Narrow Version of DIY 317 Voltage Regulator Board, with Build Files

Riki, you can't have common ground with a bridge rectifier. They must be isolated from eac other. Refer to the user notes up at the top. Common ground only works with half wave using a single diode.

Rod

Guys, addressing my ignorance, what are some uses/applications for the “common ground” option of these pcbs?  Why would I need to do this?  Thx Ted

Well for instance if you had a lighting circuit that used the engine or car frame as return, and you were using track power to feed the vreg module. This is a common ground, and half wave would be needed.

Rod

Using a diode inline with AC ground Will that work? To separate grounds?

NO!  You're confusing separating the grounds with combining the grounds.

 Thanks John. Hmmm. Ok...half wave it is.

With the boards. R1 holes could be a little larger. Pretty tight. Have to use pliers to pull thru.

TedW posted:

Guys, addressing my ignorance, what are some uses/applications for the “common ground” option of these pcbs?  Why would I need to do this?  Thx Ted

photo hi-jacked and modified from previous OGR thread

In addition to rolling-stock, it could be used track-side if you use/need insulated-rail relays for signals, crossing gates, etc.  It's just that off-the-shelf relays run about $10 each (plus shipping).  But using Rod's board configured for half-wave common-ground with 12V DC output, you could use transformer AC Accessory voltage (e.g., 16V AC) with a 1, 2 or 4 channel 12V DC relay module from eBay free shipping from Asia.  The relay modules are less than $1 per relay so you could rack up some serious $avings!

TedW posted:

Guys, addressing my ignorance, what are some uses/applications for the “common ground” option of these pcbs?  Why would I need to do this?  Thx Ted

If you do anything with TMCC or early Legacy, many of the signals and lights are returned to frame ground.  So, most of the stuff I build either uses half-wave power or has an option to jumper for half-wave power so I can use it in those situations.

Hi all. Just got finished building one of these narrow vreg boards, and I am happy to say it works exactly the same as the more rectangular original version. This build is the full enchilada with 22uH choke, bridge rectifier, 5K trim pot, 1000 uF filter cap. Output range is 1.4 vdc to 25.5 vdc; using a 19 vac brick for AC supply.

All is well as far as I can see. 

Rod

I'm 2 out of 4. Half wave. 

Replaced 317. On 2 .  No change one. Still not regulated .

The other no output.

Building 2 more. See how that goes.

Rod the ones I have r1 holes are to small.

2 more tested good. .

But. R1 holes are a pita...i don't have a drill small enough.

I did see some small drills on Amazon few months ago. Think they were for pc boards.

I also changed where I hook up. 

Neg is across from diode. Leaves the end. Of board with only wires on each end.

Dc+

And AC in.

Riki, sorry to hear you are having problems. The 2 that are not working must have something different in your component wiring, compared to the others that are working.

For the 2 that are working, are you getting the desired output voltages? What are your R1 and R2 values?

Most modern 1/8-1/4 watt resistors have small leads and will fit those R1 hole sizes. Older 1/4 watt and larger 1/2 watt resistors use slightly heavier leads and might be tight. My good friend rtr12 reported those holes being a bit tight on a board he built, but said he eventually got them through OK. I personally have not had this problem. 

One idea might be to cut the resistor leads short and bend them so that you can surface solder them on the top of the board. Or cut one short and one slightly longer and mount the resistor vertically as you had on one board. Tin the pads and leads first. No risk to the board doing it this way. 

Rod

Yes. Voltage is what I thought it should be.

Easier to solder with leads in the hole.

Using 1/4 watt resistors.

Is there anyway to check a 317 reg.  If it works or not?

But. I did replace 2. And they didn't regulate . so ..hmmm. Bad board.s. 

I 5 5 good ones.

3 that don't hold fixed voltage.

It's so unlikely that you'd run across multiple LM317 regulators that are bad that I'm having trouble coming up with enough zeros after the decimal point to express it.   I'd be looking elsewhere for the problem.

Thanks John.

Resistors check ok. Diode  checks ok. Capacitor ?. 

How about a picture of what you have built?  This is such a simple circuit I'm having a bit of trouble with it not working.  Let's see the front and the back of the board.

 I seriously doubt there is a electronic problem more then likely you put a diode or capacitor in backwards! as was said this is a really basic circuit that should not be giving you issues unless you wired the components in backwards or wrong! I noticed one board that you put a large diode in that looks  like  to me that you had the one end of the diode shorting against a lead real close by!

!!

Alan

So that no heat is used on 317. I made a 3 pin socket.

Bojack..anyone used those.?. 

Haven't tested these yet. Later. Today. 

Not sure why you're worried about heat on the LM317, I've never seen one fail from being soldered.  I used to get my lighting boards in without the thru-hole part (the LM317) installed and soldered them in myself.  I never lost one to heat, I think the first 1500 boards were done that way.  Now they're soldered in by the PCB house, but I've still never had a bad LM317.

You're solving problems that don't exist!

Never heard of Bojack before, but they are likely a repacker, buying offshore in bulk and reselling here. FWIW I have bought many offshore 317's and never had an issue. 

Rod

John. This way if I don't get a . Fixed voltage I can swap out 317. To see if it is a 317. The 3 boards that don't have a fixed voltage .solder joints look ok. But that's the only other thing that could be causing it. I can't see any solder that may have touched where it shouldn't be.

I use a magnified headset 

I did find a diagram a guy did to test the 317.

Need like 5 different resistors and switches.  Not sure why he didn't use a trim pot.

riki posted:

...

But. R1 holes are a pita...i don't have a drill small enough

If you were messing with a drill bit trying to enlarge a hole there is the possibility the pad's integrity was compromised.  That is, you have the ring on the top and bottom of the board...but there is a conductive column on the wall of the hole.  Yanking tight-fitting resistor thru the hole can also affect the pad's integrity.

As others suggest it's hard to believe you destroyed a LM317 from heat.  Presumably you clipped the 3-terminal LM317 from a non-working board and plugged the suspect LM317 into a known good board with a socket.  I'd think that would be suitable in this case for an LM317 test circuit. 

Stan. Don't have a drill that small.

Had to work and pull the resistor lead down. With pliers.

The clipped 317 I was seeing how low I could get it. That is new.

Still have not tested those boards yet.

Riki, Stan has made a good point about damaging the tiny copper sleeves through the board that connect the top and bottom pads, by pulling leads through.

The hole pointed out in this picture:

is the one that could be affected because it has a connecting trace on the top of the board. Presumably you soldered from the bottom, and its possible this trace is not electrically connected. An easy fix is a dab of solder on the top trace to assure connection. The other end of R1 is not affected because the connecting trace is on the bottom side of the board.

Regardless, the new gerbers that I did up for you have a much larger pattern for R1, with larger holes, so they should work way better. 

Rod

Ok. Thanks rod....

I was wondering about that....i will solder the top..

Thanks...

Ok..

I'm 99.9999999999999. Sure it's the voltage regs.

Did resolder . no reg.

Changed 317. Held fixed voltage.

My other boards. The ones that didn't work.. No change there. So got to be voltage regs.

The 317 that I cut. Which is new. Didn't work in either board I made with sockets.

So that. Is  My... Scientific .. Result. That some of the regs are bad.

So while everything is set up and fresh in your mind why not test the remaining 15 Lm317s in the socketed board. 

Good plan Stan.  Think I'm. Keeping those as test reg boards.

Also I'll look for different 317.  But the price was right.

But as the saying goes...  You get what you pay for. For somethings.

Riki, I think the test socket is a brilliant idea, thanks for sharing. As I understand it from our offline discussions, you used an 8-dip socket cut in half lengthwise, then cut one pin off, correct?

This would make it easy to build a one-of board for testing various TO-220 and TO-92 packages. One socket would work for 317T and 317L because they have the same pinouts. Then you could have another socket for 78xx and 78Lxx chips, and maybe others? You could also set the board up with an attached digital voltmeter to make that part easy. Hmmmmm, I think I feel another board design looming....

Rod

Yes rod. It's an 8 pin. 

Thanks for all your help.

You are more than welcome. I would be very interested in how many of the remaining Bojack 317's don't pass muster; just saying.  Stan's idea of testing all those remaining is a good one IMO.

Rod

I'm stunned that you got so many bad regulators!  I've been buying eBay parts for years as well as other cheap sources.  I don't often get a bad part, and certainly not a ton of them.

Not done yet John. Have 12 still to test.

2 out of 12 were bad.

These I know where these are made.

OK, I admit, I scanned through this thread so I'm uncertain of the details, but I am also floppin' on the dock here.  (fish outta water an all that)

I have an older Lionel caboose, with one pick up, that is blinking over FasTrack switches.  I wanna stop the blink.  What, in 1, 2, 3, etc., steps do I need to do and where do I get the stuff to do it with?

Go...

You can do it the easy way and buy a couple of these, you only need one for the caboose and some LED's to light it.  Track power to the AC connections and your LED(s) to the DC connections.

 LED Lighting Regulator