Transformer grounding questions

[veracohr]

I'm building a simple bipolar power supply, and I've got a couple noob questions about the transformer.

The schematic shows the transformer core connected to ground - stupid question, but this must mean the metal case of the transformer right? Because otherwise it only has the 2 primary and 3 secondary connectors.

Second question - if the case of the transformer is connected to ground, and is also bolted to the metal enclosure, that's OK right?

[prairiemystic]

Check the schematic that the transformer "core" to ground is not a dashed line, which would be showing an electrostatic shield ground connection. The shield is rare in most power transformers, so I doubt it's that.

For electrical code and safety reasons, the transformer case and your metal enclosure are connected to power line ground (usually using the same bolt). The theory is if the transformer failed (primary shorted to transformer case) your project would have a "live" chassis and a shock hazard exists. Don't forget to have a primary fuse too.

If your power supply is running anything else also connected to power line ground, you'll get a ground loop which makes hum. It's a big hassle and that's why some gear has only a two wire power cord (no ground).

[veracohr]

Originally Posted by prairiemystic

Check the schematic that the transformer "core" to ground is not a dashed line, which would be showing an electrostatic shield ground connection. The shield is rare in most power transformers, so I doubt it's that.

It's a double solid line. Image is attached.

I do have a fuse. Thanks!

[prairiemystic]

Yes, ground the transformer and chassis. I use a mounting lug on the transformer, which also ends up being the chassis.

I would add two protection diodes, one across C5 and the other across C6. These prevent any reverse voltages from blowing up the regulators.

[veracohr]

Originally Posted by prairiemystic

I would add two protection diodes, one across C5 and the other across C6. These prevent any reverse voltages from blowing up the regulators.

Actually I didn't even incorporate C5 and C6, but isn't that what D1 and D2 are for?

[prairiemystic]

D1 and D2 protect the IC's from reverse flow when the output has voltage but not the input (i.e. powered off and connecting a battery, larger output caps than C1 and C2 etc.).
D3, D4 protect the IC's from C3, C4 during output shorts.

I put all 6 protection diodes in. It's $0.30 to make the supply never blow up.
The first time you decide to say, charge a battery and hook it up backwards or before the power supply is switched on and poof the regulator shorts and burns up VR1 and maybe the load.
I'll email Rod at Elliot Sound and ask him to put them in, like in his project #05.

Don't leave out C5 and C6, the regulators need them for stability.

[veracohr]

Stability? I read multiple webpages regarding the same basic power supply design, and although I can't seem to find any of them now, what I read led me to believe that C5 and C6 were not that necessary, just extra filtering I think. I think for my purposes the regulation of the IC's will be fine. Besides, I've already finished the circuit and I don't want to do it a 4th time (first time screw ups abound).

[prairiemystic]

From the National datasheets:

"Although the LM117 is stable with no output capacitors, like any feedback circuit, certain values of external capacitance can cause excessive ringing. This occurs with values between 500pF and 5000pF. A 1uF solid tantalum (or 25uF aluminum electrolytic) on the output swamps this effect and insures stability."

In my experience, the chip oscillates around a 1MHz without them (C5, C6). If you have anything weird going on, try put them in. They are small caps.

[veracohr]

Originally Posted by prairiemystic

From the National datasheets:

"Although the LM117 is stable with no output capacitors, like any feedback circuit, certain values of external capacitance can cause excessive ringing. This occurs with values between 500pF and 5000pF. A 1uF solid tantalum (or 25uF aluminum electrolytic) on the output swamps this effect and insures stability."

In my experience, the chip oscillates around a 1MHz without them (C5, C6). If you have anything weird going on, try put them in. They are small caps.

Hmm...thanks for the info. I suppose if anything weird does happen, I can try putting them on the breadboard? Or would they have to be on the PCB?

[prairiemystic]

The best location for C5, C6 is close to the regulator IC's. If you don't have a lot of wire >8" between the two boards, then they could go on the other board. Good luck