5 volt regulators and layout problems

yarisboy

Alright, here is the back story. I made a board with Propeller Platform and Schmartboard features combined with battery holders on the back. Please find layout attachment. So far the board has a 2.1 mm barrel jack, a slide switch, and a regulator on it. The first regulators from Digikey put out 11.6 volts. The second batch from Digikey also put out 11.6 volts. The same regulator from Newark just got here and, you guessed it, it puts out 11.6 volts. After cleaning the board with 99.96% dry IPA the regulator still puts out 11.6 volts. The regulator is an LM2940-5.0. What did I do wrong on the layout. Do Digikey and Newark buy parts from the same place? I've never had a problem with the larger through hole versions of the same part. While people with more experience than me inspect the file I'm going to start board #2 and see if I get the same result. Yes, I got the same result on the second board. I don't have a power filter capacitor for the incoming DC. Will this keep a regulator from working properly?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
MOORE'S LAW: The capabilities of electronics shall double every 18 months.
cloyd's corollary: Hardware is easy, software is hard.

Post Edited (yarisboy) : 3/25/2010 12:47:21 AM GMT

ElectricAye

Did you check the voltage being output by the regulator while there was some sort of load on it? I think I've seen voltage regulators behave this way when they have no load whatsoever.

yarisboy

Good point. I'll solder on the 3.3 volt regulator. That will give it a load. Might be similar to the BS2. If I didn't install high resistance pull down resistors my outputs would float high.

Results: With the 3.3 volt regulator the 5 volt regulator is putting out more than 12 volts now. The 3.3 volt regulator is putting out 11.6 volts with or without a load.

I may have to scrap these boards and use the same regulators used on the Schmart board. The rest of the parts for these boards will be here tomorrow but I'm not optimistic. The 3.3 volt regulator is a LM3940-3.3/NOPB. I'll put on all the caps before I give up and open the trash bin.

Post Script: The data sheets recommend capacitor values to use to use on these regulators. They state that they must have capacitance to achieve stable regulation and transient response.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
MOORE'S LAW: The capabilities of electronics shall double every 18 months.
cloyd's corollary: Hardware is easy, software is hard.

Post Edited (yarisboy) : 3/25/2010 3:22:53 AM GMT

Brian Riley

yarisboy said...
Alright, here is the back story. I made a board with Propeller Platform and Schmartboard features combined with battery holders on the back. Please find layout attachment. So far the board has a 2.1 mm barrel jack, a slide switch, and a regulator on it. The first regulators from Digikey put out 11.6 volts. The second batch from Digikey also put out 11.6 volts. The same regulator from Newark just got here and, you guessed it, it puts out 11.6 volts. After cleaning the board with 99.96% dry IPA the regulator still puts out 11.6 volts. The regulator is an LM2940-5.0. What did I do wrong on the layout. Do Digikey and Newark buy parts from the same place? I've never had a problem with the larger through hole versions of the same part. While people with more experience than me inspect the file I'm going to start board #2 and see if I get the same result. Yes, I got the same result on the second board. I don't have a power filter capacitor for the incoming DC. Will this keep a regulator from working properly?

The Propeller Platform is laid out for VRs similar to LD1085 type where ground is one of the outer pins and IN is the middle and Out is the the other outer pin. The LM2940 series is like the LM7805 with GND in the middle and In and Out on the outers.

If you were not aware of this and just copied the pinouts from the Prop Platform, it would explain the voltages you are getting.

cheers ... BBR

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
cheers ... brian riley, n1bq, underhill center, vermont
The Shoppe at Wulfden
www.wulfden.org/TheShoppe/

Drone

Maybe the equivalent series resistance (ESR) of the cap you are using is outside what is specified in the regulator data-sheet? If so the regulator can become unstable. See National Semiconductor AN-1148 "Linear Regulators: Theory of Operation and Compensation".

www.national.com/an/AN/AN-1148.pdf

Phil Pilgrim (PhiPi)

yarisboy,

You would get more help if you could convert your .pcb file to something (e.g. .pdf, .gif, etc.) that everyone can read.

-Phil

ElectricAye

I converted his pcb file to bitmaps. I'm not sure this will help.

EDIT: Hmmm... I just looked at them on my other computer and there are no connecting lines whatsoever. I guess it's really not any help at all. Sorry.

Leon

The schematic would be more useful.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Leon Heller
Amateur radio callsign: G1HSM

hover1

PDF File

Leon

Schematic?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Leon Heller
Amateur radio callsign: G1HSM

Brian Riley

Phil Pilgrim (PhiPi) said...
yarisboy,

You would get more help if you could convert your .pcb file to something (e.g. .pdf, .gif, etc.) that everyone can read.

-Phil

I brought his file into ExpressPCB under Bootcamp on my iMac and took a screenshot


cheers ... BBR

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
cheers ... brian riley, n1bq, underhill center, vermont
The Shoppe at Wulfden
www.wulfden.org/TheShoppe/

Leon

The capacitors on the regulator outputs should be as close as possible to the devices. I always use capacitors on the inputs, also.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Leon Heller
Amateur radio callsign: G1HSM

Post Edited (Leon) : 3/26/2010 12:18:27 AM GMT

ElectricAye

hover1 said...
PDF File

Jim,
that's pretty good. How did you convert his file into PDF?

hover1

I have Adobe Acrobat Professional. I just print to a PDF instead of a printer. I use Acobat a lot because I need to pull a lot of documents into one PDF document. I here Open Office has a PDF print routine, but I have not tried it yet.

Jim

ElectricAye

hover1 said...
... I just print to a PDF instead of a printer....

Okay, I get it now. I'm surprised the details came out so well doing it that way.

thanks,
Mark

mojorizing

Stan,

looking at your pcb, I'm not seeing how you're bring 12vdc to your 5vdc reg. unless you've update the traces in the area of the 2.1mm jack.· If you have an unpopulated board still, put your jack on, power it up and see if you have 12v on your 5 and 3 volt traces.· Nothing obvious to me except the jack connection to the pcb.



Kevin

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Bad spellers of the world untie!

Phil Pilgrim (PhiPi)

Are you sure you've got connectivity between the ground pin on your LM2940 and the ground plane? Measure it with an ohmmeter. Also, the connection from the ground pin on your input cap and the groundplane is extremely tenuous -- perhaps non-existent in reality. Just because there's a skinny wisp of copper there doesn't mean it's a useful connection.

Also make sure you don't have any shorts between VIN and V50. Measure it using an ohmmeter with the power disconnected.

What is the value of the '2940's output cap? It needs at least a 22uF low (but not too low) ESR (e.g. tantalum) cap to be stable.

-Phil

yarisboy

I need to install capacitors recommended by the data sheets. The ones I ordered are not right for these regulators. Also when the slide switch is in the battery position I'll have to put on a 4.7 micro-farad cap, most likely tantalium. The electrolytic on the output of the 5 volt regulator must be 22 uf or greater and the 3.3 volt regulator needs 33 uf or larger. I'll double check the ESR spec on both before buying. If I buy the 4.7 micro in a surface mount the first three boards will still be clean on both sides (no patch jumpers or drilling). When this get posted as a finished project I'll include the rev-B layout along with the schematic and the BOM. The surface mount soldering is going well. Some people here posted some excellent youtube links on the subject. There is no solder mask on these proto-boards so one must be sparing on the solder. I also learned when to use a solid through hole to the ground plane and when to use a thermal pad to ground. That lesson learned is already on the rev-B layout. I accordance with Phil"s cautions I rang out the grounds with a meter before installing parts. I'll have the right capacitors after work tomorrow. Doing cheap boards with out the solder mask makes one install metal clad parts high (such as the crystal). I'm wondering what I would use in such areas for these three board proto trials. Nail polish?

The grounds on the regulators are punched through to the ground plane with multiple vias for both electrical and thermal reasons. I'm cautious about heat sinking with regulators that can source one amp and are less than half the size of the through hole versions I'm familiar with.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
MOORE'S LAW: The capabilities of electronics shall double every 18 months.
cloyd's corollary: Hardware is easy, software is hard.

Post Edited (yarisboy) : 3/26/2010 6:12:40 AM GMT

Leon

You can get plastic spacers for mounting crystals that need to be insulated. They also provide some shock protection.

What about the excessive track length between the capacitors and the regulators that I mentioned earlier?

A schematic would still be useful.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Leon Heller
Amateur radio callsign: G1HSM

Post Edited (Leon) : 3/26/2010 1:42:31 PM GMT

yarisboy

I installed four capacitors I picked up at Circuit Specialists today. The regulators are working fine now. The .47 uf tantalium was installed as close as possible to the input for the five volt regulator. The 1000 uf capacitor for the jack (I could get today) is a higher voltage rated one so its a tight fit in the case. When the 16 volt versions come in they will fit better and they have a better temperature rating. Its Arizona. The 22 and 32 uf capacitors are bent over to fit but the values worked as advertised in the data sheets. I should be able to load the test programs from the Propeller manual tomorrow. On a rev-B layout I'll be able to tighten up distances from the regulators to the capacitors. Because of surface mount technology I have plenty of room to move things. I may even order more surface mount components just to sharpen the skills some more. Actually, because the connections don't go through the board arranging the parts locations are easier. It seems strange and bass-ackwards to do on a surface mount board what we used to do on a bread-board. On a new layout I will avoid putting traces under the steel case of the crystal and the slide switch.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
MOORE'S LAW: The capabilities of electronics shall double every 18 months.
cloyd's corollary: Hardware is easy, software is hard.

Leon

It's the output capacitors that need to be close to the regulators, the input ones aren't so critical.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Leon Heller
Amateur radio callsign: G1HSM

yarisboy

Thanks a bunch guys. My second prototype is successfully communicating with the propeller tool. I loaded the do nothing object to RAM and then EEPROM. The first prototype may have a solder bridge somewhere or I may have just over heated the QFP-44 during my rooky SM soldering. I have about a half page of new rules to follow on the rev-B layout and guidelines for component selection while I'm surfing the Mouser catalog. When you are building to fit a given enclosure everything matters.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
MOORE'S LAW: The capabilities of electronics shall double every 18 months.
cloyd's corollary: Hardware is easy, software is hard.