[sdiy] Chris Synths polysynth

Mike Bryant mbryant at futurehorizons.com
Mon Oct 25 17:07:20 CEST 2021 

Yes I was addressing the original poster's request - and did mention the whole lot needs to be in a shielded box.  And most MCUs offer spread spectrum which would also be needed.  

But I really wouldn't knock Behringer like you seem to be doing.  Some of the very best audio engineers in the industry work for Music Tribe.  
Though quite why those engineers remain doing so given some of the things rumoured to happen there is of course another question entirely.


-----Original Message-----
From: Roman Sowa [mailto:modular at go2.pl] 
Sent: 25 October 2021 13:49
To: Mike Bryant; synth-diy at synth-diy org
Subject: Re: [sdiy] Chris Synths polysynth

I think there's some misunderstanding here. I'm talking only about using something like "PWM decoder" instead of "PWM filter/smoother". The guy already plans 100 I/O pins dedicated for PWM, generating a hell of EMI storm with nanosecond rise-fall slopes spread around huge area needed to fit at least one R and C per PWM line.

PWM decoder does not need a micro, the example from EDN was using it only as part of bigger solution described in article.

Sure, not all CVs need to be fast and ripple-free, so those can be done with passive RC filters. But pitch CVs and punchy envelopes would benefit from 1-cycle PWM decoding IMHO.
Another thing - with a decoder the actual PWM frequency may be much lower, as low as 200Hz if you want to reproduce response of vintage polysynths.

BTW it's not 1/3 of 4053, it uses all 3 switches, that's why I chose it, to minimize parts count. It can be done with transistors and NOR gates, but then the parts count is higher. One 4053 is just 3.5x2.5mm in SOT763

But anyway if something really needs 100 DACs and Behringer level cheap, then it's either a concept going wrong way, or indeed a contract work for Behringer.

Roman


W dniu 2021-10-23 o 18:47, Mike Bryant pisze:
> If you’re doing 100 DACs then it will always be extra micros.  Very 
> few have more than about 16 PWMs.  Adding in the extra opamp, 1/3 * 
> 4053 and PCB area that’s still about 30cents per DAC which is a lot 
> more than 10cents for little added performance in most use-cases.    
> Also discharging the capacitor every cycle like that will create extra 
> EMI which will have to be dealt with.
> 
> I’m not saying it’s a bad design but it really depends what the 
> application is.
> 
> *From:*Roman [mailto:modular at go2.pl]
> *Sent:* 23 October 2021 15:42
> *To:* synth-diy at synth-diy org; Mike Bryant
> *Subject:* ODP: RE: [sdiy] Chris Synths polysynth
> 
> But the micro is not needed at all. It was there because it was part 
> of the design anyway, so it could also do selfcalibration. But the 
> idea in its core does not need any micro. Like I said, 2 opamps, 4053, 
> about 10 pasives. Or a bunch of transistors and 74HC02 instead of 
> 4053. When treating PWM with traditional LPF, one needs a couple of 
> similiarly priced and sized components anyway.
> 
> Roman
> 
> 
> 
> ---- Użytkownik Mike Bryant napisał ----
> 
> Whilst the EDN circuit is obviously higher performing, the OP was 
> after 100s of DACs so adding extra MCU ports and a S/H probably isn't 
> going to meet his cost target, whereas a simple alteration to the PWM 
> encoder may give higher performance for free.
> 
> 
> 
> That said, I agree Arturia not doing something like the EDN channel 
> for a single channel output on one of the most important functions of 
> the keyboard is over cost-cutting.
> 
> 
> 
> 
> 
> -----Original Message-----
> 
> From: Synth-diy [mailto:synth-diy-bounces at synth-diy.org
> <mailto:synth-diy-bounces at synth-diy.org>] On Behalf Of Roman Sowa
> 
> Sent: 23 October 2021 08:06
> 
> To: synth-diy at synth-diy.org <mailto:synth-diy at synth-diy.org>
> 
> Subject: Re: [sdiy] Chris Synths polysynth
> 
> 
> 
> Or why not simply make sigma-delta modulator like in any audio DAC today?
> 
> Shouldn't take much of FPGA resources.
> 
> 
> 
> This morning I came up with an idea of PWM decoder giving solid output 
> after single PWM cycle using 2 opamps and 4053, but it looked so 
> obvious that I thought it had to be done before, so did some googling 
> and found this, which is basicaly the same concept, but overly complex:
> 
> https://www.edn.com/pwm-dac-settles-in-one-period-of-the-pulse-train/
> 
> 
> 
> Roman
> 
> 
> 
> W dniu 2021-10-22 <tel:2021-10-22> o15:36, Mike Bryant via Synth-diy pisze:
> 
>  > Good links. They also led to this one where the writer suggests 
> bit-reversing the counter to give a simple form of PDM replacing PWM.
> Definitely an idea I'll bear in mind.
> 
>  >
> 
>  > https://www.edn.com/three-paths-to-a-free-dac/
> 
>  >
> 
>  >
> 
>  >
> 
>  > -----Original Message-----
> 
>  > From: Synth-diy [mailto:synth-diy-bounces at synth-diy.org
> <mailto:synth-diy-bounces at synth-diy.org>] On Behalf Of
> 
>  > Steve via Synth-diy
> 
>  > Sent: 22 October 2021 14:24
> 
>  > To: synth-diy at synth-diy.org <mailto:synth-diy at synth-diy.org>
> 
>  > Subject: Re: [sdiy] Chris Synths polysynth
> 
>  >
> 
>  >
> 
>  > I found these articles interesting, but have not tried in practise
> 
>  >
> 
>  > https://www.edn.com/cancel-pwm-dac-ripple-with-analog-subtraction/
> 
>  >
> 
>  > https://www.edn.com/fast-pwm-dac-has-no-ripple/
> 
>  >
> 
>  >
> 
>  >
> 
>  >
> 
>  > Am 22.10.2021 <tel:22.10.2021> um 12:15 schrieb Vladimir Pantelic 
> via
> Synth-diy:
> 
>  >> Arturia uses a dual PWM scheme in all of the their Keystep and
> 
>  >> Beatstep products, basically following the circuit laid out here:
> 
>  >>
> 
>  >> 
> http://www.openmusiclabs.com/learning/digital/pwm-dac/dual-pwm-circui
> 
>  >> t
> 
>  >> s/index.html <http://index.html>
> 
>  >>
> 
>  >>
> 
>  >> it works nicely, BUT also introduces a >4ms slew on all CV 
> outputs,
> 
>  >> no biggie if driving an analog synth and it gets hidden in the
> 
>  >> attack, but a real nuisance if it gets sampled by a digital module...
> 
>  >>
> 
>  >>
> 
>  >> On 10/21/21 11:34 AM, Tom Wiltshire wrote:
> 
>  >>> Here’s the correction:
> 
>  >>>
> 
>  >>> Imagine you have a DAC with 4-bit resolution. If we use one with 
> a
> 
>  >>> step-between-values of 1V, and then add the output of another 
> with a
> 
>  >>> step-between-values of 1/16th of a volt, we can make an 8-bit DAC.
> 
>  >>> That’s basically what you do with the PWM. The process is helped 
> by
> 
>  >>> the very good linearity of digital-counter-based PWM as a method 
> of
> 
>  >>> producing a voltage.
> 
>  >>>
> 
>  >>> I’ve used a 3-bit PWM scaled to produce 1V per step and a 4-bit 
> PWM
> 
>  >>> scaled to produce 83mV per step to produce Pitch CVs. It worked 
> very
> 
>  >>> well and produced good CVs across the 8 octaves. The reduced bit
> 
>  >>> depth of each PWM DAC means that the output frequency can be 
> higher
> 
>  >>> without the PWM clock frequency having to be super-high, and this
> 
>  >>> improves the effect of filtering and the responsiveness of the CV 
> to
> 
>  >>> changes.
> 
>  >>>
> 
>  >>>> On 21 Oct 2021, at 09:54, cheater cheater via Synth-diy
> 
>  >>>> <synth-diy at synth-diy.org <mailto:synth-diy at synth-diy.org>> wrote:
> 
>  >>>>
> 
>  >>>> Correct me if I'm wrong, but I think summing two PWMs will just 
> add
> 
>  >>>> 1 bit of depth.
> 
>  >>>
> 
>  >>> _______________________________________________
> 
>  >>> Synth-diy mailing list
> 
>  >>> Synth-diy at synth-diy.org <mailto:Synth-diy at synth-diy.org>
> 
>  >>> http://synth-diy.org/mailman/listinfo/synth-diy
> 
>  >>> Selling or trading? Use marketplace at synth-diy.org 
> <mailto:marketplace at synth-diy.org>
> 
>  >>>
> 
>  >> _______________________________________________
> 
>  >> Synth-diy mailing list
> 
>  >> Synth-diy at synth-diy.org <mailto:Synth-diy at synth-diy.org>
> 
>  >> http://synth-diy.org/mailman/listinfo/synth-diy
> 
>  >> Selling or trading? Use marketplace at synth-diy.org 
> <mailto:marketplace at synth-diy.org>
> 
>  >
> 
>  >
> 
>  > _______________________________________________
> 
>  > Synth-diy mailing list
> 
>  > Synth-diy at synth-diy.org <mailto:Synth-diy at synth-diy.org>
> 
>  > http://synth-diy.org/mailman/listinfo/synth-diy
> 
>  > Selling or trading? Use marketplace at synth-diy.org 
> <mailto:marketplace at synth-diy.org>
> 
>  >
> 
>  > _______________________________________________
> 
>  > Synth-diy mailing list
> 
>  > Synth-diy at synth-diy.org <mailto:Synth-diy at synth-diy.org>
> 
>  > http://synth-diy.org/mailman/listinfo/synth-diy
> 
>  > Selling or trading? Use marketplace at synth-diy.org 
> <mailto:marketplace at synth-diy.org>
> 
> 
> 
> _______________________________________________
> 
> Synth-diy mailing list
> 
> Synth-diy at synth-diy.org <mailto:Synth-diy at synth-diy.org>
> 
> http://synth-diy.org/mailman/listinfo/synth-diy
> 
> Selling or trading? Use marketplace at synth-diy.org 
> <mailto:marketplace at synth-diy.org>
>

More information about the Synth-diy mailing list