7806 Regulator - 220 uf on OUTPUT??

The output capacitor is usually an NP .05 Tantalum for signal transients. It helps stabilize the regulator circuit. I guess a 220 uf one would do the same thing. It will work without it,usually.  Having said that when I convert passenger cars to LEDs, I use a 7805 circuit. I sometimes add 5, 1 farad of capacitors on the output side to keep the lights on when power is removed. This has had no ill effects on the regulator and I do not heat sink it. Not sure about the buzzing,that is a Dale Manquen question.

Dale H

Professor, I agree 220uf seems like overkill but I am not sure having a large cap would cause humming. I would first disconnect the load on the regulator and see what the resistence is. A couple of markers should not draw anywhere close to the 1.5 amps the regulator is rated for. Maybe there is a partial short somewhere in that circuit. Also, assuming you don't have a scope, you could check how much AC is on the regulator output with your multimeter set to AC. It should be only a few millivolts. Check with and without the load. 

Pete

What's the voltage rating on the 220uF?

Wild guessing here, they might have put the 220uF on the output to keep the LEDs on longer, and they put it after the regulator so they didn't have to have a higher voltage rating.

Is there a choke for the DCS signal hidden some where?

Normally the large capacitor is before the regulator with a small one after it for stabilization, as Dale said. Having a large cap after will do very little, because as soon as the voltage from the regulator begins to drop the lights will start to dim, albeit slowly in this case. Having it before provides a reserve supply for the regulator to......regulate.

Dave

Another thought, you did not mention what other components are on the board. Is there also a transistor tied to this regulator? If so, this could be set up as a switching supply.

Dave

The markers are incandescent - they are fed directly from the L7806CV, and they run HOT - too hot to touch comfortably.

There's nothing else in the board except the voltage regulator and the capacitors - and a big heat sink. I haven't measured the current draw of the markers, but the 7806 heat sink is also too hot to touch when the track is at 18v.

There is no DCS choke. Either MTH didn't bother, or maybe the main caboose incandescents (which are powered directly from the track) keep DCS happy?

I'll try to tinker with it today - see if adding different kinds of capacitance on the input does anything, or see if what happens if I replace the regulator with a switching module.  I do have 125 kHz RFID readers installed on my track, so I'm a little hesitant to be putting switching power supplies in.

If considering modifying the circuit, see this thread where cjack switched to LEDs for both interior and marker lights.

https://ogrforum.com/t...h-ca-1-marker-lights

Lighting markers should not be this complicated. I am surprised the markers get hot even if they are incandescent. You could keep the regulator circuit and switch to LEDs or use 1.5v grain of rice bulbs and run two pairs of diodes back to back across the bulb leads in series with the interior lamp(s). Thats how I have been doing it for decades. No hum or heat.

Pete

No other components? There must be at least a diode to convert the incoming AC.

If it's that hot something is amiss. Did it get hot when it didn't hum without the bulbs?

To be clear, I was not suggesting a switching supply, I was implying that the manufacturer may have set it up that way.

If it truly doesn't have any other components present, I would try a diode at the input

as a test. This will raise the input voltage, but at 18VAC track it should be within the regulators max spec.

Dave

Adding a 220 uF cap on the input eliminated the buzz.

But the 7806 is running so hot already... I'm a little afraid that a bigger filter on the input will force it to drop even more voltage.

So I'm replacing the MTH board with a switching module. No buzz, and I can turn the voltage to the lamps down a little.

Dave - sorry, there's a bridge too. 

I haven't tested the temp with the markers unplugged.  But I think the regulator should be dissipating about 2W, if each 6V lamp pulls 50mA (as mentioned in the thread Stan linked to).

Thanks for the suggestions Pete & Stan - I'll keep that in mind if I feel like tearing out the existing lamps.

Professor, did you ever determine exactly which component was buzzing?  Perhaps using a popsicle stick to touch various parts?

To an earlier point you made, going from 10uF to 220uF on the input (i.e., right after the bridge) places a greater burden (22 times larger) on the high-frequency DCS signal.  Maybe you don't plan to run this on DCS, but if you have them handy I'd put in a 10 cent 22uH inductor while you have the soldering iron hot.

I assume you're using one of those 99 cent eBayu DC-DC converter modules - many of them have at least 100uF of input capacitance so again the DCS inductor would be in order.

It was hard to tell where the buzzing was coming from, but it sounded like it was from the trucks, not the regulator or the lamps. Pressing down on the caboose reduced the buzz.

I did add the inductor before the bridge that feeds the switching module. Might the issue of high capacitance with DCS be the reason MTH designed the circuit with a small cap on the input?  But then why not just add the inductor?

I have seen cabooses with poor ground because of the blackened coating and paint on the screws, side frames, etc...  Not sure if that is in play.

MTH has used simple CV boards since the beginning to light 6V bulbs in PS-1 Diesels.  Just simple full wave bridge followed by the VR with a cap.  150-300ohm resistor if driving LED markers.  G

Do you have a way to shut this off temporarily to make sure that it isn't causing the buzz?

I'd like to hear more about the RFID and what you are using it for. Maybe start a new topic. Thanks.