From: Joerg on
Herman wrote:
> "Joerg" <invalid(a)invalid.invalid> wrote in message
> news:7n5in7F3jmflvU1(a)mid.individual.net...
>> Rich Grise wrote:
>>> On Wed, 25 Nov 2009 11:31:18 -0800, Joerg wrote:
>>>> krw wrote:
>>>>> On Wed, 25 Nov 2009 10:25:01 -0800, Joerg <invalid(a)invalid.invalid>
>>>>>> osr(a)uakron.edu wrote:
>>>>>>> Yeah, the ON circuit is crude, and the SCRs in series are a failure
>>>>>>> point. But its a start.
>>>>>>> And from a safety point of view, its got to be more of a hazard then
>>>>>>> a
>>>>>>> properly made jumper.
>>>>>>> Kinda my point.
>>>>>>>
>>>>>>> I take it this is on a linear for noise or cost reasons?
>>>>>>>
>>>>>> Two reasons. The cost of transformers is remarkably low, around five
>>>>>> bucks for 10-15VA. Can't beat that with a switcher, even considering
>>>>>> your own PWM stuff behind it because that only adds a buck or two.
>>>>>>
>>>>>> Then, temperature. Beats me but all the places I've called won't
>>>>>> endorse more than 50C or 60C for their switcher modules and that's
>>>>>> just
>>>>>> not going to fly. For some reason my own designs never had such limits
>>>>>> but I can't justify the NRE for all the agency testing on this one if
>>>>>> we rolled our own mains-connected switcher.
>>>>>>
>>>>>> The only 250V-rated selectors I see are panel mount and insanely
>>>>>> expensive, such as this one:
>>>>>>
>>>>>> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=0033.1118-ND
>>>>> How about a wall wart? You should be able to find something like that
>>>>> for $10-$15.
>>>> The only way to get one in would be to snip the end off a two-prong
>>>> Christmas extension code, solder it in and strap the whole kludge down
>>>> with cable ties. Any fire marshall would blow a gasket ;-)
>>>>
>>>> Plus I think wall warts don't like to bake in there.
>>> What exactly is it you're trying to produce? You've given several
>>> different conflicting scenarios here. Is the user supposed to be
>>> responsible for selecting the voltage? Does it get set at the factory?
>>> What code are you trying to satisfy?
>>>
>>> Please, Joerg, tell us the "big picture" here.
>>>
>> Ok, big picture follows:
>>
>> Unit in fields, outdoors, think farm. The sun is pelting it brutally all
>> day long and the electric stuff sits inside an enclosure, no vents big
>> enough to write home about. No choice, it has to be that way and it works
>> fine in a 120V version.
>>
>> Client wants users to be able to ideally just plug it into 120 or 240, and
>> not worry about a thing. That ain't going to fly because SMPS modules seem
>> to not be that great. Maybe cheap electrolytics, I don't know, but mfgs
>> don't want to endorse >60C and it sure does get hotter than that in there.
>> Just by the sun.
>>
>> So, the next best avenue would be a configurable 120/240 transformer
>> because then we can have our own (better) switcher behind that.
>> Unfortunately the cheap voltage selectors have become pariahs.
>>
>> Configuring at the factory would be ok but it has to be simple, like
>> flicking a hidden switch or reversing an insert. But it (meaning also that
>> switching element) must be code-compliant, in this case UL as well as VDE
>> (CE) plus later some others, maybe Japan. Up to now it was only UL.
>>
>> Cost is critical. Adding a buck in total is ok, adding five bucks is not.
>>
>>
>>> BTW, how do you pronounce "Joerg?" :-)
>>>
>> It's o-umlaut, pronounced just like Archie Bunker and many New Yorkers say
>> "poifect" or "point".
>>
>> --
>> Regards, Joerg
>>
>> http://www.analogconsultants.com/
>>
>> "gmail" domain blocked because of excessive spam.
>> Use another domain or send PM.
>
> See if this will work for you.
> http://s936.photobucket.com/albums/ad204/rike2/
>

Thanks, Herman, I've copied down those Molex part numbers. I had
pondered a key plug before but not looked at Molex yet. The crimping
labor would have to be added to the 79c. Possibly that can be handed off
to a place that specializes in it.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
From: Joerg on
krw wrote:
> On Wed, 25 Nov 2009 11:31:18 -0800, Joerg <invalid(a)invalid.invalid>
> wrote:
>
>> krw wrote:
>>> On Wed, 25 Nov 2009 10:25:01 -0800, Joerg <invalid(a)invalid.invalid>
>>> wrote:
>>>
>>>> osr(a)uakron.edu wrote:
>>>>> Yeah, the ON circuit is crude, and the SCRs in series are a failure
>>>>> point. But its a start.
>>>>> And from a safety point of view, its got to be more of a hazard then a
>>>>> properly made jumper.
>>>>> Kinda my point.
>>>>>
>>>>> I take it this is on a linear for noise or cost reasons?
>>>>>
>>>> Two reasons. The cost of transformers is remarkably low, around five
>>>> bucks for 10-15VA. Can't beat that with a switcher, even considering
>>>> your own PWM stuff behind it because that only adds a buck or two.
>>>>
>>>> Then, temperature. Beats me but all the places I've called won't endorse
>>>> more than 50C or 60C for their switcher modules and that's just not
>>>> going to fly. For some reason my own designs never had such limits but I
>>>> can't justify the NRE for all the agency testing on this one if we
>>>> rolled our own mains-connected switcher.
>>>>
>>>> The only 250V-rated selectors I see are panel mount and insanely
>>>> expensive, such as this one:
>>>>
>>>> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=0033.1118-ND
>>>
>>> How about a wall wart? You should be able to find something like that
>>> for $10-$15.
>>
>> The only way to get one in would be to snip the end off a two-prong
>> Christmas extension code, solder it in and strap the whole kludge down
>> with cable ties. Any fire marshall would blow a gasket ;-)
>>
>> Plus I think wall warts don't like to bake in there.
>
> I was thinking about putting the wall wart elsewhere and run your
> widget off DC.
>
> How about using your 240V transformer (or dual primary in series) with
> a secondary voltage high enough at 120V to run your switcher or
> perhaps a buck-boost switcher. Let it run at 1/2V at 120V.


I'd love to but transformers operated at much lower than rated voltage
become "soft" and you can't get the rated wattage across.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
From: Joerg on
Jamie wrote:
> Joerg wrote:
>> Rich Grise wrote:
>>
>>> On Tue, 24 Nov 2009 17:48:09 -0800, Joerg wrote:
>>>
>>>> Hello Folks,
>>>>
>>>> Never needed this before but now I do: What are those things called
>>>> that
>>>> you unplug and insert 180 degrees rotated to switch the primaries of a
>>>> transformer between 115V operation (parallel) and 230V operation
>>>> (series)?
>>>> Circuit board mount.
>>>>
>>>> Or better yet, does anyone know a source? Must be super cheap, as in
>>>> <<50c
>>>> altogether in qties, as usual :-)
>>>>
>>>> Under categories such as "voltage selector" I only found fuse
>>>> holders with
>>>> this feature in there, really expensive plus way too large. I want to
>>>> avoid a slide switch because someone could accidentally hit it and ...
>>>> *PHUT*
>>>
>>>
>>> Jumper?
>>>
>>
>> Very much against code, usually. Unless it's a wire jumper but my
>> client wants to avoid having to open the units every time an order
>> comes in for the "wrong" voltage.
>>
>
> The universal supplies are such a nice idea..
>
> ~100 .. 240 volts input. etc...
>
> Some start as low as 24 volts AC/DC
>

Yes, but when you inquire about the fine print it turns out they must be
derated at rather lowish temps, starting around 120F. That kind of rules
them out in this case.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
From: Phil Hobbs on
Joerg wrote:
> Fred Bartoli wrote:
>> Joerg a �crit :
>>> Rich Grise wrote:
>>>> On Wed, 25 Nov 2009 11:31:18 -0800, Joerg wrote:
>>>>> krw wrote:
>>>>>> On Wed, 25 Nov 2009 10:25:01 -0800, Joerg <invalid(a)invalid.invalid>
>>>>>>> osr(a)uakron.edu wrote:
>>>>>>>> Yeah, the ON circuit is crude, and the SCRs in series are a failure
>>>>>>>> point. But its a start.
>>>>>>>> And from a safety point of view, its got to be more of a hazard
>>>>>>>> then a
>>>>>>>> properly made jumper.
>>>>>>>> Kinda my point.
>>>>>>>>
>>>>>>>> I take it this is on a linear for noise or cost reasons?
>>>>>>>>
>>>>>>> Two reasons. The cost of transformers is remarkably low, around five
>>>>>>> bucks for 10-15VA. Can't beat that with a switcher, even considering
>>>>>>> your own PWM stuff behind it because that only adds a buck or two.
>>>>>>>
>>>>>>> Then, temperature. Beats me but all the places I've called won't
>>>>>>> endorse more than 50C or 60C for their switcher modules and
>>>>>>> that's just
>>>>>>> not going to fly. For some reason my own designs never had such
>>>>>>> limits
>>>>>>> but I can't justify the NRE for all the agency testing on this
>>>>>>> one if
>>>>>>> we rolled our own mains-connected switcher.
>>>>>>>
>>>>>>> The only 250V-rated selectors I see are panel mount and insanely
>>>>>>> expensive, such as this one:
>>>>>>>
>>>>>>> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=0033.1118-ND
>>>>>>>
>>>>>> How about a wall wart? You should be able to find something like that
>>>>>> for $10-$15.
>>>>> The only way to get one in would be to snip the end off a two-prong
>>>>> Christmas extension code, solder it in and strap the whole kludge down
>>>>> with cable ties. Any fire marshall would blow a gasket ;-)
>>>>>
>>>>> Plus I think wall warts don't like to bake in there.
>>>>
>>>> What exactly is it you're trying to produce? You've given several
>>>> different conflicting scenarios here. Is the user supposed to be
>>>> responsible for selecting the voltage? Does it get set at the factory?
>>>> What code are you trying to satisfy?
>>>>
>>>> Please, Joerg, tell us the "big picture" here.
>>>>
>>>
>>> Ok, big picture follows:
>>>
>>> Unit in fields, outdoors, think farm. The sun is pelting it brutally
>>> all day long and the electric stuff sits inside an enclosure, no
>>> vents big enough to write home about. No choice, it has to be that
>>> way and it works fine in a 120V version.
>>>
>>> Client wants users to be able to ideally just plug it into 120 or
>>> 240, and not worry about a thing. That ain't going to fly because
>>> SMPS modules seem to not be that great. Maybe cheap electrolytics, I
>>> don't know, but mfgs don't want to endorse >60C and it sure does get
>>> hotter than that in there. Just by the sun.
>>>
>>> So, the next best avenue would be a configurable 120/240 transformer
>>> because then we can have our own (better) switcher behind that.
>>> Unfortunately the cheap voltage selectors have become pariahs.
>>>
>>> Configuring at the factory would be ok but it has to be simple, like
>>> flicking a hidden switch or reversing an insert. But it (meaning also
>>> that switching element) must be code-compliant, in this case UL as
>>> well as VDE (CE) plus later some others, maybe Japan. Up to now it
>>> was only UL.
>>>
>>> Cost is critical. Adding a buck in total is ok, adding five bucks is
>>> not.
>>>
>>
>> Have a dual 120V primaries xfrmer with a relay switched
>> series/parallel connection.
>> The relay rest position is connecting the xfrmer in 240V mode, then
>> you have all the time you need to decide what to do with the relay.
>> Cost: one relay and 10c parts.
>>
>> But I'm confident you'll find something that won't fit ;-)
>>
>
> Yeah, two things :-(
>
> One, a primary side relay adds more than those 10 cents. The other is
> that a brown-out situation would cause the relay to partially let go ...
> bzzz ... phssss ... *PHOOMP*
>
> Of course I could make a brown-out detector and all that. But things
> would need to go fast because the transformer is without power during
> the switch-over.
>
> Did you decide which theremin to build with your daughter?
>

"Partially let go"? Given that the magnetic force drops very steeply as
the iron moves away from the pole piece, I'd have thought that a very
remote contingency...maybe it just moves too slowly for the arc to be
extinguished? Please expound.

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: Joerg on
Phil Hobbs wrote:
> Joerg wrote:
>> Fred Bartoli wrote:
>>> Joerg a �crit :
>>>> Rich Grise wrote:
>>>>> On Wed, 25 Nov 2009 11:31:18 -0800, Joerg wrote:
>>>>>> krw wrote:
>>>>>>> On Wed, 25 Nov 2009 10:25:01 -0800, Joerg <invalid(a)invalid.invalid>
>>>>>>>> osr(a)uakron.edu wrote:
>>>>>>>>> Yeah, the ON circuit is crude, and the SCRs in series are a
>>>>>>>>> failure
>>>>>>>>> point. But its a start.
>>>>>>>>> And from a safety point of view, its got to be more of a hazard
>>>>>>>>> then a
>>>>>>>>> properly made jumper.
>>>>>>>>> Kinda my point.
>>>>>>>>>
>>>>>>>>> I take it this is on a linear for noise or cost reasons?
>>>>>>>>>
>>>>>>>> Two reasons. The cost of transformers is remarkably low, around
>>>>>>>> five
>>>>>>>> bucks for 10-15VA. Can't beat that with a switcher, even
>>>>>>>> considering
>>>>>>>> your own PWM stuff behind it because that only adds a buck or two.
>>>>>>>>
>>>>>>>> Then, temperature. Beats me but all the places I've called won't
>>>>>>>> endorse more than 50C or 60C for their switcher modules and
>>>>>>>> that's just
>>>>>>>> not going to fly. For some reason my own designs never had such
>>>>>>>> limits
>>>>>>>> but I can't justify the NRE for all the agency testing on this
>>>>>>>> one if
>>>>>>>> we rolled our own mains-connected switcher.
>>>>>>>>
>>>>>>>> The only 250V-rated selectors I see are panel mount and insanely
>>>>>>>> expensive, such as this one:
>>>>>>>>
>>>>>>>> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=0033.1118-ND
>>>>>>>>
>>>>>>>>
>>>>>>> How about a wall wart? You should be able to find something like
>>>>>>> that
>>>>>>> for $10-$15.
>>>>>> The only way to get one in would be to snip the end off a two-prong
>>>>>> Christmas extension code, solder it in and strap the whole kludge
>>>>>> down
>>>>>> with cable ties. Any fire marshall would blow a gasket ;-)
>>>>>>
>>>>>> Plus I think wall warts don't like to bake in there.
>>>>>
>>>>> What exactly is it you're trying to produce? You've given several
>>>>> different conflicting scenarios here. Is the user supposed to be
>>>>> responsible for selecting the voltage? Does it get set at the factory?
>>>>> What code are you trying to satisfy?
>>>>>
>>>>> Please, Joerg, tell us the "big picture" here.
>>>>>
>>>>
>>>> Ok, big picture follows:
>>>>
>>>> Unit in fields, outdoors, think farm. The sun is pelting it brutally
>>>> all day long and the electric stuff sits inside an enclosure, no
>>>> vents big enough to write home about. No choice, it has to be that
>>>> way and it works fine in a 120V version.
>>>>
>>>> Client wants users to be able to ideally just plug it into 120 or
>>>> 240, and not worry about a thing. That ain't going to fly because
>>>> SMPS modules seem to not be that great. Maybe cheap electrolytics, I
>>>> don't know, but mfgs don't want to endorse >60C and it sure does get
>>>> hotter than that in there. Just by the sun.
>>>>
>>>> So, the next best avenue would be a configurable 120/240 transformer
>>>> because then we can have our own (better) switcher behind that.
>>>> Unfortunately the cheap voltage selectors have become pariahs.
>>>>
>>>> Configuring at the factory would be ok but it has to be simple, like
>>>> flicking a hidden switch or reversing an insert. But it (meaning also
>>>> that switching element) must be code-compliant, in this case UL as
>>>> well as VDE (CE) plus later some others, maybe Japan. Up to now it
>>>> was only UL.
>>>>
>>>> Cost is critical. Adding a buck in total is ok, adding five bucks is
>>>> not.
>>>>
>>>
>>> Have a dual 120V primaries xfrmer with a relay switched
>>> series/parallel connection.
>>> The relay rest position is connecting the xfrmer in 240V mode, then
>>> you have all the time you need to decide what to do with the relay.
>>> Cost: one relay and 10c parts.
>>>
>>> But I'm confident you'll find something that won't fit ;-)
>>>
>>
>> Yeah, two things :-(
>>
>> One, a primary side relay adds more than those 10 cents. The other is
>> that a brown-out situation would cause the relay to partially let go ...
>> bzzz ... phssss ... *PHOOMP*
>>
>> Of course I could make a brown-out detector and all that. But things
>> would need to go fast because the transformer is without power during
>> the switch-over.
>>
>> Did you decide which theremin to build with your daughter?
>>
>
> "Partially let go"? Given that the magnetic force drops very steeply as
> the iron moves away from the pole piece, I'd have thought that a very
> remote contingency...maybe it just moves too slowly for the arc to be
> extinguished? Please expound.
>

It's an old lesson that I learned after building my first RF power
amplifiers (kilowatt-class). The T/R switch relay had to be controlled
with gusto when doing very fast morse code sequences where you get close
to realtime switching, else its contacts would erode rather fast. IIRC I
went through almost a dozen Potter&Brumfields over ten years back then.

A transformer presents to some (small) extent an inductive load which
likes to arc a bit on anything that mechanically opens and closes. So in
this case you'd need some circuitry that cuts the relay loose instantly,
like a threshold detector. It's all doable but at some point the costs
kind of add up.

The main problem with power relays is often when they want to come back
on. A contact that sluggishly leans into its counterpart can easily
cause a brief "bzzzt". If that happens a lot it's like a pinging engine,
it causes wear. This is why the HV guys spend so much effort to speed up
the action in their contactors. It's a whole science in itself. At the
power electrics institute of my university I had seen some really nifty
prototypes, such as contactors where a pressure capsule was used to give
them gusto. Every time they switched you heard a bang from that
pressure-assist mechanism, to the point where they required people to
wear ear protection.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.