Troubleshooting RFI from switch-mode PS

I have a Samlex 120VAC-13.5VDC at 35A switch mode power supply for my amateur radio equipment. Working great for last 7 years. Recently it's now producing "swooshing" noise in my radio receiver, especially on frequencies from 1.8-10MHz. Took it apart and noticed blown tops on the 3 output filter electrolytics, so I replaced those. But that did not help.

I could replace the 2 larger input filter electrolytics but they "look" OK and I wonder if maybe I should be thinking of also replacing the switcher MOSFET - I don't know where else the noise could be coming from?

Any ideas greatly appreciated.

 

On Sunday, November 20, 2016 at 2:49:41 PM UTC-5, laga...@gmail.com wrote:

I have a Samlex 120VAC-13.5VDC at 35A switch mode power supply for my amateur radio equipment. Working great for last 7 years. Recently it's now producing "swooshing" noise in my radio receiver, especially on frequencies from 1.8-10MHz. Took it apart and noticed blown tops on the 3 output filter electrolytics, so I replaced those. But that did not help.

I could replace the 2 larger input filter electrolytics but they "look" OK and I wonder if maybe I should be thinking of also replacing the switcher MOSFET - I don't know where else the noise could be coming from?

Any ideas greatly appreciated.

I've never seen a power mosfet do anything but work normally or become zero ohm jumpers.

Are you sure it's the supply? Put a scope to the output and see what's going on. As far as the capacitors "looking" OK, remember that electrolytics need not be vented or bulged to be bad. Either put an ESR meter to *all* of them in the supply (including the primary) or scope them (preferred method).

 

[Charles]

wrote in message
news:0f362b7e-4756-4027-be47-79014bc937f5@googlegroups.com...

I have a Samlex 120VAC-13.5VDC at 35A switch mode power supply for my
amateur radio equipment. Working great for last 7 years. Recently it's now
producing "swooshing" noise in my radio receiver, especially on frequencies
from 1.8-10MHz. Took it apart and noticed blown tops on the 3 output filter
electrolytics, so I replaced those. But that did not help.

I could replace the 2 larger input filter electrolytics but they "look" OK
and I wonder if maybe I should be thinking of also replacing the switcher
MOSFET - I don't know where else the noise could be coming from?

Any ideas greatly appreciated.

If the supply is normal in all other ways, it could be that your replacement
caps are not suitable. Filtering switcher hash requires very low esr caps.

 

[whit3rd]

On Sunday, November 20, 2016 at 11:49:41 AM UTC-8, laga...@gmail.com wrote:

I have a Samlex 120VAC-13.5VDC at 35A switch mode power supply for my amateur radio equipment. Working great for last 7 years. Recently it's now producing "swooshing" noise in my radio receiver, especially on frequencies from 1.8-10MHz. Took it apart and noticed blown tops on the 3 output filter electrolytics, so I replaced those. But that did not help.

I could replace the 2 larger input filter electrolytics but they "look" OK and...

The fast rectifiers on the low voltage side can fail/become leaky, and when they
do, they can take out capacitors. Check those.

 

[Jeff Liebermann]

On Sun, 20 Nov 2016 11:49:38 -0800 (PST), lagagnon@gmail.com wrote:

I have a Samlex 120VAC-13.5VDC at 35A switch mode power supply for my amateur radio equipment. Working great for last 7 years. Recently it's now producing "swooshing" noise in my radio receiver, especially on frequencies from 1.8-10MHz. Took it apart and noticed blown tops on the 3 output filter electrolytics, so I replaced those. But that did not help.

I could replace the 2 larger input filter electrolytics but they "look" OK and I wonder if maybe I should be thinking of also replacing the switcher MOSFET - I don't know where else the noise could be coming from?

Any ideas greatly appreciated.

I can't tell from here where the noise is coming from. Usually, I put
a small loop antenna on my spectrum analyzer and save it around the
power supply while under load. That usually locates the general area.

Much more interesting is why three capacitors would decide to bulge.
Bulging is usually caused by excessive ripple current through the
filter capacitors. High ripple could be coming from the diodes, or
simply from undersized capacitors that can't hold a charge between
switching cycles and therefore draw more current than normal.

I suggest you put a scope on the output while under load and see if
there's any ripple present. Compare the value with the manufactures
specs. Also, use an IR thermometer to measure the case temperature of
the capacitors. If they are getting warm, then there's ripple current
going through them.

--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

 

[pfjw@aol.com]

On Sunday, November 20, 2016 at 2:49:41 PM UTC-5, laga...@gmail.com wrote:

I have a Samlex 120VAC-13.5VDC at 35A switch mode power supply for my amateur radio equipment. Working great for last 7 years. Recently it's now producing "swooshing" noise in my radio receiver, especially on frequencies from 1.8-10MHz. Took it apart and noticed blown tops on the 3 output filter electrolytics, so I replaced those. But that did not help.

I could replace the 2 larger input filter electrolytics but they "look" OK and I wonder if maybe I should be thinking of also replacing the switcher MOSFET - I don't know where else the noise could be coming from?

Any ideas greatly appreciated.

a) Replace all the caps. Unless you have both X-ray eyes and an ESR meter, you have no effective way of discerning whether they are good or not.
b) If it works otherwise - that is, produces the correct voltage - unlikely that the MOSFet is bad.
c) It becomes a close-run thing whether the cost/time involved in repairs exceeds the value of the device if it is more than just the caps.

Peter Wieck
Melrose Park, PA

 

On Monday, November 21, 2016 at 1:29:40 AM UTC-5, Jeff Liebermann wrote:

Much more interesting is why three capacitors would decide to bulge.


--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

In the last 15 years, I've changed literally thousands of bulged electrolytics in consumer electronics for no other reason except they were just crappy parts at the end of their limited lifetimes. Most of those that bulge or vent are higher value types, say those beyond 330uf or so. Smaller electros fail at the same rate of their big brothers but don't generally visually vent. In the old days, capacitors didn't bulge unless they were tortured. Today it's mediocre chemistry that causes this, not circuit issues.

If you were to take apart 10 working flat screen TVs made in the last few years, half of them would show some sign of bulged or top vented capacitors despite running normally.

 

[pfjw@aol.com]

On Monday, November 21, 2016 at 1:07:20 PM UTC-5, Jeff Liebermann wrote:

SNIPPAGE
Ok, time to vent...

I blame CAD (computer aided design) for the problem. CAD has allowed
designers to accurately predict the life of some components including
electrolytics. If the operating parameters are known (voltage, temp,
ripple current, starting ESR, etc), the lifetime can be calculated.
Having a component last longer than the warranty period is considered
wasteful by many manufacturers. Therefore, they select the cheapest
possible part that will work up to when the warranty expires but no
longer. The result is a product where every component blows up at the
same time, just after the warranty expires. The 10 flat screens in
your example will not show one bad component. Instead, there will be
dozens, all different brands and values. This is not due to crappy
manufacturing, but rather due to careful selection for the MINIMUM
lifetime of the component.

I would refine this somewhat. When this happens, the average consumer will change brands as most (well, not the typical Trump supporter) recognize that something failing just after the warranty period is the manufacturer wanting to maintain the demand stream. I have found that such electronics - based on average hours used and other factors - tent to last about 25% beyond the warranty period - that is, if they last the first six weeks. Just long enough for the consumer to forget the warranty period.

There are exceptions: We have a Sony CRT-type TV purchased in Saudi in 2002 that has probably 13,000 hours on it that has survived two moves, two ocean voyages, kids, cats, dogs and such without a hitch. The Panasonic Plasma with about 6,000 hours on it is doing fine (so far) as well. Frigidaire refrigerators have about 5 years + one month before they crap out. GE units of US origin (in our experience) are pretty bullet proof. Bosch appliances, also, are pretty bullet-proof. It's all according to a mix of decent manufacture and pure blind luck.

Peter Wieck
Melrose Park, PA

 

[pfjw@aol.com]

Live fast.
Die young.
Leave a good looking corpse.

Peter Wieck
Melrose Park, PA

 

[Jeff Liebermann]

On Sun, 20 Nov 2016 22:29:35 -0800, Jeff Liebermann <jeffl@cruzio.com>
wrote:

I can't tell from here where the noise is coming from. Usually, I put
a small loop antenna on my spectrum analyzer and save it around the
power supply while under load. That usually locates the general area.

Is this the one or is there some reason you didn't bother disclosing
the model number? Notice that the SEC-1235 is a 12V 30A device and
current limits at 35A. How many amps are you drawing?
<http://www.samlexamerica.com/products/ProductDetail.aspx?pid=66>

See page 10 for RFI information:
<http://www.samlexamerica.com/documents/manuals/11001-SEC-1235-1235M-1114.pdf>

--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

 

[Jeff Liebermann]

On Mon, 21 Nov 2016 08:46:43 -0800 (PST), ohger1s@gmail.com wrote:

In the last 15 years, I've changed literally thousands of
bulged electrolytics in consumer electronics for no other
reason except they were just crappy parts at the end of
their limited lifetimes. Most of those that bulge or vent
are higher value types, say those beyond 330uf or so.
Smaller electros fail at the same rate of their big
brothers but don't generally visually vent.

I know the feeling:
<http://802.11junk.com/jeffl/pics/repair/slides/bad-caps.html>
That's about a 5 month collection, mostly from motherboards, power
supplies, and LCD monitors. I've cut back on the component level
repairs in the last year, so this years collection is about 1/3 the
size. I have 4 different ESR testers and use them often.

In the old days, capacitors didn't bulge unless they were
tortured. Today it's mediocre chemistry that causes this,
not circuit issues.

In the bad old daze, the electrolytics were running at 60Hz, not
60KHz. The power (heat) dissipated in the electrolytics is a function
of frequency.

If you were to take apart 10 working flat screen TVs made
in the last few years, half of them would show some sign of
bulged or top vented capacitors despite running normally.

Ok, time to vent...

I blame CAD (computer aided design) for the problem. CAD has allowed
designers to accurately predict the life of some components including
electrolytics. If the operating parameters are known (voltage, temp,
ripple current, starting ESR, etc), the lifetime can be calculated.
Having a component last longer than the warranty period is considered
wasteful by many manufacturers. Therefore, they select the cheapest
possible part that will work up to when the warranty expires but no
longer. The result is a product where every component blows up at the
same time, just after the warranty expires. The 10 flat screens in
your example will not show one bad component. Instead, there will be
dozens, all different brands and values. This is not due to crappy
manufacturing, but rather due to careful selection for the MINIMUM
lifetime of the component.

--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

 

[Phil Allison]

Jeff Liebermann wrote:

In the bad old daze, the electrolytics were running at 60Hz, not
60KHz. The power (heat) dissipated in the electrolytics is a function
of frequency.

** The heat loss in an electro is related to ripple current and ESR at the frequency in use. ESR is much HIGHER at 60Hz or 120Hz than it is at 60kHz - so losses are LOWER at high frequencies.

Also, ESR goes down with increasing temperature - by up 5 times at the max rated temp of the cap.

Running in a hot environment is the main cause of short lifetimes, eg with nearby heatsinks and no fan.




...... Phil

 

On Monday, November 21, 2016 at 1:07:20 PM UTC-5, Jeff Liebermann wrote:

In the bad old daze, the electrolytics were running at 60Hz, not
60KHz. The power (heat) dissipated in the electrolytics is a function
of frequency.

They've been using high frequency smps since the late 70s (Sony was the first that I can recall (and boy were they *squirrely*). And while we replaced capacitors on the wholesale level even back then, they didn't bulge. The vented cap syndrome (VCS?) has really only been an issue since about 2004 or so. Around 2008 was the real big explosion of top bulged caps, when we'd see TVs in warranty with bulged/vented caps. There's been a big slowdown in bulged caps in recent years.

 

[Dave Platt]

In article <l3d63c93jobjuht7hpsd04i5fghuf66pu4@4ax.com>,
Jeff Liebermann <jeffl@cruzio.com> wrote:

Having a component last longer than the warranty period is considered
wasteful by many manufacturers. Therefore, they select the cheapest
possible part that will work up to when the warranty expires but no
longer. The result is a product where every component blows up at the
same time, just after the warranty expires. The 10 flat screens in
your example will not show one bad component. Instead, there will be
dozens, all different brands and values. This is not due to crappy
manufacturing, but rather due to careful selection for the MINIMUM
lifetime of the component.

That's an "honorable" American tradition that dates back at least a
century. The story goes that Henry Ford has his engineers buy and
disassemble scrapped Ford cars, to see which parts had failed and
which ones were still in good shape. He pointed out that the company
could often save money by redesigning (more cheaply) parts which
rarely failed... so the new versions would wear out and fail at about
the same time as everything else.

I remember one of the Lord Darcy stories by Randall Garrett, which
involved a (magic-based) process somewhat like this for the human
body. The spells would draw upon the health of long-lived parts of
the body, to support those which were more prone to obvious aging.

The recipient of this treatment might have the apparent health and
vigor of one in his thirties, and sustain this for several decades
(e.g. up until the age of 60 or so). At that point, all of the body
systems would collapse simultaneously, and the patient would die of
old age in the space of an hour or two.

 

[Phil Allison]

Jeff Liebermann wrote:

Notice that the SEC-1235 is a 12V 30A device and
current limits at 35A. How many amps are you drawing?

** Transceivers tend to draw a LOT LESS current while in receive mode - which is the only time PSU generated RF noise might be a problem.

SMPSs often send a fair amount of RFI back up the AC power cable, even when running at low power outputs which makes them troublesome when used with radio receivers - especially in the lower part of the HF band.

The OP's supply may have deteriorated components ( plastic film caps most likely ) in the filter fitted to the incoming AC line. Only way to know for sure is to get another, similar supply and try it.

If the OP has a linear 12V supply that can run the transceiver in receive mode, he should try that too.



...... Phil

 

I will respond to some of the posts above, without discussing the side topic of poorly manufactured capacitors etc.

1) the PS is rated at 35A, but I rarely draw more than 18A - and that is not a continuous draw.
2) I run the PS in a cool room, and also as a result of point 1 above the cooling fan has never come one (but has been tested to work). In other words the PS runs very cool
3) nothing has changed for 7 years until recently so this is definitely recent noise coming from the PS as I can switch to a deep cycle battery instantly and the RX noise disappears.

I have ordered the two 1000uF, 200V input caps and should replace them in about 3 days and will let you all know the results.

 

[Dave Platt]

In article <9234021a-116c-431f-ba5b-7cb11400d692@googlegroups.com>,
<ohger1s@gmail.com> wrote:

The vented cap syndrome
(VCS?) has really only been an issue since about 2004 or so. Around 2008 was the
real big explosion of top bulged caps, when we'd see TVs in warranty with
bulged/vented caps. There's been a big slowdown in bulged caps in recent years.

Sounds like this might be the "capacitor plague" problem? Bad
water-based electrolyte?

https://en.wikipedia.org/wiki/Capacitor_plague

 

[Jeff Liebermann]

On Mon, 21 Nov 2016 19:20:43 -0800 (PST), lagagnon@gmail.com wrote:

I will respond to some of the posts above, without discussing
the side topic of poorly manufactured capacitors etc.

Sorry for the topic drift, but my intuition tells me that the noise is
somehow related to the bulging caps.

1) the PS is rated at 35A, but I rarely draw more than 18A -
and that is not a continuous draw.

That's nice. Now, for the 2nd time, will you kindly disclose the
model number of the power supply. Is it this one?
<http://www.samlexamerica.com/products/ProductDetail.aspx?pid=66>

2) I run the PS in a cool room, and also as a result of point
1 above the cooling fan has never come one (but has been tested
to work). In other words the PS runs very cool

I didn't suggest that the power supply was overheating. I suggested
that the capacitors might be overheating from too much ripple current.
I also suggested that you use an IR thermometer to measure the case
temperature of the capacitors.

3) nothing has changed for 7 years until recently so this is
definitely recent noise coming from the PS as I can switch to a
deep cycle battery instantly and the RX noise disappears.

Nothing has changed is what all my customers say. Nothing means
nothing that is obvious to them. In most cases, that's a polite way
for them to suggest that I not pry too deeply into what the customer
has been doing with the device. If I do pry, I often find
questionable practices. I had one like that 2 days ago. Nothing had
changed in years before the computer crapped out. Too bad the fan and
most of the insiders were packed with cat hair.

I have ordered the two 1000uF, 200V input caps and should replace
them in about 3 days and will let you all know the results.

I rarely have any problems with the input caps. It's the lower
voltage caps that blow.

--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

 

On Tuesday, November 22, 2016 at 12:15:58 AM UTC-5, Jeff Liebermann wrote:

I rarely have any problems with the input caps. It's the lower
voltage caps that blow.

Not common, but not unheard of. I've replaced a bunch of 450V PFC capacitors in smps that don't show signs of venting, but they're dead and when removed, feel like half the weight of what they should be. They also "rattle" inside as the assy inside has shrunk.

 

On Monday, 21 November 2016 21:15:58 UTC-8, Jeff Liebermann wrote:

That's nice. Now, for the 2nd time, will you kindly disclose the
model number of the power supply. Is it this one?
http://www.samlexamerica.com/products/ProductDetail.aspx?pid=66

Yes, that's the one. Does that make a difference to your assessment?

I didn't suggest that the power supply was overheating. I suggested
that the capacitors might be overheating from too much ripple current.

OK, I understand, and I wonder whether the output caps which I've replaced have
blown because the input caps are not as effective at removing ripple?

... If I do pry, I often find
questionable practices. I had one like that 2 days ago. Nothing had
changed in years before the computer crapped out. Too bad the fan and
most of the insiders were packed with cat hair.

How is that relevant to my post and my situation?

I rarely have any problems with the input caps. It's the lower
voltage caps that blow.

See my suggestion above....