SE/30 Capacitor addendum

[techknight]

C1, C7, C12, C13 are 47uf 16v, C11 is a 220uf at 16v. These capacitors form a supply filter network on the 5V source. (If bad, will cause simasaimac condition. if noise floor and glitches exceed safe limits, will fry logic ICs)

C10 is a 47uf at 16V. This in conjunction with L11 and C25 forms a tuned pi filter network on the 12V rail. (will cause noisy audio, floppy disk drive issues)

C2 470uf at 16v is the audio power supply filter capacitor. (will cause noisy and weak/distorted audio)

C9 is 47uf at 16v, This forms the -12v supply filter network. (will cause noisy audio in speaker only, -12v is used in the op-amp, and floppy disk drive issues)

C8 is same as C9, forms the -5v filter network. (will cause serial communication or other issues.)

C4 is 47uf and the Right output DC blocking capacitor. (will cause weak/distorted or missing audio in right channel) (If shorted, will drive DC into your headphones/speakers, and that... is.... BAD)

C3 is 47uf and the Left output DC blocking capacitor. (will cause weak/distorted or missing audio in left channel)

C5 is 47uf and is the output reference voltage filter capacitor. (will cause missing audio)

C6 is the Right channel DAC output filter capacitor. 1uf. (will cause noisy or missing audio)

Funny thing is, C24 is the same capacitor for the left channel, BUT its a different value! 0.1uf. Maybe a design error?

Also, C3 and C4 are INLINE CAPACITORS AND ARE NOT GROUNDED!!!!!!!!! These are in the audio circuitry as i mentioned above. Ground these, and you will lose audio for sure.

Anyway, there you go.

 

[Mk.558]

I'll update the reference image tomorrow to reflect these new additions!

Also, C3 and C4 are INLINE CAPACITORS AND ARE NOT GROUNDED!!!!!!!!! These are in the audio circuitry as i mentioned above. Ground these, and you will lose audio for sure.

This means I should reference the pinouts to something other than ground in case someone has to run a wire?

I'd call that an "oops" :-x

 

[techknight]

Yes... C4 runs to R12. and C3 runs to R13. Both are 180ohm resistors. The other ends of the resistors run into a 3 position RF choke/ferrite bead.

Edit: had to make a small correction. Also the Sony ICs are just DACs. thats all.

 

[phreakout]

It's at times like these where tracking down the original design team (the actual people who created every last detail on this model's logic board) would be the craziest, yet most awesome thing to do. I'd just love to sit down and pick their brain for hours and hours about this. Heck, why not the same for other models?

73s de Phreakout. :rambo:

 

[techknight]

Ya, that would be nice. But for me, it would be about the software and how they planned out all the integration they did.

Hardware wise, i pretty much already know how it works. Its really no different than the other systems from the time period, which was brought forward from the homebrew computer designers.

 

[Mk.558]

I'd be happy to list on whatever they can dish out. While I don't understand electronics theory and operation very well (might take some courses on it someday) I'm sure I could learn a thing or two: like the following...

There's obviously an input side and an output side to a resistor. In the Apple drawings, this is shown as 1 and a 2 on it. Now how to find out the right side, as we don't want people running wires around the resistor, but into the resistor.

Oh okay, I found out the answer: 181 printed on the SMD resistor means its 18x10^1, which is 180 Ohms (which is what the drawing said) and so set the DMM to 200. Hold one probe on the capacitor (-) side, and probe the left side of the resistor. About a couple of Ohms. Measure the other side: Ahh 178-ish Ohms. So that means folks should solder a wire onto the LEFT side of the resistor, which is the input side.

Right? If so, how would I "say that" in electrospeak? Also, what would be a good ground hookup in case wires have to be run someplace? I also was told that the whole +5V supply filters are just power decouplers, right?

Edit: multiple edits, in fact...but: You're sure C3 is the Left DC output blocking capacitor, and C4 is the Right? If you look at the BOMARC schematics (Floppy, SWIM, audio) you'll see it says UB10 is the Right audio and UB11 is the Left audio. C3 is married to UB10, and C4 is associated with UB11. Pardon me if I'm all over the place in this post, I was monitoring the local Sheriff when someone decided to beat it on a bicycle in an RV park (and apparently got away).

 

[phreakout]

Mk,

Resistors normally don't have a preference of which end is the entrance and exit points. You can install them backwards, if you'd like, and they still work. It's capacitors (except for ceramic-type), diodes, transistors, and other sensitive electronic components, that do need to be connected the right way. Otherwise, if you don't, a so-called "reflective power" will exceed that component's tolerance and cause arcing or short circuit.

Capacitors act both as a temporary rechargeable battery and be applied as a filter or conditioner. It discharges and builds up a charge just like a battery, they can be wired in parallel to act as a filter (such as a crossover circuit for a loudspeaker) of audio and radio frequencies, also you can wire them in series and/or parallel to smooth out power spikes within a power supply or other power feeds.

Diodes are designed to allow current to flow in one direction, but not the other. If you wire these the wrong way, they will tolerate a reverse direction of current only so much before they short circuit and either open or pass the current through. In circuits with power supplies, where rectifier diodes are used, you have 2 to 4 diodes in one rectifier that will convert AC (alternating current) into DC (direct current); 1 or 2 diodes make sure AC doesn't short or mix with DC and the remainder diodes will pass the DC polarity through 1 or 2 other diodes, depending on whether you use a half-wave or full-wave rectifier.

Transistors and other semiconductors, being they are a DC component and sensitive like diodes, must be wired the right way or they will short, open, or pass the current. In a transistor, you apply a forward current to change the physical properties of the silicon or crystal, so it can allow electrical flow across 2 points. Think of a transistor as being a flood gate on a river. The pathway the water flows through is the base to the collector. The gate that opens is the emitter side. In some transistors, the gate only opens when current is applied to the emitter. When the current stops, the gate closes. Also vice-versa; the gate only opens when current stops being applied to the emitter.

One other note: It is normally good practice to take each resistor out of circuit to get their correct values. If you just check resistance while still wired in a circuit, you are getting the resistance values from other components. In series circuits, the resistance values are added up; in parallel, they are inversely proportional to the total. When 2 resistors in parallel are the same value, the resistance will actually be half of the total through that leg of the circuit (product over sum; R1 * R2 / R1 + R2).

73s de Phreakout. :rambo:

 

[techknight]

Yes, the APPLE schematics show UB11 as Right, which goes through C4. I think I am going to go with apple over BOMARC any day. Because apple made it. lol.

 

[Mk.558]

Yes, the APPLE schematics show UB11 as Right, which goes through C4. I think I am going to go with apple over BOMARC any day. Because apple made it. lol.

You got it.

"So that means folks should solder a wire onto the LEFT side of the resistor, which is the input side.

Right? If so, how would I "say that" in electrospeak? Also, what would be a good ground hookup in case wires have to be run someplace? I also was told that the whole +5V supply filters are just power decouplers, right?"

Any remarks for this?

 

[techknight]

Sorta. The electrolytics are the reserve tanks, and hold the supply from power fluctuations, etc...

I will give you an example of how important those filters are. Well lets say that you have switching currents going on between logic ICs, it causes power fluctuations on the supply line. If the fluctuation is "great" enough, it drops the outputs of the logic ICs which will cause the demand to lower, and raise the supply voltage again. Then once that happens, it sags again.. This is called a parasitic oscillation. This could happen if the filters got bad enough. Where the noise gets into its inputs/outputs and affects the over all operation.

low-value decoupling capacitors are designed to stop parasitic oscillations in the high frequencies. The high value electrolytics are required to stop parasitic oscillations in the lower frequencies.

The small value ceramic capacitors that are in parallel, are the decoupling bypass caps.

 

[Mk.558]

Very well, now how should I "say" that since C3 and C4 run to a resistor, that the wire should be on the side of the resistor that is the "input" into the circuit? If we run a wire on the opposite side of the resistor, we bypass it, which defeats the purpose of having one there. The current version (2.0 is up, 2.1 is in the works) is absolutely stuffed with text, and there's pretty much no room for a 25 word "explanation" without deleting something somewhere else; perhaps I say "Input into resistor"? people are going to be confused by that...primary side of resistor? that's also not clear to the average man. Guess I might have to squeeze something in somewhere: let's see I already shoehorned the voltage rating for the caps...

 

[techknight]

Well, make the picture bigger

But yea the side opposite of that going to the ferrite bead assembly. If you trace the opposite side of the resistor, it goes to the ferrite bead. Maybe you just gotta draw it.

 

[Mk.558]

Well.

It turns out that I had a PM indicating I was in error. We don't do that here! So instead of coming forthright about it, I'm going to try to backpedal and uhmm-ahhh through the whole process with a grim pursed lips look as if I was a politician.

Image

Even better is when I pull the corporate silence/deny/silently fix it/dodge the issue/make you sign NDAs/do just like what Apple and many other corporations do and say nothing and deny everything. ("There's a problem with our iBook GPUs on the logic boards!" Our statement: [insert raw silence, until the class-action lawsuit] "My iPhone 4 got really hot and caught fire!" Our response:[insert silence] "What is the clock rate and the RAM of the iPod touch 4th gen?" Our response: [insert silence]) While you are mulling that over, here is my Public Relations Officer:

We here at the 68kMLA group always strive for the highest quality. We work rigorously and tirelessly to weed out errors, factual mistakes, blunders and discover the truth behind everything we do. This requires excellent dedication, commitment, and values to exceed in our profession. We will work to correct this issue immediately.

There. Now the reality is that somehow the .cwk version is not found, and I did not check the prints for the C8 to C10 section before I published it (I just copied the stuff from the 1st revision). Checked both of my backup volumes. Not there. The iBook that it was made on has been thoroughly wiped. I'm pretty sure I made a backup of it, I don't know where it is. I know what it is called, it is called "untitled 11b.cwk".

This means I must re-create the thing all over again, well, at least most of it. At least there is no fancy dithering or anti-aliasing that I have to deal with when I use the lasso tool to fish out the good stuff. In the meantime the image has been pulled from the server so that it gives PB some time to delete it from their servers while I re-upload it in the same name (otherwise it won't replace it).

Technically, my DMM gets continuity between C10+ and Pin 5 of J12 (is actually supposed to be Pin 14) but that's not factually correct, because that's basically saying it runs ground-capacitor-ground.

 

[techknight]

Now you lost me, except for the last part.

Anyway, I could have sworn you sent me the workfile for that diagram once. But I dug through my emails and PMs and didnt see where you did. But i thought you did.

 

[Mk.558]

I'm pretty sure I didn't send it off to anybody, either.

Ahh, whatever. I already know exactly how I made it, it's a little easier than starting from scratch since I already knew how I did things.

 

[Mk.558]

Finished.

Back on up.

Made some changes to highlight the importance of not gouging the board, TKO'ing traces near C8, C9 and C10...do that and you'll probably loose SCSI, for sure...

Should be A+ now. For some reason, even though both are set to be US Legal in Page Setup, the pixel size is slightly off. Oh well. Close enough.

Once again, if anybody wants the .cwk version, just shoot me a PM. Secret trick about installing AppleWorks on an Intel machine: Install AppleWorks on a PPC machine, under OS 9 if required. Update to 6.2.9, which makes it Carbon also. Then .sitx or something like that the application folder, to retain all the hidden files and what not. Then copy onto the Intel machine, unstuff and hit the launch button. (You can't run Classic mode on an Intel machine, so...yeah.) I'd like to get away from AppleWorks, but I don't like iWork and anything else has caveats.