Build a Dual 15V Power Supply Using Off the Shelf Modules for Under $50




Introduction: Build a Dual 15V Power Supply Using Off the Shelf Modules for Under $50

About: Ex electronics tech, now sales rep living in the snowy region. Always tinkering fixing or building something in the workshop. I play guitar, I fix guitars, I build guitars. Any project that has wood, electroni…


If you're a hobbyist that deals with audio, you'll be familiar with dual rail power supplies. Most low power audio boards such as pre-amps require anywhere from +/- 5V to +/- 15V. Having a dual voltage power supply makes it just that much easier when prototyping designs or just general repairs.

This power supply is easy to put together as it generally uses off the shelf module boards with the exception of the regulator board, which you'll have to build yourself. However there is a reason behind that which I'll come to later on.

The regulator board used boasts voltages from +/- 1.25V to 37V (depending on your input voltage). I only need +/- 15V, so an input power supply a couple of volts above that (around 19V) is fine. LM317 and LM337 voltage regulators can also pump out around 1.5A ea (depending on how much voltage they're dropping), so the input power supply's current rating also needs to be higher than this. That's why I chose two laptop power supplies to supply the input voltages. They output 19V and around 3.4A, which is more than enough to supply the regulator board. Not to mention they're cheap as chips.

I also wanted a linear power supply as they generally have less DC ripple on the output (although not as efficient as a full switch-mode power supply). Using a switch-mode input power supply to drop the 240VAC to 19V is cheap and effective. Their switching is also generally above the audio band so doesn't affect power supply noise going into your test pieces. The linear regulators will filter out most of the residual DC ripple. So, you pretty much get the best of both worlds.

The meters used can measure voltage and current (0-100V and 0-10A), are dual colour for easy reading.

With a few modifications, you can turn a bunch of parts into a very useful bench top power supply.

Note: One thing this power supply doesn't have and that is a constant current regulating control. The LM317/337 regulators themselves have some over current protection, however I wouldn't run them too long this way. That's why the load switch was placed in this project. So if that's import, you could use a different regulator board to suit your needs.

Step 1: A Word of Warning & General Notes

240V Wiring & Laptop power Supplies:

As this project uses high voltages (240V), they can be quite lethal if you get it wrong. If you're unsure on how to hook up high voltage components, or are not comfortable working on live equipment, my suggestion would be not to attempt this. I take no responsibility if you kill yourself. I don't want to hear from you after you've died saying Pete, I electrocuted myself and now I'm dead - OK??

Now with that being said, you have a couple of other options:

1. Just use the laptop power supplies in their supplied form and use some DC power connectors on the back of the box. It just means you have to plug in two laptop power supplies - but it's a much safer option. However, you will need to find another solution to powering the LED meters as they require separate supplies as well.

2. You could mount the laptop supplies in the case, and just cut the 240V plugs off and wire them directly to an IEC socket on the back. However, you will need a bigger case than what I've used and again, it's got live connections so still not that safe..

LED Panel Meters + Supply Voltages:

There are several types of LED meters on the market. They all essentially do the same thing, however their connections are not always the same. Going off the gauge of wire is not always guaranteed. When ordering, try and get their wiring diagram. Generally the two thick wires will be the current shunt meter. The other three will be meter power (to power the display which is red/black) and a yellow voltage sense wire to measure the voltage.

What you will notice with the meters is that they have a common earth or 0V point (the black wires are connected together internally). For this particular project that is no good. This is why the meters are powered separately via two small power supply boards (240VAC to 12VDC module board). You also have to use two boards to power, otherwise you'll be shorting outputs when using the power supply. One other essential reason is that the LED meters require a minimum or 4.5V to run. So if you turn your output down to 1.25V from the regulator board, the meters won't power on.

Step 2: Bill of Materials

This is what you'll need. You can buy all of this off Ebay, Amazon or Aliexpress. I bought it all off Ebay

- Plastic Case (I've used a plastic instrument case) - $12-15

- 1x LM317/337 Regulator Kit Board - $10

- 2x 19V 3.42A Laptop Power Supplies - $6.75 ea

- 2x 240VAC to 12VDC 450mA switch-mode step down transformer boards - $1.50 ea

- 2x Voltage/Current Panel Meters 0-100V/0-10A - $3.50 ea (cheaper in bulk and available in different colours)

- 2x 10K ohm multi turn pots + knobs to suit - $2 ea (you can use the supplied pots, but multi-turn are easier to set)

- Miscellaneous and general hardware: 240VAC switch (I used one with a 12VDC LED light), binding terminal posts (6), IEC socket, fuses and fuse holders (3), small off cut of aluminium angle (2), stand offs (6), general lengths of wire and heat shrink - probably another $5-10

Note 1: The fuses to use will depend on how much current you intend to use. I'd suggest 1-1.5A for the two regulator boards and 0.5A for the 240V supply. You could go lower as well as you won't be drawing 7A from both supplies.

Note 2: The most expensive part of the build is the case. So if you can find one cheaper or want to roll your own it will save you a few bucks.

Note 3: There are a few brands of multi-turn or precision pots available The one sent was a Bochen branded pot, that has specific knobs available and doesn't use standard rough knurl knobs. It doesn't particular matter which type you use, only that you're able to get knobs to suit.

Note 4: I bought these laptop power supplies as they were only about $6ea. Save a few bucks again if you happen to have a few old ones laying about.

Step 3: Schematics & Wiring Diagrams

First image is the original schematic for a stock regulator board, with input caps and rectifier included, using an AC 12V-0V-12V transformer to power the board (for this power supply we don't use)

Second image is the wiring diagram for all individual boards to connect together

Third and Fourth images are stock wiring diagrams for the panel meters (I used) showing different configurations to power and measure. Essentially in this project, we're using the fourth diagram.

Step 4: Laptop Power Supply

Why 19V Laptop Power Supplies?

The reason for this is that the regulator board originally was designed to run off a dual 12V AC transformer (12V-0-12V). However, if you look at the cost of one of these either from ebay or in you local electronics shop - they're around $30AU. Two laptop supplies come in at half that.

If you want a higher voltage out of the regulators, just use a higher input power supply. Remember the regulator boards will output +/- 37V, so the input can be a few volts above that. Just remember though, the higher the voltage differential (input to output), the more heat that is produced by the regulators. For example: if input voltage is 35V and output is 5V, there's going to be a lot of heat developed and you might need bigger heat sinks and/or a fan.

Prepping The Laptop Supplies

For my build, I took the supplies out of their cases as I needed them to fit in the instrument case. If you're just going to use the laptop supplies as is and use DC connectors you can skip this step.

What you'll need to do is crack the plastic case. Use a flat screw driver and carefully pry the edge until the top comes off. Then remove the circuit board assembly.

In the 2nd photo, I've drilled a piece of angle aluminium and drilled some holes into the side of the supply (I believe I used the existing holes in the supply). Be careful not to damage any components while you're doing this. I've also drilled some extra holes to screw mounting posts to it and attach the assembly to the bottom of the plastic case. Using the angle made it a little more sturdy than just using mounting posts.

The wires coming out of the board looked a little light, so I've changed them over to heavier gauged wire. De-solder the old wires, insert the new wires through the top of the board and solder them in place on the bottom of the board (in hindsight I should have used a lighter gauge but longer lengths as it was difficult having so many wires connect to the same points).

The laptop supplies also have a power LED. They're not needed, however you can keep them in if you want confirmation that each supply is effectively working (they'll die out if there's an issue with the supply or the amount being drawn). I kept them in for easier fault finding.

Note: You should use the same type of laptop power supply. The reason being that if the voltages are out by a little, they can tend to sink current into themselves and run away and then blow. Generally, it shouldn't be a problem if you use the same supplies. However, if you're worried or want extra protection, you could place a couple of power diodes (such as IN4004 or IN5404) reverse biased across the outputs of each supply (so cathode to positive, anode to negative). This will stop each supply from sinking any current from the voltages being slightly off or if one supply powers up before the other.

Step 5: Building the LM317/337 Regulator & Initial Test

The regulator board comes in kit form, meaning you have to solder it up yourself. There are a few suppliers that will sell them pre-assembled for a few extra dollars. Sometimes, removing components from these types of boards can accidentally rip tracks off. You'll need to remove some components anyway, so it's just as easy to built them in the first place without them.

The first picture shows a completed board (which is what it should look like if you made it stock). The second photo however shows the modifications with the input caps and rectifier removed. I've added links instead to change the input terminal block to accept the +/-19V and direct it to the input of the regulators. You could keep the input caps if you want but they're not necessary as the laptop supplies are pretty good.

You'll also note I've put in terminals for the LED power light and also the pots just to make it easy to remove boards if necessary.

So just assemble the board as in their instructions except for the modifications above.

Once completed, hook it up to a working power supply and verify the output of each regulator stage. Remember, if using a single input supply to test, +/- in (on the +/0V terminals) +/0V out of regulator board. +/- in (on 0V/- terminals), 0V/- out of regulator board. Make sure you can adjust the output voltage (last picture showing external test power supply).

Step 6: Prepping the Case

Measure out how you want your components to sit on the rear of the front and back panels. Remember, it's going to be back to front (I made that mistake myself). Actually, I wanted the mirror image on the front panel. But luckily, I hadn't done the back panel yet, so I just made it fit to the front (or I might have just turned it around 180 deg).

Drill holes using small drill bits first. Then enlarge with a bigger drill bit. If you don't have a large enough drill bits (as I have), you can use a reamer to enlarge the holes (very handy tool).

Once all the holes are drilled, pop the cut outs for the meter panels and file down just enough so the meter and IEC socket fit.

I've also added some labels to the front (using letter sheets). You can get these online, or you can print your own on clear printer paper. Then I just sprayed some protective lacquer over the top.

Step 7: Mounting the Hardware

Once the front and back panels have had time to dry, mount all of the hardware on the front and back panels.

The two laptop power supplies can be mounted to the bottom of the case. Remember to leave room for the IEC socket, fuse and wires to run to the switch at the front. Alternatively, you could mount a switch to the back if you prefer.

Mount the regulator board.

Last but not least, as the 240V/12V power supplies for the meter panels doesn't have anywhere for them to be screw mounted, I've used a blob of silicon to hold them in place. Just make sure you've added input and output wires first!

Step 8: Wiring It All Up

Start by wiring up the 240V wiring from the IEC plug to the switch and also the input fuse holder. Then connect all 240V wiring to the two laptop power supplies and two meter board supplies. Insert a fuse and at this stage, it's probably a good idea to check your wiring and power up, just to make sure all voltages coming out of the laptop supplies are correct (should be 19V each)

Connect the pots and LED to the front panel controls from the regulator board. I've used 2-pin sockets and pins to make disassembly easier at the regulator board.

Now connect up the outputs of the laptop supplies and connect to the input of regulator board. You can also connect the power to the meters. Remember, that the positive of one supply goes to the negative of the other to create a virtual zero voltage point. Again, power up and make sure voltages are as expected - you should have 38V between the input voltages, +/- 19V between 0V at the inputs and some nominal voltage on the output of the regulator board (depending on where the pot is set).

Connect the output of the regulator board to the output fuses and the load switch. Connect the meter current lines (as per the wiring diagram) and then the voltage sense lines from the meter. Insert some fuses and again, test and see if the meters are reading a voltage. Fingers crossed, you haven't let the magic smoke escape!

Note: The meters are probably the hardest thing to get running. Just remember that the current part of the meters runs from positive to negative. Same will happen with the negative voltage - it flows from 0v to negative voltage!

Step 9: Testing and Calibration Adjustments

Once you've verified that smokes not escaping, hook up a reliable meter and check the output voltages on both the positive and negative outputs. You'll most likely find that the LED meters are slightly out (as in photos 2 + 4). As these meters can be slightly out at either end of the spectrum, calibrate them to the voltage you're generally going to use the most or in the middle of a range of voltages. For example if you use 12V a lot, calibrate them to 12V. If you go between 5V and 15V regularly then calibrate at 10V.

If you have two multimeters you can do the voltage and current adjustments together. Otherwise, connect a nominal load to the output, adjust the voltage, then disconnect the meter and put in series with the power supply and swap the multimeter lead (if your meter has separate voltage and current terminals) to measure the current.

On the back of the LED panel meters, there will be two small trim pots to adjust voltage (v-adj) and current (i-adj) (see photo one). It's generally a good idea to load the output with a resistor when calibrating as the output voltage may move a little when loaded.

So adjust the v-adj until the voltage reads the same as the meter. The trimmers are a little sensitive and a small turn can go past where you want it. Just persevere until it's correct

For the current adjustment, I'd recommend using a large heat sinked resistor to calibrate (photo 6). Just make sure it's not lower than what the supply can put out. Each side of the regulator board can supply 1.5A. Calibrating it at around 1A should be sufficient.

Using ohms law V=IxR - so (V/I=R) 15V/1A=15ohms. 15 ohm resistors are a little hard to come by so 2x 8ohm resistors in series will give 16ohms. Measure the resistors - the two I've got measure 8.3 and 8.1 ohms = Total 16.4 ohms.

So, plug the numbers in again (V/R=I) 15V/16.4ohms = 0.914634A - that's the number we'll calibrate to. You should find the meter should display this as well as a double check of your meter.

You'll also have to calculate the power being put into the resistors as you don't want them frying! So, ohms law again P=VxI - 15Vx0.91463=13.72W. Make sure your resistors are each bigger than this value - 25W is good. I've used a couple of 100W which is gold (see photo 6). You can get these off ebay for about $8 for two.

To measure current out of the supply, you'll need to put your meter in series with the power supply and the load resistors. Doesn't matter if the meter is first or the resistors, just make sure the current flowing through the meter is positive to negative (so positive & 0V terminals - positive/negative on the multimeter current terminals). The negative side of the supply should be measured from 0V to negative with positive of the meter going to the 0V and negative of the meter going to the negative of the power supply. If that just confused you - look at the last photo.

Once connected you should see both a voltage and current on the front panel meter. Adjust the current pot on the back of the panel meter until it's reading the same as your multimeter. If you have two meters, have one to measure current (in series) and one to measure the voltage (in parallel).

Now you're good to go.

Step 10: Final Thoughts

While everything fit in the case, I could have played around with the internal layout a little and perhaps moved the IEC socket over to allow for the two laptop supplies to sit 90 deg to where they currently are. The layout should have been mirrored also, as I generally like everything to go from left to right. I've included a sketch of what potentially I should have done.

I used 7.5A 240VAC wiring from a mains cord (because that's what I had laying around). Being that this is such a confined space, I probably should have used lighter gauge 240V wire as the project doesn't draw a lot of current.

I also didn't notice that one of the case screws went straight through where the 240V switch was. In retrospect I should have moved the switch over slightly and probably should have installed the 240V fuse holder on the front panel as well to avoid unnecessary wiring. With a little shuffling, I probably could have put the output fuse holders on the front panel as well, but the front panel was already fairly crowded.

At the end of the day, it supplies the +/- 15V that I require, easy to adjust, is reliable and uses readily available parts.

Future projects

I've also got another dual 0-30V/3A power supply in the works, although this might end up as two separate power supplies (again depending on spacing). This one does have constant current features. I bought these boards at the same time as I couldn't make my mind up which one I wanted so I got both!

There's also going to be the mother of all power supplies - a dual low/high voltage power supply using two regulator boards per side (4). It will switch from a low range 0-30V to a high range 30-90V and 5A! This will be used for testing dual voltage power amplifier boards. Again, it might end up as two separate power supplies depending on spacing.

Be the First to Share


    • Fruits and Veggies Speed Challenge

      Fruits and Veggies Speed Challenge
    • Make It Modular: Student Design Challenge

      Make It Modular: Student Design Challenge
    • Stone Concrete Cement Contest

      Stone Concrete Cement Contest



    7 months ago on Step 10

    You just saved me a shed load of trouble and money. I have a 16 channel posh recording studio mixer that I bought new years ago. It has given me great service until recently when the purpose built, specially designed power supply died. Of course the machine is obsolete, of course the manufacturer doesn't want to know and of course there are no spare PSU boards anywhere, but I did find a schematic that helped me greatly. I am a retired marine engineer and as such, not one to give up easily. I managed to get the +48V and the +10V working, but there is an internal problem in the transformer on the 15,0,15 windings. I am not confident enough to break down the transformer and rebuild the secondary windings so I was looking at a 18-0-18 transformer to supply the + & 1- 15V regulators - problem is that such a beast is too big to fit inside the case and it would have to hang outside like a colostomy bag. I found your article using laptop PSUs and ordered 19v units that fit very nicely inside the mixer case - they are already surge and EMI protected, low heat, no noise - they just sit there and quietly give me volts. I live in Spain and anything remotely luxury is expensive, so your $6 AUD PSUs cost me 18 EU each, but it gave me my mixer back, without any embarassing dongles. Thanks pal - much appreciate your article. Keep up the good work.


    Reply 7 months ago

    Glad you managed to adapt it for your purpose. You might want to also put a couple of large diodes reversed across the -/0V and 0V/+ terminals (so anode -ve to cathode 0V, anode 0V to cathode +ve) just in case one of the power supplied drifts off a bit and runs away! The diodes should generally be rated 2x the max current each supply will draw. Just to be safe. But yes, you can pick up old laptop supplies pretty cheap even if you have to buy two. The generally come in way under buying a ferrite transformer these days. Last time I priced a torroid transformer they were up over $50.


    Question 1 year ago

    Hello! I know you mentioned that this system would have a low ripple voltage. Do you have any estimate about what exactly the value is for that ripple? Thanks!


    Answer 1 year ago

    I couldn't find any spec's from the module sellers. However if you look up the datasheet for the regulators (LM317 or LM337) at max, the typical load regulation is 0.5% of output voltage with a 10uF output cap. Obviously, the higher the output voltage the more noise is introduced. But regulators are are generally pretty quiet.


    2 years ago on Step 10

    I built this but had trouble with the meter(s). There are 28v 1w pilot lights across the outputs so as to keep a minimal current for proper regulation before a load is applied. They are off at 1.5v and glow normally at 27.5v which is high as it will go with my 22-0-22 transformer. Each meter has its own power supply as recommended. My regulator board is the same with the exceptions of LEDs across the DC input, and there are no diodes across the regulators.

    If I turn the positive up to max, its ammeter reads 36mV as expected. But as the negative is turned up, the positive ammeter proportionately reduces in value, and goes to zero when the negative is maxed out.

    The ammeters and voltmeters operate properly only when one of the outputs is left at around 1.5v.


    Reply 2 years ago

    Yeah that would probably work if you were using two separate regulators (and not the dual PS in this project).
    I think you're right though. I just had a bit of look at mine, and low and behold mine's doing something similar (if you put a dual load between +/0V/- only one of the ammeters will work at a time. Looking back at the schematic I can see why (it's where the ammeters placed on the 0V). The problem is you can't have the pos ammeter on the positive line or it senses both the pos & neg current (which is what I had problems with when I first constructed it).
    I'm not too sure whether you could use diodes - I don't think so. Unless you treat the pos and neg sides as two separate PS and put a diode off the pos return to 0V and a separate terminal straight off the 0V for the negative side.
    The only way I can see it may work, is to put the ammeters before the actual regulators (and you'll sense some of the quiescent current also).
    I suppose the simplest solution is not to use ammeters!
    Thanks for the fault detection.


    Reply 2 years ago

    It would definitely work since the current paths would be completely isolated from one another. As far as the meters are concerned, they would "think" they are both positive, yet from the standpoint of the common output terminal on the front panel, there would be a pos and neg on each side. This scheme would also work if a load connected in series from the pos to neg terminals, if you need double the voltage. So the project would not change at all, except the use of positive regulators for each polarity and the appropriate wiring according to the diagram I posted. A minimal load should be always connected to each side to prevent oscillation if loaded in series. Resistors may be used, but require a fairly low value so the regulators will adjust smoothly if there is no load attached to the output terminals. That's why I chose 28v incandescent pilot lights. The resistance is about a tenth of rated when the regulator(s) is adjusted to the lowest setting. Mine is 1.5v. With my scheme, two separate power supplies, like the SMPS ones you used are mandatory. Separate AC step down transformers may also be used. Speaking of those, if the output is DC as with the SMPS, then the rectifiers on the regulator boards should be eliminated. It would also be a good idea to check how much noise is coming out of the front panel terminals since all SMPS supplies have high frequency noise. The LM317 should get rid of it I would imagine, but it should be checked under the full range of load conditions. Otherwise, two separate step down transformers are necessary. I am going to open my 22-0-22 and see if the common tap may be separated. If not, then I would probably use two of those enclosed Radio Shack transformers I see on eBay all the time. The 273-1512B is good for this bench-top supply. It is 12.6-0-12.6, for a total of 25.2VAC without using the center tap. It is also rated at 2A which would yield a DC capacity of over 1A, and would make the regulator output very quiet.


    Reply 2 years ago

    The whole point of using the SMPS from old laptops was to keep the cost down (and to sort of use off "the shelf" parts and modules - so you didn't have to stuff around too much). A 12-0-12V @2A transformer over here runs you about $30 and not much cheaper (with shipping) off ebay. The SMPS were only about $6-7ea at the time, so two came in at half the price of an AC trannie. But the original schematic for the regular board (which is what I posted) suggests using an AC trannie. And yes, in the project I ripped out the rectification off the regulator board and jumped them over (as in the photos).
    The output caps should provide a small load anyway to stop oscillations. The LM317/337 are pretty good from that standpoint. And yes, you're right - The SMPS do have some higher frequency noise, but that's cleaned up pretty well with the linear LM317's (infact a lot of the regular modules clean up noise pretty well if you have a slightly noisy input PS).
    As for your schematic - it will work and infact this is exactly what I do when I've made other dual supplies with positive only regulator boards. However, I'm not too sure whether you could easily modify the current board to do this (without cutting tracks and making a mess of it).
    When I get time, I might have another look at the project and see if I can't make it a bit more reliable with the ammeters. I didn't really need current sensing (as it was suppose to be a standalone PS for testing preamps with split supplies). But there wasn't much price difference between just a volt meter and a volt/amp combo at the time. The output switch is in the wrong position too and should be between the volt meter and output (which it isn't and I noted it at the end of the project).


    Reply 2 years ago

    Agreed the AM VM is not absolutely necessary, although it is very convenient. Well just some things to think about for in the future perhaps. The caps load nothing once they reach full charge; at that point current stops. They improve ripple in the regulator output and keep things steady with varying loads. The oscillation I mentioned is only a problem when two positive regulators are hooked up in series (with a common connection as in my diagram) and a single load is connected to the outer terminals. That's where individual loads on each supply are needed to prevent oscillation. The separate regulators don't know where ground is, and that causes the problem, in layman's terms anyway. The idea of using the laptop supplies is a good one. I just happened to save a tranny from an old, blown amp, so it was free. But as I said, there is high frequency noise coming out of those so it should be checked on an oscilloscope. One final and probably the most important is the DC supply feeding the regulators should be significantly higher than the voltage you will be applying to your circuits; the more the better. If there is not a significant voltage drop across the regulator, some low frequency noise may affect your audio circuits. If you are using transformers, you would want the AC secondary voltage to be higher than the DC that you will use. For instance, if you need 12VDC, a 15VAC transformer secondary would yield around 21VDC after its rectified and filtered with the caps. This keeps the regulator noise down to safe levels, especially in preamps. I wish you success.


    Reply 2 years ago

    I had a bit of a look and there's some V/A meters with a current sensing ring. I think that's the easiest for this sort of application (where you have split rails). I don't think there's an easy way around it other than sensing rings.
    As for the input voltages before a regulator, they only need to be to 1.5V above the output in most cases. The rest just turns to heat (which is what you get when you go from 19V to a 5V output) ;)
    But thanks for the heads up on the meters problem. I'll put out a MK2 at some stage!


    Reply 2 years ago

    The regulators need more than that so their noise remains out of the audio band. If the preamp is +/- 12v the 19v supplies are good. +/- 15v risky. It also depends on the PSRR of the preamp. Preamps don't draw so much power that would increase heat. The sensing rings are current transformers and are for high currents. They will not be accurate at low current, except the meters we used can be calibrated to make a particular range fairly accurate. Two LM317 regulators cost me $10 shipped with the tax. I chose the kit form so nothing has to be de-soldered. There are also very low dropout regulators but those are more expensive.


    Reply 2 years ago

    I'd be betting you've connected the meters up wrong. Have a look at the
    wiring diagram in section 3 again. Notice where the volts/amps meters are
    connected on the 0V line. I'm sure I had the same issue when I first
    connected the meters up (although it's been a few years since I built
    this). Draw out your wiring and think of it as two separate power
    supplies when connecting the meters (i.e +15V/0V, 0V/-15V). So the volt
    meters go in parallel, and the ammeters in series (+ to -, even on the
    negative side of the PS). Also, forget about the output switch for the
    moment too until you figure out the meter wires. It's most likely the
    negative side meter that's the problem. And you used separate PS for
    each meter (240VAC to 12VDC or similar)?? It won't work if you just use
    something like a buck module as they're negatives aren't isolated (which
    is the whole point of using the AC to DC PS). Let me know how you go.


    Reply 2 years ago

    I found it. The negative to common load at the output terminals of the supply passes through the positive meter's current sensing loop--in the opposite direction. The only way to prevent this is isolated DC sources and two separate regulators. The only point of connection would be the common output terminal on the front panel.

    I read somewhere that some load needs to be on at least one of the regulators to prevent oscillation. My pilot lights should do that. Does this sound good?


    Reply 2 years ago

    I thought you said you were running the power for the meters off separate power supplies anyway? That was one of the issues I had in that each meters negative's are connected together and you can't separate them (the negative current sense and negative meter power). That's why the power for each meter needs to be a separate power source so they don't interfere with each other. It's not an issue if you only have one PS, but an issue when you put two together and make a +/- dual supply. It might be possible to do it with diodes, but the small 240AC to 12VDC modules are only about $1.50ea so we're not breaking the bank for two of them.


    Reply 2 years ago

    Oh my, I meant the main power supplies. The meter supplies are separately wired to each meter, and isolated from input to output internally.
    What is happening is the return path of the negative supply must pass through the positive meter's current loop--it's the only path for the negative supply to return to the common on the regulator board. For an example, each pilot light is a 28v 1w incandescent; if the negative is adjusted to the lowest voltage setting, the positive meter reads both current and voltage properly. But as the negative is turned up, the current from that side passes through the positive meter on its way to the common. Since it is traveling in the opposite direction, whatever current the negative draws is subtracted from the positive ammeter reading. This doesn't matter if only one polarity is used at a time, but I would like to see accurate, individual amp readings. Kindly check my wiring below. Thanks.

    Dual AM VM Wiring.PNG

    Reply 2 years ago

    One ps is an LED driver; the other is a cell phone charger. Checking with a Fluke 87, there is infinite resistance from primary to secondary. This was checked from each primary lead to each secondary lead.

    I actually figured out the ammeter wiring before seeing this post. The recommended wiring uses positive polarity. So maintaining the same current flow direction, the same meter connected to a negative source and a common (or zero) would necessarily have to connected as if the common is the positive. This means there should be one black and one red on the common, which there is. The yellow voltage sense wires are connected as such: the positive meter to the positive terminal; the negative meter to the common terminal. The wiring must be correct as both meters work properly on an individual basis.

    There obviously is a bleed somewhere. The only place I can think of is the regulator board. The only differences between my board and the one you used is the pre-regulator DC supply LED indicators, and the reverse protection diodes across the regulators. The LEDs could be popped and the diodes added easily. I would prefer to figure it out first, though.

    Thank you for responding. Your project was presented quite nicely.


    2 years ago

    I think this scheme will eliminate the problem. It requires two separate positive regulators, and obviously two separate power supplies. The image must be expanded to see all of it.

    Dual AM VM Wiring with POS Regulators.PNG