Newb Thinks He's Figured the Optimized PSU

[glitched]

Ok, so after reading some stuff on the internets, I think I've figured out the optimized PSU thing, at least conceptually.  Can you help me confirm some things?

(Remember, any information you pro's can give me will benefit future newbies because I'll be writing a detailed, step-by-step guide on the whole CORE+SID construction process.)

Schematic:

To me, the schematic is linear; I visualize it as four separate tracks: 1) +5v (pin 5) 2) GND (pin 2), 3) +9VAC (pin 6), 4) +9VAC (pin7).

From left-to-right, the 5V "track" goes to a switch, the positive side of the 2200 16v capacitor, then to a 100nf cap, and finally out as 5VDC. 

The GND track goes to the negative side of the 2200 16v cap, then to the 100nf cap, and out on the other pin, which will be ground.

On the other side, we have the 9VAC (pin 7) track, which goes to the squiggly (~) of the rectifier, out the positive side (+) to the positive side of the 2200 25v cap, then to the 330nf cap, into the input side of the 7809 regulator, and finally out as 14VDC*. [*The + side of the 2200 16v cap (from the 5v track) is connected to the 330nf cap.]

The other 9VAC line (pin 6) goes to a switch, through the ~ side of the bridge rectifier, out the negative side (-), to the negative side of the 2200 25v cap, then to the 330nf cap, and the GND pin of the regulator.

The 14v goes to J1 on the SID and the 5V (and GND) goes to J2 on the CORE (the 7805 regulator on the CORE is not stuffed).

Conceptually, is that it?

Physically, I just mount the components and connect them, as above, as "tracks" I guess.  A 5v side and a 14v side. 

By the way, I've been using this site as a guide:

http://www.danielprice.org.uk/synth-diy/sid_2.htm

If I'm right about all this, it's a breakthrough because I had never even looked at a schematic, picked up a multimeter, or scorched myself with a soldering iron before this project.

Thanks.

-d

[Narwhal]

Yes this is a pretty good way to think about it as tracks, or signals that are each processed separately.

If you want to further understand what each of the components is doing in this circuit, I recommend http://en.wikipedia.org/ as an excellent source.

Here is how the rectifier turns the AC in DC: http://en.wikipedia.org/wiki/Rectifier

The device in the schematics you are looking at that has the squigly connections and the plus and minus outputs is a full-wave rectifier.  It take an alternating current that has no net force in any particular direction and converts it into a pulsing positive current with a net forward or positive direction.  http://en.wikipedia.org/wiki/Rectifier#Full-wave_rectification

You can think of AC current as a force pushing and pulling a boat that is sitting the water.  Whereas the full-wave rectified DC current is a force pushing a boat in pulses such that the end result is that the boat moves forward.  But since the boat is receiving pulses of full-wave rectified DC, it also pulses forward with lots of starts and stops.  We need to smooth these pulses by taking a bit of the strong part of the pulse and filling in the stops.  Since capacitors store up charge and release it this is where they come into play.

The capacitors in the circuit are smoothing out the pulse in this full-wave rectified DC current, so that the regulator will receive a steady flow.

http://en.wikipedia.org/wiki/Capacitor

The capacitors in that circuit are in parallel which means that they add together to form one big capacitor.  They both have voltage ratings that ensure that they will be less likely to blow up giving the voltages that they will see.  If you were to hook up an oscilloscope to the outputs of the capacitors, you will see voltage flows in almost steady DC, but with minor ripples in it.  This is where the regulator comes in.

Then you have the regulators making sure that what comes out is the DC voltage that you wanted.  It removed the ripples.

http://en.wikipedia.org/wiki/Voltage_regulator

In some ways a regulator can also be thought of as a protector.  It can shut down if it sees too much current draw at its output, or blow out if it sees too much voltage at its input.  This might not actually "protect" anything else in your circuit, but I have doubt that things could get much worse if it wasn't there in the first place.

Hope that helps a bit.

[glitched]

Yes, that's great info.  Here's another great wiki link on electronics: http://en.wikibooks.org/wiki/Electronics.

By the way, the first step is complete: +5.15VDC on the multimeter on the 5V side!

On to the other side...

UPDATE:  It works!  Amazing!  So happy when I saw the 14.19VDC on the meter.

I really didn't understand it until I visited a couple sites that put it into simple terms.

Here's another one of those:

http://www.zen22142.zen.co.uk/Prac/vero_circ/vero.htm

Highly recommended.

Ok, next step!

[Wisefire]

im glad to hear you figured it out.. nice going!

[nebula]

The capacitors in the circuit are smoothing out the pulse in this full-wave rectified DC current, so that the regulator will receive a steady flow.

http://en.wikipedia.org/wiki/Capacitor

The capacitors in that circuit are in parallel which means that they add together to form one big capacitor. They both have voltage ratings that ensure that they will be less likely to blow up giving the voltages that they will see. If you were to hook up an oscilloscope to the outputs of the capacitors, you will see voltage flows in almost steady DC, but with minor ripples in it. This is where the regulator comes in.

Then you have the regulators making sure that what comes out is the DC voltage that you wanted. It removed the ripples.

http://en.wikipedia.org/wiki/Voltage_regulator

In some ways a regulator can also be thought of as a protector. It can shut down if it sees too much current draw at its output, or blow out if it sees too much voltage at its input. This might not actually "protect" anything else in your circuit, but I have doubt that things could get much worse if it wasn't there in the first place.

Hope that helps a bit.

I don't think it's a good idea to think of the two paralleled capacitors as 1 big capacitor in this case. After all, one is 2200 uF and another is 330 nF - do we really need exactly 2200.33 uF? Could we leave out the 330 nF if we could find a capacitor that was 2200.33 uF?

Not really. The reason we have two capacitors in this case is that the big one will filter out power supply ripple, while the small one will filter noise. Big capacitors are SLOW, so they don't do a good job of filtering high frequencies. And the small capacitors don't have the girth to handle ripple down to 60 (or 50) Hz.

Also: never count on a regulator to shut down with excessive current draw. I have seen them do all sorts of strange things when there's too much current going through them, and I generally would never want to get it to that point. Usually they will overheat to dangerous temperatures, the output voltage will start to fluctuate, or they may even start to oscillate. That can do more damage than whatever is causing the excessive current draw!

Cheers

[glitched]

I don't think it's a good idea to think of the two paralleled capacitors as 1 big capacitor in this case.  After all, one is 2200 uF and another is 330 nF - do we really need exactly 2200.33 uF?  Could we leave out the 330 nF if we could find a capacitor that was 2200.33 uF?

Not really.  The reason we have two capacitors in this case is that the big one will filter out power supply ripple, while the small one will filter noise.  Big capacitors are SLOW, so they don't do a good job of filtering high frequencies.  And the small capacitors don't have the girth to handle ripple down to 60 (or 50) Hz.

You guys really know your stuff!

I was thinking about this, the inclusion of the 330nf cap...it's such a low value, how can it affect anything?  It makes sense that a bigger cap would take a longer time to charge up and discharge, thus affecting lower frequencies ("slower" waveforms).

[mat_s]

hi, i am building my psu just right now, and i have a question.

about the following connection:

[*The + side of the 2200 16v cap (from the 5v track) is connected to the 330nf cap.]

would it be ok to make the connection almost anywhere on the relevant tracks (after the switch and bridge rectifier basically) ? or does it has to be specifically from the 2200 16v cap to the 330nf cap ?

i was thinking of connecting the ground pin of the regulator to the +5V side of the 100nF cap.

is it alright ?

sorry for the n00b question  :-\

cheers :)

[Narwhal]

If you mean like this image..  yes that should be fine.

[mat_s]

thanks :)

[MRE]

damn..  I had a nice long post.. but the internet ate it..  an (much) abreviated form

http://www.tpub.com/neets/book2/3c.htm

A cap has a charge and discharge time. The charge time is always MUCH longer than the discharge time.

The large cap 'fills in the gaps.'  When the regulator output is low, the cap discharges to take up the slack.. when the regulator output is peak, the cap charges.  Think of it as an ultra fast battery.  In fact, so called mega or super caps are used just like this.  They have charge times of nearly an hour, and discharge times at high currents of up to 2 or 3 seconds (often used in low voltage clocks and the like to trigger a high current event once an hour)  Tripping a solenoid once an hour on a double A will drain it in a day or two. but charging a cap over an hour, and using the cap to trip the solenoid will make the battery last for weeks.

Large caps are used in monitors and TVs to build up enough current at a high voltage to fire the tube, and in stun guns to.. well.. stun you!

To understand the small guy, you need to know noise, and more importantly, that noise is spectrum bursts.

The regulator does a lot of switching inside.  Switching is noise:

ALL signals, are actually compunds of sine waves. Even square waves and sawtooths.  A square wave is actually made up of a fundamental wave (a sine wave at the intended frequency of the square) PLUS several (if not nundreds) of additional frequency waves in order to make up the smooth top and bottom. Likewise, triangles and sawtooths are compunds of the base frequency, plus multiple frequencies at different amplitudes, thus creating the wave form.

So, when a digital signal switches, it bursts out all these waves out into the spectrum.  It EXPLODES them, like... (description deleted for the safety of children).  If we dont do something, these bits of noise will get into your tunes, or out into the open..  (ever notice that sticker on the back of electronics that says "This device will not interfere with other equipment.." ?)

baby cap says 'screw that noise.."  Because its size is ultra small (im not talking package, im talking farads) its charge/discharge time is very tiny (thousands/millions of times a second).  It takes a beating.

Side note:  Check out a professional product power supply schematic.  Older computers like the C64 are perfect.. notice the scematic has a LOT of caps.. most at one or two specific sizes, but a few at apparently random choices..  the common size ones are usually attached as close as possible to all ICs on the boad.. the others are usually just bunched together somewhere near the supply circuit.  The common size ones handle the noise 'at the source'  while the rest filter out the random event noise that could potentially cause interfereance with other devices.

[stryd_one]

Nice post man :) Sorry you lost the 1st one!

[Wilba]

Warning about that image, the bridge rectifier is strange pinout, typically the + and -  are diagonally opposite pins. Someone blew up their C64 PSU by not noticing this.

[Therezin]

Warning about that image, the bridge rectifier is strange pinout, typically the + and -  are diagonally opposite pins. Someone blew up their C64 PSU by not noticing this.

Yeah, when I realised that I put a warning next to the diagram on the page. When I redesign the page I think I'll put a better board layout on there, and put the warning on the layout image.

[glitched]

Warning about that image, the bridge rectifier is strange pinout, typically the + and -  are diagonally opposite pins. Someone blew up their C64 PSU by not noticing this.

Right.  Always check the data sheet for your part, but that isn't always fool-proof.

I had the cylindrical type with the four legs with opposing polarities (I assumed).  Only one leg was marked as positive (+) and the data sheet said nothing of the other legs.  Before soldering anything, I looked at the part on the CORE board for comparison and guessed that this part should be the same--that is, with the opposing polarities diagonally across from one another.  I guessed right, but in order for the circuit to work, I had to "tuck" one of the ~ legs under so that the signal went to it first, then to the + side, etc...

As I have just gotten a new camera, I can start making a mini-tutorial on the c64 optimized PSU.

[MRE]

Nice post man :) Sorry you lost the 1st one!

I think it was the interweb's way of telling me I am long winded.

[glitched]

By the way, here's a pic of my power supply board:

Nevermind the unattached wires; they should be on the +5V side of the board (near the bottom of the pic).

It looks huge, but it's actually half the size of the core board.

I used one of those "box" type capacitors, which took up most of the room on the +9V side.  Next time I'll find more space-saving parts.

Also, you may notice that I had to bend one of the legs (~) of the bridge rectifier.  In this case, finding a PDIP-type rectifier is a better idea, but I didn't have any clue what I needed when buying parts from mouser.

Anyway, the SID is working wonderfully and without a flaw.

Thanks.

-d

[Smithy]

Oh man,

just by looking at this thread i know im after getting in way over my head with the project.  :-\

Im just going to have to learn as much info as i can on electronics before i even start putting the thing together.

Thanks a lot for linking that site which shows how to make the Optimised PSU.

I was trying to find something which shows a noob how to build it on this site from scratch but I couldnt find anything.

Id be lost without it!

[Mr. zip]

Oh man,

just by looking at this thread i know im after getting in way over my head with the project.  :-\

Im just going to have to learn as much info as i can on electronics before i even start putting the thing together.

Thanks a lot for linking that site which shows how to make the Optimised PSU.

I was trying to find something which shows a noob how to build it on this site from scratch but I couldnt find anything.

Id be lost without it!

For a lot of this, just follow the schematics and you'll be ok. Knowing stuff about electroncis helps, but when you've got a schematic or a kit, you really just need to know how to put it all together. *Gets back to making the cheapest transitor-ladder filter possible*

[Smithy]

Hi guys, another noob question here.

Edit: Shortened post!

I want to buy the components for the circuit from a local Maplin store.

One component i couldnt find on their site is the:

2200uf / 25v Capacitor  :-\

The closest they have is a 2200uf / 35v. Will this be okay or not?

Heres their >> List

Any help would be greatly appreciated.

[Goblinz]

Hi guys, another noob question here.

I want to buy the components for the circuit from a local Maplin store.

One component i couldnt find on their site is the:

2200uf / 25v Capacitor  :-\

The closest they have is a 2200uf / 35v. Will this be okay or not?

Heres their >> List

They do have the 2200 / 16v Cap.  :)

Next up is the 100nF Cap.

Maplin have one named as "Monores Cap 0.1uF" Here, which is the same thing right?  ??? (100 nanos = 0.1 micros). Just search for RA49D

I think ive found the 330nF here

Its the 3rd one down = Tant 0.33uF 35V  (Tantalum Bead Capacitor)

Heres the datasheet for their Voltage Regulator:

http://www.maplin.co.uk/Media/PDFs/Module%207937.pdf

TS7809 id imagine.

and finally the data sheet for the Bridge Rectifier

http://www.maplin.co.uk/Media/PDFs/Module%2046435.pdf

The MBS8 seems to be the equivalent of the B40C800.

Oh and of course i need a 200? resistor, and the pins, and cables

Could ye please let me know if the 35v capacitor will do, or if theres something similar on the site.

For my annoyance ill think i put a List Of Parts For C64 Optimised PSU section in the wiki. ;)

Along with the correct codes for Maplins.

It will be usefull for other noobs like me!

Keep reading and when you get stuck ask questions and people will help... I'm relying on that too!

I find that Maplins is very expensive for componants, this seems to be the case with most componant shops in the UK. I used to get all my parts through a small electronics shop which had an account with RS componants, unfortunately they've shut down now. I've recently found one called rapid online (http://www.rapidonline.com/), they seem more reasonable than Maplins and will deal to the public, check your componant prices against there's. I haven't ordered from them myself, but it may be worth taking a look at their site (anybody here used rapid before?).

Another good way to get things cheaper would be to find somebody with an account with RS, I have a friend who's just started (proffessionally???) circuit bending kids toys and persuaded him to get an account, if you ask around you may be able to find someone the same... SOund/light tecjnicians may be a good bet if you know any.

          G

[Smithy]

Okay, sorry for the long post, but i really do need to know if the 35v capacitor is acceptable.

Everything else im pretty happy with, and i reckon it should work.

As for maplins, theyre only 5 mins walk from my house, it would be handy to get them from them.

I only use maplins for small things i need, beacause yeah, they are a bit of a rip off (to say the least).  :P

Im pretty sure the 35v cap will not do because its 25v in the schematic for a reason.

Im not really sure where i could read to find a solution.  :-\

Thanks for the help anyway goblinz.

[DavidBanner]

I use rapid all the time - they are great!!!

[bugfight]

Okay, sorry for the long post, but i really do need to know if the 35v capacitor is acceptable.

...

Im pretty sure the 35v cap will not do because its 25v in the schematic for a reason.

overrating v on a cap is ok, as long as you have space for it

(larger v rating usually means larger cap)

underrating is not ok...

[Smithy]

overrating v on a cap is ok, as long as you have space for it

(larger v rating usually means larger cap)

underrating is not ok...

Ah cool so its rated upto 25v!

Then all is cool, thanks.  :)