Pedal power supply design problems

Discussion in 'General Electronics Chat' started by dabbistandard, Mar 18, 2016.

  1. dabbistandard

    Thread Starter New Member

    Mar 18, 2016
    5
    0
    Hey guys I'm designing my own guitar pedal power supply. And it needs to have 8x 9v outputs and one 15v output. The 15v output needs to have 400-500mA and the 9v outputs need to have around 6-800mA capability.
    Anyway, I got lm7815 and lm7809 regulators. A 2x12v transformer with 2A capability. I was thinking of using 1n54xx diodes for the rectifier. Now first off I was going to take 24v out of the transformer and use it to power the regulators. After some research I noticed that these 24v aren't exactly 24v anymore after they've gone through the rectifier. More like 32v with regard to the 2v-ish drop of the rectifier. I didn't realize this when I bought the transformer but this high voltage is going to make the lm7809 really hot and it will probably loose current. So a friend suggested that I use the transformer in parallel, ergo 12vAC. After the rectifier that would be around 16,3vdc which is totally fine for the lm7809 but not the lm7815.
    Now the datasheets claim that the lm7809 needs at least 11.5v to be able to provide steady 9 volts. and the 7815 needs 17.5 volts.
    I'm not so good at reading and understanding the data sheets but as I understood it, it generates 5° celcius for each watt plus the room temperature. So if I were to use the 32v I am thinking if the formula would be something like this. (for the lm7809) 32v(voltage it gets)-11.5v(voltage it needs)= 20.5v(unused) 20.5v*1amp=20.5 watts and therefore
    20.5w * 5°celcius= 102.5° plus the room temperature, am I understanding this correctly? If so it's a huuge heat generation.
    Anyways
    So here are a few ideas to solve this matter:

    Using 24vac (around 32vdc) and using something like a voltage divider or a zener diode to drop the voltage down to say, around 18 volts. But I'm thinking, will this generate enormous heat over the two resistors or the a 18v zener diode?

    Another idea: Return the 2x12v transformer and get a 2x9v transformer and use 18 volts ac( around 23v Dc) which will still be a bit hot for the lm7809. Or even using a 2x7.5v transformer ergo 15v Ac(around 19v Dc) would that be to hot for the lm7809? It is perfect for the 7815.

    Yet another idea: how about those experiment transformers that give 6, 9, 12, 15, 18, 24 volts and then take the zero tap and 12v through one rectifier for the 7809 channel and the zero tap and 15v through another diode bridge. Therefore both channels are using the same ground and hopefully eliminate ground loop problems I guess. Or do those transformers just don't work like that. I'm also thinking if those transformers have isolated taps or is it maybe just the same wire making a tap and continuing to the next tap.

    What do you guys think?
     
  2. paulktreg

    Distinguished Member

    Jun 2, 2008
    611
    120
    Look into Buck Convertors.

    Something like this would do all your 9V outputs with another for the 15V.
     
    Last edited: Mar 18, 2016
  3. dabbistandard

    Thread Starter New Member

    Mar 18, 2016
    5
    0
    Maybe this looks like a good idea but I will still have to get another transformer which is closer to the 15 volts so the lm7815 wont generate to much heat I think
     
  4. InspectorGadget

    Active Member

    Nov 5, 2010
    211
    42
    Linear supplies from 12V to 5V or 24V to 15V are just a big waste. That's why you get all that heat.

    Better just to buy a switching supply from 110/220 to 5V/15V at the currents you need. There are lots of them on the market.

    Now if you must build your own, realize two things.

    1. You just need some massive heat sinks and parts that can stand the peak input voltage.

    2. The rated transformer voltage is RMS and you're measuring peak voltage integrated on the capacitors. When you load it it'll fall closer to the RMS rated voltage and you'll get a lot of 60 cycle hum. But the linear regulators will filter that out well.
     
  5. dabbistandard

    Thread Starter New Member

    Mar 18, 2016
    5
    0
    No. Sorry but a switchmode is not an option for guitar effect (audio) pedals because of the oscillation of the switching which results in a siren (police car) sound inside the beautiful buzz
     
  6. dabbistandard

    Thread Starter New Member

    Mar 18, 2016
    5
    0
    Ok but you can hold that voltage up, right? with low amount of ripple and huge capacitors after the rectifier
     
  7. blocco a spirale

    AAC Fanatic!

    Jun 18, 2008
    1,438
    368
    Noise shouldn't be a problem with well designed equipment; lots of audio devices use switching power-supplies these days. Buy a quality brand and avoid cheap un-branded PSUs from China and you should be ok.
     
    Last edited: Mar 19, 2016
  8. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,540
    1,251
    At 800 mA per 9 V output you need 7.2 amps of 9 V. That's a lot, and you can't get there with any single-chip linear regulator. Ant there's the heat. Linear regulators are inherantly simple and low noise, but the do make heat. Selecting the transformer can get this down to a reasonable amount, but you still will need heatsinks, and possibly a fan for relable operation. Also, the lower the transformer voltage (and regulator heat), the larger the filter capacitor must be.

    A possible option is one or two switching pre-regulators adjusted to 12 or 13 V, followed by 8 7809's, one per output. You still will need a heatsink, but not nearly as large as before.

    800 mA for a pedal? Really? 7.2 W seems about right for a small amplifier (my brother has a little 5 W guy he travels and practices with).

    ak
     
  9. dabbistandard

    Thread Starter New Member

    Mar 18, 2016
    5
    0
    Excuse me I ment 800mA total for the 9 volts. All outputs derived from the same 1 amp regulator
     
  10. #12

    Expert

    Nov 30, 2010
    16,298
    6,809
  11. #12

    Expert

    Nov 30, 2010
    16,298
    6,809
    I think I had a better idea:confused:
    Oh, by the way, that capacitor formula only works for the first capacitor. The rest of them are according to the datasheet for the regulator chips.;)
    Oops. Slipped a digit on the last capacitor size.:oops:
    There. I fixed it.:p
     
    Last edited: Mar 19, 2016
Loading...