Maker Pro
Maker Pro

failure of the low-ESR electrolytic capacitor in second winding

E

Eeyore

Jan 1, 1970
0
Winfield said:
What's your Vcc voltage and load current?

He seems reluctant to talk about this. There are 2 x ER3B diodes feeding this
output. These are 3 A diodes. That suggests an output current in excess of 3A.

This doesn't bode well where the capacitor only has around a 1A ripple current
rating.

The reason for the caps failing seems pretty clear to me now.

Graham
 
E

Eeyore

Jan 1, 1970
0
Winfield said:
Phil's suggestion of a bad electrolyte is sadly a real
possibility. But this scene has been most common with
inexpensive substitutes for premium-grade conductive
polymer or organic-semiconductor high-conductivity-
electrolyte capacitors, like Sanyo's OS-CON types.

Eh ?

The 'bad caps' were standard low-ESR electrolytics.

Gigabyte has even gone as far as moving entirely to polymer types for its
motherboards to AVOID the suspicions of bad caps.

Graham
 
L

legg

Jan 1, 1970
0
I met a big problem that the DC-DC converter circuit failure happend
frequently

I use UC3843B to control the Planar TRf 101 transformer to provide
multiply DC output;

and almost every failure has the same feature:

The failure on the low-ESR electrolytic capacitor at the second wiring
of the transformer;

it's kind of Low-ESR electrolytic capacitor, 1000uf/10V (ESR=0.07
ohm).

The UC3843B working at 100KHZ.


Is there anybody met that problem before?

What's the possible reason to lead to the failure?
If these things aren't burned in or the product of carefully evaluated
and tested design, the caps could be inserted reverse polarity. Even
the artwork could be wrong, so everything looks ok 'per dwg'.

Electrolytics don't always fail immediately or catastrophically when
mis-applied in low voltage circuitry. With your double section filter,
there need be no immediate indication of misbehavior on the outputs,
if you're measuring ripple/regulation at final test. Cross-regulation
might show an exageration or shifted centering.

I've seen batches of capacitors fail because they were mislabelled.
There's no defence against this other than burn-in and supplier
vetting

On multiple output flybacks, you are at the mercy of the end-user when
it comes to operating stress - end-use loading can grossly exceed the
(rapparent) 2A ouput rating of this circuit. Assuming a flyback as
it's a capacitive input filter and a single ended switch controller.

I'm surprised to see the relative distribution of the components - the
bulk of the capacitance budget should be in the first stage of a
flyback output.

RL
 
J

Jamie

Jan 1, 1970
0
Eeyore said:
Jamie wrote:




Classic Jamie nonsense !
Really.

I've never seen anything original from you to date.

Every thing you post is just a retake from a few here that
really do know their profession or x profession.

I notice how you try to take credit for other's work. You
conveniently step in after the answer is obvious and make a
mockery of it all and then try to stake a claim.

You can fool the unknowing, but all are not as you see perceive
them. (Fools). The only fool here is?
 
E

Eeyore

Jan 1, 1970
0
Jamie said:

Yes !

I've never seen anything original from you to date.

I haven't seen anything from you that shows even the tiniest sign of knowledge
or experience of electronics to date.

Shorted electrolytic capacitor my arse !

Do please explain the shorting mechasnism won't you ?

Graham
 
S

Simon

Jan 1, 1970
0
Thanks Eeyore and Jamie,

You two gave me some start point to re-evaluate my
problem.

Actually, this product had been stable (with same design,
but the capacitor changed due to additional connector size
requirement) for 3 to 4 years.

Only lately from last year, began to met these problems.

Thanks and hope you two have good day.
 
S

Simon

Jan 1, 1970
0
Thanks Windfield.

1) VCC is around 5 V
2) According to the transformer requirment, the curren
will be less than 1.7ADC (here, we depart to 0.3A + 1.5A (1.5 A is the
trip value of fuse, not the practical load) )
3) I think the circuit edited is right
4) UC383B working frequency 100KHZ

5) 1040 mA is the ripple current rating of that EXR
capacitor, not the real ripple current, i need to measure that

6) I need to measure the voltage ripple on the capacitor.

7) How to calcurate the needed ripple current rating, if
say the voltage ripple is some value, say, 60mV?


Thanks

Simon
 
R

Robert Baer

Jan 1, 1970
0
Simon said:
I met a big problem that the DC-DC converter circuit failure happend
frequently

I use UC3843B to control the Planar TRf 101 transformer to provide
multiply DC output;

and almost every failure has the same feature:

The failure on the low-ESR electrolytic capacitor at the second wiring
of the transformer;

it's kind of Low-ESR electrolytic capacitor, 1000uf/10V (ESR=0.07
ohm).

The UC3843B working at 100KHZ.


Is there anybody met that problem before?

What's the possible reason to lead to the failure?


Some body may say because of the high temperature, but it happends on
the product even working in less than 30 cent degree ambient operation
temperature.
Put a current sensor in series with the capacitor and then you will
know the answer (think about it, and you should then know the answer).
 
W

Winfield

Jan 1, 1970
0
Thanks Windfield.

1) VCC is around 5 V
2) According to the transformer requirment, the curren
will be less than 1.7ADC (here, we depart to 0.3A + 1.5A (1.5 A is the
trip value of fuse, not the practical load) )
3) I think the circuit edited is right
4) UC383B working frequency 100KHZ

5) 1040 mA is the ripple current rating of that EXR
capacitor, not the real ripple current, i need to measure that

6) I need to measure the voltage ripple on the capacitor.

7) How to calcurate the needed ripple current rating, if
say the voltage ripple is some value, say, 60mV?

Thanks

Simon

The ripple current is equal to your load current.

I had stated, "This would limit your supply to well
under 2 amps, say 0.75 to 1.0A max." in which I used
a 2x safety margin. But I didn't know about your use,
and I failed to take into account the greater safety
margin needed for an extended operating lifetime, as
Spehro Pefhany pointed out. It appears your failures
are due to pushing the capacitors beyond their ability,
rather than bad parts. You need more serious caps.
 
E

Eeyore

Jan 1, 1970
0
Simon said:
Thanks Windfield.

1) VCC is around 5 V
2) According to the transformer requirment, the curren
will be less than 1.7ADC (here, we depart to 0.3A + 1.5A (1.5 A is the
trip value of fuse, not the practical load) )

With a load current of 1.5 - 1.7A, your 2 x 1000uF capacitors connected in
parallel will not have an adequate ripple current rating.

The ripple current is GREATER than the load current. The exact value depends on
the waveform but a multiplier of say 1.6 - 2 ought to allow you to determine a
suitable rating.

So, each output cap (assuming you continue to use 2 in parallel) would need a
ripple current rating of at least 1300 - 1700 mA.

An alternative might be to use 3 of them in parallel. Or use a better brand
capacitor.

Graham
 
S

Simon

Jan 1, 1970
0
With a load current of 1.5 - 1.7A, your 2 x 1000uF capacitors connected in
parallel will not have an adequate ripple current rating.

The ripple current is GREATER than the load current. The exact value depends on
the waveform but a multiplier of say 1.6 - 2 ought to allow you to determine a
suitable rating.

So, each output cap (assuming you continue to use 2 in parallel) would need a
ripple current rating of at least 1300 - 1700 mA.

An alternative might be to use 3 of them in parallel. Or use a better brand
capacitor.

Graham



Dear Eeyore and WinField,

Thanks for your help!

I am sure i totally understand the situation now.


Thanks and don't forget to enjoy the coming weekend

Simon
 
L

legg

Jan 1, 1970
0
Thanks Windfield.

1) VCC is around 5 V
2) According to the transformer requirment, the curren
will be less than 1.7ADC (here, we depart to 0.3A + 1.5A (1.5 A is the
trip value of fuse, not the practical load) )
3) I think the circuit edited is right
4) UC383B working frequency 100KHZ

5) 1040 mA is the ripple current rating of that EXR
capacitor, not the real ripple current, i need to measure that

6) I need to measure the voltage ripple on the capacitor.

7) How to calcurate the needed ripple current rating, if
say the voltage ripple is some value, say, 60mV?

The ripple current for a 1A7 output flyback will depend on duty cycle
and secondary inductance. If you're quoting measured secondary rms
current, however, then the capacitor rms ripple current should equal:

Icaprms = ( Isecrms^2 ­ Ioutdc^2 )^0.5

...............and the actual output current will be less than 1A7
under those conditions.

Note that the ripple current in the capacitor is not greater than the
output current, but the ripple current in the transformer winding is.
Perhaps this is where some of the other responders are confused.

You should make the measurements at high line, full load, as this
gives the worst case output capacitor ripple in mixed-mode circuits,
with the output current being more discontinuous in nature.

If the circuit is unstable, or cycle-skips, the output ripple current
will also increase. It doesn't look like you're worried about
cryogenic conditions or stability.

I think most of this is right - I tend to have to look it up anyways
nowadays.

RL
 
L

legg

Jan 1, 1970
0
Win !

Wash your mouth out.
This is a reasonably assumed limit.

In fact it's always less.

It's the winding rms current that exceeds the output current.

Time for a refresher, buddy.

RL
 
E

Eeyore

Jan 1, 1970
0
legg said:
Note that the ripple current in the capacitor is not greater than the
output current

The *average* current may be the same, but the RMS value most certainly is NOT
the same.

It's the RMS value that matters here.

Graham
 
E

Eeyore

Jan 1, 1970
0
legg said:
This is a reasonably assumed limit.

In fact it's always less.

Utter and complete nonsense.

It's the winding rms current that exceeds the output current.

Time for a refresher, buddy.

You first sunshine.

What I stated is 100% correct.

I'm staggered quite frankly that so few people here know how to design something
as simple as a PSU competently.

I mean SERIOUSLY ! Not to know how to determine the required ripple current
rating of a capacitor ? Bizarre.

Graham
 
S

Stig Carlsson

Jan 1, 1970
0
What happens in a no-load condition? Does the voltage across the capacitor
go above 10V?
In the company where I worked some years back there was a design of a
flyback converter that did just this, increased the voltage with no load
connected.

Stig
 
W

Winfield Hill

Jan 1, 1970
0
Eeyore said:
Utter and complete nonsense.


You first sunshine. What I stated is 100% correct.

I'm staggered quite frankly that so few people here know
how to design something as simple as a PSU competently.
I mean SERIOUSLY ! Not to know how to determine the
required ripple current rating of a capacitor ? Bizarre.

We don't know what kind of smps Simon has, do we?

But, taking for example a boost converter (or an
equivalent flyback), operating in continuous mode
(which is much more efficient than discontinuous),
RMS current = 1.0 * load current if the duty cycle
is 50%. It's lower for duty cycles higher than 50%,
and higher below 50%, e.g., rising to 1.5 Iout for
D = 25%, according to a quick notepad calculation.

But one thing, since power loss goes as I^2 R, a
low duty cycle, with very high charging currents
(e.g. 3x load current for D = 25%), will actually
cause more heating than the simple RMS calculation
implies, and means a greater safety margin should
be required.
 
E

Eeyore

Jan 1, 1970
0
Winfield said:
We don't know what kind of smps Simon has, do we?

From the power level I'd expect it to be a flyback type. He talks of a second
winding which I take to mean secondary winding. It certainly looks like one to
me.

You edited the circuit yourself and Simon seems to think it's right. It has the
secondary winding feeding a rectifier (actually 2 in parallel) and this charges
up the 2 x 1000 uF storage caps. This WILL have an RMS ripple current greater
than the DC load current.

The filter that follows this ( with 200 uF ) will not take much burden off the
main reservoir caps.

But, taking for example a boost converter (or an
equivalent flyback), operating in continuous mode
(which is much more efficient than discontinuous),
RMS current = 1.0 * load current if the duty cycle
is 50%. It's lower for duty cycles higher than 50%,
and higher below 50%, e.g., rising to 1.5 Iout for
D = 25%, according to a quick notepad calculation.

But one thing, since power loss goes as I^2 R, a
low duty cycle, with very high charging currents
(e.g. 3x load current for D = 25%), will actually
cause more heating than the simple RMS calculation
implies, and means a greater safety margin should
be required.

Pardon. I thought the whole point of RMS is that the I2R heating effect will be
accurately represented.

Graham
 
Top