Odd output from Elenco XP-15K Power Supply

[tomherrick]

Bought the kit and put it together yesterday. First step afterward was to check the VDC output. When the pot was turned all the way up to the 15V mark I got -0.53V. Yup, negative. When I turn it all the way back to zero Volts it reads -0.65. Voltage across P2/P1 reads 17VAC; same reading across P2/P3. When I unplug the unit, the LED slowly dims to off and the multimeter value dropped from -0.65 to -0.32 over the course of a minute or so and continues to drop very slowly.
I've been over and over the instructions and rechecked the following: polarity of the diodes, caps and LED; output connections from the transformer to P1, P2 and P3; ensured that the regulator was oriented properly. I went through the test procedure and shorted across the output; the unit did not turn off.

The error is likely mine and I keep looking past it when I QC my own work... Suggestions???

Thanks.

[tpowell1830]

Hi Tom, welcome to the forum.

Take some very crisp pictures of the board from both sides. Also, a shot of the schematic and post them in another post so that we can see your work and the schematic. You have a size limit for your jpg's (I can't remember what size), but the uploader will complain if they are too big.

[tomherrick]

Here are photos of the circuit board and the schematic.

[T3sl4co1l]

Easy enough to inline as well. Click the attachment, right-click the image, copy link.  Paste into an img tag:



Hah, looks like it's drawn in Multisim of all things.

Tim

[Zero999]

As posted in the other thread:

Diodes D5 and D6 make up a poor man's zener diode and the voltage between ground and the point where the diode connects to R4 should read -1.4, give or take.  Black wire of you meter goes on ground.  This negative voltage is what allows the regulator to get down to 0V.

I don't know the AC voltage between the blue and yellow wires of the transformer but about 1.4 times that value should be measured at pin 3 of the LM317 as a DC voltage.

Pin 1 of the LM317 should move between some negative value, like -1.4V, up to some positive value, nearly 15V,  but I haven't calculated the exact range.

Yes, this is a classic way of getting near 0V from the LM317, which can normally only go down to 1.25V.

It could be a problem with the potentiometer.

Bought the kit and put it together yesterday. First step afterward was to check the VDC output. When the pot was turned all the way up to the 15V mark I got -0.53V. Yup, negative. When I turn it all the way back to zero Volts it reads -0.65. Voltage across P2/P1 reads 17VAC; same reading across P2/P3. When I unplug the unit, the LED slowly dims to off and the multimeter value dropped from -0.65 to -0.32 over the course of a minute or so and continues to drop very slowly.
I've been over and over the instructions and rechecked the following: polarity of the diodes, caps and LED; output connections from the transformer to P1, P2 and P3; ensured that the regulator was oriented properly. I went through the test procedure and shorted across the output; the unit did not turn off.

The error is likely mine and I keep looking past it when I QC my own work... Suggestions???

Thanks.

What are the voltages from pin 3 to 0V and pin 1 to 0V?

[tomherrick]

"What are the voltages from pin 3 to 0V and pin 1 to 0V?"

Pretty sure I know what Pin 1 and 3 refer to, but am lost at 0V unless it's Pin 2...

I'm still boggled by getting a negative DCV reading at the output measuring around half a volt.

[rstofer]

0V refers to 'ground'.  P2 on the transformer or the test jack (black) on P4.

In general, any number we give you will be in reference to 'ground' because it is too hard to calculate from a different reference.

Pin numbers are referring to the pins on the LM317

Pin 1 = the adjustable control voltage - variable from about -1.4V to perhaps 15V - these aren't exact numbers
Pin 2 = output voltage
Pin 3 = the voltage going into the regulator.  This might be on the order of 18V

We're trying to dissect the problem.  Is there about 18V on pin 3?  If so, everything to the left of the LM317 probably works.  It is still possible the -1.4V (give or take) on the junction between R4 and D5 is not correct.  So, there are two measurements.  They are the first numbers required for troubleshooting.

Next we're looking at pin 1 and expecting to be able to vary the voltage between some negative small value (like -1.4V) and some positive higher value - perhaps 15V.  There's another set of measurements.

Finally, we want to know if the regulator works by measuring pin 2.  I assume it is not working because you have said it was stuck low.

We don't know how you took your original measurements but would probably guess that you just measured the output across the jacks.  Measuring a small negative value might be realistic if pin 1 is -1.4V.  Just make your measurements with reference to the black output jack.

So, make some measurements and post them.

[tomherrick]

I've taken these measurements with photos of results:

Voltage from P4 to P3 is 17.1V


Voltage from P4 to P1 is 17.1V


Output voltage with pot turned to 0V is -0.65V


Output voltage with pot turned to 15V is -0.53V

[rf+tech]

Why is an LM337 installed, when the schematic calls for an LM317?

RF+ Tech

[tomherrick]

I installed the components provided in the kit. It appears to be an LM317T.

[mvs]

I've taken these measurements with photos of results:

Voltage from P4 to P3 is 17.1V

Voltage from P4 to P1 is 17.1V

P4 is connected to P2, so you are measuring AC voltage from the transformer.
Have you switched your DMM to DC voltage mode to measure DC voltage at the output?

Check DC voltages across capacitors. You may post only values, no need to have them on photos.

[tomherrick]

A previous respondent indicated wrote: "We don't know how you took your original measurements...", hence the photos to make it clear. I switched the DMM from VAC to VDC to take the output readings; guess I should have included the DMM selector switch in the photos.

[rf+tech]

Thanks for the clear photo. That middle digit sure looked like a 3.

What is the voltage on the LM317 adjust pin, at minimum and maximum?

Please refer to the datasheet for correct pin identification.

RF+ Tech

[tomherrick]

MVS, not sure how to properly measure voltage across the caps. In VAC or VDC? The only steady reading I've gotten so far is across C1 at 49.5VAC. When I switch to VCD readings I get a myriad of results that keep changing. Am I looking for AC or DC values? Should I expect the values to change/degrade?

rf+tech:  On the LM317, Pin 1 is labeled ADJ. When pot is set to 0V the value is -0.57V; set to 15V it's -1.30V

[rstofer]

Just to make sure everything is clear, measure AC voltage when looking at the transformer, DC voltage everywhere after the 4 diodes.

For fun, measure between blue and yellow and blue and the other yellow.  They should be the same but post these as the AC voltage results.  I think you did this and got 17.1 VAC but I just wanted to be certain.  Pretty sure that's valid.

49.5 VAC on C1 is invalid because it is a DC voltage - it's after the rectifier diodes.  I would expect a number less than 20 VDC.

Voltage across ALL capacitors is DC.  They are after the rectifier diodes.

Is the regulator getting hot?   One thing that will cause the DC output voltage to vary is the built in self protection from overcurrent.  This will reduce the voltage substantially until overload is removed.

[rf+tech]

What is the DC voltage across pins 1 and 2 of the LM317?

To be clear, place the DMM black lead on pin 1 and the red lead on pin 2.

This is expected to be 1.25 Volts DC, regardless of the potentiometer setting.

Also, several requests have been made for the DC voltage on LM317 pin 3. Please measure this, with the DMM black lead connected to P4 common/ground/minus terminal. And for all future measurements, unless requested otherwise.

[tomherrick]

DC voltage across pins 1 and 2 of the LM317 is -0.10.
DC voltage across pin 4 and pin 3 is -0.83 @ 0V setting; -0.39 @ 15V setting.

[rstofer]

Pin 3 to the black terminal should be 1.4 * 17.1V (assuming this is AC voltage between the blue and yellow or about 24VDC.  This is really high for a 15V power supply.  Make sure you measured 17.1VAC and check pin 3 as VDC.

One other thought:  The regulator takes between 3.5 and 10 mA of load current in order to regulate.  At very low voltages, say 0.1V, if your device needs 10 mA, you need a 10 Ohm load.  At very high voltage, say 15V, for 10 mA,, you need a 1500 Ohm load.  Notice that the load resistor varies widely over voltage.  You surely wouldn't want to use a 10 Ohm load on 15V as that would be 1.5 Amps.

I used 10 mA for those numbers because that is worst case.  Ordinarily I would expect your actual load to be adequate.

[tomherrick]

Here are the measurements:

P2 to P1: 17.0VAC


P2 to P3: 17.0VAC


P4 output negative to LM317 pin3: -0.51VDC

[rf+tech]

Now we're getting somewhere. The AC input side looks as expected and the focus is now from D1, D2, D3, D4 and to the right side of the schematic.

Please answer re is anything getting hot, especially the LM317.

Measure DC voltage from P4 output to D1 cathode (banded end) and from P4 output to D3 cathode.

Edit:

A second look at the back side seems to show a suspicious looking solder joint at P4, as noted in attached photo.

[tomherrick]

OK, now it's getting weird...  DMM on VDC, neg to P4, pos to cathode side of D1, I get a momentary reading that differs each time then is replaced with the reading you see below. Momentary readings were 15.45, 17.24, 11.72, and 17.50 on D1. Momentary readings on D3 were 14.89, 19.82, and 19.21. All were shown for less than a second then the DMM display shows "1  .  ".  Golly, I hope that helps...   

[rstofer]

That measurement on pin 3 of the LM317 is very wrong.  Unfortunately, troubleshooting beyond this point may require some disassembly and that is often problematic on single sided boards.  The pad tends to lift off the board.

If I were making the tests, I would remove the LM317 and then check the Input and Adjust pads.  The Adjust won't change because it doesn't have the feedback from the output voltage but it should be about -1.4VDC.  And the Input should be about 24VDC.

Before we worry about regulating, we need to get the Input level correct and we need to know that we have that negative rail for the Adjust circuit.

SolderWick is your friend and I would use the wider variety.  Make sure there is plenty of solder on the pad before you apply the wick.  I know, counterintuitive.  If you have to buy SolderWick, buy some paste flux as well.  Put that over the solderball on the pad to increase the amount of flux beyond what SolderWick supplies.

Were it mine, I would cut the leads off the LM317 and remove them one by one.  Less chance of damage that way.  Of course, I would need to buy a couple of spare LM317s...  I would absolutely do it by cutting the leads.  I have destroyed too many pads...

https://www.digikey.com/products/en/soldering-desoldering-rework-products/desoldering-braid-wick-pumps/265?k=solder+wick&k=&pkeyword=solder+wick&sv=0&v=473&pv1698=6&sf=0&FV=ffe00109&quantity=&ColumnSort=0&page=1&pageSize=25

I would use the 0.1" braid for this operation.

https://www.digikey.com/products/en/soldering-desoldering-rework-products/flux-flux-remover/266?k=paste+flux&k=&pkeyword=paste+flux&sv=0&v=473&pv183=3336&pv771=256&pv771=259&sf=0&FV=ffe0010a&quantity=&ColumnSort=0&page=1&pageSize=25

LM317

https://www.digikey.com/product-detail/en/texas-instruments/LM317T-LF01/LM317T-LF01-ND/3640769

[rstofer]

Since the voltage should be 24VDC, I would expect it to overload a 20V scale.

[rf+tech]

A less invasive technique would be to take a few more measurements. Up to this point, all DC measurements have been from the back side of the negative output binding post. And there appears to be a suspect solder joint where the wire from this binding post connects to the board.

The next step would be to measure DC voltage across C1 (2200 uF), with the DMM black lead to the striped side of C1 and red lead to the other side of C1.

Set the DMM to the 200 VDC range.

[tomherrick]

These will have to wait a bit. I have a long list of tasks that I haven't gotten to yet. This 12V power source was simply intended for use with a marine GPS chartplotter to load maps and waypoints. This has turned into a stumbling block rather than an interesting intellectual exercise.

That said, I hope to get back to it soon and thank everyone for their thoughtful input. Very nice of you all.

I'll be back soon to pick up where we left off...