All MY big electrolytics were "twist-lock" with a chassis insert, so they all blew out the bottom.
My worst assignment (at my father's TV repair shop) was to clean-up a chassis where lightning blew all the potting out of the lead holes of a transformer, filling the chassis, then solidifying ;-)
...Jim Thompson
--
| James E.Thompson, P.E. | 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 |
I love to cook with wine Sometimes I even put it in the food
But when you watch MacGyver as an adult, don't you just have to cringe a bit? :-)
I believe they used an old Drake HF rig as an "ultrasonic oscillator" once... In the same show they use the AM receiver in a Jeep and a handheld AM/FM radio all of about 20 feet away to triangulate the position of the bad guys... who are using what appear to be 5W CB walkie talkies... :-)
Oh, and the handheld radio has run down its 9V battery, so they tap into a cactus to provide it with power. :-)
It was a good show, certainly.
I suspect Jim would feel a few more munitions might have made some of the solutions simpler. :-)
Never watched MacGyver or most of the other nonsense on network TV.
I liked Columbo, and I remain a fan of Monk and CSI... in the summer of 2007 I actually took a week-long CSI course at UC Riverside (part of the MIT "Road Scholar" travel series)... great fun, except for the smell of bodies in the morgue ;-)
...Jim Thompson
--
| James E.Thompson, P.E. | 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 |
Obama... Recklessness cloaked in Righteousness
That looks like a reasonable design, though I would prefer a P-fet pass device, granted at probably a higher component cost. My concern would be some "weirdness" (glitch) keeping the charge pump running. A P-fet pass design would be more "analog."
Not really. After moving to the US we stayed in an apartment until escrow closed. Cable TV. I loved to watch a episode of Gilligan's Island once in a while. My wife began to question my sanity ...
Yeah, it's all show, but funny.
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
These caps had solder pins but very strong ones. The bottom was crimped around a base plate with the solder pins in there. Ripped it all out and took off in a straight vertical line.
Nobody would go to that length anymore these days :-(
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
I had a look at those--they look pretty good but cost $4. The TL431 is nice and cheap and works well, but needs at least 400 uA to regulate properly, which makes the dropping resistor dissipation a little embarrassing at high input voltage.
I'm looking at a series pass solution using an Infineon BSP149 depletion MOSFET (which gives me some capacitance multiplication as a bonus). The whole thing comes to about $1.50 in quantity 100. It should be reasonably immune to overvoltages of up to about 200V.
(Schematic is at
formatting link
Critiques welcome. (As I said, I haven't done this particular job in a loooong time.)
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
But it depends on how much overvoltage surge you expect.
There are also 431 variants that are happy with 100uA but not all of those have a large voltage range, some can't be taken past 6V or so. An example:
formatting link
Nice penmanship. Don't look at my sketches before they go into CAD, they look more like Bob Pease's.
Careful, this will drop a few volts and could become toasty if large currents are drawn. It's customary to use P-channels there unless you have a charge pump or can switch the negative rail.
I assume the cap on the gate is 10nF as in 0.01uF. Should be ok but I usually don't like an MMBT3904 to yank down a cap bigger than 1000pF or so. Per Murphy's law you'll need a bigger one some day for more cap multiplier action, get tempted to assume that it'll be ok and one day ... tsssk ... *PHUT*.
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
--
| James E.Thompson, P.E. | 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 |
Obama... Recklessness cloaked in Righteousness
Looks like my first reply went lalaland, so here goes again:
Phil Hobbs wrote: > qrk wrote: >> On Tue, 16 Jun 2009 10:19:56 -0700, Joerg >> wrote: >>
Take a look at these, much cheaper:
formatting link
But it depends on how much overvoltage surge you expect.
There are also 431 variants that are happy with 100uA but not all of those have a large voltage range, some can't be taken past 6V or so. An example:
formatting link
Nice penmanship. Don't look at my sketches before they go into CAD, they look more like Bob Pease's.
Careful, this will drop a few volts and could become toasty if large currents are drawn. It's customary to use P-channels there unless you have a charge pump or can switch the negative rail.
I assume the cap on the gate is 10nF as in 0.01uF. Should be ok but I usually don't like an MMBT3904 to yank down a cap bigger than 1000pF or so. Per Murphy's law you'll need a bigger one some day for more cap multiplier action, get tempted to assume that it'll be ok and one day ... tsssk ... *PHUT*.
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Doesn't 10nF defeat your purpose of fast turn-off during a transient?
...Jim Thompson
--
| James E.Thompson, P.E. | 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 |
Obama... Recklessness Cloaked in Righteousness
--
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
I don't need it to turn off that fast--there'll be a volt or so of kick and then the FET becomes a regular source follower as the voltage drops.
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
This looks like overkill of a problem that may not need to be solved. A polyfuse and a transzorb is plenty of protection. If you used a wall-wart and no protection at all, it's unlikely that any customers will ever have any problems.
Figure maybe 25 cents per part purchasing+placement cost for low-volume production.
You're probably right. (It's partly an exercise.) OTOH, after doing some math to get the transzorb/polyfuse solution to work at moderate overvoltage without the transzorb unsoldering itself, it's a bit more subtle than it first appeared.
Does it really cost a quarter to put a resistor on a board?
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
Tube days. An apartment house TV distribution amplifier. I'm sure, now, that some of those repairs I had to make, were to ensure my "education" in the school of hard knocks... I even know how grind valves (*) in situ ;-)
(*) NOT tubes... automobile engines ;-)
...Jim Thompson
--
| James E.Thompson, P.E. | 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 |
Obama... Recklessness Cloaked in Righteousness
Well, at some point the user blew it up. I assume that any abusive voltage will be applied suddenly and will (usually) open the polyfuse before the transzorb melts. But it could be carefully teased to fry stuff. That's improbable. Worst case, the transzorb shorts and protects the expensive stuff downstream.
One other comment about your circuit: it still has an all-silicon path from the outside world into the guts. A good ESD zot could take out the fet or the schottky. I like to avoid any all-silicon zap paths. The polyfuse+transzorb+capacitor is rugged and is a first-level RFI filter too.
That's the number we use for purchasing/kitting/placement/inspection. That really changes your perspective on design.
** The majority of switching wall warts have over-volatage failure protection in them already - usually just a zener. But others do not, as you found out.
** All wall warts are designed to be fail safe when loaded with a short - transformer based ones have thermal fuses in the transformer and switching ones have current limiting in the switching drive circuit. There will also be a fuse or fusible resistor in the AC supply side of switching types.
If you have diode across the DC input your " gizmo " to save it from reverse polarity and an SCR crow bar to save it from over-voltage - that is plenty. Damn the silly wall wart if is faulty or wrong.
The scenario that has ME worried is the failure of an electro in a switching wall wart, so electrolyte leaks and bridges the mains side to output side barrier or else pieces of aluminium foil ( when one explodes from over-voltage) do the same.
I'm expecting to draw about 500 mA. The FET should drop about 1.3V, I think. I could use a P-channel there and save some power, it's true...I see that there are quite a few for less than 50 cents apiece. Good call, thanks.
Hmm. I'm not meaning to drag the 10 nF down all that fast--the maximum base current on the 3904 at the 25V trip point is 100 uA, even if one transistor were to hog all the base current. Even assuming a beta of
500, that's a maximum current of 50 mA for 5 us, which should be well within the 3904's limitations, I think. If I were hitting the base harder, I might vapourize a bond wire. The 1000 pF rule of thumb is worth my remembering for other things, though.
Thinking about it again, I doubt I'll get any cap multiplier action to speak of--the FET is in its linear range, so there won't be much voltage feedback happening. I suspect I'd be better off doing it the vanilla way.
Thanks again,
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
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.