Jump to content

latigid on

Frequent Writer
  • Posts

    2,524
  • Joined

  • Last visited

  • Days Won

    149

Posts posted by latigid on

  1. The spacing is a bit tight with 700 mil between the pots. I've enlarged the board to 100 mm and spaced the pots 900 mil/22.9 mm apart.

    With a bit of suggestion from Altitude I think it will be better to use an LM4040 reference in place of the linear Vreg.

    I will also include 4427 chips on the control board to optionally select 12 V for two (or maybe all four) gates... Edit: because I will use the same board as an output module for AOUT (+atten.), CV range switch, and gates with LEDs

  2. Nice work. I like that you can switch between different input voltages. Very cool.

    Last week I stumbled over the mutable instrument schematics as well. I prefer the the attenuverting scaler from the braids schematic. Should work with a MCP6002.

    Christoph.

     

     

    Many thanks!

     

    That's a good idea for sure, but for now I prefer to keep it simple. Also attenuation to zero is much easier with a unipolar control. I don't yet know how TK will handle the software side of things and perhaps the levels will be limited in any case.

     

     

    Here's another installment: a control board to work with the AIN board above.

     

     

    med_gallery_5453_5_39310.png

     

    med_gallery_5453_5_24219.png

     

     

     

    My idea is to panel mount the AIN board via the pots, and have this at a right angle above. It includes:

     

    4 sockets for AINs, connected via a ribbon to the bottom board

    4 switches (ON-OFF-ON) to control the input ranges. I chose 2x 16-way IDCs which will span over two of the jumper spaces below.

    4 outputs from a DOUT module. These will be either four CLOCK OUTS, or two CLOCK OUTS, one START signal and one CLOCK IN (see below). These are connected to 3.5 mm sockets but I have also used Schmitt triggers as LED drivers. The 74LVC14A chip (SMD) runs off 3.3 V but has 5V tolerant inputs, so it can be run directly from the DOUT.

    1 CLOCK IN routed to two Schmitt triggers in series (logic level preserved) and the same LED as above. This is jumperable so the correct function (CLOCK IN or OUT) can be selected. I think 100k resistors should provide enough current limiting to protect the chip. The 3.3 V signal can then be connected to a DIN module for clock functions.

     

     

    If I've missed anything, please let me know.

  3. Hey, is this an original build? 

     

    If you want a SEQ V4, I highly doubt anybody will trade down to a non-working V3. Presumably you have a veroboard CS and not the Wilba PCB installed? The good news is that you can swap out the Core and have a V4 right away using all of the original control surface. There is a page on uCapps.de detailing the upgrade.

  4. Well... of course the pot can't drive an op amp as the impedance is too high... I'll move them to the output. Anyone know off hand recommended pot values? I.e. what's good to drive the GPIO correctly?

     

    Think again: the first op amp stage is a better place to put the pot.

  5. Can the free DOUT pins be assigned to LEDs for the seven clock outs? It would be nice to have some visual indication here.

    My current idea is to have a front panel with the following controls:

    Clock out with clock led (8th socket for clock in, LED for this too?)

    CV in

    CV range switch (0-5, 0-10, -5-+5).

    Input attenuator for CV. Possibly this can also act as a fixed 0-5 V control (i.e. just a pot) if nothing is plugged into the CV input.

    Feature request: CV inputs can act as momentary or latching "switches" to enable outputs or operate internal controls. As a crude example, see Doepfer's A-151 sequential switch module. (This just distributes a signal to one of four outputs stepped through by a clock.)

    http://www.doepfer.de/a151.htm

  6. Case(s) in point from MI:

     

     

     

    BAT54S diodes in Frames:

    post-5453-0-50474300-1416003265_thumb.pn

     

     

     

    Rail-to-rail op amp (AD8534) in Braids:

    post-5453-0-46711500-1416003553_thumb.pn

     

     

     

    Rail-to-rail op amp (MCP6004) in Tides (also Grids):

    post-5453-0-62048200-1416004286_thumb.pn

     

     

     

    Rail-to-rail op amp (MCP60042) in Edges:

    post-5453-0-62635000-1416005063_thumb.pn

     

     

     

    The diodes are a bit easier to integrate but their leakage current could mean the ADC inputs go a bit too far over the allowed limits. The op amps will ensure that any over/undervoltages will be clipped to the supply (3.3 V and ground) but the input range is limited to about 0.025-3.275 V, or about 0.06-3.292 for the 8534.

     

    Any preferences?

  7. Hey FFW, Of course, I don't have to buy them. And I'm not after the BLM, so I can't fulfil your package sale. I do think it's a little wrong to profit off Smash's PCBs though, even if you add convenience for somebody. Best of luck with your sale, looks like you have a buyer already!

  8. You can put anything through a recitifier (AC or DC) but you have to deal with the diode drop in both cases. Study the PCB layout on the wiki and draw a circuit diagram for a start. Also keep in mind that several of the pins in the DIN connector are not used... If you're serious (and crazy) about using two separate supplies there's nothing stopping you having two different configurations wired to separate pins of the socket.

  9. Alt, do you mean 1 amp draw for a regular MB-6582 or with your switching solution? I am using your recommended 5 V switching voltage regulator (http://www.mouser.ch/ProductDetail/Texas-Instruments/78SR105HC/?qs=sGAEpiMZZMtwaiKVUtQsNWWYJJUUxYbPRFCRyYymPj0%3D) where I am drawing about 200 mA with all LEDs lit. This is without SIDs so far, hopefully this weekend!
     
    I found the back was a better place to mount it, this way the 7809 can heatsink on the huge ground plane. This works better for the 5 V pinout too, where the (redundant?) smaller smoothing caps are now on the output.
     
    med_gallery_5453_5_327266.jpg
     
     
    med_gallery_5453_5_1539854.jpg
     
     
    To Blatboy: as Wilba designed this with a C64 brick in mind, you have to solder a few jumper wires to get the power in off a single supply. You can for instance use the DIN7 socket and forgo redesigning the back panel. Check this thread for Altitude's work:
     
     
     
     
    Just a note: your rectified 9 VAC will actually be near to 12.6 VDC once smoothed out.
     
    9 VAC * 1.4   = 12.6 VDC  rectifying and smoothing
    12.6  - 2*0.7 = 11.2 VDC  two diode drops
    
    Should be enough for a 7809.

     

×
×
  • Create New...