jojjelito

Ahh, I see. I guess I was too quick when just taking a peek while at work. There is no such thing as a free lunch... I dread doing the transistor matching just because I figured I wanted instant gratification :whistle: BUT, it will take a while before I get to that stage. Feel free to share to the CEM linky. I'm not sure at the moment how many 3379 chips I have - I stashed some 15 years ago being young and foolish. Now I'm not that young anymore :hmm: Take care, Johan

Hi Jonas, Ahh. At first I thought you were going to port the old app across, but you're right. It's probably easier and faster just to strip out not needed bits out of the full MB-SEQv4 app. Cheers!

Hi Jonas and Peter! I see that Yves (YUSynth) uses a transistor array for his take on the famous VCF instead of manually matching discrete transistors the way that René Schmitz and later Jörgen Bergfors of Bergfotron fame did it. That would save work! After thinking about it over some sleep I decided to keep it as simple as possible: Either use SSM2044 or some other IC I have, or a very simple array-based filter as this will be polyphonic. Also it takes 4xAOUT_NG (expensive, innit?) for control. Add a bipolar PSU and maybe VCAs. Either I will use Seppos VCAs (as I already planned ahead and got some PCBs) or I'll use what's already in the CEM chips if I go that route. Or I can attempt a 1-transistor VCA the way Korg did it in the PS-series of synths, but that's less than optimal for sound and control. Good for the low parts count though. But, I'll make up a tutorial once I get there for sure. At the moment I'm at the graphing paper stage. Then I need to do some breadboarding... Maybe Stophlex has made some progress as I know that he was also looking at an external VCF/VCA box? However, the JB Weld has cured enough. Time to do the encoders and LEDs!

Haha, just add the 808 cowbell and hi-hats and you're almost good to go :thumbsup: But, the 808 has cool toms - they sound almost bongo-like unlike the 909. The main irritation if you build a 9090 is that what seems like half of all the components are used by the toms which are a bit boring. Good job on porting the MB-808 variant of MB-SEQ to LPC-17! It's time for a more modern micro that costs less. Best of luck, it will be interesting to see the results of your work.

Lately I've been working on the proper control surface of the 6582 having run the mighty 6582 in a somewhat completed state with a perfboard control surface just for lulz. I followed Hawkeye's excellent tutorial on the CS construction but mixed and matched with the older methods from Wilba's construction hints where applicable just to be non-conformist fat rebel. :bug: Now I'm at teh "waiting for the glue to cure" stage. :sleep: I wish one could just use superglue or Goop and be done with it. But oh noes, clamping and waiting seems to be the soup du jour... Guess I should detent the encoders or make up a nicer cable to my LCD meanwhile. Meanwhile, this got me thinking about that elusive external filterbox with VCAs (ze famous EEFV, of which there are only whispers of a handful, being rare as hen's teeth). Wonder which filter to use: Transistor ladder a la Moog, diode ladder a.k.a ARP stylee, CEM3372 in the Oberheim fashion, Wonky ladder 303 stylee, or IR3109 in the Jupiter 8 fashion. Mebbe I should just stick to LPF using SSM2044's there or CEM3379 and call it a day? Anyone got any suggestions? Gimped late/early MS-20 perchance? SVF based on a SSM2164 quad VCA like the Shruthi is doing it? Decisions, decisions... :frantics: I smell experiments and a sudden uncotrollable urge to slap something together. It also needs pre/post filter distortion if I can get away with it :sorcerer:

Great idea Nebula! Always go for quick and dirty (doesn't everyone have a bench supply, what's up with that?). Here we go for a quick and reliable solution. Add a 7905 regulator for some -5V fun to the -9V side (and reverse polarity protection diodes if you're ultra-picky!): TFTFY :flowers:

Google voltage inverter and you'll find some (noisy) circuits based around a 555 timer. Or you use a dedicated IC for lesser noise like a LM2687 for instance. The best way to do this without noise or various switchers is to use a bipolar supply. Please be aware that you can't use an inverting opamp to get -12V from a +12V supply. It will never go below GND at any event. You can use a trick like connecting +6V to a local virtual ground, use 0V as -6V, 12V then becomes 6V. However, as soon as you connect this to the outside world where 0V is ground you'll be out of luck unless you do some fancy AC coupling trickery. Also, in real life you won't get 5V out from most opamps when they are fed +5V DC. The max output will be 5V - 0.7V (one diode drop) = 4.3V unless they are of rail-to-rail type which can go full swing. In short: Get a bipolar supply to be safe.

Ahh, nice and trippy! Here's some power-up mushrooms and a couple of extra lives :flowers:

Hear hear! If I knew then I would have used yellow LEDs all over and all that. Now that the rest of my 6582 is green it would look like it was aesthetically challenged with a yellow OLED. But, you're right. It must be OLED now! Plus I want it to display a falling rain-effect like in the Matrix when idle. And use Cyrillic characters since those are inherently cool. And... meh! Just OLED please :D

Eeek! Warum gelb, das ist böse! Mangelhaft! Apart from that ahem, detail, that OLED looks like the business! Wish it was green :whistle: Burp!

Very SID-modulesque choice of port. Either alternative should be fine. Who carries those chips? Vintage Planet or any other suspect? Better look around... I've only given thought to the CEM3396 connected to the 83C53/54 timer IC before. EDIT: Found 'em here: Analogue Renaissance I guess one of the gurus like TK is best suited to answer the extendability question - but some performance measurements won't hurt once you get to that stage. As for the DCO and AOUT I don't see why that shouldn't be possible. The issue here is rather which port to tie the AOUT to. I've seen both J10 and J7 been used in different applications.

In theory you could control the HC595 to output a square at the right frequency, but that means that you'll have to write to it each time you need the output to go either low or high. This in turn means that the CPU must take care of this very exact timing and issue the command exactly when due. This could run off a timer on an interrupt but that won't be more exact than using the 82C53 or 82C54 timer (which also won't use up a ton of CPU cycles better used elsewhere) plus other interrupts will make this very jittery. So, the 82C53/54 is connected somewhat like a SID instead. The other pins can be controlled by a second 74HC595 (look at how the DOUT module can connect 4 of those). For prototyping you can use the J10 timer output (easier to do range switching), then move over to dedicated oscillators (since you may need to implement range switching between 2,4 and 8MHz via controlling some flip-flops). Then we'll follow the 1 osc per timer IC idea as above. The AOUT or AOUT_NG are connected to either J10 or J6 on the old core module or to J19 on the Core32. The SID uses a 24-bit phase accumulator design using dedicated, programmable hardware. You could wire this up, but it would be a large hot mess for sure. See the Bob Yannes interview for some in-depth info. Or you could start using a 24-bit or even 32 bit counter and do it all in software (see DDS on the Electric Druid pages) but with earlier discussed caveats. Anyhow, the larger number of bits for the counter plus the usage of dedicated hardware is the key to precision in the SID case. For us, we could attempt to use a higher base clock for some troublesome parts of the scale -> halved errors for double frequency (which the timer chips won't accept). Or we switch to a more suitable time base. The effects of that could be studied using a spread-sheet.

What you need to do is to find out beyond the shadow of a doubt is which two cables are the 0V (GND) connection. Then connect those two together to the GND of the transformer PSU input of the board. The plus of secondary one then goes to one of the 15V inputs that's connected to either plus or minus of the diode bridge. The plus of secondary two goes to the other 15V input. Both of the secondaries will be AC when they are output from the transformer so which is plus, which is minus of the diode bridge won't matter, as long as you have a common ground and connect the two outputs to the two different sides of the diode bridge. Your picture is on the small side, but it looks like the dots indicate the plus side of the transformer windings. But, please measure before making the connection! Rosch's picture above shows a common dual primary transformer. The cool thing about this is that you can use either input for US voltage, or connect the first zero to neutral, bridge the two 115V and the 0V in the middle, and connect the rightmost 115V to a common 230VAC voltage. OR, you wire up a switch (careful!). The lower side shows the two secondaries. Connect the GNDs of both together for a double voltage (+-9VAC) output and then make sure that the plus outputs goes to either side of the diode bridge.

Ahh, this way we can get 8x8 CVs for sure. There was a S/H module in olden days of MBHP, the SHX8 but it came with latency. Also it used EOL chips from JRC (NJU7304). Those can be substituted with the usual 4051 and 8 external caps we see in other schematics. Still, as you said: it would take some research in order to performance test for latency as well as droop rate which would then yield the size of caps needed etc. Next is the voice architecture: Your example above is the bare minimum. It's then possible to add things like a mixer with a few VCAs for different waveforms going into the VCF, PWM for the pulse-wave, CV for FM, cross-mod, output panner using a cross-fader, digital control for sub-oscillator octave selection, digital control for applying oscillator sync, external signals or noise as needed. The issue there is to lab what *you* deem useful assuming it's possible to use for instance 8CVs plus 8 bits of control signals per VCO. The good thing is that there's lots of classic and modern analog synths to look at for inspiration.

Sounds nice in theory. But sadly it's not easy to use a the divide by n counter (n=1,2,3...) and prescaler in order to get a still very large frequency that is close to what we had from the beginning. Assume fXtal=40MHz. Then divide by 5 and we get 8MHz. Obvious. How do we divide with something that gives us 7.98MHz? The possible frequency steps become too big for small values of n. Obviously we end up with much more subtle differences if we divide the base clock frequency with 15879 in one case and 15878 in another. We can get base clock differences of 100kHz if we want 2MHz (40/20 40/21 40/19 etc). If we want to be hardcore with this we can either use a crystal with some very high frequency which is not very practical, or resort to inherently unstable (compared to crystals) LRC-networks where temperature differences and component tolerances will do the work of introducing some jiggly bits. Or resort to more exotic spread-spectum techniques or frequency shifting or use PLLs. However, any responsible parent or teacher ought to teach that it's generally a bad idea to open a can with a stick of dynamite. However, before anyone gets to clocking there's still the main task of generating square waves with the desired frequencies plus controlling waveshapers, filters and other bits. Be that with timers, PWMs which we sadly lack, or other means such as the garden variety V/oct VCO, direct synthesis etc. The base clock design is a (nice to have) diversion later down the line. Toodles!

Waves hand: These are not the droids you're looking for... The delay is the same, but therein lies not the magic! The thing is that the phase offset is there and will be continously variable over time as we can safely assume that the different clocks will behave differently. Hence, the sound will be less static. We will have all sorts of errors (although small) in both frequency and phase. A little dose of chaos is good. It's not going to be more fun than the fully static condition (where everything is in phase to begin with) if we have just one phase offset that is unchanging. Clear as mud? Otherwise I hafta resort to scope shots or samples :thumbsup: Cheers!

As you observed the phase offset can be anywhere between infinitesimally close to 0 and 100 percent regardless of frequency :sorcerer: BUT you won't exactly have much time to register the phase offset at those high frequencies. The thing is that the offset remains the same (percentage-wise) when we divide down the base frequencies unless we have different signal propagation delays in different dividers. Thus it too becomes audible. The issue is that we may *want* some frequency drift and that comes from using several timing sources, each with their own drift that's independent of the others. Think of it as a subtle chorus effect. Once you try it you won't look back. Or something :ike: Think "super-saw" and you'll get it at its most extreme.

I got some Lelon caps, some Xicon last time I ordered. I've been using them interchangeably in circuits with very high demands on tolerances and temperature stability with no ill effects so far. That's just my limited exposure. YMMV.

Queue please!

There I was, deep in... is the camera on?! Ahh! Got mine today! Henceforth, let it be known that your efforts are deeply appreciated! Thanks Phunk :flowers: This will add some sizzle to the shnizzle that is the 6582 ferschizzle! Must build moar control surfaces :frantics: Can't! Resist! da Uuuurge!

The CPU can be one of several possibilities for the base clock generator. At least it's programmable so it's easy to redefine in software should the need arise. But this way we get three different pitches (in phase wrt to each other when the timer starts) per PWM. There are those who argue that the Matrix 1000 sounds less good than the Matrix 6r due to the latter not using a crystal-derived (or MCU-based) clock but rather gets its base frequency from a high-speed RC oscillator. If we want polyphony with a more lush sound of oscillators beating against each other we want to use several base clocks, up to one per timer chip so that they are out of phase with respect to each other. This comes with the added bonus that it ought to shut up some sceptics who are analog purists. Maybe. I've even heard people complaining that the Juno 60 sounds somewhat thin (wait, what were the smoking?) due to a shared root clock scheme. More schemes to look at than the usual 3396 and Roland suspects would be nice. Maybe something from Crumar or Elka? The Bit 99/Bit 01 are underrated faves from the past for instance. Cheers!

Interesting idea, it's certainly another option. However, you can only get one output divided by N (from 3 to 15999) so it's not full 16-bit division accuracy we're getting. But, it's cheaper. And the package contains the same friendly number of pins. You can get the 82c54 as SOP from Toshiba. If that's hard it can be had from Intel, NEC and Samsung as well, possibly in several package variants. There's lots of ICs out there to explore for sure, the old 4541 contains a full 16-bit divide by N counter, but that IC only works up to 100kHz. These days it seems that the development seems to trend towards low-cost - thus people will have to stick with what's internal in the MCU. The issue then is that we can abandon polyphony unless we have heaps of PWM outputs or use lots of MCUs which may well be cheaper at some point. Take care, J