Help with fuses blowing at startup

[Paul Birkeland]

Yeah, a backwards diode is a dead diode. 

The part is specifically a UF-4007.  A 1N4007 would also work, but not quite as well.

An NTE-575 would likely be available locally to you at a retail outlet selling electronic components.  (IE Fry's or your favorite local joint)

[Ankaret]

Im in Seattle, anybody have good recommendations of places to look for a replacement diode?

Seems like Radioshack is about all there is in town...

If I need to order the UF4007, is there anything else I should grab while Im at it that you think would have a high likelihood of needing replacing in my situation?

Thanks for the help as always.

[kgoss]

Since you are in Seattle getting them directly from Bottlehead should be quick. That's what I would do if I lived that close to BH headquarters.

[caffeinator]

Another thing comes to mind, particularly after seeing the long lead in one of the photos.

I had a very similar problem while building my FPIII.  My build was pretty sanitary, good-looking wiring, resistance checks all normal, but it would instantaneously blow fuses as soon as the power was turned on.  I checked it over and over, tried some higher amperage fuses and still they'd instantaneously blow.

I was starting to tear it down when I found a possible problem - a solder drip that had left a kind of solder stalactite (or stalagmite for those in Australia) that came within about 1mm of the the chassis plate.  As I verified by trying once more with the lights out, the solderlactite was close enough to allow an arc that blew the fuse.  Since it wasn't contacting, it didn't throw the resistance check, making it a pretty confounding problem.

So, if it's not bad wiring or blown components, look all over very carefully for an errant lead, solder blob or any other cause of intermittent shorting contact.

[rlyach]

One more thing to consider. If you did have one of the diodes in backwards, chances are that two diodes were damaged. The full wave rectifier uses two diodes to steer the current in the right direction through the rest of the circuit. Then during the other phase the other two diodes steer the current. If one of the diodes is in backwards, it provides a direct current path back to the transformer without going through the rest of the circuit. Consequently, all the high current would have gone through two diodes, and you should replace two or even all four of them. Since your transformer got very hot, it would confirm a direct short across the secondary winding, and a backwards diode is a reasonable explanation for this. A metal short from long leads is another.

[Attached is the diagram of the current flow through the full wave bridge rectifier with proper diode orientation (Good). The positive current flows in the direction that the diode symbol points. During the positive cycle, the current follows the red path. During the negative cycle, the current follows the green path. The bad circuit with the reversed diode shows that during the positive cycle the current simply flows through the two diodes and back to the transformer, creating a high current short. Both of these diodes should be replaced. You don't have to replace the other two if you know which one is reversed and compare it to the circuit. If you just want to be safe, replace all the diodes.]

[Ankaret]

I requested 4 replacement diodes from bottlehead today, so hopefully that invoice will come through soon and I can try replacing them in hopes of that solving things.

Also, I believe it was Josh(?) that I talked to, and he mentioned that I should double check I got the right transformer (120V, even though he thought mine was accidentally labeled as 240V). One way to check he said was to look for about 12ohm resistance between transformer terminals 1 and 2. But Im getting a reading of only 6 ohms here, what do you all think? Something wrong with my power transformer?

[rlyach]

I just measured my transformer, and I get ~12 ohms of DC resistance on the primary winding of the mains transformer. If you are getting 6 then I would assume you have half the windings and thus you probably have a 240V transformer. I will let one of the guys at BH confirm this.

[fullheadofnothing]

That is incorrect. A 240V PT-3 measures about 45Ω. I think Kyrill has a 120V transformer, and the exact reason that it is measuring lower than normal is unknown (quite possibly related to the method of measurement). What is important to know here is that it is not a higher reading, indicating that it is a 240V.

I knew that a few 120V PT-3 were mislabeled, and I was asking for confirmation that this was why a 240V label was visible.

[rlyach]

If that is the case... Then would a 6 ohm reading indicate a partially shorted primary winding?

[fullheadofnothing]

Some meters don't do well measuring low resistances. Some don't do well measuring inductive resistances. Placement of probes on the terminals has a major influence on the measurement, particularly when not using clip leads. Solder and flux on the terminals will influence measurement. There will be some variation from sample to sample of what a transformer will measure.

It is not measuring high, and that was my concern.

[Ankaret]

Thanks for weighing in Josh, and catching the reason to test this out just in case in the first place. Ill rest assured that Ive got the right transformer and wait for the diodes to arrive and keep my fingers crossed this does the trick.

[Ankaret]

Got replacement diodes (thanks for the speed Bottlehead) and replaced all four just to be safe, all oriented correctly, carefully soldered etc. and STILL won't start without blowing the fuse...

What makes sense to try next? Getting a little flustered...any help is greatly appreciated.

[NightFlight]

You replaced the cap that was powered the wrong way around?   I can't see from the photos how the power switch is wired. One pin should have L on it and the other go to pin 1 of the transformer. I pretty certain that is what you have, but I couldn't tell for sure.

[Ankaret]

It was actually a rectifier diode that was wired the wrong way. And yes, Ive replaced all four of the diodes, now positioned correctly. For the power switch, one pin goes to 1 on the transformer, and the other is connected to a power cord socket pin. Should be correct.

[Ankaret]

In terms of build errors, from my original post this might be noteworthy:

"I found that I had forgotten to solder one end of the 270 5w resistors (15L)" and something started to smoke around that area when I first tried to power it up. What might need replacing or checking there? If something was smoking, probably a good chance another component there is blown?