Broken rectifier/filter, blows fuse, schematics included

The most likely bet is either one of the 4 diodes in the bridge, or C4 being short circuit. The fuse should probably be a " T " rated type. An " F " type will almost certainly blow at power up, even with no fault present, due to the high inrush current that a cap with a value like C4's, will have. As a first move, with the supply disconnected from power, you should check with an ohm meter across each of the four diodes in turn. You should read somewhere in the region of 600 to 700 ohms, depending on the meter, in one direction, and infinity or thereabouts, in the other. A reading of zero, or just a few ohms either way across any diode, would indicate that it is faulty. Across the cap, you should read initially, a low figure, which will climb as the cap charges from the meter. A steady low reading would indicate a faulty cap. After that, you would be into disconnecting various potential current paths, and troubleshooting them individually.

Arfa

Being a "new apprentice" or novice - you need to have a basic tool - a VOM or DVM. With this you can test for shorts (continuity); diodes and bipolar transistors.

You are blowing a fuse - due to unacceptable current draw -- which is likely due to a short caused by a wire or actual physical short or a failed component.

gb

In data Sat, 24 Jun 2006 13:59:38 +0200, Johan Winge ha scritto:

Where's that 5a fuse?

--
Inty.Evolution
intiglietta@email.it -> per email normali e con immagini Jpeg

looks like it is on the primary side of 8v xformer

I was assuming that these were two windings on a transformer with a ( not drawn ) common primary, and that the fuses were wrongly drawn as the elements that look a bit like switches, but with only one little circle ?? Perhaps the OP could confirm this, as it potentially makes a difference to the diagnosis.

Arfa

It's inside the transformer (which is external to the unit, and thus not included in the schematic). The blue text was added by me in a try to show that.

-- Johan Winge

Ah, yes, there's a common primary. I realised it is shown on the original schematic, but I didn't have room for it when I did the scanning... Sorry for that.

I believe though that the things that look like switches, are indeed switches. The inner workings of the transformer are not shown, probably since it was made by another manufacturer. I believe the various retailers provided their own transformer, or something like that.

-- Johan Winge

Quite right. The blown fuse is marked "T 5A 250V".

I had made some measurements on the bridge, and the results were somewhat... irregular, but I was a bit unsure about how to understand the results, since I was thinking that the capacitor may be screwing up the results. To be entirely sure, I disconnected one side of each of the diods, and indeed, I found that two of them transmitted current in either direction!

Yep, that's what I see, so no problem there it seems.

I'll replace the diods on Monday, and we'll see if that solves it. I'll post my findings.

-- Johan Winge

What I found was that my local electronics shop had shut down, so I will order some parts from the web instead. This also means I have some more questions for you... ;-)

I couldn't find any diodes called only R250, but the parts list for the unit states that it should be specced for 50V, 6A. Based on this, I found one called 6A10, which was the closest I came at one particular vendor. Data-sheet can be found at

. Is it possible that this could be a valid replacement for "RTFR R250 50V 6A", as it is stated in the parts list? Which characteristics are relevant and important to take note of?

Similarly, for another unit, I will have to replace some of the 1N5400. However, most vendors seem to only carry 1N5401, which is specced for 100V instead of 50V ("Maximum recurrent peak reverse voltage" and "Maximum DC blocking voltage"). However, while this being over-kill in this case, I presume it won't have any adverse consequences?

-- Johan Winge

Diodes are very easily substituted and this application is low frequency so not critical at all. You need something with equal or greater amperage PIV (DC blocking, reverse voltage, etc) so the ones you found will work fine.

In data Sat, 24 Jun 2006 17:25:26 +0200, Johan Winge ha scritto:

try to remove everything after the 78xx regolator output, and see if fuse blowns yet.

--
Inty.Evolution
intiglietta@email.it -> per email normali e con immagini Jpeg

I would also recommend that all four diodes in the bridge are replaced with similar types, irrespective of how many are actually blown, and that you add

1000pf disc ceramic caps across them.

Arfa

"Johan Winge" hath wroth:

Yep. The power on surge current charging the 4700uf and 10,000uf capacitors might blow the fuse. These capacitors are a nice short circuits while charging on initial power on. A resistor between either the xformer and diodes, or between the diodes and the capacitors, should limit the surge current. My guess is about 1-2 ohms should help. You can also use slow-blow fuses.

If the unit is really 25 years old, you might want to check the capacitors for leakage. If there is any evidence of bulging or case expansion, they're blown. Big electrolytics deteriorate. If they're defective or dying, that might also be the cause. A crude way to check this is to substitute an ammeter for the 5A fuse and see if the power on current approaches 5A.

As others have suggested, it might be a blown diode. These are easily checked with a volt-ohms guesser.

Incidentally, the design of the circuit is truely an abomination. It vaguely resembles something I vaguely recall from Atari game products. The 7805 and 7812 regulator can only output about 1amp without shutting down. However, the game required more than 1A. So, the designers added an unregulated emitter follower in parallel with the

78xx regulators to boost the output current. I'm not sure, but I think you've drawn the schematic incorrectly. The regulator input line and the MJE2955 PNP xistors don't look quite right.

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

The circuit diagram looks perfectly drawn to me. The current paths to the regulators' inputs are taken through the 56 ohm (5R6 would probably be more appropriate) resistors. The transistors aren't in emitter follower arrangements; they are common emitter circuits. Taking the +5V circuit as an example; the transistor is turned on in response to the current through R1, which is the current being passed through the regulator. Until the voltage across R1 increases to the point that the transistor turns on, the regulator passes the full load current. When the voltage exceeds the turn-on potential of the transistor, the transistor turns on, passing more current to the load. The more current demanded by the load, the more the transistor turns on, thereby keeping the regulator from exceeding its current rating. Full voltage regulation is maintained.

The circuit isn't an abomination; in fact, this arrangement is recommended in the data sheets as a method to increase current capacity of the regulator.

--
Dave M
MasonDG44 at comcast dot net  (Just substitute the appropriate characters in the 
address)

Some days you're the dog, some days the hydrant.

"DaveM" hath wroth:

The typical circuit on the LM340/78xx data sheet shows: R = 0.9 / I(in) where I(in) is the input current to the regulator.

0.9???. I would expect 0.6v. See page 13 of the LM340 data sheet at:

True. The ones in the Atari power supply were emitter followers. Sorry if I didn't make that clear. The circuits are similar, but not identical.

to

regulator.

on,

Thanks for the explanation. I didn't see how it worked. The 56 ohm resistor had me confused as 1A through 56 ohms is 56 watts which isn't going to happen. Sorry.

I stand corrected and the circuit will work with the correct resitor value.

However, 14VAC from the power xformer is barely enough to produce the minimum voltage required to supply the input of the 7812. The full wave bridge produces 14VDC. The regulator requires 12V plus at least

1.6V input-output differential. Add 0.6V for the MJE2955 EB junction for a total of 14.2VDC required. The 14VAC xformer just isn't going to supply that.

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

Off load, the 14v AC winding will provide around 19.5v DC at the output of the bridge. Whilst I accept that this will start to dip from this peak value, when the bridge comes under load, I would have expected a decent quality 4700uF cap to have been able to hold the rail up to a level enough for the dips to still be higher than the dropout voltage of the regulator ?

I have a little project on the go at the moment, where a small 9v output torroid that " came to hand " out of the junk box, is supplying a 7810 that's being thrashed quite hard. I also have non critical circuitry ( some relays and a fan ), being fed from the input side. The resevoir cap that also came to hand, is a mere 1000uF. Agreed, the voltage across the cap is quite ripple-y, but it never drops below the point where any disturbance or lack of regulation is seen on the output of the 7810.

Arfa

The 14V secondary will produce about 19.75 V at the peak. Without doing a lot of math, the rule of thumb for the filter capacitor is about 2000 uFd per amp for 2V P-P ripple. That means that the trough of the ripple waveform is at about 17.75 V. That gives the 12V regulator over 5V of headroom, more than adequate for regulation. Even after subtracting the .65V drop from the B-E junction of the transistor, that still leaves approx. 17V; again, more than adequate. One thing that does appear to be marginal is the transformer's output at low line voltage. With a 10% drop in mains voltage, the DC input to the regulator drops to about 15VDC. Getting close to the regulator's dropout point, but still within ratings. If operated within its specs, this power supply should perform well. Cheers!!!!

--
Dave M
MasonDG44 at comcast dot net  (Just substitute the appropriate characters in the 
address)

They call it PMS because Mad Cow Disease was already taken.