I screwed up (....again)! Need cap help.

Started by mpm October 12, 2018
The circuit is a 52kHz simple switcher (LM2575HVT-5.0)
The current demand is 150mA (min), and 750mA (max possible), with normal being about
220 mA (upwards of 95% of the time).

The input voltage is about 50 volts (it's really a stack of four 12-volt batteries
under charge, but I bleed off a little so as not to run right up to the part's max
voltage limit).

Anyway, the datasheet calls for a 100 uF aluminum electrolytic cap on the input.
My problem is I put a 330 uF, 100-Volt, 3.5mm cap lead spacing (footprint) on the
through-hole PCB.  Classic case of the BOM not matching the footprint.

I would like to have at least a 100-volt rated cap.
CHIME IN IF YOU THINK I COULD LIVE WITH LESS THAN THAT...

About the closest I can find and still "fit" the existing board is an 80-volt, 62uF
cap.  Any chance that will work long term?  I'm trying to avoid super-height caps. I
have the vertical space, but it would stick out high enough off the board to be a
concern.

We're only talking 50 boards, not yet stuffed.  I guess I could remake them??  Ughh.
 Folding the cap over isn't really an option, though there is plenty of space to do
so.

I guess another option might be to parallel a PAIR of smaller caps, one on the
component side, and another on the solder side. (??)

On Thursday, October 11, 2018 at 11:06:38 PM UTC-4, mpm wrote:
I forgot to mention a 330 uF, 3.5mm space, 100-volt cap does not exist.
Not any neighboring uF value for that matter.  Hence my question..
On Thursday, October 11, 2018 at 11:06:38 PM UTC-4, mpm wrote:
> The circuit is a 52kHz simple switcher (LM2575HVT-5.0) > The current demand is 150mA (min), and 750mA (max possible), with normal being
about 220 mA (upwards of 95% of the time).
> > The input voltage is about 50 volts (it's really a stack of four 12-volt batteries
under charge, but I bleed off a little so as not to run right up to the part's max voltage limit).
> > Anyway, the datasheet calls for a 100 uF aluminum electrolytic cap on the input. > My problem is I put a 330 uF, 100-Volt, 3.5mm cap lead spacing (footprint) on the
through-hole PCB. Classic case of the BOM not matching the footprint.
> > I would like to have at least a 100-volt rated cap. > CHIME IN IF YOU THINK I COULD LIVE WITH LESS THAN THAT... > > About the closest I can find and still "fit" the existing board is an 80-volt,
62uF cap. Any chance that will work long term? I'm trying to avoid super-height caps. I have the vertical space, but it would stick out high enough off the board to be a concern.
> > We're only talking 50 boards, not yet stuffed. I guess I could remake them??
Ughh. Folding the cap over isn't really an option, though there is plenty of space to do so.
> > I guess another option might be to parallel a PAIR of smaller caps, one on the
component side, and another on the solder side. (??) I'm not understanding your problem. but have you searched digikey with whatever the critical parameters/ spacings are? George H.
On Thu, 11 Oct 2018 20:06:33 -0700 (PDT), mpm <mpmillard@aol.com>
wrote:

>The circuit is a 52kHz simple switcher (LM2575HVT-5.0) >The current demand is 150mA (min), and 750mA (max possible), with normal being
about 220 mA (upwards of 95% of the time).
> >The input voltage is about 50 volts (it's really a stack of four 12-volt batteries
under charge, but I bleed off a little so as not to run right up to the part's max voltage limit).
> >Anyway, the datasheet calls for a 100 uF aluminum electrolytic cap on the input. >My problem is I put a 330 uF, 100-Volt, 3.5mm cap lead spacing (footprint) on the
through-hole PCB. Classic case of the BOM not matching the footprint. What's the problem? Height? Lead spacing?
> >I would like to have at least a 100-volt rated cap. >CHIME IN IF YOU THINK I COULD LIVE WITH LESS THAN THAT...
A 50 volt aluminum cap will be fine running at 50 volts. If you want some safety margin, a 63 or 80 volt cap is even better.
> >About the closest I can find and still "fit" the existing board is an 80-volt, 62uF
cap. Any chance that will work long term? I'm trying to avoid super-height caps. I have the vertical space, but it would stick out high enough off the board to be a concern.
> >We're only talking 50 boards, not yet stuffed. I guess I could remake them??
Ughh. Folding the cap over isn't really an option, though there is plenty of space to do so.
> >I guess another option might be to parallel a PAIR of smaller caps, one on the
component side, and another on the solder side. (??) 100 uF sounds like a lot. 33 or 47 should be enough. ESR will matter, if there are no ceramics in parallel. -- John Larkin Highland Technology, Inc lunatic fringe electronics
On Friday, 12 October 2018 04:06:38 UTC+1, mpm  wrote:

> The circuit is a 52kHz simple switcher (LM2575HVT-5.0) > The current demand is 150mA (min), and 750mA (max possible), with normal being
about 220 mA (upwards of 95% of the time).
> > The input voltage is about 50 volts (it's really a stack of four 12-volt batteries
under charge, but I bleed off a little so as not to run right up to the part's max voltage limit).
> > Anyway, the datasheet calls for a 100 uF aluminum electrolytic cap on the input. > My problem is I put a 330 uF, 100-Volt, 3.5mm cap lead spacing (footprint) on the
through-hole PCB. Classic case of the BOM not matching the footprint.
> > I would like to have at least a 100-volt rated cap. > CHIME IN IF YOU THINK I COULD LIVE WITH LESS THAN THAT... > > About the closest I can find and still "fit" the existing board is an 80-volt,
62uF cap. Any chance that will work long term? I'm trying to avoid super-height caps. I have the vertical space, but it would stick out high enough off the board to be a concern.
> > We're only talking 50 boards, not yet stuffed. I guess I could remake them??
Ughh. Folding the cap over isn't really an option, though there is plenty of space to do so.
> > I guess another option might be to parallel a PAIR of smaller caps, one on the
component side, and another on the solder side. (??) If it wants 100uF for 1A, 75uF would give the same ripple/etc at 0.75A. 62uF is pretty close. Lead acids have some capacitance. No-one can guarantee it for you but I'd expect it to be fine. I don't suppose you could guarantee it either unless you test them all once stuffed. If you need the guarantee maybe you could add another 13uF somewhere. Or even select the caps that fall within the tolerance limits for 75uF. Bear in mind also the specified 100uF is for operation from who knows what source impedance. Lead acids are relatively stiff at under an amp. And operating at the top end of V_in should reduce I_in thus reducing the required C some. The datasheet says 60v max operating, 4x 12v lead acids should stay within that at all times unless unduly aggressively charged. Eliminating any input dropper should improve stiffness & enable the chip to make more use of the battery capacitance. NT
On 2018-10-12, mpm <mpmillard@aol.com> wrote:
> The circuit is a 52kHz simple switcher (LM2575HVT-5.0) > The current demand is 150mA (min), and 750mA (max possible), with normal being
about 220 mA (upwards of 95% of the time).
> > The input voltage is about 50 volts (it's really a stack of four 12-volt batteries
under charge, but I bleed off a little so as not to run right up to the part's max voltage limit). 750mA@5V or 750mA@50V? - I guessing the latter - that chip only does 1A so about 80mA avg in from a battery in 52Khz pulses - pulses are about 1.6uC so to cut the input ripple to 2% (or 1v) all you need is 1.6uF
> Anyway, the datasheet calls for a 100 uF aluminum electrolytic cap on the input.
Datasheet is probably assuming the worst case supply. you've got batteries which are near ideal voltage sources. I'd try a 4.7uF MLCC if I could find one with the right lead spacing
> I would like to have at least a 100-volt rated cap. > CHIME IN IF YOU THINK I COULD LIVE WITH LESS THAN THAT...
after 60V your chip explodes... but yeah extra voltage headroom can get extra life, There's no way you need 100uF in this application -- Notsodium is mined on the banks of denial.
On 12/10/2018 04:06, mpm wrote:
> The circuit is a 52kHz simple switcher (LM2575HVT-5.0) > The current demand is 150mA (min), and 750mA (max possible), with normal being
about 220 mA (upwards of 95% of the time).
> > The input voltage is about 50 volts (it's really a stack of four 12-volt batteries
under charge, but I bleed off a little so as not to run right up to the part's max voltage limit).
> > Anyway, the datasheet calls for a 100 uF aluminum electrolytic cap on the input. > My problem is I put a 330 uF, 100-Volt, 3.5mm cap lead spacing (footprint) on the
through-hole PCB. Classic case of the BOM not matching the footprint.
> > I would like to have at least a 100-volt rated cap. > CHIME IN IF YOU THINK I COULD LIVE WITH LESS THAN THAT... > > About the closest I can find and still "fit" the existing board is an 80-volt,
62uF cap. Any chance that will work long term? I'm trying to avoid super-height caps. I have the vertical space, but it would stick out high enough off the board to be a concern.
> > We're only talking 50 boards, not yet stuffed. I guess I could remake them??
Ughh. Folding the cap over isn't really an option, though there is plenty of space to do so.
> > I guess another option might be to parallel a PAIR of smaller caps, one on the
component side, and another on the solder side. (??)
>
Voltage derating capacitors varies a lot between the types - i.e. you might need 3X for solid tantalum. For wet aluminum electrolytics not so much - if at all. If you look at the Illinois Capacitor life-time calculator you'll see that they only apply voltage derating above 160V parts so I reckon a 80V or even 63V part will be just fine for you. The real killer and with huge derating factors is temperature. In the app you describe ESR is probably more important than actual capacitance value. piglet
Electrolytics don't need to be rated more than the voltage they see.

63VDC is fine.

Also consider a polymer type if low ESR is required (regular electrolytics 
can be quite good in smaller values, honestly; look around for both).

Tim

-- 
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Design
Website: https://www.seventransistorlabs.com/

"mpm" <mpmillard@aol.com> wrote in message 
news:2db1fa90-3872-4e22-a5c2-30385ba0189b@googlegroups.com...
The circuit is a 52kHz simple switcher (LM2575HVT-5.0)
The current demand is 150mA (min), and 750mA (max possible), with normal 
being about 220 mA (upwards of 95% of the time).

The input voltage is about 50 volts (it's really a stack of four 12-volt 
batteries under charge, but I bleed off a little so as not to run right up 
to the part's max voltage limit).

Anyway, the datasheet calls for a 100 uF aluminum electrolytic cap on the 
input.
My problem is I put a 330 uF, 100-Volt, 3.5mm cap lead spacing (footprint) 
on the through-hole PCB.  Classic case of the BOM not matching the 
footprint.

I would like to have at least a 100-volt rated cap.
CHIME IN IF YOU THINK I COULD LIVE WITH LESS THAN THAT...

About the closest I can find and still "fit" the existing board is an 
80-volt, 62uF cap.  Any chance that will work long term?  I'm trying to 
avoid super-height caps. I have the vertical space, but it would stick out 
high enough off the board to be a concern.

We're only talking 50 boards, not yet stuffed.  I guess I could remake 
them??  Ughh.  Folding the cap over isn't really an option, though there is 
plenty of space to do so.

I guess another option might be to parallel a PAIR of smaller caps, one on 
the component side, and another on the solder side. (??)