Hi, WOuld some kind soul help me diagnose my switch mode power supplies?
Stephen
I have three identical Headend UHF channelised amplifiers that all use
identical switch mode power supplies. One is working, the other two are
faulty but appear to have an identical fault as the output voltages are the
same low values on both units.
So I've got my screwdrivers out and a multimeter. Upon taking the backs of
all three units there is a 8 wire ribbon lead between the SMPS board and the
main board, comprising of 1 black wire and 7 grey ones.
There is a 8 wire ribbon cable that connects the SMPS to the main board,
which are the wire numbers in the middle of the image.
There are two columns of output voltages as tabulated below:
Probing for grounds and voltages gives me on both units:
Pin number Good unit 2 Bad Units
pin 1 ground ground (this has zero
resistance to the metal case)
pin 2 ground ground (this has zero
resistance to the metal case)
pin 3 -3.42v 0.4v
pin 4 0v 12.85v
pin 5 8.73v 0.76v
pin 6 18.32v 7.4v
pin 7 0v 8.8v
pin 8 0v 8.5v
There is approx 0.5Hz ripple on the secondary voltage outputs of the two
bad units and the green power indicating LED can be seen to ripple with the
naked eye..
None of the secondary side capacitors are bulging or have brown crud on them
and I'm already aware of the bad caps issue, so I replaced all the secondary
side electrolytics with Panasonic Low ESR FM series capacitors. This has not
made any difference.
I have traced the circuit schematic from the transformer to the output pins
of the SMPS board and the primary side from mains input to the transformer.
I don't know the exact winding configuration of the transformer, and I have
not yet traced the optoisolator board, but this is based on a 3 chip design,
TL400, CNY17 and UC3842AD, One is an opto isolator, one appears to be an
error amplifer/feedback chip and the third is a precision voltage reference.
I'd really appreciate some comment amongst any SMPS experts amongst you all
as to where the problem lies so I can get the two faulty units repaired.
The Primary side is available on
https://picasaweb.google.com/109749938109115860428/Apr242011?authkey=Gv1sRgCKzmsu6muqPFQA#5599104701271247186
and the numbers on the right refer to the pins on the transformer primary
The secondary side is available on
https://picasaweb.google.com/lh/photo/AHQo6tOf_NwmsTV_8owu9IbByfimUi_ZLKkVo27AQGI?feat=directlink
and the numbers on the left hand side refer to the order of connections on
the
transformer secondary
The whole of this side of the circuit is isolated from the primary side,
except for the one feedback wire off pins 1&2 which go off to an
optoisolator on a small daugherboard on the SMPS PCB.
Any comment?
Also there is a diode between two of the secondary windings that would seem
to short ciruit the winding back out onto itself? I have checked my PCB
tracing and its definately the case.
I'm of the view that the output of pin 4 should be 5 volts due to the
LM7805, so why is the working unit producing 8.73V at this point?
Give the 2.2K series resistor for the LED, and assuming a 10mA current, its
suggesting that the voltage coming out at this point should be 24.2 or is
there something subtle I'm missing? Also there are noi marked values for any
of the inductors.
Regards,
Stephen.
Uffe Bærentsen
"Stephen" <i.wan...@spam.com> skrev i en meddelelse
news:rtGdnTSRqdpvlSnQ...@brightview.co.uk...
> Probing for grounds and voltages gives me on both units:
>
> Pin number Good unit 2 Bad Units
> pin 1 ground ground (this has zero
> resistance to the metal case)
> pin 2 ground ground (this has zero
> resistance to the metal case)
> pin 3 -3.42v 0.4v
> pin 4 0v 12.85v
> pin 5 8.73v 0.76v
> pin 6 18.32v 7.4v
> pin 7 0v 8.8v
> pin 8 0v 8.5v
What is these resistances:
pin 1 to pin 2
pin 1 to pin 4
pin 1 to pin 7
pin 1 to pin 8
I think you have an issue with the ground/0V.
--
Uffe Bærentsen
Stephen
the PSU board?
If so which end should I measure?
"Uffe Bærentsen" <leiti_...@hotmail.com> wrote in message
news:4db41709$0$56778$edfa...@dtext02.news.tele.dk...
Stephen
"Stephen" <i.wan...@spam.com> wrote in message
news:zdadnXF1H7GrsCnQ...@brightview.co.uk...
> Would this be with the ribbon cable disconnected between the main board
> and the PSU board?
>
>>
>> What is these resistances:
>> pin 1 to pin 2 0.05 ohms. this is as expected as pins 1 and 2 are
>> soldered together at the main board
>> pin 1 to pin 4 380kohm
>> pin 1 to pin 7 1.4 Mohm
>> pin 1 to pin 8 1.2 Mohm
With the ribbom cable disconnected, measuring at PSU board end:
>> What is these resistances:
>> pin 1 to pin 2 0.05 ohms. this is as expected as pins 1 and 2 are
>> soldered together at the PSU board
>> pin 1 to pin 4 180Kohm
>> pin 1 to pin 7 1 kohm
>> pin 1 to pin 8 1 kohm
With the ribbom cable CONNECTED, measuring on the ribbon cable:
>> What is these resistances:
>> pin 1 to pin 2 0.05 ohms. this is as expected as pins 1 and 2 are
>> soldered together at the main board
>> pin 1 to pin 4 340 kohm
>> pin 1 to pin 7 1 Kohm
>> pin 1 to pin 8 1 Kohm
All of the above readings are not steady, so I geuss there are capacitors
charging/discharging via the multimeter.
Wicaksono S.
that the UC3842 is repeatedly goes on & off. There is an small
elecrolyte capacitor at UC3842 power supply pin that maybe went dry,
please replace it.
> The Primary side is available onhttps://picasaweb.google.com/109749938109115860428/Apr242011?authkey=...
> and the numbers on the right refer to the pins on the transformer primary
>
> The secondary side is available onhttps://picasaweb.google.com/lh/photo/AHQo6tOf_NwmsTV_8owu9IbByfimUi_...
Stephen
daugherboard? It is actually near to the connecting pins between the PSU
board and the daughterboard.
reason I ask is that there are no electrolytics at all on the daughterboard
which contains the UC384x, all there are is the three IC's and a load of
surface mount components?
Regards
Stephen.
"Wicaksono S." <w_s...@lycos.com> wrote in message
news:3efc4e2d-0c00-42b8...@w9g2000prg.googlegroups.com...
Kripton
search for a small capacitor 10-47µF next to the uc3842 and change it...
regards,
--
---
Kripton
Kripton
you can make a quick esr analog meter to check all your caps
sure the small chemical capacitor next to the uc3842 is dry ...
my web site contains easy and more complex esr meter to build yourself
regards,
http://kripton2035.free.fr/esr-repository.html
--
---
Kripton
Stephen
So here are the correct values:
> Probing for grounds and voltages gives me on both units:
>
> Pin number Good unit 2 Bad Units
> pin 1 ground ground pin 2
> ground ground
> pin 3 -3.42v -0.4v
> pin 4 30.2 12.85v
> pin 5 8.73v 0.76v
> pin 6 18.32v 7.4v
> pin 7 23.7v 8.8v
> pin 8 23.7v 8.5v
Also pins 7 & 8 are soldered together at the PSU end, pins 1 & 2 are
soldered toegether at both the PSU and the main board.
The UC3842AD, TL400 and the CNY17 are on a small daughterboard, which also
has a load of surface mount tantalum capacitors and resistors.
Asfar as electroyltics go, there is one on the main board, which is a capxon
22uF 50V job. is this likely to be the capacitor thats gone dry?
"Stephen" <i.wan...@spam.com> wrote in message
news:rtGdnTSRqdpvlSnQ...@brightview.co.uk...
Plai...@yawhoo.com
power supply are prime suspects for a problem like this. As a matter
of course, I replace them all. At $.30 each, it's cheap insurance.
PlainBill
On Sun, 24 Apr 2011 16:44:02 +0100, "Stephen" <i.wan...@spam.com>
wrote:
Stephen
be low ESR? if so I'' go for the Panasonic FM series that I used to replace
the secondary output electrolytics.
If a capacitor has gone dry, wouldn't there be a clue like bulging tops,
leaking brown crud, and a horrible smell when they get hot?
Regards,
Stephen
<Plai...@yawhoo.com> wrote in message
news:edr8r69d9s70hr405...@4ax.com...
who where
wrote:
>OK, I'll get some 22uF 50V 105 degree C ones ordered in then Do they have to
>be low ESR? if so I'' go for the Panasonic FM series that I used to replace
>the secondary output electrolytics.
>
>If a capacitor has gone dry, wouldn't there be a clue like bulging tops,
>leaking brown crud, and a horrible smell when they get hot?
Those symptoms occur some time *after* the cap's performance degrades
to the point of compromising the PSU.
Agree with the others, the small cap on the UC3842 is the most likely
candidate. Note that it *may* be a surface mount type and not an
aluminium electrolytic.
Kripton
> If a capacitor has gone dry, wouldn't there be a clue like bulging tops,
> leaking brown crud, and a horrible smell when they get hot?
>
> Regards,
>
> Stephen
you can make a quick esr analog meter to check all your caps
the small capacitor next to the uc3842 aways fails to an esr meter
even a cheap build from scratch analog one.
JW
in Message id: <I8Odncf6apym2CnQ...@brightview.co.uk>:
>capxon
^^^^^^
Shoot on sight. Capxon makes crap. Also Lelon, Luxon, and GSC off the top
of my head.
Stephen
I bought two new new 22µF 50V 105°C capacitors today, total cost 42 pence.
(About 0.5? or 50 cents for the rest of the world)
took the first faulty unit, swapped over the 22µF 50V capxon, tested it and
it worked.
I then took the 2nd faulty unit.... Swapped over the 22µF 50V capxon, tested
it and it worked.
Thank you, Wicaksono, Kripton, Plain Bill and Who Where, you were all on the
nail.....
As a token of gratitude, I think I should give some background to the
story.....
I am in the reception area for 4 TV transmitters in four different
directions and want to get TV channels from all four. Simple aerial group
combiners with grouped Yagi aerials would not work as the frequencies used
at each transmitter overlap at least one of the other 3. One way of solving
this is to used a channelised head end UHF amplifier.
These units have 10 frequency agile filters, with independently adjustable
gain, and can accommodate up to 3 UHF aerials as well as an FM radio aerial
and a Band III aerial.
These units are still available for £400 each ( around ?400 or $400 each.)
via wholesalers.
So I bought 2 off Ebay complete with a programmer. One worked, the other
didn't. After haggling with the seller, I got half the money back and was
allowed to keep the faulty unit, so it cost me £100 all told.
I kept an eye for more on Ebay. Another one came up, I was the only bidder
and bought it for £5 (about ?5 or $5) This was too also faulty with an
identical fault. I didn't raise a dispute given it was beer money, and if
the worst came to the worst, it could be cannabilised for spare parts.
I contacted the Spanish manufacturers of these units enquiring about a PSU
repair kit. "sorry, these units are no longer being made, you need to buy
the new version 2 unit. These are £500 each and the only difference is the
addition of a SD memory card slot allowing you to copy settings from one
unit to another."
So now thanks to the advice given here, I now have 3 working units instead
of 1 working and two faulty.....
So now I could combine up to 9 UHF aerials, 3 FM aerials and 3 Band III
aerials all pointing in different directions, combine the three units
outputs using a passive combiner, and feed that lot into my 17x16
multiswitch.... :-)
I also plan to add CCTV at some point in the future so these units will
allow me to combine digital TV with analogue CCTV too....
Thanks again,
Stephen
"JW" <no...@dev.null> wrote in message
news:tfjar6ho20o7mae6c...@4ax.com...
Arfa Daily
We need to get some posting consistency here. With half the respondents top
posting, and half bottom posting, it's becoming impossible to follow the
thread. Please all stick to bottom posting, as is the normally accepted
protocol in usenet newsgroups.
Right, Now that's out of the way. The voltages being quoted make no sense
against the drawn schematic that the OP provided. If pin 4 of the ribbon is
connected directly to the output of a 7805, then the voltage at that point
*must* be +5v on the unit that is working. It cannot possibly be at +30.2v -
period. This indicates that what you have picked as your reference ground to
measure against, isn't a valid secondary side ground. Bear in mind that the
primary side ground and secondary side ground on a switchmode power supply,
are very different. The valid ground for your measurements, should be taken
as the centre pin of the 7805. You might think that other places on the
board or within the circuit configuration are the same ground, but without a
manufacturer's schematic, you just can't be sure, so use a point that you
can be sure of or at least 99% sure of. I say only 99%, because
occasionally, you will see the output of a three terminal monolithic
regulator 'jacked' a little by the insertion of a diode or two in the ground
leg, but it's pretty rare on proper commercial equipment.
When you have reached a point where you can read a steady +5v on the
righthand pin of the 7805, then remeasure the other voltages, and let's see
if they make some more sense,
Arfa