is this circuit modification right ?

[debi]

Hi guys,

i started an electronic course few months ago. Besides of that, im building eurorack modules to make my own synth and fx based on schematics i find on internet (they ll be working with -12/+12vdc and some with only +9Vdc. It s time to build the PSU : i found a great schematic but my input source is not the same as shown. It would be a laptop psu 19V 2A instead of a 12VAC 1A.

I asked help from my teacher but he doesn't want me to go to fast and in the same time, being stuck in the processus make me impatient.

based on the schematic below is it ok to use a LM7818 and a LM7918 to get my +12 and -12VDC (without recalculating other components specs ?)and also get a 9V rail from the +12Vdc by adding a LM7809 circuit (found 12VDC to 9VDC converter, i ll show it also)

i wish it s that simple but id be surprised if its the case.

thanks for your time !

Deb

[rs20]

Your -12V output rail isn't going to work if you switch to +19V DC input.

The circuit shown relies on the input being AC, because the positive peaks of the sinewave (let through by the diode D1 and temporarily stored in C1) provide the +12V rail, and the negative peaks of the sinewave (let through by D2 and stored by C4) provide the -12V rail. With pure +19V DC input, there are no negative pulses, so C4 will never charge up, and the whole bottom half of that circuit will be dead (specifically, sitting at 0V, "switched off").

There are techniques for generating a negative rail from a positive rail. If you rename the subject of your message "how to generate negative rail from positive rail", then you'll attract replies on that particular challenge. (One easy albeit relatively expensive option is to just get as 12V-to-12V isolated DC-DC converter, example digikey search [in AUD sorry])

[debi]

My bad,

thank for your reply, it's obvious now.

In case i put my hands on the product you mentioned, the rest of the circuit mod is fine ?

Cheers mate

[CJay]

As RS20 says, you really do need the AC input but if you have 2 of the 19V adapters and they're not earthed then you may be able to use them.

It may be a typo or I may have misunderstood what you meant but you suggested using 7818 and 7918 devices which are 18V so you'd end up with + and - 18V

You can use the 7809 to get a 9V rail from the 12V rail, it's just about in spec.

For future reference you need to pay attention to dropout voltage, it's the minimum difference between input voltage and output voltage, if you don't have at least that much difference then the regulator might not work as expected.

[debi]

Hi CJay,

yes it s a 7812 and a 7912 needed. 

i only have one 19V psu... i think im going to look for a 12V AC/AC instead as shown normally because a -/+12v DCDC converter is too expensive for the project.

Thank you guys for your reactivity, it's really helpful !

[debi]

Hi found this link http://nozoid.com/diy-eurorack-power-supply/

I actually have 2 of these modules at home (LM2596). It might be the right combination with my laptop psu finally.. unfortunatly some comments say the circuit is incorrect or cannot work as it's built. ill give it a go anyway and see if i can get a cheap psu for my diy eurorack.

 

[Ian.M]

Should work.  See https://iw0ffk.wordpress.com/2014/09/19/negative-voltage-from-lm2596-dc-dc-ebay-module/ for more detail on using a cheap LM2596 buck module for negative supplies, and also the LM2596  datasheet.

N.B. most chinese 'LM2596' modules nowadays are fake.  See https://www.eevblog.com/forum/beginners/lm2596-help/
They'll quickly overheat at 1A and I wouldn't trust them long-term over 500mA.  Also the negative rail (-12V)  'LM2596' 'sees' an effective input voltage of 31V, which may be pushing your luck for a 'pure chinesium' fake 'LM2596' chip.  Therefore if you are unsatisfied with their current capability, reliability or even just their switching noise, you may need to replace the fake 'LM2596' chips on the modules with genuine T.I. ones direct from a major western distributor.

[debi]

big thank you for all the details you provided. 

A guy told me i can stick with the 12 VAC mentioned on the schematic by taking a part  my 19v psu and just get my +12/-12 after the transformer... i don't want to play too much with 220V so ill try the LM2596 in the first place. 

[ledtester]

Here's a video which also explains how to use the module as an inverting converter:

https://youtu.be/r49BwY0eFPA

One possible problem, though, is the large initial current required to start up the module in this configuration (which could be the reason for a lot of reported failures). The video shows how to modify the board to mitigate this issue (discussion of this starts at 5:40).

[rstofer]

A simple 12VAC transformer is pretty cheap and the original schematic provided the power necessary for synthesizers which are primarily based on op amps with bipolar rails.

https://www.digikey.com/en/products/detail/triad-magnetics/WAU12-1500/4915293

I would think that 12VAC 1.5A would be adequate but you could go a little larger - say 2A.  Put a fuse on the project board.

[Terry Bites]

My first question is why you have no choice but to use a lap top PSU instead of a 12VAC source?

An AC supply going to save you a lot of time, pain and money. If you've got AC mains for your laptop PSU then you can power a 12v transformer no?
To make or buy an effective 19v to ±12V is going to cost a lot more than a 12V transformer.  Use a fuse in the primary side. (par. ex VTX-146-015-212 Toroidal Transformer 15VA 0-12V Vigortronix) And you can feed a 9V regulator to your 12V DC as you've shown so long as you make allowances for current and thermal ratings. If you are not used to working with 230V get some one who is to check your wiring.


Take the path of least resistance!

[CJay]

big thank you for all the details you provided. 

A guy told me i can stick with the 12 VAC mentioned on the schematic by taking a part  my 19v psu and just get my +12/-12 after the transformer... i don't want to play too much with 220V so ill try the LM2596 in the first place. 

I'd be very wary of that, 19V psus are usually laptop or other IT chargers and they're not a standard transformer inside, they can be very dangerous if you don't know which bits are safe to touch.

[debi]

Hello,

i have found 2 transformers in a trash for +12/-12 rails but one is 23VA and this other is 3VA : from this link https://syntherjack.net/modular-synth-wall-wart-psu/ , it says i need 1 amp or greater. From what i heard, to know how many amp do 23VA, i need to do 23/12 = 2 amp. So this one should be ok if im right. The other one only make 3/12 = 250mA...

the 23VA is one to go with right ?

Cheers 

[CJay]

Spot on, the 23VA one is the one to go for.

[Ian.M]

Your 23VA transformer may not be sufficient.   When calculating the permissible secondary I_RMS for a transformer, you can't use the DC output voltage after the regulator, or even the DC voltage before the regulator, but must use the nominal AC RMS output voltage.  If the transformer has multiple identical secondaries, the calculated current must be divided equally between them. 

Then there is a topology dependent derating factor going from from I_RMS to I_DC to consider.  For a bridge rectifier directly feeding a reservoir capacitor, that factor^* is 0.62, so for 1A @12V DC out you need approx 1.6A RMS.

The regulator requires headroom to operate, typically at least 2V for a non LDO regulator, and the input ripple trough must be at least that much above the output voltage.  If you do the math (or crunch the numbers in a SPICE sim), by the time you've allowed for diode drop in the rectifier, and assuming reasonable size reservoir capacitors you need something like a 13.5V secondary to ensure it can still deliver 12V at full load during low line conditions (i.e. when the mains supply is at its minimum permitted voltage).

Therefore *IF* your 23VA transformer is suitable for a 12V PSU (and it may not be if its secondary voltage is too low or too high) at best, assuming a 13.5V nominal secondary voltage, it will be good for 23/13.5 * 0.62 A DC = 1.06A TOTAL.  If the secondary voltage is higher the limit will be reduced proportionately.  If the voltage is lower, there will be an increasing tendency to drop out of regulation at full load resulting in ripple breakthrough.

 If the nominal secondary voltage isn't marked, measure its secondary voltage unloaded, and in the absence of any other data, assume 10% regulation, so take the nominal voltage as 90% of the measured voltage, and recalculate the max continuous DC current accordingly.  N.B. line frequency transformers tolerate short term overloads very well for minutes at a time, so if it doen't quite meet the peak current requirement for your design it may not be  problem as long as the average current is well under the calculated limit.


* Hammond Transformer division's Design Guide for Rectifier Use gives derating factors from I_RMS to I_DC for various transformer secondary, rectifier and filter capacitor typologies.   Save a copy and study it!

[debi]

Hello guys,

the transformer is doing great ! it should be ok for 7 modules in total. CHeers !