DC Power Supply - Choke for regulated power supply

[Stuart Hossack]

Hi,


This my first post on this forum. I’m not a beginner but not intermediate either. I have some large theory and maths gaps for the issue I am attempting to tackle. I have loads of electronics books, and the internet, to help but I can’t seem to find an answer to my question.

Any help would be appreciated.

I am building a 10 output isolated power supply unit to give 9, 12 and 18 volts (depending on the value of voltage regulator I use). It's a simple design which uses multiple individual transformers to give multiple isolated power outputs. The unit is for guitar effects pedals and so ultra clean and smooth power is paramount to avoid hum and ripple noise issues for some of the older and simpler circuits where noise on the supply (coupled with high gain) is easy to produce, amplify and hear.

The psu design is from General Guitar Gadgets and works nicely (I have bootstrapped it together to test voltages) but I want to add a choke before the voltage regulator for “belt and braces” noise reduction. And of course I want to resolve this before I etch the 10 boards.

I have attached a pic of a similar psu design (off the internet) to the one I'm using, which shows a choke but doesn't give a value(s). I have seen values of 10 henries for valve amplifier supplies but the size is quite large for the project – and I’d need ten of them so I'm not even sure if an effective choke would be small enough or available to use.

Hope someone can point me in the right direction and offer advice.


Thanks, Stuart.

 

[Colin Mitchell]

No-one has explained this to you:

If you only use the voltage that is well below the ripple, you will not get any hum.
Just supply 24v and only use 18v.
Use 2,200u for each amp.
The choke is not needed.

 

[AnalogKid]

What is the transformer AC secondary voltage (RMS or peak to peak)?

What is the load current?

You probably are using a 78xx regulator (or one of is derivatives). Consider using an LM317. It is a little bit more complicated with the two output setting resistors, but it has a lower noise output, attenuates input ripple better, and you don't need to stock three different IC's, just three different resistors. You can put all 3 on the pc board with a 2x3 pin header for shunts and have only one assembly for all three outputs.

ak

 

[Stuart Hossack]

No-one has explained this to you:

If you only use the voltage that is well below the ripple, you will not get any hum.
Just supply 24v and only use 18v.
Use 2,200u for each amp.
The choke is not needed.

Thanks Colin, I appreciate your time and help.

Would the 2200 uF cap replace the 220uF? and when you say "amp" do you mean each supplied outlet?

Thanks again for your help.

Stuart.

 

[Stuart Hossack]

What is the transformer AC secondary voltage (RMS or peak to peak)?

What is the load current?

You probably are using a 78xx regulator (or one of is derivatives). Consider using an LM317. It is a little bit more complicated with the two output setting resistors, but it has a lower noise output, attenuates input ripple better, and you don't need to stock three different IC's, just three different resistors. You can put all 3 on the pc board with a 2x3 pin header for shunts and have only one assembly for all three outputs.

ak

Hi AK,

Thanks for your thoughts. The regulator is indeed a LM78"", I bought the LM78""s because they seemed simpler to use and I didn't have any clue about the noise/ripple advantage of the LM317. I see your point about having switched resitors, it would have been usefull. I can still do it though and shelve the LM78"s

I will have to get the transformer data sheets out to see if I can answer the AC secondary voltage question (RMS/peak to peak) - I'll get them out tomorrow and reply. I do think they have 18V outputs though so 18 volt supply will be a non starter - unless I double up the secondaries (they have 2 x 18 volt windings).

Thanks again, Stuart.

 

[AnalogKid]

A transformer with a secondary rating of 18 Vrms can drive a regulator with an 18 Vdc output. Depending on the output current, the bulk filter capacitor might be a bit large, but nothing extreme or expensive.

 

[Stuart Hossack]

A transformer with a secondary rating of 18 Vrms can drive a regulator with an 18 Vdc output. Depending on the output current, the bulk filter capacitor might be a bit large, but nothing extreme or expensive.

Thanks again AK,

I will get the transformer data sheet out later - I'm out all day and won't be back 'til late.

When I bootstrapped the circuit together, I did measure (from memory, I didn't really register consciously) 32???? volts after the rectifier (may even be after the first cap, not sure). I will check the data-sheet.

Searched for LM317 on eBay and found (albeit from China) fully constructed variable voltage boards for 99p (1.4 dollars??) each, shipped but where's the fun in that ) - downloaded the data sheet and it looks a nice regulator.

On a tangent, what's with the Maslow reference? This is hierarchy of needs fella' isn't it? - I remember studying him for Banking exams of all things (

Thanks again for your time, patience and help.

Stuart.

 

[Stuart Hossack]

Thanks again AK,

I will get the transformer data sheet out later - I'm out all day and won't be back 'til late.

When I bootstrapped the circuit together, I did measure (from memory, I didn't really register consciously) 32???? volts after the rectifier (may even be after the first cap, not sure). I will check the data-sheet.

Searched for LM317 on eBay and found (albeit from China) fully constructed variable voltage boards for 99p (1.4 dollars??) each, shipped but where's the fun in that ) - downloaded the data sheet and it looks a nice regulator.

On a tangent, what's with the Maslow reference? This is hierarchy of needs fella' isn't it? - I remember studying him for Banking exams of all things (

Thanks again for your time, patience and help.

Stuart.

A transformer with a secondary rating of 18 Vrms can drive a regulator with an 18 Vdc output. Depending on the output current, the bulk filter capacitor might be a bit large, but nothing extreme or expensive.

Hi again AK,

The data sheet didn't show peak to peak just 18 volts per secondary so this (measured with a meter) is 18Vrms so peak to peak is just over 50 volts, load currents vary but typically under 150ms but some pedals draw nearer 200ma

Stuart.

 

[AnalogKid]

18 Vrms, after a full wave bridge and at only 200 mA, will be close to 24 Vdc. A 1000 uF filter capacitor will be more than enough to keep the negative peaks of the ripple voltage above the required minimum input voltage for an 18 V regulator.

ak

 

[Stuart Hossack]

18 Vrms, after a full wave bridge and at only 200 mA, will be close to 24 Vdc. A 1000 uF filter capacitor will be more than enough to keep the negative peaks of the ripple voltage above the required minimum input voltage for an 18 V regulator.

ak

18 Vrms, after a full wave bridge and at only 200 mA, will be close to 24 Vdc. A 1000 uF filter capacitor will be more than enough to keep the negative peaks of the ripple voltage above the required minimum input voltage for an 18 V regulator.

ak

Thank you AK,

I'll test-bed the 1000uF cap(s) "point to point" on a single circuit and then compare the LM317 with the LM78**s - the advantage of easily switching at least some of the voltages for convenience is a convenience I can't ignore. Many thanks for your help.
I'll report back in a couple of weeks when I've got it all together.

Thanks again.