Transformer question

[simonbaker]

Has anyone seen this...?

I'm troubleshooting an appartment 3-wire intercom.

The power supply is a 115 volt to 16 volt AC-AC wall transformer (according to the writing on the module).

When I disconnect the transformer and observe the secondary (16 volt nominal) voltage on my o-scope, it does not look sinusoidal. Instead, it looks "trapezoidal" -- straight line ramp up and down, and flat tops at around +/- 24 volts (say 48 volts p-p). (I should have taken a photo before reinstalling but didn't). I think the flat tops were maybe 50% or less of the cycle.

I thought maybe core saturation, but descriptions usually showed other kinds of distortion like spikes and rapid drop to zero when the core saturates. This waveform just clips, like an amp hitting its rails.

Does anyone know what this is?

Is it possible there is an overvoltage protection circuit inside the wall wart in case the transformer goes bad (like zener diodes, etc)? I suppose it is possible some primary windings shorted causing less of a step-down ratio or something like that.


-SB

[Skippy]

Small transformers of low power rating (less than 10VA) can have very poor load regulation, and a 40% increase above nominal would not be surprising.with no load. There could be a transzorb type device (bidirectional power-zener) in it to protect the subsequent circuitry from mains spikes/surges. Does it seem to give normal voltages when the rest of the circuit is connected? If you suspect a power-related problem, it might be the large smoothing electrolytic cap(s).

[simonbaker]

Small transformers of low power rating (less than 10VA) can have very poor load regulation, and a 40% increase above nominal would not be surprising.with no load. There could be a transzorb type device (bidirectional power-zener) in it to protect the subsequent circuitry from mains spikes/surges. Does it seem to give normal voltages when the rest of the circuit is connected? If you suspect a power-related problem, it might be the large smoothing electrolytic cap(s).

The circuit seems to function OK. It was the shape of the no-load waveform that seemed unusual to me, and I just want to make sure the transformer isn't giving out and about to cause a major problem. I don't think I looked at the waveform under load conditions -- maybe it becomes more sinusoidal.

The designers of the circuit decided to use the most minimal power filtering you can imagine -- a single diode with two capacitors separated by a ripple-drop resistor. It boggles my mind that the circuit functions adequately. There is a distinct hum in the intercom when using it, but given its purpose, it does the job fine. It is a nice example of not over-designing for the application .

What amazes me is that the apartment speakers are used as unamplified microphones over 50 to 75 feet of twisted pair to the main amplifier -- and it still works!

I'll post my LTSpice circuit for fun once I get it cleaned up a little. There is beauty in primitive things.

-SB

[golfnut]

I saw this in the UK years ago. I was told by a guy that it was the fact that loads on the main shape it.

Millions of loads with rectifiers in them (DC consumer items) clip the peaks of the main.

How correct he was Im unsure

S

[simonbaker]

I saw this in the UK years ago. I was told by a guy that it was the fact that loads on the main shape it.

Millions of loads with rectifiers in them (DC consumer items) clip the peaks of the main.

How correct he was Im unsure

S

Interesting, thanks.

BTW, the intercom uses 45 ohm speakers as the mic and speaker. I think they are about 4 inch size. I'm trying to find some economical replacement for the front speaker which has been battered by weather -- but these things are scarce as hen's teeth and/or expensive! If anyone comes across some reasonable ones (say around $5.00 - $8.00 US) I'm interested.

-SB

[Skippy]

This might be of interest:
http://www.jameco.com/webapp/wcs/sto...001_2099593_-1 or you could try a 16 ohm one with a 22R (or similar) series resistor, or maybe a 32R speaker (&10R res) or how about using one of those audio transformers, as used in PA systems? Connect the speaker to "8 ohms" secondary tap, and use the "16 Ohm" tap as the output, you get the idea - experimentation needed.

[simonbaker]

This might be of interest:
http://www.jameco.com/webapp/wcs/sto...001_2099593_-1 or you could try a 16 ohm one with a 22R (or similar) series resistor, or maybe a 32R speaker (&10R res) or how about using one of those audio transformers, as used in PA systems? Connect the speaker to "8 ohms" secondary tap, and use the "16 Ohm" tap as the output, you get the idea - experimentation needed.

Good ideas, I'll experiment with that, especially transformer impedance matching.

I think the main reason for 45 ohm speakers is they make better microphones. The circuit would probably drive 8 ohm speakers as is.

-SB