Author |
Message |
NuttyMonk
Joined: Jun 30, 2020 Posts: 62 Location: UK
|
Posted: Tue Oct 20, 2020 7:51 am Post subject:
|
|
|
Lots of good comments. Thanks dk.
Quote: | I'm not sure why you would want center detent pots on a unipolar CV, unless you have the Arduino programmed so that 2.5V means no change in CV? Just note that some center detent pots have a tolerance where the dentent is, meaning you might need to program a "dead zone" around 2.5V to keep from having slight differences in step length. |
Yes, 2.5v equals no change in CV. I was thinking i'd do a software adjustment on each individual pot to equalise them, as long as they aren't too far out. Otherwise i could try matching them and buy more than i need or a combination of both.
Quote: | There are also a few things that seem like an odd workaround - namely using and AND gate to square up the All Gates and then an opamp buffer afterwards. You'd need less parts if you just ditch the 4081 and make the opamp a comparator instead of a buffer. Unless you meant to AND the All Gates signal with the Clock instead of 12V, so that you'd get a proper pulse out of each step? Cool Wink |
Using the AND gate was to get a 12V gate coming out of it, no matter what the voltage coming off the switches was (due to losses from LEDs), as long as it was over 6V. But you are right. I could just do that with a comparator instead which would also buffer the signal.
Quote: | For the CV summing, I think you'd be better off using summing resistors instead of diodes. The diodes will make the first part of the pot rotation non-responsive, as nothing under 0.6-0.7V will make it through the diode (most of the whole first octave of range in a 1V/Octave system!). I'd use 100K summing resistors and make the resistor in the feedback loop of the summing amp switchable if you want to be able to change the CV range. Out of curiosity, why have the attenuator at the end if you're also going to be switching the gain of the summing amp? Wouldn't it be easier to just have a switch that lets you choose presets (3V, 5V, 12V, etc) and a pot (for variable range)?
|
I never thought about the forward voltage of the diodes. That would be a problem. Summing resistors and a switchable feedback resistor in the opamp sounds better. Do you mean to have a pot in line with the switch at the point of the feedback resistor for variable gain? I just thought it was getting messy in there, especially with designing the PCB. Putting a pot at the end to attenuate the signal seemed easier to me. And the switch at the end to turn the CV row On/Off just seemed like something i might need. It'd be easier than removing the jack and putting it back in if i wanted to quickly check what something sounds like with the CV on or off.
Thanks for all of the comments and suggestions. Will be making some changes today. Thankfully my PCB design won't have to change much as my PC runs really slow when i have a big chunk of components in the PCB design part of EasyEDA. Still trying to decide whether to have each row as a separate PCB or combining them all into one. Would be easier to make changes at a later date if they were separate and a lot quicker to work on in EasyEDA. Just means i need to sort out IDC cables for each row back to the main board behind or be accurate enough with the design to use headers to connect the 'row' and 'main' boards.
Now, it's 3PM, time for my breakfast
Cheers dk.
NM |
|
Back to top
|
|
|
NuttyMonk
Joined: Jun 30, 2020 Posts: 62 Location: UK
|
Posted: Tue Oct 20, 2020 9:00 am Post subject:
|
|
|
The comparator idea for the gates doesn't work well. I set it up with +12V on +VCC and GND on -VCC. I get 11.4V on the HIGH signal and 1.3V on the LOW signal. Trying the AND gate i got exactly 12V on the HIGH signal and 20mV on the LOW signal which is probably just noise in the power rails. Unless there is some way to get a clean 12V on HIGH and 0V on LOW, the comparator just doesn't seem like the right way to go. The AND gate is far more accurate even if it does mean an extra IC on the PCB.
So often electronics doesn't quite work the way it's explained in 'ideal conditions' when you try to build it in real life.
Cheers
NM |
|
Back to top
|
|
|
dk
Joined: Feb 12, 2019 Posts: 115 Location: Europe
|
Posted: Tue Oct 20, 2020 10:11 am Post subject:
|
|
|
What are you using for a chip for the comparators? If you're running it single supply, it needs to be a chip that can run off single supply and (nearly) reach 0V (LM324, LM358, etc). If you used a TL074 or the like there, they don't drop below 1-2V on single supply...
For the feedback switch, I meant that you'd have your largest resistor (most gain) permanently fixed in the feedback loop, and you'd be able to either engage fixed resistors or a pot in parallel to lower that value/make it variable. I'll draw it in a bit when I have a chance. It would be obviously simpler to calculate without having the fixed resistor, but I don't think it would be particularly healthy to have the opamp jump to open loop gain while switching _________________ Horrors Of Dial-Up! on Facebook
Horrors Of Dial-Up! on Instagram |
|
Back to top
|
|
|
NuttyMonk
Joined: Jun 30, 2020 Posts: 62 Location: UK
|
Posted: Tue Oct 20, 2020 10:32 am Post subject:
|
|
|
I was wondering about the opamp i was using. Using one that can reach 0V makes more sense. As long as it can get closer to the +VCC as well. 11.4V was all i could get from the TL072 i was using.
I hadn't thought about what happens to the opamp during switching. That could be producing weird results if it wasn't designed right.
Thanks dk. Look forward to the schematic of the switching/gain pot on the opamp.
Cheers
NM |
|
Back to top
|
|
|
|