On paper the LM1117 is a much better (*) regulator than e.g. a 78L05, so as long as you stay within the power limits you’ll be fine. It is a bit trickier to replace the onboard regulator, though 
*) but here you’re regulating regulated 12 V, so the differences don’t matter that much.
The linked thread goes through that, with a detailed analysis for several Nano variants.
i’ve just find this for gate to Nano
and Maxhirez schem proposal for CVin to Nano
470 ohm or 20K , the Gate send more electricity that a CV ?
Doepfer recommends a 1k resistor with BAT42 diodes. I don’t think the resistor value is critical-like on a passive mixer, as long as it’s there. (I wouldn’t mess with megohm values though.)
thanks, i’m just afraid of frying the arduino with a too powerful CV above 5V
I would probably use this quantizer mainly with S&H which can send little as much I think, I would have to calculate all this …
thanks ; A in is directly a pin of arduino ?
Refers to any of the analog pins that you’d send CV to, I presume?
Yes it is.
I’d be using the dac/ino board I just got the other day and haven’t built yet. I’ve breadboarded the circuit, pretty much, but not tested the board, so it is not necessarily in final working form. The software seems to do the right thing, but hasn’t been used on the final hardware. Anyway, you can have a look at https://github.com/holmesrichards/dac_ino and https://github.com/holmesrichards/QuantizerModule for schematics and code as they are so far.
I’d say 20k is too high and 470 is maybe too low – the Arduino’s sampling machinery wants a bit more current than the 20k can give you to reliably charge the sample & hold capacitor. The datasheet recommends max 10k impedance on the way to the pin (i.e. sum of output impedance and the input resistance should be < 10k). I’d go for 1-2k or so, to limit the max current flow to a few milliamps at reasonable input levels.
(Shottky and zener are very different things, btw, so not entirely sure what they mean there 
)
I think they are there to prevent too high voltages. So if the cv goes above 5v. Or below 0V But I don’t know if that is how to do it…
Yeah, using shottkys as “clamping diodes” is a common pattern (see e.g. the BAT42s and 1N5817s mentioned earlier). They have a low forward voltage, so the input signal won’t get very far above or below the supply rails before it’s shorted away. Zeners have a somewhat different behaviour, though; they’re used backwards, and open when the reverse voltage exceeds a certain voltage.
(but I guess I just answered my own question – with shottkys, the one connected to ground will start conducting when the input voltage goes below −0.3 V or so, but with zeners the diode in the same position will start conducting when the input voltage goes above the zener voltage.)
(Note btw that the Arduino pins have internal clamping diodes, so if you can limit the current you can just rely on them. External ones are perhaps more robust, and definitely easier to replace, though )
EDIT: Below −0.3 V, that is – i.e. one shottky diode drop from 0 V. Similar in the other direction.
thanks a lot all the crew !!! 
After other tips I come back to the construction of this Quantizer.
for the CV IN i do not have 1N5817 or BAT42, but I have some other zener diode …
could one of them do the trick for this application plz ?
1N4001
1N4007
1N4004
Zener 1N759 ATR
Zener 500MW 10V DIO ?
THX
Short answer: none is an exact replacement, but they all provide a bit of protection; any 1N400x is the best option, I think. Or you could leave them out and rely on the ones built into the Arduino, which will handle reasonable mistakes on their own (you’ve built enough gear that I can assure you that you have a ton of stuff that’s more fragile than this circuit without these diodes).
Longer answer:
The 1N5817 is a schottky rectifier with a forward voltage of ~0.2 V at low currents, so they’ll shunt out voltages above VCC+0.2 and below −0.2 V. They can handle 1 A which with a 1k in series means ~1000 V (but at that point everything will break for other reasons, so don’t feed 1 kV into your synth please ).
The 1N400x gang is a bunch of rectifier diodes, with a forward voltage of 0.7 V at low currents, so they’ll handle voltages above VCC+0.7 and below −0.7 V. They can handle 1 A, and the 1N4007 can even handle 1000 V, but I still wouldn’t recommend feeding 1 kV into your synth 
.
(for 1N400x the last digit is max voltage, otherwise these guys are all the same; for all practical purposes, you can always replace a 1N400x with one with a larger last digit, and if the schematics require anything above 1N4001 for low voltage work or 1N4004 (400V) for mains side work it’s either wrong or you should be very careful).
Your zeners are too big for a 5 V gate; 1N759A is a 12 V zener with a forward voltage of 1.1 V so it’ll work as a diode long before it works like a zener, and will thus handle voltages above VCC+1.1 and below −1.1 V. Same/similar for the other zener.
Finally, the built-in ones have ~0.5 V forward voltage, so will shunt out voltages above VCC+0.5 and below −0.5 V. They cannot handle a lot of current, though (said a voltage here before but I cannot math so deleted that) (cannot english either, sheesh)
thanks a lot @fredrik for those info, allways very precise !
it doesn’t work, I put it aside and moved on but I don’t understand the problem.
already even without any signal in the CV IN, the CV OUT outputs a voltage that I can adjust with the trimmpot from 0 to more than 8 V …
i check also with multimeter i have 4.94 V at the arduino power … all seem good … but .
I’ll wait for that of @analogoutput for my next case
I decided to proceed with a revised version of dac/ino, just submitted that to JLCPCB along with a corrected front panel. There’s been lots of delays but maybe in a week to two…