Short Circuit Protected Constant Current Supply

John Larkin <jlarkin@highland_atwork_technology.com> wrote:

> On Fri, 7 Jan 2022 12:40:07 -0000 (UTC), "Arnie Dwyer (ex Jan Frank)"
> <spamme@not.com> wrote:

[...]

>>This causes a serious droop in the +HV supply. As the constant current 
>>demand increases, the +HV drops significantly. This makes it impossible
>>to reach the desired 20 mA maximum with a large load impedance. 
> 
> Sure. It depends on what the current limit is for. If it's to prevent
> explosions if the hv supply gets grounded, there's probably a value
> that works.

It would still increase the impedance of the +HV supply and degrade the
voltage regulation under load. I need the current source to supply as high
a voltage as is reasonable in order to drive current into a high
resistance load at max current. As it is, the voltage drops from 330V to
286V at 20mA, which is acceptable. 

This limits the maximum load resistance to 286/20e-3=14,300 ohms at max
current. This is acceptable, since the area of the electrodes will be
fairly large in order to achieve a low current density. The cell will be
the largest I have ever built, so there is some uncertainty in the
results. 

During the electrolysis process, the conductance of the cell will increase
by a factor of 10 or 20 as silver ions enter the solution. The current is
limited to restrict the combination of silver and hydroxide ions in the
Nernst Diffusion layer to produce silver hydroxide, AgOh. This is
undesirable, since it limits the concentration of silver ions to low
values, such as 4-5 ppm. It takes care to reach 18 ppm, which is about the
best you can do at home. 

The current supply will be connected to an external electrolysis cell, so
the leads will be exposed. The current limit is to protect the supply in
the event the positive lead accidentally strikes ground during the
connection. 

Normally the supply will be turned off and the capacitors discharged while
it is being connected, but there could be some other accident that shorts
the supply. For example, a DVM could be connected to monitor the voltage
across the load to check for saturation. The leads could be connected
together for an instant and short out the supply. 

Also, the 1N4007 rectifier diodes have a peak surge current of 30A, but
I'd rather not deal with that kind of current even for a half-cycle. The
power supply switch is a miniature DPDT and I'm certain the contacts would
not handle it. So the old uA723 style Vbe current limiter seems the best
choice. It is only there for transients, so I don't need to go into
voltage foldback and all its complications. 

Thanks for your help.

Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

[...]
 
> One common approach is to do resistive or capacitive current limiting 
> inside a feedback loop.  The two fight each other, and the point where 
> the feedback loop loses is the current limit.

Thanks, but sounds complicated. All I need is protection against accidental 
transient short circuits. The old ua723 Vbe limiter seems to be the optimum 
solution. 

I don't know of any other current limiter that can handle 30mA at 330V, at a 
price I can afford:)
 
> Cheers
> 
> Phil Hobbs
> 

On a sunny day (Fri, 7 Jan 2022 21:05:50 -0000 (UTC)) it happened "Arnie Dwyer
(ex Jan Frank)" <spamme@not.com> wrote in
<XnsAE18A2F78FF27idtokenpost@144.76.35.252>:

>Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:
>
>[...]
> 
>> One common approach is to do resistive or capacitive current limiting 
>> inside a feedback loop.  The two fight each other, and the point where 
>> the feedback loop loses is the current limit.
>
>Thanks, but sounds complicated. All I need is protection against accidental 
>transient short circuits. The old ua723 Vbe limiter seems to be the optimum 
>solution. 
>
>I don't know of any other current limiter that can handle 30mA at 330V, at a 
>price I can afford:)

?
No idea about this project but I used BU??? HV 400V power MOSFETs in the eighties
to do just that, on a big heatsink.