rated at 20A (or less, depending on the surge or instantaneous
current rating), or you may find that you need more.

Post a copy of your circuit someplace and folks will critique it for you.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
From: WangoTango on
In article <OdOdnTwYH-NUeRLWnZ2dnUVZ_vqdnZ2d(a)web-ster.com>,
tim(a)seemywebsite.now says...
> lerameur wrote:
> > On Mar 3, 9:20 pm, Hammy <s...(a)spam.com> wrote:
> >> On Thu, 04 Mar 2010 03:03:26 +0100, Sjouke Burry
> >>
> >>
> >>
> >> <burrynulnulf...(a)ppllaanneett.nnll> wrote:
> >>> lerameur wrote:
> >>>> hello,
> >>>> I have a very simple circuit here, I am charging a 900uF capacitor to
> >>>> 600v. (900v max capacity)
> >>>> I want to discharge this at 0.01seconds interval. I have 30 ohm
> >>>> resistance, without so from my calculation I should have 20 amp max.
> >>>> All this element are in series with each other, I want to add a
> >>>> transistor (or mosfet) to make and break the intervals I need.
> >>>> DO I need to get a 600v and a 20 amp rated transistor ?
> >>>> Thanks
> >>> Always apply a safety factor, I would use 800 and 30.
> >>> Whether the 30 amp is needed depends on your duty cycle,
> >>> if the dutycycle is low, 20 and even 10 might be enough.
> >>> Just check the allowed peak current.
> >> If he is saying
> >>
> >>> " (900v max capacity)"
> >> Shouldnt he be useing a 1kV-1.2kV fet or IGBT with maybe a drain
> >> clamp?
> >
> > The capacitor can take up to 900v, but I am only charging it to 600v..
> > I thought maybe if I can get some transistors in parallel to drain
> > the current, but using transistor rated below 600v. But I do not think
> > it will work.
> >
> Certainly not in parallel -- a parallel connection sees the full voltage
> across each device.
>
> In theory you could connect transistors in series, with appropriate gate
> or base drivers. But it'd be a lot cheaper to just get a 900V transistor.
>
>
What about a big assed 'puck' style SCR / thyristor?
I don't know about how fast you can trigger the things, but don't they
use them for big motor control apps?
I know you can get them with hold off voltages in the 1000V+ range and
dump hundreds of amps through them. Might get a bit hot, but what about
the heat from the cap ESR???

From: lerameur on
On 4 mar, 12:35, WangoTango <Asgar...(a)mindspring.com> wrote:
> In article <OdOdnTwYH-NUeRLWnZ2dnUVZ_vqdn...(a)web-ster.com>,
> t...(a)seemywebsite.now says...
>
>
>
> > lerameur wrote:
> > > On Mar 3, 9:20 pm, Hammy <s...(a)spam.com> wrote:
> > >> On Thu, 04 Mar 2010 03:03:26 +0100, Sjouke Burry
>
> > >> <burrynulnulf...(a)ppllaanneett.nnll> wrote:
> > >>> lerameur wrote:
> > >>>> hello,
> > >>>> I have a very simple circuit here, I am charging a 900uF capacitor to
> > >>>> 600v. (900v max capacity)
> > >>>> I want to discharge this at 0.01seconds interval.  I have 30 ohm
> > >>>> resistance, without so from my calculation I should have 20 amp  max.
> > >>>> All this element are in series with each other, I want to add a
> > >>>> transistor (or mosfet) to make and break the intervals I need.
> > >>>> DO I need to get a 600v and a 20 amp rated transistor ?
> > >>>> Thanks
> > >>> Always apply a safety factor, I would use 800 and 30.
> > >>> Whether the 30 amp is needed depends on your duty cycle,
> > >>> if the dutycycle is low, 20 and even 10 might be enough.
> > >>> Just check the allowed peak current.
> > >> If he is saying
>
> > >>> " (900v max capacity)"
> > >> Shouldnt he be useing a 1kV-1.2kV fet or IGBT with maybe a drain
> > >> clamp?
>
> > > The capacitor can take up to 900v, but I am only charging it to 600v...
> > > I thought maybe if I can get  some transistors in parallel to drain
> > > the current, but using transistor rated below 600v. But I do not think
> > > it will work.
>
> > Certainly not in parallel -- a parallel connection sees the full voltage
> > across each device.
>
> > In theory you could connect transistors in series, with appropriate gate
> > or base drivers.  But it'd be a lot cheaper to just get a 900V transistor.
>
> What about a big assed 'puck' style SCR / thyristor?
> I don't know about how fast you can trigger the things, but don't they
> use them for big motor control apps?
> I know you can get them with hold off voltages in the 1000V+ range and
> dump hundreds of amps through them.  Might get a bit hot, but what about
> the heat from the cap ESR???- Masquer le texte des messages précédents -
>
> - Afficher le texte des messages précédents -

I do not want to discharge all the gap,
Ok Bob mentionned the time delay, I may have to reduce the resistance,
(increase the current) BUt I need a spark in the circuit.
So capacitor, resistance, spark gap and transistor.
Ok so I need several kV, lets say 3000v , but now lets say I use a 10
ohm resistance, that gives use a wapping 300 amp.
Is there such a transistor rated at 3kV and 300 amp?

That was my original idea, but when I saw outstanding figures I
decided to decrease the voltage and start low.

JP
From: Tim Wescott on
WangoTango wrote:
> In article <OdOdnTwYH-NUeRLWnZ2dnUVZ_vqdnZ2d(a)web-ster.com>,
> tim(a)seemywebsite.now says...
>> lerameur wrote:
>>> On Mar 3, 9:20 pm, Hammy <s...(a)spam.com> wrote:
>>>> On Thu, 04 Mar 2010 03:03:26 +0100, Sjouke Burry
>>>>
>>>>
>>>>
>>>> <burrynulnulf...(a)ppllaanneett.nnll> wrote:
>>>>> lerameur wrote:
>>>>>> hello,
>>>>>> I have a very simple circuit here, I am charging a 900uF capacitor to
>>>>>> 600v. (900v max capacity)
>>>>>> I want to discharge this at 0.01seconds interval. I have 30 ohm
>>>>>> resistance, without so from my calculation I should have 20 amp max.
>>>>>> All this element are in series with each other, I want to add a
>>>>>> transistor (or mosfet) to make and break the intervals I need.
>>>>>> DO I need to get a 600v and a 20 amp rated transistor ?
>>>>>> Thanks
>>>>> Always apply a safety factor, I would use 800 and 30.
>>>>> Whether the 30 amp is needed depends on your duty cycle,
>>>>> if the dutycycle is low, 20 and even 10 might be enough.
>>>>> Just check the allowed peak current.
>>>> If he is saying
>>>>
>>>>> " (900v max capacity)"
>>>> Shouldnt he be useing a 1kV-1.2kV fet or IGBT with maybe a drain
>>>> clamp?
>>> The capacitor can take up to 900v, but I am only charging it to 600v..
>>> I thought maybe if I can get some transistors in parallel to drain
>>> the current, but using transistor rated below 600v. But I do not think
>>> it will work.
>>>
>> Certainly not in parallel -- a parallel connection sees the full voltage
>> across each device.
>>
>> In theory you could connect transistors in series, with appropriate gate
>> or base drivers. But it'd be a lot cheaper to just get a 900V transistor.
>>
>>
> What about a big assed 'puck' style SCR / thyristor?
> I don't know about how fast you can trigger the things, but don't they
> use them for big motor control apps?
> I know you can get them with hold off voltages in the 1000V+ range and
> dump hundreds of amps through them. Might get a bit hot, but what about
> the heat from the cap ESR???

Turn-off becomes an issue. I know that there are ways to turn off an
SCR on a DC supply, but it's some wacky thing involving an auxiliary
inductor that snaps the current off briefly so the SCR can turn off --
it all looks pretty smoky to me.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com