> On Jan 26, 7:06 pm, Joerg <inva...@invalid.invalid> wrote:
>> Tim Wescott wrote:
>>> On Thu, 26 Jan 2012 12:42:53 -0800, Fred Bloggs wrote:
>>>> On Jan 25, 12:44 pm, Tim Wescott <t...@seemywebsite.com> wrote:
>>>>> I've got a 24V incoming supply, and a need for a number of supplies
>>>>> (+12, +5, -5, and +3.3). The +/- 5V supplies are for a bit of analog
>>>>> circuitry on a largely digital board, so they have low current
>>>>> requirements. My plan for the +5V supply is to just hang a 78xx05 off
>>>>> the +12V. +12V and +3.3V are going to come from switchers off of +24V.
>>>>> Anyone have any suggestions for a clever way to get -(something) from
>>>>> one of the positive supplies? I'm thinking that if I can get some
>>>>> negative voltage for cheap from one of the existing switchers that I
>>>>> can then regulate it to -5V with a 79xx05. I expect I'll end up using
>>>>> a charge pump -- but suggestions are welcome.
>>>>> --
>>>>> My liberal friends think I'm a conservative kook. My conservative
>>>>> friends think I'm a liberal kook. Why am I not happy that they have
>>>>> found common ground?
>>>>> Tim Wescott, Communications, Control, Circuits &
>>>>> Softwarehttp://www.wescottdesign.com
>>>> The 34063 is false economy with its low frequency operation, internal
>>>> heat dissipation, and primitive PWM architecture.
>>> I hadn't realized just how primitive its PWM architecture is until I
>>> looked closely. Now I know...
>> Yeah, it's ghastly. However ... it is one of very few chips that is
>> truly multi-sourced. And that counts for a lot. 15c a pop is also very
>> hard to beat but in my cases the multi-source argument usually matters more.
>>
>> --
>> Regards, Joerg
>>
>> http://www.analogconsultants.com/- Hide quoted text -
>>
>> - Show quoted text -
>
> It's Stone Age stuff compared to the modern offerings. Even for a
> minimalist design, you have to use L's and C's 10x larger than what's
> needed switching at a relatively modest 500KHz- more expensive parts
> and more board space.
True. However, since the advent of large numbers of Chinese inductor
manufacturers that problem has largely gone away. And board space is
rather cheap if you use ... gasp ... phenolic :-)
No kidding, I have samples of UHF stuff on phenolic here. With
microstrip and all that on there.
--
Regards, Joerg
http://www.analogconsultants.com/
Reply by legg●January 27, 20122012-01-27
On Thu, 26 Jan 2012 13:12:58 -0800, Joerg <invalid@invalid.invalid>
wrote:
>MarkK wrote:
>> "Joerg" <invalid@invalid.invalid> wrote in message
>> news:9obcq6Fh32U1@mid.individual.net...
<snip>
>> If the - (something) is low current and you are going to post regulate it,
>> and you have a switcher already operating, you can pretty simply AC couple
>> the switching off of the switcher and use 2 diodes to rectify it to a
>> negative voltage = to the P-P of the switching waveform (less the diode
>> drops). Its basically a charge pump put you already have the waveform.
>>
>> Check the full range of loads and line though.
>>
>
>I've done that when I needed a voltage above the input voltage. The idea
>was implemented but was met with disgust, grunting and scoffing at the
>design review :-)
You should have spun it as being quasiresonant and cost-free. They
would havw oohed and aawed.....
RL
Reply by Fred Bloggs●January 26, 20122012-01-26
On Jan 26, 7:06=A0pm, Joerg <inva...@invalid.invalid> wrote:
> Tim Wescott wrote:
> > On Thu, 26 Jan 2012 12:42:53 -0800, Fred Bloggs wrote:
>
> >> On Jan 25, 12:44 pm, Tim Wescott <t...@seemywebsite.com> wrote:
> >>> I've got a 24V incoming supply, and a need for a number of supplies
> >>> (+12, +5, -5, and +3.3). =A0The +/- 5V supplies are for a bit of anal=
og
> >>> circuitry on a largely digital board, so they have low current
> >>> requirements. =A0My plan for the +5V supply is to just hang a 78xx05 =
off
> >>> the +12V. =A0+12V and +3.3V are going to come from switchers off of +=
24V.
>
> >>> Anyone have any suggestions for a clever way to get -(something) from
> >>> one of the positive supplies? =A0I'm thinking that if I can get some
> >>> negative voltage for cheap from one of the existing switchers that I
> >>> can then regulate it to -5V with a 79xx05. =A0I expect I'll end up us=
ing
> >>> a charge pump -- but suggestions are welcome.
>
> >>> --
> >>> My liberal friends think I'm a conservative kook. My conservative
> >>> friends think I'm a liberal kook. Why am I not happy that they have
> >>> found common ground?
>
> >>> Tim Wescott, Communications, Control, Circuits &
> >>> Softwarehttp://www.wescottdesign.com
> >> The 34063 is false economy with its low frequency operation, internal
> >> heat dissipation, and primitive PWM architecture.
>
> > I hadn't realized just how primitive its PWM architecture is until I
> > looked closely. =A0Now I know...
>
> Yeah, it's ghastly. However ... it is one of very few chips that is
> truly multi-sourced. And that counts for a lot. 15c a pop is also very
> hard to beat but in my cases the multi-source argument usually matters mo=
It's Stone Age stuff compared to the modern offerings. Even for a
minimalist design, you have to use L's and C's 10x larger than what's
needed switching at a relatively modest 500KHz- more expensive parts
and more board space.
Reply by Joerg●January 26, 20122012-01-26
Tim Wescott wrote:
> On Thu, 26 Jan 2012 12:42:53 -0800, Fred Bloggs wrote:
>
>> On Jan 25, 12:44 pm, Tim Wescott <t...@seemywebsite.com> wrote:
>>> I've got a 24V incoming supply, and a need for a number of supplies
>>> (+12, +5, -5, and +3.3). The +/- 5V supplies are for a bit of analog
>>> circuitry on a largely digital board, so they have low current
>>> requirements. My plan for the +5V supply is to just hang a 78xx05 off
>>> the +12V. +12V and +3.3V are going to come from switchers off of +24V.
>>>
>>> Anyone have any suggestions for a clever way to get -(something) from
>>> one of the positive supplies? I'm thinking that if I can get some
>>> negative voltage for cheap from one of the existing switchers that I
>>> can then regulate it to -5V with a 79xx05. I expect I'll end up using
>>> a charge pump -- but suggestions are welcome.
>>>
>>> --
>>> My liberal friends think I'm a conservative kook. My conservative
>>> friends think I'm a liberal kook. Why am I not happy that they have
>>> found common ground?
>>>
>>> Tim Wescott, Communications, Control, Circuits &
>>> Softwarehttp://www.wescottdesign.com
>> The 34063 is false economy with its low frequency operation, internal
>> heat dissipation, and primitive PWM architecture.
>
> I hadn't realized just how primitive its PWM architecture is until I
> looked closely. Now I know...
>
Yeah, it's ghastly. However ... it is one of very few chips that is
truly multi-sourced. And that counts for a lot. 15c a pop is also very
hard to beat but in my cases the multi-source argument usually matters more.
--
Regards, Joerg
http://www.analogconsultants.com/
Reply by Tim Wescott●January 26, 20122012-01-26
On Thu, 26 Jan 2012 12:42:53 -0800, Fred Bloggs wrote:
> On Jan 25, 12:44 pm, Tim Wescott <t...@seemywebsite.com> wrote:
>> I've got a 24V incoming supply, and a need for a number of supplies
>> (+12, +5, -5, and +3.3). The +/- 5V supplies are for a bit of analog
>> circuitry on a largely digital board, so they have low current
>> requirements. My plan for the +5V supply is to just hang a 78xx05 off
>> the +12V. +12V and +3.3V are going to come from switchers off of +24V.
>>
>> Anyone have any suggestions for a clever way to get -(something) from
>> one of the positive supplies? I'm thinking that if I can get some
>> negative voltage for cheap from one of the existing switchers that I
>> can then regulate it to -5V with a 79xx05. I expect I'll end up using
>> a charge pump -- but suggestions are welcome.
>>
>> --
>> My liberal friends think I'm a conservative kook. My conservative
>> friends think I'm a liberal kook. Why am I not happy that they have
>> found common ground?
>>
>> Tim Wescott, Communications, Control, Circuits &
>> Softwarehttp://www.wescottdesign.com
>
> The 34063 is false economy with its low frequency operation, internal
> heat dissipation, and primitive PWM architecture.
I hadn't realized just how primitive its PWM architecture is until I
looked closely. Now I know...
--
My liberal friends think I'm a conservative kook.
My conservative friends think I'm a liberal kook.
Why am I not happy that they have found common ground?
Tim Wescott, Communications, Control, Circuits & Software
http://www.wescottdesign.com
Reply by Joerg●January 26, 20122012-01-26
MarkK wrote:
> "Joerg" <invalid@invalid.invalid> wrote in message
> news:9obcq6Fh32U1@mid.individual.net...
>> HardySpicer wrote:
>>> On Jan 26, 6:44 am, Tim Wescott <t...@seemywebsite.com> wrote:
>>>> I've got a 24V incoming supply, and a need for a number of supplies
> (+12,
>>>> +5, -5, and +3.3). The +/- 5V supplies are for a bit of analog
> circuitry
>>>> on a largely digital board, so they have low current requirements. My
>>>> plan for the +5V supply is to just hang a 78xx05 off the +12V. +12V
> and
>>>> +3.3V are going to come from switchers off of +24V.
>>>>
>>>> Anyone have any suggestions for a clever way to get -(something) from
> one
>>>> of the positive supplies? I'm thinking that if I can get some negative
>>>> voltage for cheap from one of the existing switchers that I can then
>>>> regulate it to -5V with a 79xx05. I expect I'll end up using a charge
>>>> pump -- but suggestions are welcome.
>>>>
>
> If the - (something) is low current and you are going to post regulate it,
> and you have a switcher already operating, you can pretty simply AC couple
> the switching off of the switcher and use 2 diodes to rectify it to a
> negative voltage = to the P-P of the switching waveform (less the diode
> drops). Its basically a charge pump put you already have the waveform.
>
> Check the full range of loads and line though.
>
I've done that when I needed a voltage above the input voltage. The idea
was implemented but was met with disgust, grunting and scoffing at the
design review :-)
--
Regards, Joerg
http://www.analogconsultants.com/
Reply by Fred Bloggs●January 26, 20122012-01-26
On Jan 25, 12:44=A0pm, Tim Wescott <t...@seemywebsite.com> wrote:
> I've got a 24V incoming supply, and a need for a number of supplies (+12,
> +5, -5, and +3.3). =A0The +/- 5V supplies are for a bit of analog circuit=
ry
> on a largely digital board, so they have low current requirements. =A0My
> plan for the +5V supply is to just hang a 78xx05 off the +12V. =A0+12V an=
d
> +3.3V are going to come from switchers off of +24V.
>
> Anyone have any suggestions for a clever way to get -(something) from one
> of the positive supplies? =A0I'm thinking that if I can get some negative
> voltage for cheap from one of the existing switchers that I can then
> regulate it to -5V with a 79xx05. =A0I expect I'll end up using a charge
> pump -- but suggestions are welcome.
>
> --
> My liberal friends think I'm a conservative kook.
> My conservative friends think I'm a liberal kook.
> Why am I not happy that they have found common ground?
>
> Tim Wescott, Communications, Control, Circuits & Softwarehttp://www.wesco=
ttdesign.com
The 34063 is false economy with its low frequency operation, internal
heat dissipation, and primitive PWM architecture.
Reply by MarkK●January 25, 20122012-01-25
"Joerg" <invalid@invalid.invalid> wrote in message
news:9obcq6Fh32U1@mid.individual.net...
> HardySpicer wrote:
> > On Jan 26, 6:44 am, Tim Wescott <t...@seemywebsite.com> wrote:
> >> I've got a 24V incoming supply, and a need for a number of supplies
(+12,
> >> +5, -5, and +3.3). The +/- 5V supplies are for a bit of analog
circuitry
> >> on a largely digital board, so they have low current requirements. My
> >> plan for the +5V supply is to just hang a 78xx05 off the +12V. +12V
and
> >> +3.3V are going to come from switchers off of +24V.
> >>
> >> Anyone have any suggestions for a clever way to get -(something) from
one
> >> of the positive supplies? I'm thinking that if I can get some negative
> >> voltage for cheap from one of the existing switchers that I can then
> >> regulate it to -5V with a 79xx05. I expect I'll end up using a charge
> >> pump -- but suggestions are welcome.
> >>
> >>>
>
If the - (something) is low current and you are going to post regulate it,
and you have a switcher already operating, you can pretty simply AC couple
the switching off of the switcher and use 2 diodes to rectify it to a
negative voltage = to the P-P of the switching waveform (less the diode
drops). Its basically a charge pump put you already have the waveform.
Check the full range of loads and line though.
Mark
Reply by Martin Riddle●January 25, 20122012-01-25
"John Larkin" <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote in
message news:4fl0i7psgj6um168fi77ri84vo42ucsjdq@4ax.com...
> On Wed, 25 Jan 2012 11:44:25 -0600, Tim Wescott <tim@seemywebsite.com>
> wrote:
>
>>I've got a 24V incoming supply, and a need for a number of supplies
>>(+12,
>>+5, -5, and +3.3). The +/- 5V supplies are for a bit of analog
>>circuitry
>>on a largely digital board, so they have low current requirements. My
>>plan for the +5V supply is to just hang a 78xx05 off the +12V. +12V
>>and
>>+3.3V are going to come from switchers off of +24V.
>>
>>Anyone have any suggestions for a clever way to get -(something) from
>>one
>>of the positive supplies? I'm thinking that if I can get some
>>negative
>>voltage for cheap from one of the existing switchers that I can then
>>regulate it to -5V with a 79xx05. I expect I'll end up using a charge
>>pump -- but suggestions are welcome.
>
> You can charge pump off the switcher node of one of the positive
> switchers to get negative volts, essentially a classic "voltage
> doubler" capacitor-diode thing. It will make roughly -24, which is a
> lot to regulate to -5. A small series resistor is a good idea... some
> switchers don't like this being done to them.
>
> But if it's low current, you may as well buy a standard charge-pump
> voltage inverter chip and run that off +5.
>
> The LTC "LTM" module series can be used to make -5 from +24. Not real
> cheap, but easy. Or build a Cuk switcher.
>
> Another possibility: buy one of those cheap SIP dc/dc converters, $4
> roughly.
>
> John
>
>
>
If there is board space a MAX232 can be used for it's +/- 10 v charge
pump.
Cheers
Reply by Joerg●January 25, 20122012-01-25
HardySpicer wrote:
> On Jan 26, 6:44 am, Tim Wescott <t...@seemywebsite.com> wrote:
>> I've got a 24V incoming supply, and a need for a number of supplies (+12,
>> +5, -5, and +3.3). The +/- 5V supplies are for a bit of analog circuitry
>> on a largely digital board, so they have low current requirements. My
>> plan for the +5V supply is to just hang a 78xx05 off the +12V. +12V and
>> +3.3V are going to come from switchers off of +24V.
>>
>> Anyone have any suggestions for a clever way to get -(something) from one
>> of the positive supplies? I'm thinking that if I can get some negative
>> voltage for cheap from one of the existing switchers that I can then
>> regulate it to -5V with a 79xx05. I expect I'll end up using a charge
>> pump -- but suggestions are welcome.
>>
>> --
>> My liberal friends think I'm a conservative kook.
>> My conservative friends think I'm a liberal kook.
>> Why am I not happy that they have found common ground?
>>
>> Tim Wescott, Communications, Control, Circuits & Softwarehttp://www.wescottdesign.com
>
> Unless you are looking for really low ripple then buck/boost is the
> modern way to go for everything. Far more efficient - up in the 95%
> odd.
>
WRT ripple the MC34063 is absolutely horrid because of its regulating
method. But sometimes it's good enough. I wouldn't use it in sensitive
anlog apps.
--
Regards, Joerg
http://www.analogconsultants.com/