It’s not much but it’s all relative. I spent a little less than 14 EUR for the regular transformer (which works great), and only 6 EUR for the toroidal (which was sold used by a local audiophile who was clearing out his old stuff).
1 Like
I’m surprised Analogoutput has not chimed in yet, he’s pretty good at this sort of stuff. Is he ok?
3 Likes
Chip90125/PSU: creating a PSU A newer version. trying to add a 5v rail bt going wrong. Is there a Kicad file which has all 3 raills?
These prices are from local shops, not Tayda/AE.
1 Like
Hi Eddy, hope you doing well.
I was thinking about some stuff like rebuilding some of my older modules, and i also made a new psu according to your schematics using the LM317 and LM377.
Now i was wondering what would happen if the 10k trimpot to set the desired voltage, goes broke?
I use 15vac toroidal transformers and sometimes they dont output that whats on the sticker. I have 4 of them. 1 outputs 14,2 volts. 2nd outputs 17volts and the last one outputs 18,44 volts. So i order 15vac transformers and i get ones with higher output voltages. No problem i would say but after rectifying, the voltage becomes 23volts which then will be transformed by the voltage regulator to 12volts. The rest will be dissipated in the form of heat.
Also no problem, but what if the 10k trimpot gets broken? Does that mean that the LM’s are outputting the full rectified voltages destroying my modular synths and burn out everything inside?
Hi Marty,
That’s actually a very good question. If you look at the formula that determines the output voltage then a higher resistance would mean a higher output voltage.
What you should do is set up a little test circuit on a breadboard or piece of stripboard and test it out.
The Voltage regulators are protected internally against short circuit and overheating so it’s unlikely to break if anything goes wrong.
I’ll do a little test in Falstad to see how it reacts. Interesting question. Please keep me informed of your findings and if I find anything out about this, I will notify you.
1 Like
About the transformers, I use a transformer in my synth that has 17VAC and 21VAC output and it’s no problem. The rectifiers and regulators can handle it fine but it does indeed produce a lot of heat.
Linear power supplies are by their nature not very efficient.
Btw, I did a little test in Falstad and if you disconnect the potmeter from the regulator, the voltage does indeed spike. Maybe some overvoltage protection could be built in with a zener diode or something. Just a thought.
1 Like
Hi Eddy.
Thanks for trying it out. Didnt have the time yet but i will soon as i have ordered new LM317’s.
Expecting to receive them next week or so.
About the regulators. Im not worried about them as im aware that they can handle the spike in case the trimpot breaks but i dont think that my 12volt setup likes it🤪
I mean, the TL074 opamps and other IC’s can handle 15volt max right? So if trimpot breaks. Poef weg synth haha.
Protecting them with zener diodes sounds like a good protection but prevents voltage adjustments. Using LM3xx regulators are maybe doubtfull to use. Using 7812 and 7912 which have fixed voltage outputs are a better solution for protection saving your modules.
Im gonna look up on the www if i find some info about power overvoltage protection methods
adjustable/trimmable zenerdiodes dont exist right?
Actually they do. Not as stand alone components but I have a schematic for an adjustable zener diode made up of two resistors and some side components.
I’ll try and find it for you.
1 Like
The voltage will drop to whatever is set by the fixed resistor that is used between pins 1 and 2, which is something like 1.2V given that the fixed resistor is typically something like 240Ω. If the potentiometer fails while the modules draw much current, worst that will happen is that the voltage drop across the LM317 (and 337) will get much higher, potentially initiating a thermal shutdown of the regulators. At the very very worst, you might get two busted regulators, but the modules should be safe as they will be getting less, not more, voltage (and even that, will only be momentary in the worst case scenario).
I would advise you to stay with the tried and tested design of the datasheet 
and avoid adding Zener diodes and what not.
That’s because the “sticker” value is only valid under the specific load that the transformers have been designed for. Whenever you put them under a lighter load (as it usually happens), the output voltage is typically a little higher.
Again, you are overthinking it 
That’s what we use the regulators for down the line. To regulate the voltage to a specific value irrespective of the load. With proper heatsinking, a little more voltage on the input should not be something to worry about. Make sure that the filter capacitors after the transformer are rated for that voltage though!
1 Like
Hi,
thanks for your extended clarification. Now i’m more at peace, convinced that nothing will happen if such situation occurs. thanks
I wouldn’t rely on self-adhesive PCB pillars to take the weight of that toroidal transformer. The adhesive can soften when it gets warm. I’d bolt it down (but avoiding making an effective short circuit coil by ensuring the bolt does not complete a conducting loop, natch)
1 Like
Thanks for your advice but i dont think that the regulator heatsink will reach plastic melting temperatures.
I said nothing about heatsink temperatures, and nothing about plastic melting.
I said the adhesive can soften. And it has to take the weight of the transformer.
If you have ever seen the damage a heavy detached transformer can do, it get’s ugly.
Anyway, it’s up to you…
1 Like
Ok. Im sorry. Apperantly i didn’t understand it what you meant
1 Like
What adhesive? Standoffs? What weight needs to take off? Again. Sorry for not understanding.
A few words on voltages.
The nominal European single-phase mains voltage is 230V, with a legally permitted range of approximately 207V to 253V (230V, +10% / -6%). This harmonized voltage range allows countries that historically supplied 220V to still operate within the European standard, and for the UK and other countries that historically supplied 240V, to also remain compliant by having a wider tolerance range that overlaps with 230V.
This means that a standard 230V to 12V transformer could actually output between 10.8V RMS and 13.2V RMS.
But these are NO LOAD RMS votages; to get the peak voltage going into a rectifier, you have to multiply by 1.4142 (square root of 2). This gives you a possible peak input to the rectifier between 15.27V DC and 18.66V DC.
Then you have the voltage drop across the rectifier diode of around 0.7V * for a bog-standard silicon diode. So the voltage across the smoothing capacitor after the diode can be anything between about 14.57V and 17.96V. Again, this is the no-load voltage.
What people seem to be missing is that the capacitor will be discharging under load, so the voltage will droop. At 50Hz supply this will be 100 times a second. The voltage droop will depend on the capacitor value and the current drawn. If you consider a, say, 12V regulator supplying say, 1 Amp, the load is roughly like a 12 Ohm resistor. This will give you a time constant RC of about 12 x the Capacitace value in Farads.
What needs to be avoided is the voltage drooping to below the minimum input voltage requirement of the regulator, as this would result in a lot of ripple on the output.
To simplify things, assume that the voltage droop is linear (it is actually exponential).
A 7812 requires at least 14v input across it’s temperature range. In the worst case above with a voltage charging the capacitor to only 14.57V, droop of more than 0.57V will start to cause the regulator output to ripple. So the capacitor has to be big enough to prevent droop of more than 3.9%.
You may be lucky enough to live somewhere where the mains voltage is reliably well above the regulatory minimum of 207V RMS, but it is good practice to design for worst case normal operating conditions, plus a healthy margin.
So, I prefer to use 15V transformers with 12V regulators.
* In some configurations of bridge rectifier, there are two diodes conducting, resulting in a drop of about 2 x 0.7V i.e. 1.4V.
Edited to clarify a few points.
1 Like
The photo (now deleted?) showed a PCB-mounted transformer, standing on (four?) nylon legs (the white things). These have a self-adhesive base. it is this adhesive that can soften when it gets warm. They can ease loose over time. I would suggest machine screws, nuts and/or spacers to secure that PCB, depending on what you are fixing to.
I can’t remember who posted the photo now! But anyway… another thing to consider is that if the transformer becomes loose and goes walkies it may result in a mains wire becoming loose, which is not a good idea.
1 Like
Hi,
yes metal standoffs/screws, nuts and bolts are better I know.
transformer looks bigger on the photo then in real life 
its not heavy.
All the weight pushes down preventing moving (not build up-side down)
I can’t see a reason or think of how the adhesive can get hot. it never touches any hot parts.
My modular system isn’t travelling, having a fixed place in my studio.
But lets wait and see when things get loose, I’ll keep an eye open and find a solution then. Its a living modular system so there’s always room for improvement.
Hi,
wow, thanks for taking the effort to write this comment. very useful and another lesson learned.
yeah. I could trim the trimpots to 15v dc to power my modular synth. less excess voltage to dissipate so less heat.
2 Likes