What's the 'best' way to get bipolar power from a wall-wart?

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, luketeaford, Joe.

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Wed Feb 17, 2021 5:48 pm

plushterry wrote:
Wed Feb 17, 2021 4:01 pm
That is really really helpful thanks!

so you're sort of funnelling all the grounds from the switcher together and sending them back to the main ground as one, rather than letting them scatter out and find their own way back.

I like the AM radio trick, will defo try that.
Exactly !!! Love your 'finding their own way back' comment - this is exactly what gives so many SMPSes a bad name. If you do this then it's (usually) fine to ground fill the top and bottom, provided the ground fills go back to the same place. But remember this single point of connection to the main ground must be really solid. Don't then feed it through 10mm of 25mil track :-)

Bare in mind with the right inductor and capacitor values you can even sometimes lose the regulators in many applications, but best to leave that for your next design.

User avatar
plushterry
Common Wiggler
Posts: 240
Joined: Mon Oct 02, 2017 6:52 pm
Location: Cornwall

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by plushterry » Thu Feb 18, 2021 7:14 am

Nice one Mike

After doing some more research it seems like a capacitance multiplier circuit would probably be more effective than the regulators at removing switching ripple.

User avatar
plushterry
Common Wiggler
Posts: 240
Joined: Mon Oct 02, 2017 6:52 pm
Location: Cornwall

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by plushterry » Thu Feb 18, 2021 7:22 am

I wondered last night if it would be a good idea to get a second PCB fabbed that's the same dimensions as my PSU board with just copper fill on it. I could mount it on standoffs above the PSU, and attach via a thick wire to the ground on the PSU board to act as a shield.

It's going to affect the amount of air circulation, and therefore make the board run hotter but I would imagine it could help lower the amount of EMI that gets out.

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Thu Feb 18, 2021 10:05 am

plushterry wrote:
Thu Feb 18, 2021 7:14 am
Nice one Mike

After doing some more research it seems like a capacitance multiplier circuit would probably be more effective than the regulators at removing switching ripple.
You've been on SynthDIY ;-). If you add a zener across the capacitor it both regulates and smoothes ripple reasonably well
Last edited by MikeDB on Thu Feb 18, 2021 10:52 am, edited 1 time in total.

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Thu Feb 18, 2021 10:06 am

plushterry wrote:
Thu Feb 18, 2021 7:22 am
I wondered last night if it would be a good idea to get a second PCB fabbed that's the same dimensions as my PSU board with just copper fill on it. I could mount it on standoffs above the PSU, and attach via a thick wire to the ground on the PSU board to act as a shield.

It's going to affect the amount of air circulation, and therefore make the board run hotter but I would imagine it could help lower the amount of EMI that gets out.
Just try it with a bit of aluminium cut to size. That will conduct more heat away as well

User avatar
plushterry
Common Wiggler
Posts: 240
Joined: Mon Oct 02, 2017 6:52 pm
Location: Cornwall

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by plushterry » Thu Feb 18, 2021 12:46 pm

MikeDB wrote:
Thu Feb 18, 2021 10:05 am
You've been on SynthDIY ;-). If you add a zener across the capacitor it both regulates and smoothes ripple reasonably well
Haha, no it was an eevblog video that YouTube has been nagging me to watch for ages.

It's almost as if they know what I want. How strange :hmm:

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Thu Feb 18, 2021 2:14 pm

plushterry wrote:
Thu Feb 18, 2021 12:46 pm
MikeDB wrote:
Thu Feb 18, 2021 10:05 am
You've been on SynthDIY ;-). If you add a zener across the capacitor it both regulates and smoothes ripple reasonably well
Haha, no it was an eevblog video that YouTube has been nagging me to watch for ages.

It's almost as if they know what I want. How strange :hmm:
This is the one referred to on SynthDIY two days ago which I bet is the same

User avatar
KSS
Super Deluxe Wiggler
Posts: 4803
Joined: Mon Jan 25, 2016 7:28 am

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by KSS » Thu Feb 18, 2021 2:18 pm

plushterry wrote:
Thu Feb 18, 2021 12:46 pm
It's almost as if they know what I want. How strange :hmm:
Attachments
Googl.jpg
Googl.jpg (89.51 KiB) Viewed 204 times

User avatar
plushterry
Common Wiggler
Posts: 240
Joined: Mon Oct 02, 2017 6:52 pm
Location: Cornwall

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by plushterry » Thu Feb 18, 2021 6:30 pm

MikeDB wrote:
Thu Feb 18, 2021 2:14 pm
This is the one referred to on SynthDIY two days ago which I bet is the same
yeah that's the one!!

User avatar
plushterry
Common Wiggler
Posts: 240
Joined: Mon Oct 02, 2017 6:52 pm
Location: Cornwall

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by plushterry » Thu Feb 18, 2021 6:42 pm

I poked the layout around a bit and came up with a new version. I think that I managed to incorporate most of the suggestions. I changed the jumpers for wire links so there is now uninterrupted ground surrounding the switcher. managed to get it all on the top layer so all of the grounds for the switcher, except the input decoupling cap go back through a group of vias right under the IC.

the board has come out a bit longer, 70×40mm now, which means the regs have a little more room to breathe at least.
pcb0v4layers.png
pcb0v4render.png
pcb0v4renderrear.png

rich_de
Learning to Wiggle
Posts: 40
Joined: Thu Nov 19, 2020 3:00 pm

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by rich_de » Thu Feb 18, 2021 7:28 pm

Near-abducting the thread, but as the 12V AC solution was already mentioned: Has anyone had issues with a rail not powering up? I've built a Moog-designed EM Theremin that uses such a 12V wall wart, half-wave rectifying, caps, and a 7812/7912 pair. Schematic PDF everywhere on the net.

After some sweat the circuit itself works lovely, but initially, with a good chance, the power supply does not want to ramp up the regulated +12V. The -12V is always there. When that happens, I can switch it off, and most of the time a blip on the power LED flashes. After turning it on again when the blip was seen, it always works. I haven't deeply looked into this strange behaviour, the primary DC is there, but the output of the 7812 stays at zero. Thought I'd ask around while the topic is up: Bad regulator, or a particular trap with this layout?

Edit: mention the negative rail always works

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Fri Feb 19, 2021 12:19 am

Do you have the reverse protection diodes from input to output of the two regulators ? If the regulator sees a voltage less than it's ground pin already on the output, some decide that's a short circuit and so don't switch on. Even with the diodes, I suspect this might still be the problem, it all depends on what your load is like.

No problem with abducting the thread. It's been moreorless an IM but hopefully somebody else has found it useful, or will do in future.

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Fri Feb 19, 2021 12:22 am

plushterry wrote:
Thu Feb 18, 2021 6:42 pm

the board has come out a bit longer, 70×40mm now, which means the regs have a little more room to breathe at least.
You seem to have vias in the pads of C1-3 and under the regulator tabs. Some PCB fabs won't like that or charge more for them. Best to move them slightly

rich_de
Learning to Wiggle
Posts: 40
Joined: Thu Nov 19, 2020 3:00 pm

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by rich_de » Fri Feb 19, 2021 9:32 am

MikeDB wrote:
Fri Feb 19, 2021 12:19 am
Do you have the reverse protection diodes from input to output of the two regulators ? If the regulator sees a voltage less than it's ground pin already on the output, some decide that's a short circuit and so don't switch on.
Mindlessly tacked a diode (1N4148) across the 7812 O-to-I, still refused to always power up. THEN thought about it and figured out that this diode won't help with the output going below ground. Measured the output again to be sure, there we had the -800mV. Hooked up another diode (again 1N4148) between GND and regulator output. No improvement. Swapped the 1N4148 for a BAT46 (Schottky, Vf<400mV up to 10mA), bingo! Works perfectly now.

The original circuit is minimal, exactly as in the original Moog plans, 2 rectifier diodes, 2 large electrolytics, 2 regulators, 2 small output caps, done. Regulators are the TO-220 from ST. (The schematic lists LM79L12, which would've been original NS at the time, in TO-92). Lesson learned: Even the master himself could not foresee that ST's second source regulators would behave like that.

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Fri Feb 19, 2021 9:57 am

rich_de wrote:
Fri Feb 19, 2021 9:32 am
MikeDB wrote:
Fri Feb 19, 2021 12:19 am
Do you have the reverse protection diodes from input to output of the two regulators ? If the regulator sees a voltage less than it's ground pin already on the output, some decide that's a short circuit and so don't switch on.
Mindlessly tacked a diode (1N4148) across the 7812 O-to-I, still refused to always power up. THEN thought about it and figured out that this diode won't help with the output going below ground. Measured the output again to be sure, there we had the -800mV. Hooked up another diode (again 1N4148) between GND and regulator output. No improvement. Swapped the 1N4148 for a BAT46 (Schottky, Vf<400mV up to 10mA), bingo! Works perfectly now.

The original circuit is minimal, exactly as in the original Moog plans, 2 rectifier diodes, 2 large electrolytics, 2 regulators, 2 small output caps, done. Regulators are the TO-220 from ST. (The schematic lists LM79L12, which would've been original NS at the time, in TO-92). Lesson learned: Even the master himself could not foresee that ST's second source regulators would behave like that.
That's good to hear. Despite some people thinking the 79xx is just a mirror of the 78xx, it's not and is a totally different circuit, and the original attempts weren't even that stable or reliable. So the newer versions have evolved and as you've noticed the diodes are often now needed in bipolar systems. Don't know why they don't just integrate them, would hardly affect the cost.

rich_de
Learning to Wiggle
Posts: 40
Joined: Thu Nov 19, 2020 3:00 pm

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by rich_de » Fri Feb 19, 2021 11:53 am

MikeDB wrote:
Fri Feb 19, 2021 9:57 am
Despite some people thinking the 79xx is just a mirror of the 78xx, it's not and is a totally different circuit, [...] Don't know why they don't just integrate them, would hardly affect the cost.
Soldered the BAT46 in. Case closed (literally, too!).

In my theremin, it was the positive side playing up (a L7812ACV to be precise), the negative rail seems good despite having an insufficient output cap as per datasheet: The LM79L15 requires 100nF ceramic, or 1uF electrolytic, but the TO-220 variants (both NS and ST) call for 10 times as much (or a 1uF tantalum from supply to ground... ouch). Maybe I should scope it to check for HF oscillations.

Despite the trouble, I like the scheme. I'd use it for small scale eurorack or similar stuff. It is very simple and does not use any switching, which is another can of worms, especially with audio. I think the single 12VAC wall wart with linear regulator is a good choice for, say, up to around 250 mA. Just don't forget your diodes. (I think they leave them out for process reasons; as simple example: the different pinouts between 78xx and 79xx seem to come from the need to have the substrate most negative, and their TO-220 packaging needs to have the substrate on the center pin, obvious only if you know both...).

One significant downside is that 12V AC wall plugs are rather hard to get these days (a proper big and heavy transformer means expensive), the world and dog has mostly standardized on 5V DC switchers. Above 250 mA it's pretty much the Meanwell RT65B for up to 1A on -12V, or a transformer with two secondaries, but that leaves the thread scope of "wall-wart".

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Fri Feb 19, 2021 12:21 pm

rich_de wrote:
Fri Feb 19, 2021 11:53 am
One significant downside is that 12V AC wall plugs are rather hard to get these days (a proper big and heavy transformer means expensive),
... and the sale of them is banned in some EU countries, with the rest supposedly getting round to implementing REGULATION (EC) No 278/2009 sometime.

User avatar
plushterry
Common Wiggler
Posts: 240
Joined: Mon Oct 02, 2017 6:52 pm
Location: Cornwall

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by plushterry » Fri Feb 19, 2021 2:28 pm

rich_de wrote:
Thu Feb 18, 2021 7:28 pm
Near-abducting the thread
no worries! still on topic. glad you got your problem sorted.

User avatar
plushterry
Common Wiggler
Posts: 240
Joined: Mon Oct 02, 2017 6:52 pm
Location: Cornwall

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by plushterry » Fri Feb 19, 2021 2:33 pm

MikeDB wrote:
Fri Feb 19, 2021 12:22 am
You seem to have vias in the pads of C1-3 and under the regulator tabs. Some PCB fabs won't like that or charge more for them. Best to move them slightly
Nice one Mike, you're a legend for dishing out all this info.


I'll let you know if it works :hihi:

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Fri Feb 19, 2021 2:52 pm

plushterry wrote:
Fri Feb 19, 2021 2:33 pm
Nice one Mike, you're a legend for dishing out all this info.
Hehe, been there, f***ed that up ! :-)

rich_de
Learning to Wiggle
Posts: 40
Joined: Thu Nov 19, 2020 3:00 pm

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by rich_de » Fri Feb 19, 2021 5:54 pm

MikeDB wrote:
Fri Feb 19, 2021 12:21 pm
the sale of [12V AC wall plugs] is banned in some EU countries, with the rest supposedly getting round to implementing REGULATION (EC) No 278/2009 sometime.
RS still stock them in different varieties (e.g. 139-1764 would be suitable for the theremin), point out in their description that they are pretty conformant, and don't hint at regional restrictions. I had a quick look at the regulation, it specifically lists AC supplies with less stringent requirements, which somehow seem to be made to let a plain transformer pass if it is of decent quality. (e.g. the RS part has less idle wattage spec'd than the limit allows).

Which countries ban these, and through which limits will the EU regulations kill them off? I knew of the limit for (internal) linear supplies above 75W (which they wanted to lower, but in the end didn't), but haven't been aware of external AC supply elimination.

User avatar
MikeDB
Veteran Wiggler
Posts: 525
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by MikeDB » Fri Feb 19, 2021 8:47 pm

As you can see are pretty strict limits for efficiency in the directive, and although it states a no-load input power of 0.3W on the page, which is legal, the actual spec sheet says 35mA which is almost 9 Watts, so one of these numbers is wrong ! If the latter is correct it doesn't meet Annex I which specifies a limit of 0.5W for AC-AC convertors, and neither number meets the design target of 0.1W set out in Article 6/Annex III.

However if you look at the product page, it specifies it meets ERP which although applicable, isn't the strictest relevant spec for driving electronic devices within the EU. My guess is this thing was approved as a halogen lighting convertor as the output is 12V and those were exempted since you always turn a light off at the mains, hence the transformer would be turned off. But that doesn't make it legal for one to use it for other applications. It also may have export limits on it which aren't shown. I know many German distributors pulled all their stock of such things in 2015.

The 75W limit, which indeed should have been much lower, is for conducted harmonics, a totally different problem.

The problem has been there have been some truly poor switchmode supplies on the market which have given SMPSes a bad name. The last time I used a linear supply was in 1976 as a well designed SMPS will always give superior performance, be it for a low noise mic amp or a 2kW PA amp, and everything in between. It's far easier to get rid of a switching frequency above 20kHz than it is to get rid of 50/60 Hz, provided you design it correctly.

Thomas73951
Learning to Wiggle
Posts: 3
Joined: Sat Dec 07, 2019 11:13 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by Thomas73951 » Sun Feb 21, 2021 7:54 am

Portabella wrote:
Tue Feb 09, 2021 5:56 pm
I'm using some cheap boards from china and they do the job:

https://www.ebay.de/itm/AC-DC-to-5V-15V ... 2749.l2649

however those TO220 regulators resp. the heatsinks on them get kinda got.
I meassured 63 degree Celsius .. is that okay for a voltage regulator TO220?
It says these accept AC or DC in. I've seen quite a few circuit schematics that use the LM7815 & LM7915 but all of them have just an AC voltage input.

Do you know if you can get ±15v out of this with just a 18-36v DC input? The listing makes it sound like it can, but I'm not quite sure how it manages to do it, just looking at the board (with limited knowledge).

If it can do it I think I'll get one because my current solution is rather janky.

Thomas

User avatar
plushterry
Common Wiggler
Posts: 240
Joined: Mon Oct 02, 2017 6:52 pm
Location: Cornwall

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by plushterry » Sun Feb 21, 2021 8:37 am

Thomas73951 wrote:
Sun Feb 21, 2021 7:54 am
It says these accept AC or DC in. I've seen quite a few circuit schematics that use the LM7815 & LM7915 but all of them have just an AC voltage input.

Do you know if you can get ±15v out of this with just a 18-36v DC input? The listing makes it sound like it can, but I'm not quite sure how it manages to do it, just looking at the board (with limited knowledge).

If it can do it I think I'll get one because my current solution is rather janky.

Thomas

either DC or AC will work, there's a rectifier at the input so it doesnt matter. you are going to get an extra diode drop across the rectifier, but that's not the end of the world, you just need to make sure you have enough voltage at the input to allow for that.

Thomas73951
Learning to Wiggle
Posts: 3
Joined: Sat Dec 07, 2019 11:13 am
Location: UK

Re: What's the 'best' way to get bipolar power from a wall-wart?

Post by Thomas73951 » Sun Feb 21, 2021 10:17 am

plushterry wrote:
Sun Feb 21, 2021 8:37 am
either DC or AC will work, there's a rectifier at the input so it doesnt matter. you are going to get an extra diode drop across the rectifier, but that's not the end of the world, you just need to make sure you have enough voltage at the input to allow for that.
Ah ok cool thanks :)

Post Reply

Return to “Music Tech DIY”