Prev: Nichia red leds
Next: Battery buffer charging/maintenance
From: Joel Koltner on 9 Aug 2010 21:22
"John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in message
news:4i8166dkk2tof9ent2379luhfhg9ifvrpq(a)4ax.com...
> Really, somebody needs to make a good current limiter chip. Low drop,
> resistor programmed, periodic retry, SOT23, 3-15 volts, cheap. They'd
> sell zillions of them.

Sounds a lot like the controller ICs built into DC fans: If you block the
blades with your finger, they'll usually stop and then pulse a bit every few
seconds trying to restart...

From: Jim Thompson on 9 Aug 2010 21:37
On Mon, 9 Aug 2010 18:22:16 -0700, "Joel Koltner"
<zapwireDASHgroups(a)yahoo.com> wrote:

>"John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in message
>news:4i8166dkk2tof9ent2379luhfhg9ifvrpq(a)4ax.com...
>> Really, somebody needs to make a good current limiter chip. Low drop,
>> resistor programmed, periodic retry, SOT23, 3-15 volts, cheap. They'd
>> sell zillions of them.
>
>Sounds a lot like the controller ICs built into DC fans: If you block the
>blades with your finger, they'll usually stop and then pulse a bit every few
>seconds trying to restart...

Easy enough to do, but customers won't PAY for such smarts.

...Jim Thompson
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |

Spice is like a sports car...
Performance only as good as the person behind the wheel.
From: Phil Hobbs on 9 Aug 2010 22:17
John Larkin wrote:
> On Mon, 09 Aug 2010 08:25:14 -0400, Phil Hobbs
> <pcdhSpamMeSenseless(a)electrooptical.net> wrote:
>
>> John Larkin wrote:
>>> On Sat, 07 Aug 2010 15:28:47 GMT, nontelo(a)dico.no (Mino) wrote:
>>>
>>>> Hello, what happens when too much current is drawn from an LM317
>>>> configured for 3.3V output? Does the output voltage decrease?
>>> It decreases a little bit as you load it, at any current [1]. At some
>>> large current you hit either its current limit or its thermal limit.
>>> At the current limit, voltage starts to drop steeply, such as to keep
>>> the load current constant. If it get too hot before that happens, most
>>> 317s will just shut off until they cool down some.
>>>
>>> Right now, we really need a low-voltage-drop current limiter, to
>>> protect a circuit powered by a dc/dc converter. Something like 150 mA
>>> would be good. Polyfuses suck. Discrete circuits, like the things
>>> James was playing with, would work, but I don't have room for all the
>>> parts. We're considering using an adjustable LDO and running it
>>> wide-open, so only the current+thermal limits are in use.
>>>
>>>
>>> John
>>>
>>> [1] I suppose a 317 could have a slight increase in voltage with load,
>>> if the thermal situation were right, but most will droop in real life.
>> If you can afford 1.2V, there's the LT3092. Otherwise maybe a
>> beta-graded BJT that comes out of saturation when you hit the limit?
>> You could use one comparator and 12 resistors to see when any one of the
>> supplies comes off the peg, and use that to turn off the base bias so
>> you don't roast the BJT. If it had some AC hysteresis, it could even do
>> the retry function, or you could use a processor to do the turn off and
>> retry function once the BJT was doing the fast reacting.
>>
>> Cheers
>>
>> Phil Hobbs
>
> Looks like I'll have to do this:
>
> ftp://jjlarkin.lmi.net/DCL8.JPG
>
> It's a lot of parts to protect a $4 dc/dc conveter, but it does all
> the stuff we need and it's orthogonal, in the sense that all its
> important parameters are settable independently. I could go for, say,
> 240 mA trip, 1 millisecond delay, and 1 second off. Or whatever.
>
> The LM317L works fine to protect the dc/dc brick, but I can't really
> tolerate the voltage drop.
>
> The ideal way to do this would be just the mosfet and the resistor.
> Have our main uP scan the currents and drive the fet gates, and do the
> protections in software. But we don't have enough ADC inputs or port
> pins to do 12 channels.
>
> Really, somebody needs to make a good current limiter chip. Low drop,
> resistor programmed, periodic retry, SOT23, 3-15 volts, cheap. They'd
> sell zillions of them.
>
> John
>
>

You could sum the voltages into one comparator and turn all of them off
when there's a problem, then turn them back on one by one until you find
the one with the problem. I've use daisy-chained 74HC595s driven by
SPI for that sort of thing.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net
From: John Larkin on 9 Aug 2010 22:49
On Mon, 09 Aug 2010 22:17:27 -0400, Phil Hobbs
<pcdhSpamMeSenseless(a)electrooptical.net> wrote:

>John Larkin wrote:
>> On Mon, 09 Aug 2010 08:25:14 -0400, Phil Hobbs
>> <pcdhSpamMeSenseless(a)electrooptical.net> wrote:
>>
>>> John Larkin wrote:
>>>> On Sat, 07 Aug 2010 15:28:47 GMT, nontelo(a)dico.no (Mino) wrote:
>>>>
>>>>> Hello, what happens when too much current is drawn from an LM317
>>>>> configured for 3.3V output? Does the output voltage decrease?
>>>> It decreases a little bit as you load it, at any current [1]. At some
>>>> large current you hit either its current limit or its thermal limit.
>>>> At the current limit, voltage starts to drop steeply, such as to keep
>>>> the load current constant. If it get too hot before that happens, most
>>>> 317s will just shut off until they cool down some.
>>>>
>>>> Right now, we really need a low-voltage-drop current limiter, to
>>>> protect a circuit powered by a dc/dc converter. Something like 150 mA
>>>> would be good. Polyfuses suck. Discrete circuits, like the things
>>>> James was playing with, would work, but I don't have room for all the
>>>> parts. We're considering using an adjustable LDO and running it
>>>> wide-open, so only the current+thermal limits are in use.
>>>>
>>>>
>>>> John
>>>>
>>>> [1] I suppose a 317 could have a slight increase in voltage with load,
>>>> if the thermal situation were right, but most will droop in real life.
>>> If you can afford 1.2V, there's the LT3092. Otherwise maybe a
>>> beta-graded BJT that comes out of saturation when you hit the limit?
>>> You could use one comparator and 12 resistors to see when any one of the
>>> supplies comes off the peg, and use that to turn off the base bias so
>>> you don't roast the BJT. If it had some AC hysteresis, it could even do
>>> the retry function, or you could use a processor to do the turn off and
>>> retry function once the BJT was doing the fast reacting.
>>>
>>> Cheers
>>>
>>> Phil Hobbs
>>
>> Looks like I'll have to do this:
>>
>> ftp://jjlarkin.lmi.net/DCL8.JPG
>>
>> It's a lot of parts to protect a $4 dc/dc conveter, but it does all
>> the stuff we need and it's orthogonal, in the sense that all its
>> important parameters are settable independently. I could go for, say,
>> 240 mA trip, 1 millisecond delay, and 1 second off. Or whatever.
>>
>> The LM317L works fine to protect the dc/dc brick, but I can't really
>> tolerate the voltage drop.
>>
>> The ideal way to do this would be just the mosfet and the resistor.
>> Have our main uP scan the currents and drive the fet gates, and do the
>> protections in software. But we don't have enough ADC inputs or port
>> pins to do 12 channels.
>>
>> Really, somebody needs to make a good current limiter chip. Low drop,
>> resistor programmed, periodic retry, SOT23, 3-15 volts, cheap. They'd
>> sell zillions of them.
>>
>> John
>>
>>
>
>You could sum the voltages into one comparator and turn all of them off
>when there's a problem, then turn them back on one by one until you find
>the one with the problem. I've use daisy-chained 74HC595s driven by
>SPI for that sort of thing.

Yes, that would work. I could sum them with schottky diodes and a
shared pulldown, or just use a comparator per and wire-OR the outputs.

I'll talk that over with the boys.

John

From: Joel Koltner on 10 Aug 2010 12:57
"Jim Thompson" <To-Email-Use-The-Envelope-Icon(a)On-My-Web-Site.com> wrote in
message news:a8b166h0jbitd5esvoabf0t7pdh27agehg(a)4ax.com...
> On Mon, 9 Aug 2010 18:22:16 -0700, "Joel Koltner"
> <zapwireDASHgroups(a)yahoo.com> wrote:
>>Sounds a lot like the controller ICs built into DC fans: If you block the
>>blades with your finger, they'll usually stop and then pulse a bit every few
>>seconds trying to restart...
> Easy enough to do, but customers won't PAY for such smarts.

Apparently Larkin's will? :-)

What I was really thinking was that perhaps those fan controller ICs could be
pressed into service as current limiters, but I've never looked at a data
sheet for one so that is wild speculation.

---Joel

First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4
Prev: Nichia red leds
Next: Battery buffer charging/maintenance