DC relay buzz

You need a filter capacitor as well as the rectifier, start with a 220uf/35V one.  Also, if you're supplying it with 12VAC to the bridge, I'd use a plain diode as you'll be providing significantly more than 12 volts DC to the relay with a bridge and a capacitor.  If you're powering this from the track voltage and you have DCS, you'll also want a 22uh choke in series with the power to prevent DCS signal degradation.

 You sure you have it wired correctly?   The BR 's plus and minus should go to the relay coil. That leaves the relays contact to do what ever you want. I've used a number of RS 12 volt relays for signalling and can't remember any buzzing.... What power are you using to feed the Bridge rectifier

I've just put together a bunch of relay modules for signals on our modular layout.  Without a capacitor, the relays chattered.  It all depends on the characteristics of the relay, many smaller signal relays have response times of only a couple milliseconds, those will likely chatter without a capacitor filter.

I agree but are we talking chattering or Buzzing? Buzzing might indicate AC rather than DC on coil ?

 I've hooked up a lot of stuff wrong.

Actually, buzzing happens with rectified DC, at least it has for me with several different types of relays.  Chattering is typically a poor connection on the track if these are for signals, that's just dirty track, dirty wheels, etc.

Fair questions regarding wiring correctly.  That was the first thing I checked.  "+" and "-" are run through to the coil.  To be fair, I even ran a DC powerpak to the coil.  As soon as power was applied it switched the signal and was quiet as well.  The buzzing only occurs with the bridge.  The connection from the transformer to the bridge is wired through the "~" pins.  The relay is connected to the "+" and "-".  As a test to ensure I had the hook up correct it was a simple MTH Z-500.  The plan was to have a 10 V constant voltage to the bridge once installed.  

Not sure I understand what a diode will do here.  To be fair the wiring scheme I think came from this forum a while back.  There was no mention of anything other than the bridge rectifer and mini relays.  I purchased the parts from RS based on that.

Are you using insulated track(outside rail) to activate the relay thus governing your track side signals

I don't know what thread you're talking about, but I can assure you that adding the capacitor filter will solve the issue.

John

What is a capacitor filter?  Something easily located at radio shack?

Henry

Here you go, this one will do the trick.  Just connect it across the relay, and observe the polarity.

http://www.radioshack.com/prod...p;numProdsPerPage=60

John, with a large filter, what's the difference between full- and half-wave rectification?

To be fair, I even ran a DC powerpak to the coil.  As soon as power was applied it switched the signal and was quiet as well.

If you're using insulated track to activate the relay , you could use The Dc power pack.   One leg  of the  dc pact  (+) power would go directly to one side of the relay coil. The other leg(dc-) would go the common outside rail. Same rail that gets the AC outside rail feed.  The -dc is  just there for the ride and doesn't hurt anything.

The insulated section would have a wire going to the other side of the coil. The train wheels complete the circuit and energize the relay's coil. Use the contacts on the relay for whatever you want.(signals)

Depends on the current draw of the relay.

Are you using this Radio Shack SPDT relay (12VDC, 1A, #275-241, $4.49)?

If so, that's a 320 ohm coil.  If using GRJ's 220uF capacitor, that's an R-C time-constant of 320 ohms x 220uF = 0.07 sec.  A 60 Hz AC line cycle is 0.016 sec, so there's plenty of filtering to handle half-wave rectification where the voltage is "lost" for 0.008 sec (half the line cycle).

I like Gregg's idea of using your DC power pack if it's available. This would allow you to drive additional block signals or other activation chores with no additional parts (except the relays).

Stan, that's the one.  Gregg, I am going have to try that as well.  Thanks for the help guys.  I do love these winter nights with nothing on television.  Let's me devote time to this stuff.

Henry

The diode is required to prevent the capacitor from discharging through the bridge when the voltage goes to zero.

I respectfully disagree. The bridge contains 4 diodes as edited into your photo below.  The upper two diodes, which are in the same "direction" as your discrete diode, block any discharge path from the capacitor back into the bridge.

Gotta' go with Stan here, I have never seen or heard of a bridge having a reverse discharge path as long as it's a fully functional part.

Right… what I'm not clear on is how useful half-wave is to reduce the voltage to the relay.

If the filter is too small, then you get buzzing.  Too large, and I would think you would be delivering constant peak voltage.  So only in some range of capacitance, dependent on the coil current, would you achieve voltage reduction without buzzing.

But maybe I'm thinking about it the wrong way.

I've always noticed that the voltage is higher with the bridge, so since I'm trying to keep it lower, I use the plain diode.  Besides, it's cheaper and smaller.

There is no path back through the bridge,since there are opposing diodes,I think? The diode is still a good idea because the capacitor will get the reverse voltage. The diode should go across the coil,at least that is the way it is usually done.

Dale H

You can do it either way. With full wave rectification you just need less capacitance to do the same work. DC relays do not like pulsed current although they work that way. . The contacts lose power 60 times per second. This can be hard on the contact points especially with half wave rectification.

Dale H

Good point Dale, the diode would actually be useful across the relay to suppress the inductive kickback.

Well, I'll have to disagree here too.  You would not have "inductive kickback" in this circuit because there is no switch that rapidly removes the current from the coil.  Instead you have a rather large capacitor across the coil which slows any rapid changes in the coil's current.  The reverse voltage or "-L * di/dt" is inconsequential in this configuration.  In my opinion of course... 

Makes sense Stan, I'm so used to putting diodes across the coils that I don't think much about it.

All I can say it is standard practice to put a diode across the coil on anything with PC chips.  Pinball machines,for example started using them in the 1970s on all coils when PC boards were put in them.   For AC relays a resistor- capacitor is often used across the coil. Some relays,like the ones I use,have them built into it. Most modern automotive relays also have them built in. If it does not have one it costs about 2 cents to put one on.

Dale H

This has been mentioned, and a way that I like to think about it is how much ripple voltage will there be in that power supply with that load. The relationship is V=i*t/c where i is in ma, t is in ms and c is in uF.

In the relay example above, the current drawn by the relay is 12/320 or 37.5 ma. That's assuming that the dc voltage average applied to the relay is approximately 12 vdc. The time t is 8.33 ms for a full wave rectifier for 60Hz  and the capacitance suggested is 100 uF. Then the ripple voltage is 3.125 vpp. That's 3.125 volts peak to peak variation in the dc voltage applied to the relay. That's pretty significant and will probably cause an audible hum. It's clear that you can reduce that by better than half by increasing the capacitor to say 220 uF. And it's clear how changing the values in the relationship above will affect the ripple voltage.

It's been my experience that even a fairly small capacitor will solve the issue with the relay buzzing.  I know that even with a bridge, the small relays we were using for signals would buzz and actually open the contacts. 

Maybe ok with 100 uF...start with that and add another one if needed. I've been thinking about this issue lately since I was fighting the buzz from a Lionel semaphore and finally ended up with a really quiet dc supply. A relay is probably less buzzy than a solenoid.

Meanwhile, back at the ranch...

So necrails, does the capacitor solution make sense?  I see Radio Shack in-store carries radial-lead 35V electrolytic capacitors with values 100uF, 220uF, 470uF...any of which should work for you.  As mentioned earlier you just hook this capacitor across the two wires from your bridge rectifier output to the relay coil.  The capacitor has a + and - lead; as shown in the RS product photos the "-" side is a stripe with a bunch of "-" symbols.  The "+" and "-" leads go to the corresponding "+" and "-" leads of your bridge rectifier.

The correct capacitor would be one that removes the ripple enough so that the sine wave does not fall below the drop out voltage of the relay coil. For full wave rectification 100 uf should be sufficient for most relays. For half wave rectification I would use at least 220 uf.

For chatter removal in block detection I use a larger capacitor. 470-uf to 1000uf to hold in the coil when train wheels make intermittent contact. Shown here

www.jcstudiosinc.com/BlogShowThread?id=410

For timing circuits I use large capacitors up to 10000 uf across the coil. For the relay I use this will hold the coil in around 1 second per 1000 uf used. This is useful if say we want a semaphore to release, maybe 10 seconds after the train sits on the block or a soft start and stop circuit. When using large caps I put in a small resistor,maybe 18 ohm, in series to one of the leads to reduce arcing when the capacitor charges.

Dale H

Funny you should mention this Dale, I just build a couple of relay modules for signals on our modular layout, and I went with larger caps to kill the chatter, they work great now.  I'm going back to make the other ones larger.

Great John

If you get some outside wheel sparking just add the small resistor. This was a tip I got for Dale Manquen years ago.

Dale H

The resistor is already there, thought of that when I installed the originals.

Out of curiosity, is the application to switch between red/green standard 14V incandescent bulbs?  If going through the trouble of wiring up diodes, capacitors, resistors, etc., why not use a few transistors and skip the relay entirely.  You don't need mulit-Amp switching capacity.  With transistors the control currents become 100x or more smaller so capacitors can be 100x smaller to achieve the same time delays (if needed).  A 10,000uF cap can be quite large and expensive.  If the relay is taken out, all components might even fit in the cavity underneath a rail section.

Or maybe someone already sells such a widget?

I use a 40 cent capacitor, a diode, and a 100 ohm resistor.  The relays are salvaged units from control boards, they have a 5ma coil current at 12 volts.  The module works great and gives me several seconds of delay on the signals. 

I know there are many ways to skin this cat, but since we have a number of the 7-light MTH PRR signals that have died with cooked electronics for no reason, I'm doing the low tech solution here with bulletproof parts.  None of my relay modules have died, and the MTH signals seem to drop like flies!   The relay modules cost me about $1 in parts, the terminal block and the relays are free salvaged parts.

So I did not get to a radio shack yet but re-did all of the wiring in the off chance I did something wrong.  Here is a picture of the diagram I followed.  It is one of several that all show the same thing but seems to be the clearest.  Despite the rewiring the buzz continues so the capacitor is the next step.  

If you get a 220 uf 35 volt capacitor and put it across the relay coil IN PROPER POLARITY,it should solve the problem. Use the 10 volt tap of the Z4000 instead of the 14 v one. Otherwise you have too much voltage on the 12VDC coil.

Dale H

 What if you used center rail track power instead of the transformers aux post to power the  one side (hot) of the BR. ?   I had a similar problem trying  to use a BR with my KW and using 14 volt accessory posts