Maker Pro
Maker Pro

240VAC Transformer based 12 VDC PSU - Problem

peterlonz

Feb 11, 2010
28
Joined
Feb 11, 2010
Messages
28
Hello,
I have just discovered that my "trusty" DIY PSU no longer outputs 12V.
I see in fact about 0.5V
The unit has been very reliable over the 6 year in intermittent operation; until now.
I removed the cover & quickly found (under no load) the LM1085 IT-12 regulator (3A, fiixed 12V LDO output) was getting very hot.
I had fitted an aluminium heat dissipation sink (75mm x 50mm) & this quickly became too hot to touch & I could smell trouble & switched off before magic smoke appeared.
According to the data sheet this regulator has thermal protection & current limiting, so I was surprised at what seems a failure of this component.
Then I saw that it is temperature rated to 125 C; much higher than I would ever have expected.
Of course there may be upstream problems, the rectifier bridge maybe ................
I have not checked yet the two capacitors, but they look OK.
Anyway I ask for guidance, maybe I should rebuild from the transformer & include a bigger heat sink?
 

Bluejets

Oct 5, 2014
6,901
Joined
Oct 5, 2014
Messages
6,901
You need to provide a few more details on the supply.
Perhaps a few pics also.
 

peterlonz

Feb 11, 2010
28
Joined
Feb 11, 2010
Messages
28
OK, sorry about delay in responding.
Mislaid my original file on this build.
It's point to point wired so pics may not help much - working on that anyway.
Meantime details are:
upload_2017-3-14_11-1-30.png
As-built details:

Wiring: Point to point (no Circuit board)
Mains supply 3 Wire, 240VAC.
Mains fuse Replaceable fuse, (500mA Fast) in 240VAC supply
Rocker switch: Power On/Off; Rated 250V; 10A. (with red neon indication).
Transformer output: not yet measured
Rectifier: Type KBU605; 6A / 200V rated, (direct bolt to small aluminium heat sink)
Smoothing caps: 2 x 4,700 mF aluminium electrolytics, 25V / 105 C rating.
Regulator: LM1085 IT-12 (LDO, fixed output). Bolted direct to heat sink, no insulator bushes or washers.
LED Indication: Across output.
Cooling: Natural convection.
Earthing: Removable aluminium cover & transformer chassis earthed via mains earth wire.
Output: 12VDC, as measured by non-calibrated DMM.
Output power: Driving head unit judged about 10/12 watts typically.
Ripple: Not measured but supply proved suitable to drive auto radio head unit with no (audio) noise.

Apologies for @borrowed circuit, as yet I have not found an EE free drawing program that I can quickly master.
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
I have just discovered that my "trusty" DIY PSU no longer outputs 12V.
I see in fact about 0.5V
The unit has been very reliable over the 6 year in intermittent operation; until now.

a sign of either a failed regulator or something past the regulator loading it down

so when you measured the output at 0.5V was the PSU connected to something ?

measure the voltage going into the regulator ( no load on the output of the PSU)
disconnect the input to the regulator and again measure the voltage across the smoothing capacitor


Of course there may be upstream problems, the rectifier bridge maybe ................

maybe but more likely downstream


I have not checked yet the two capacitors, but they look OK.

looks can be deceiving



but before we continue answer the earlier Q's of mine
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Ldo regulators can also oscillate. It might be worth sticking some 0.1uF caps between common and input, and also between common and output.

Given that the regulator is not insulated from the heatsink, there may be something touching the heatsink or anything electrically connected to it that is shorting the output (I'm not sure what the case of this regulator is connected to)
 

peterlonz

Feb 11, 2010
28
Joined
Feb 11, 2010
Messages
28
Ldo regulators can also oscillate. It might be worth sticking some 0.1uF caps between common and input, and also between common and output.

Given that the regulator is not insulated from the heatsink, there may be something touching the heatsink or anything electrically connected to it that is shorting the output (I'm not sure what the case of this regulator is connected to)
The case cover (top & sides) is aluminium; chassis front,back & base is 16 mm plywood.
The aluminium portion is connected to mains earth for safety as is transformer chassis.
I doubt this is an LDO oscillation problem, or that the heat sink was being touched because this was a fairly sudden failure after many years of service with zero problems.
I still have to follow up on some info requested.
 

peterlonz

Feb 11, 2010
28
Joined
Feb 11, 2010
Messages
28
" but before we continue answer the earlier Q's of mine "
Ok here goes:
1) Transformer Vout (no load) = 12.9-13.0 VAC
2) Regulator, under no load: Vin = 16.8VDC; Vout12VDC
3) Across smoothing caps (Reg disconnected) = 17.4VDC
PSU on only long enough to take readings but no heat build up at Reg or Bridge.

I noticed these transformer markings:
CEFBW 663200-1
I/P: 240VAC; 50Hz
O/P 14.0VDC; 1.6A
Yes 14.0VDC no idea how DC gets here?

Hope this helps, & I apologise for my delayed response.
I am planning to recheck assembly & terminal blocks, etc. & see if problem disappears.
If not I will rebuild from transformer up but use 35V caps.
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
noticed these transformer markings:
CEFBW 663200-1
I/P: 240VAC; 50Hz
O/P 14.0VDC; 1.6A
Yes 14.0VDC no idea how DC gets here?

maybe a typo ... that would be pretty unusual tho

did you actually try to measure the output using DC mode on your meter ?

now when you say transformer are you referring to an actual transformer or a complete plugpack ?

photo please and another pic of the whole PSU sharp and well lit :)
 

peterlonz

Feb 11, 2010
28
Joined
Feb 11, 2010
Messages
28
Output of tranni measured on AC scale
DC scale not considered.
I am referring not to a plug pack but a conventional transformer.
Clear well lit pics very hard on ordinary phone camera, particularly with point to point wiring - it looks a bit like a rat's nest. My next build I will allow more space & consider wiring layout more carefully.
I have a few however, will attempt a clean up, select best & send tomorrow.
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
I doubt this is an LDO oscillation problem, or that the heat sink was being touched because this was a fairly sudden failure after many years of service with zero problems.

That isn't a good analysis .... you may have just been lucky over the years. but after years of use, component values change and it may have reached a condition where oscillation now occurs, it would easily explain the loss of voltage and the increase of heat in the regulator
 

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
Why not look at the datasheet for the regulator:
1) The DC input you have is very close to its maximum allowed voltage and maybe your electricity voltage is a little higher today than before, mine is.
2) The datasheet says that a 10uF tantalum capacitor must be added to the output to prevent oscillation. Yours has none.

The metal tab on the regulator is connected to the output so if the heatsink is grounded then the regulator output is shorted.
 

peterlonz

Feb 11, 2010
28
Joined
Feb 11, 2010
Messages
28
I had read the specs but to some extent they are contradictory (on the tantalum output caps) & I note only the spec summaries show Max Vout as 18V.
Compare page 2 spec sheet statement:
"Maximum Input to Output Voltage Differential: LM1085-12 =18V" (IE a differential).
Anyway I am new to this so I'm probably wrong.
Some not very good camera phone pics attached.
 

Attachments

  • 12V x 3A PSU AC-DC general open side view.jpg
    12V x 3A PSU AC-DC general open side view.jpg
    96.8 KB · Views: 110
  • 12V x 3A PSU General View 01.jpg
    12V x 3A PSU General View 01.jpg
    105.1 KB · Views: 133
  • 12V x 3A PSU General View 02.jpg
    12V x 3A PSU General View 02.jpg
    115.6 KB · Views: 105
  • 12V x 3A PSU General viiew of build .jpg
    12V x 3A PSU General viiew of build .jpg
    103.8 KB · Views: 178
  • 12V x 3A PSU view general arrangement open.jpg
    12V x 3A PSU view general arrangement open.jpg
    104.4 KB · Views: 114
  • 12V x 3A PSU view of build arrangement showing heat sinks dimantled.jpg
    12V x 3A PSU view of build arrangement showing heat sinks dimantled.jpg
    73.4 KB · Views: 179

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
You are correct, I was wrong. The maximum input to output is 18V so the maximum allowed input is 30V.
You should add a 10uF tantalum capacitor to the output.
 

peterlonz

Feb 11, 2010
28
Joined
Feb 11, 2010
Messages
28
Been wondering why I ever chose an LDO Reg in the first instance.
Would not a plain Reg do the job, although I guess regulation losses would be about the same?
No idea yet if a tantalum would still be required?
Also to spread the loss inevitable in such a design would it not be wise to now fit an in line diode before the Reg input?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Would not a plain Reg do the job

Yes. And it would be more stable. However an input and output capacitor are still recommended. LDO regulators can be quite stringent on their requirements, normal regulators far less so.
 

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
Years ago I used tantalum capacitors because they were small. Every single one of them blew up and disappeared.
They were Name Brand European ones, not the junk we get from Asia today.
 

peterlonz

Feb 11, 2010
28
Joined
Feb 11, 2010
Messages
28
It appears that opinions vary widely on the use of tantalum caps.
If they are recommended in application notes from a recognised manufacturer, I can't see a good reason to steer clear, particularly as here I think we are still talking about a 12V LDO regulator?
 

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
The capacitor is said to be a 10uF tantalum or a 47uF electrolytic. I use the electrolytic but not in a freezer.
 
Top