I read different instructions how to made a ACI cable, but sorry, I am still not sure how it works correct.
On the Appl-1 side it's clear, there are the GND and the signal tab, but on the iPod/iPad side have I to connect the left and right channel together to one signal?
I use this connectors (see picture)
... is described in my "Tips & Tricks" pdf every buyer of my famous IC kits gets. But to make it short, you connect the grounds, and you connect the "Tips" and don't connect the "Ring" on the stereo connector. The mono connector has no "Ring" and is plugged into the ACI.
The idea of the whole thing is to avoid shorting one stereo channel (the one on the "Ring") to ground. This would happen if a mono plug is plugged into a stereo outlet, and the audio player could get damaged. This type of cable avoids the hazard.
regarding stereao audio cables, I remember a discussion on a ZX-spectrum forum where they suggested to create WAV/AIFF files with left channel inverted respect to the right. This would help increase the signal strength coming out of smartphones when the two channels are mixed back to mono. Is this something worth trying ?
I'm not certain what kind of connection the Spectrum users were talking about, but from a purely naïve perspective, I don't understand how that would work. The issue is: what do you mean by "mixed"? If you mean an actual active mixing device that sums an L and an R channel onto an output bus, then opposite phases will cancel, yielding a null output.
If you are talking about a stereo output jack (tip, ring, and shield) where the ring and shield are shorted together by inserting a mono cable (tip and shield only), the signal in the R channel is simply forced to ground (zero). Remember, every signal needs to have a reference to which it is compared. The shield is the reference for line level (unbalanced) cables.
If your connecting cables were specially wired (not off the shelf!) and you connected L (tip) and R (ring) of a stereo plug, to L (tip) and shield of a mono plug respectively, the effect would be to make the R channel of the source appear as the reference of the destination. This would make the signal at the destination be the difference L-R from the source. So yes, opposite phase of L compared to R would double the voltage output in this one bizarre case. However, it's not good practice generally to connect a ground signal to an actively driven signal on the other side of the cable, for noise and electromagnetic interference reasons. So this cable construction should not be used.
P.S. Just to clarify my potentially ambiguous wording above, when I said "and you connected L (tip) and R (ring) of a stereo plug, to L (tip) and shield of a mono plug respectively", I meant:
Nothing is being shorted in this situation as long as the source uses a stereo jack and the destination uses a mono jack respectively, and the source and destination don't share an earth connection (at least one of the two is floating). When a music player or computer is running from batteries and has no other I/O plugged in, it's floating.
It is a bad idea from a noise and RFI standpoint because whatever is on the source's R channel is going onto the ground plane of the destination and radiating energy.
In post #4, robespierre wrote:
"If your connecting cables were specially wired (not off the shelf!) and you connected L (tip) R (ring) of a stereo plug ..."
Uncle Bernie warns:
Don't do this. When you short tip and ring of a stereo plug you short the left and right channel. This is no good and may cause damage, as the two output drivers now may fight against each other.
The whole idea of my MONO/STEREO cable (used for PLAYBACK / TAPE READ) is to leave the "ring" disconnected so as to avoid shorts. This means that the "R" channel is not used. However, note that this cable is meant to plug into a STEREO signal source such as an iPod, which is heinously expensive, so we need to protect it from damage. These media players usually provide a mono AIFF or WAV recording in both output channels, so the ACI will happily see the signal anyways.
To make recordings from the ACI to a media player or sound card, you can use regular MONO/MONO or STEREO/STEREO cables. Just to avoid pitfalls, again in other words: from the ACI "TAPE OUT" to the media player (or cassette recorder "MIC IN") you can use any cable you want. But one channel will be "dead". The signal will only be recorded in the LEFT channel of a stereo recording. Most cheap cassette recorders are MONO, though, so this does not matter. If you use a media player for recording, do not use any compression like MP3, instead record the raw audio, and then edit the resulting file with a good tool, bring up the volume as needed, and save it as a MONO (single channel) AIFF or WAV file. I use the free tool AUDACITY which is excellent as it has a built-in oscilloscope (waveform viewer) so you can read the "0" and "1" in a record per eyesight, and you can judge the quality of the recording.
HOWEVER, your "Uncle Bernie" recommends you to make two custom MONO/STEREO cables with the non connected "ring" and ban all other audio cables in the vicinity of your Apple-1.
The reason is that if you are lazy and only make one MONO/STEREO cable, one bad day, or in a dark night with a poor light, the MONO/MONO or STEREO/STEREO cable will be confused with the custom cable and that ordinary cable will be plugged into the OUTPUT of your media player, which then will have a shorted output amplifier on one channel, and then good luck.
So there is a deeper reason behind everything I recommend. Don't ask how I know about all these pitfalls. Please don't.
About the Sinclair trick:
You would connect the cable like this:
Stereo L channel (tip) to tip of mono plug
Stereo R channel (ring) to ground of mono plug
Stereo ground not connected
Now, if you generate a synthetic stereo audio file where the L channel is the opposite phase (phase turned 180 degrees, or signal inverted) from the R channel, you get a signal of twice the amplitude on the mono plug.
This may help with a typical Sinclair system where everything is small in size and nothing is grounded, but it's an ugly trick to compensate for inadequate sensitivity of the tape read input of the microcomputer. Again, this boils down to the volume being just right.
(And despite of the deficiencies of his notorious "products", Clive was knighted to be "Sir Clive Sinclair". Alas, a title does not mean money, and so he went on to squander the ill gotten gains from his notorious "computers" on his electric vehicle, a death trap which flopped, hereby saving many lives. Still, many fine programmers cut their teeth on Sinclairs. So at least he did some good to mankind).
IMHO, the whole topic of microcomputer cassette interfaces is a highway to hell, littered to the left and right with car wrecks, caused by those who wanted to go too fast (high data rate) and so they crashed and burned. The slow ones (low data rate) never made it either, because the ride (the loading process) took too long and the occupants died from thirst before reaching the next water hole.
This is why cassette interfaces quickly disappeared when the 5.25" floppy disk drive came out. Apple, thanks to the technical genius of Woz, and the greed for profit of Steve Jobs, had the cheapest floppy disk system in the world, as far as manufacturing costs go. Nobody could compete with that. Only the notorious Clive Sinclair could, with his cheap "microdrive" (But remember: you get what you pay for).