buzzing relay

Yes this needs to be fixed and not just put out of earshot!!

The yellow item is a capacitor, in front of it is the reservoir capacitor which I think is probably defective. The rectifier, is to the right of the black capacitor with 4 legs one on each corner. The 8 legged device is the heart of the controller. Acquire a similar replacement capacitor of a similar value and at least the same voltage rating and solder it temporarily in parallel with the existing one observing the correct polarity (grey line with minus signs on it) If the chattering stops, then whip out the old one and fit the new again making sure it is the right polarity.

hth Bob

You can skip the parallelling, just replace the part. Polarity needs to be right. If your junkbox doesn't contain one of the same capacity/voltage, in creasing either by 50% is no problem - if it still fits.

NT

Thanks both. Easily fixed.

So all OK now? Result!

Er...not quite. I meant it will be easily fixed (I hope)! :-)

It worked?

I might be wrong but the 024-1Zs suggests the relay coil is designed for

The symptoms rather suggest that one of the bridge rectifier elements has failed short circuit, degrading it from full wave to half wave rectification which, in this marginal PSU cct will produce very high 50Hz ripple and a significant reduction in the average voltage feeding the larger 8 pin dc/dc converter chip.

There's an example of this type of compact mains to low voltage DC supply circuit shown here:

Watch out for line wrap.

Hmm...thanks, or not, as the case may be. I hope you're wrong, as I can't see me replacing the rectifier if it's shot. Far too small and fiddly! Lets see what replacing the cap does first.

====snip====

If you've got a suitable spare, handy to drop in, that's worth a shot. However, removing that smoothing cap and the big yellow dropper cap will give you much better access to replace the rectifier chip if it tests faulty.

If you've got a multimeter to test with, you should be able to probe the pins to test for short circuits between the pins if googling the chip number confirms my suspicion that it's a rectifier chip. The nice thing about testing for bridge rectifier diode short circuits in this circuit is the lack of a low resistance dc shunt path from a transformer secondary winding which normally requires that one of these connections be opened to properly test the rectifier.

The resistance of the load usually looks high when testing with a multimeter with the test voltages typically applied to measure resistance. If you're testing with a classic moving coil multimeter set to its x1 ohms range (driven from a single uncladded AA zinc carbon cell), silicon diodes will give a reading close to mid scale when testing the forward 'resistance' of a single diode. You should be able to identify a shorted diode element in the bridge rectifier chip without having to disconnect it from the circuit so it's a test worth applying if you have a multimeter of one sort or another to test with.

A DMM uses a constant current source and measures the volt drop to produce an "Ohms" reading, essentially translating the voltage directly to ohms, eg 10mA test current will produce a 500mV reading when testing a

I would expect that DMMs that offer 'autoranging' in the ohms measurement mode could become confused by this non-linear behaviour. They may either use this behaviour to automatically identify or warn you that you're measuring a diode or else ignore this and leave it entirely to you to manually select the 'diode test' function (often associated with a continuity test buzzer feature).

Thanks for all that. I have a little DMM with a diode test function, as you describe. I'll test the rectifier before I set about replacing the cap. I should be able to get the test prods in there without too much difficulty.

Check the voltage at the relay coil. Most relays have their coil operating voltage marked on them.

If it is low, buzzing is a likely outcome.

only for an ac driven relay, I presume this one is dc. If a meter reads it as low it'll be a bad Cap or Diode.

NT

Interim Report:

I've tested the bridge rectifier using the 'diode' range on my little DMM, and AFAICT it's OK with no shorts. I'm glad it's OK, because it's _very_ small, an MB 6S type, and I wouldn't even attempt to replace it myself.

I've checked the DC output of the rectifier: 29 VDC when the relay isn't activated; 18VDC when it is, i.e. it drops significantly when the relay is activated. No significant AC on the output of the rectifier. The relay coil is 24 volt.

The voltage applied to the relay coil when activated is 16.6 VDC

The big yellow capacitor is 0.27 µF, 275VAC.

I'm awaiting delivery of a suitable reservoir capacitor to replace the possibly faulty one, as I don't have one suitable.

Many thanks for all the help and suggestions so far.

Not unreasonable sounding measurements. 16.6 volts is a bit low/marginal for a nominal 24v relay. I'd expect to see significant AC on the rectifier output when the relay is energised. I would still have reasonable expectations of the replacement electrolytic capacitor sorting it out. Bit of a uncommon value for the yellow capacitor 0.22, 0.33 or 0.47 more commonly found values - maybe they had a job lot of 0.27uF in stock.

Bob

Yes but no .... often tends to be the quoted rating of those X rated types for continuous across the ac mains use. I've just looked through my stock and find lots of 275v and also 305v both AC voltage and some other non branded ones marked 630 v which I presume to be the DC peak voltage - I think they came out of some 3 phase filters.

Bob

It looks like the PSU circuit is even simpler than I assumed. It's basically a 'lossless dropper' (large yellow 270nF 275vac X rated capacitor providing a 12K reactive impedance) feeding a half amp 600vac rated bridge rectifier which can tolerate a non-repetitive 35A surge over a half cycle (8.3ms) of 60Hz ac sinewave.

Like all such simple series impedance dropper power supplies, these can only be used with a constant current load. In this case the two zenner diodes in series across the rectifier's output terminals are most likely a nominal 15v 500mW type incorporated to clamp the the output voltage at a nominal 30v and shunt the 18mA or so of current that would otherwise raise the no-load voltage to a destructive level.

The 35A non-repetitive peak rating of the rectifier bridge is important in that it determines what value of inrush limiting resistor would be required (if any - the time constant created by a 15 ohm resistor and a

In fact in the absence of such a resistor and assuming a mains supply impedance of half an ohm, the time constant drops to 135nS at a worst case peak of some 1300A (circuit disconnected at the peak of a negative cycle (-336v) then switched back on at positive peak (+336v) a total charging voltage of 672v applied across a half ohm impedance hence the

Generally, if the manufacturer plots test data beyond the half cycle surge time, the peak current capacity goes up as the time interval reduces. Eventually a limiting maximum current value is reached where the diodes will be destroyed no matter how narrow the peak. Where that ultimate limit lies, I couldn't say offhand. Plus there remains the effects of inductive series reactance acting to limit surge current spikes measured in nanoseconds. It's quite possible that when such a low value lossless dropper capacitor value is used, the need to include a discreet inrush current limiting resistor disappears. There may or may not be one in this circuit.

Now that you've been able to provide some more information about the components and the results of your tests, it looks to me as though either the capacitor has dried out and its ESR has shot up or else the relay winding has developed a partially short circuited winding fault. I'd expect the relay winding resistance to be close to 1333 ohms based on the component values used in the 30v DC supply. If it tests noticeably less than this, say 900 ohms or less, you may find yourself looking for a replacement relay.

If you do test the coil resistance, try reversing the test leads if you get a very low value. It's quite common for the circuit to include a diode across the coil to suppress high voltage back emf spikes when the driving current flow is otherwise abruptly interrupted.

Since the capacitor is a good candidate for high ESR failure and is also the cheapest and the easiest component to swap out, This would be my first target component to replace (even if the relay coil resistance looked a little lower than expected - there may be other parallel paths to confuse your measurements when testing in-circuit).

The balance of probabilities lies with it being a faulty capacitor. You'll know soon enough once you've replaced it. :-)

====snip====

In most cases that would be true. Using a larger value than the minimum mandated by most circuit design requirements generally improves the performance of the circuit. However, you have to remember that in this case picking a slightly larger or smaller value would be like replacing a

In this case, a 220nF cap wouldn't be able to supply enough voltage to the relay and the use of a 330nF may require the use of 400mW zenner diodes in place of 300mW ones. The next available rating up from 300mW may jump to a more expensive 500mW type.

Whilst it may be possible to use higher wattage zenner diodes *and* be able to safely dissipate the extra heat, this design choice (330nF and

In any case, the choice of 270nF is taken from the E12 range which, for this capacitor type is standard and not uncommon. Electrolytics, otoh,

Relays can be undervolted to reduce heating & power consumption, and give b etter tolerance of tiny enclosures. They can also run on PSUs that deliver a moment of greater voltage, dropping to reduced v under load. If the PSU r esistance gets too high I expect this could cause chattering.

NT