blast from past
in [Design]
|
Prev: Baxandall class D oscillator, can it produce a triangle like waveform?
Next: How to fix PS conducted interference triggers RCD?
From: Jim Thompson on 25 Jul 2010 20:29 On Sun, 25 Jul 2010 17:19:23 -0700, Joerg <invalid(a)invalid.invalid> wrote: >John Larkin wrote: [snip] >> >> The 1117 is basically an LM317 architecture with a little less >> dropout. The final pass transistor is an NPN, so the open-loop output >> impedance is relatively low. They are rugged, cheap, and stable if you >> put a tantalum or aluminum on their output. >> > >At least they state a range from 0.3ohms to 22ohms for this one instead >of some obscure graph. I'd still prefer ceramics plus resistor but >that's more real estate which in your case often matters a lot. > [snip] All these problems just to save a package pin, or do the design right in the first place. ...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... Only as good as the person behind the wheel.
From: krw on 25 Jul 2010 20:31 On Sun, 25 Jul 2010 15:10:47 -0700, Joerg <invalid(a)invalid.invalid> wrote: >krw(a)att.bizzzzzzzzzzzz wrote: >> On Sun, 25 Jul 2010 14:39:46 -0700, Joerg <invalid(a)invalid.invalid> wrote: >> >>> John Larkin wrote: >>>> I spent most of the day building this. >>>> >>>> ftp://jjlarkin.lmi.net/V220_Test_Front.jpg >>>> >>>> ftp://jjlarkin.lmi.net/V220_Test_Back.jpg >>>> >>>> which is a fixture to help test/cal a new VME module. It's powered by >>>> a 48-volt wall wart. Those old-fashioned teevee type terminal strips, >>>> and rotary switches, are pretty cool. This really deserves a more >>>> macho knob, but that's all I could find around the place without >>>> stealing one from something in the antique collection. >>>> >>>> The thing it connects to has 13 ARM processors and one FPGA on the >>>> board. Time warp. >>>> >>>> ftp://jjlarkin.lmi.net/V220_top.jpg >>>> >>> Do I see three tantalums on the bottom right? Oh-oh ... :-) >> >> Your replacement is? > > >Ceramics. Cheap. If an LDO is used (I always recommend not to do that) >and it requires a particular ESR range you can place resistors in >series. At least then you can be sure what the ESR is. Ceramics are pretty limited. I've tried eliminating our Tantalums, but have only managed to find replacements for about half of them (by part number).
From: Joerg on 25 Jul 2010 20:58 krw(a)att.bizzzzzzzzzzzz wrote: > On Sun, 25 Jul 2010 15:10:47 -0700, Joerg <invalid(a)invalid.invalid> wrote: > >> krw(a)att.bizzzzzzzzzzzz wrote: >>> On Sun, 25 Jul 2010 14:39:46 -0700, Joerg <invalid(a)invalid.invalid> wrote: >>> >>>> John Larkin wrote: >>>>> I spent most of the day building this. >>>>> >>>>> ftp://jjlarkin.lmi.net/V220_Test_Front.jpg >>>>> >>>>> ftp://jjlarkin.lmi.net/V220_Test_Back.jpg >>>>> >>>>> which is a fixture to help test/cal a new VME module. It's powered by >>>>> a 48-volt wall wart. Those old-fashioned teevee type terminal strips, >>>>> and rotary switches, are pretty cool. This really deserves a more >>>>> macho knob, but that's all I could find around the place without >>>>> stealing one from something in the antique collection. >>>>> >>>>> The thing it connects to has 13 ARM processors and one FPGA on the >>>>> board. Time warp. >>>>> >>>>> ftp://jjlarkin.lmi.net/V220_top.jpg >>>>> >>>> Do I see three tantalums on the bottom right? Oh-oh ... :-) >>> Your replacement is? >> >> Ceramics. Cheap. If an LDO is used (I always recommend not to do that) >> and it requires a particular ESR range you can place resistors in >> series. At least then you can be sure what the ESR is. > > Ceramics are pretty limited. I've tried eliminating our Tantalums, but have > only managed to find replacements for about half of them (by part number). What kind of limitations did you encounter? Nowadays you can get tens of uF in X7R at really low cost. Of course, one has to stay away from very large sizes in portable equipment or anything that can vibrate. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Tim Williams on 25 Jul 2010 22:27 "Muzaffer Kal" <kal(a)dspia.com> wrote in message news:blrp46hp87j383026vqmeefa4jgr58noqr(a)4ax.com... > Which dielectric did you find to give +80/-20% ? Z5U is often rated as such. The dismal ratings give way to equally dismal temperature and voltage coefficients, in the opposite direction (something like -80% at full rated voltage or temperature). Tim -- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms
From: John Larkin on 25 Jul 2010 23:14
On Sun, 25 Jul 2010 17:19:23 -0700, Joerg <invalid(a)invalid.invalid> wrote: >John Larkin wrote: >> On Sun, 25 Jul 2010 15:08:52 -0700, Joerg <invalid(a)invalid.invalid> >> wrote: >> >>> John Larkin wrote: >>>> On Sun, 25 Jul 2010 14:39:46 -0700, Joerg <invalid(a)invalid.invalid> >>>> wrote: >>>> >>>>> John Larkin wrote: >>>>>> I spent most of the day building this. >>>>>> >>>>>> ftp://jjlarkin.lmi.net/V220_Test_Front.jpg >>>>>> >>>>>> ftp://jjlarkin.lmi.net/V220_Test_Back.jpg >>>>>> >>>>>> which is a fixture to help test/cal a new VME module. It's powered by >>>>>> a 48-volt wall wart. Those old-fashioned teevee type terminal strips, >>>>>> and rotary switches, are pretty cool. This really deserves a more >>>>>> macho knob, but that's all I could find around the place without >>>>>> stealing one from something in the antique collection. >>>>>> >>>>>> The thing it connects to has 13 ARM processors and one FPGA on the >>>>>> board. Time warp. >>>>>> >>>>>> ftp://jjlarkin.lmi.net/V220_top.jpg >>>>>> >>>>> Do I see three tantalums on the bottom right? Oh-oh ... :-) >>>> Yes. LM1117's like tantalums on their outputs; the ESR is just right, >>>> and will hold up at -20C. They are all 22u, 10v, 11 cents each, on the >>>> outputs of 3.3, 2.5, and 1.25 volt regulators. That should be safe. >>>> >>>> Why make rules if you can't break them once in a while? >>>> >>>> The 1.25 is FPGA core voltage; that's an LM1117 with its adj pin >>>> grounded. The 2.5 is FPGA VCCaux, another LM1117 with its adj pin tied >>>> to the 1.25 rail. I would happily show the programming resistor >>>> values, except there aren't any. >>>> >>> Well, if you have good experience with the LM1117 maybe it works. I just >>> don't trust LDOs anymore, too much grief. Only one was my own design and >>> only because the client absolutely insisted on a particular LDO to be >>> used. Then, sure enough ... *PHOOF* ... *POP* >> >> The 1117 is basically an LM317 architecture with a little less >> dropout. The final pass transistor is an NPN, so the open-loop output >> impedance is relatively low. They are rugged, cheap, and stable if you >> put a tantalum or aluminum on their output. >> > >At least they state a range from 0.3ohms to 22ohms for this one instead >of some obscure graph. I'd still prefer ceramics plus resistor but >that's more real estate which in your case often matters a lot. > > >> I use them as power amps, too, driving the adj pin with an opamp. >> > >I have done that with the LM317, using it as a modulator. It doesn't get >much cheaper than that because the transmit stage needed a regulator >anyhow. So it was essentially a free modulator. And you get thermal limiting and current limiting for free. John |