Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


foft last won the day on August 23 2018

foft had the most liked content!

Community Reputation

1,320 Excellent


About foft

  • Rank

Profile Information

  • Gender
  • Location
    Nyon, Switzerland

Recent Profile Visitors

22,499 profile views
  1. Yeah I think TKII works pretty well with pokeymax if you flash a recent core. I have the XEGS to test with.
  2. Ok, I see. An on-off bit for spdif and each dac output is possible. I’ll add it in the next version.
  3. What are you hoping to achieve?
  4. There are 4 analog outputs. One used for pin 37, two used for line level stereo and the other mostly unused (used in xel install for 2nd pokey pin 37).
  5. I’m not sure I understand what you are trying to achieve. Do you mean using spdif only and disabling the analog outputs for some reason?
  6. Just to let you know, besides that @santosp and I are working on other projects, the fpgas are currently unobtainable due to the chip crisis. So will likely be mid next year at the earliest.
  7. Sio tape sounds are mixed by the analog mixer. So leave that connected. It’s just mixed at a lower level so loud tape sounds are not clipped and the interference isn’t picked up much.
  8. I have this added to the core and it sounds fine to do it digitally. I'll release it probably next week, I want to fix a two tone issue that @Synthpopalooza found/reported first.
  9. I think the plan is still to make v2 for the basic pokey+covox only options. Though the max10 is impacted badly by the chip shortage.
  10. Ok, I’ll add it as an option in the next core.
  11. I guess since its mostly interference from data in, this could also be added digitally in the core. The digital option would also have the benefit of working on v2. I did need (want) the analog mixing (or an adc) to mix audio in from tapes though.
  12. I tested this and it works fine to me. I used a 500K pot to test, which allowed adjusting the volume of the 'silent' beeps from nothing to overpowering. With 10/500 ratio it sounds close enough to the original to my ear. i.e. wired like this: pin 3 sio --- 490K ohm --- pokeymax audin1 --- 10k ohm --- gnd pin 11 sio --- pokeymax audin2 The exact values are not important, but the ratio of ~2% seems good. Also large enough not to disturb sio data transfers. Of course more improvements would be possible, but this gets the core working well enough. Its also a heck of a lot easier than trying to mod the pokeymax. One bonus of this setup: You'll get the sio interference on devices without an SIO cable, such as FujiNet! @DuddieCould you include this in the user manual install process please?
  13. An idea came to mind: sio audio in -> one pokeymax audio input sio data in ->potential divider -> other pokeymax audio input i.e we explicitly mix in some noise. Much easier to build (two resistors and a wire).
  14. I've done some analysis on this tonight. It is not going to be a trivial modification. In fact I think it may be best done either in a v4 board or in an external adaptor, which scale and bias the input first. The circuit on pokeymax v3 for the second two channels is a non-inverting summing amp. https://microchipdeveloper.com/asp0107:non-inverting-summing-amp Essentially this is a passive averager + an op-amp with a fixed gain. The gain is set to 3x. The averager scales the fpga output by 2/4 and the sio and pbi inputs by 1/4. Then it sums them. Giving overall fpga output * 1.5 (i.e. 3.3v->5.0v) and sio in * 0.75 (or connect to both inputs for *1.5). We need instead a gain of ~6x for the sio input and need to maintain 1.5x for the fpga output. Also we need to decouple and bias the channel input since its not realistic to digitally bias at this scale. Ratios between channels are set with RN4 (ext audio in), R8 and R9 (the two channels). Currently RN4 is 1M and the other two 500K. Gain on channel 3/4 is set with half of RN6 and RN7. Gain = 1+RN7/RN6. The half of RN7 used is 1-8 and 2-7. The half of RN6 used is 3-6 and 4-5. RN6 is currently 10K and RN7 is currently 20K. So, this is one potential solution: RN4: 120ohm -> ~10% fpga, ~45% per channel RN6 (half): 1.5K ->gain ~14x Then the audio input should be decoupled then biased to 2.5v. The other audio input should be disconnected.
  • Create New...