The 74HC595 devices are daisy-chained - so writing values will 'push' the data to the second chip... So you just need to write sufficient data for both devices before latching the data. I haven't used this - but the MAX7219 works in a very similar manner.
Martin

Found the issue - unplugging the PICKit and the speed increase at which it works is dramatic - I changed to using 32MHz internal - and running happily at 100kHz (with mods as per previous post)

Martin

Sounds an interesting project - let us know how it goes.

I just tested changing the ISR to just toggle the state of 'on' - and in the main loop have pin = on - still reflects the right time period, but still doesn't gp above 5kHz

Martin

Something to try: The ZCD should be capable of much higher speeds than this (although maybe not toggling a pin - which will take quite a few cycles) - try incrementing a counter in the ISR instead of the pin set - then output the count every 1s to a UART (via a FTDI convertor) or a display. A capaci...

I wired up a PIC16F18877 to a signal generator and tried the ZCD hardware. Note - I tested with a 20MHz external crystal - if you have use the internal oscillator at 32MHz you will need to set OSCCON. (I did a blinkie first to test settings correct!) This very small sample code - I enable the ZCD in...

Yes, you would need to use some C to enable the ZCD module - this would be in either a code block or in a custom interrupt. See section 21 of the datasheet. It looks relatively straightforward - but if using it to measure mains zero crossing then great care is needed with the wiring (not something f...

An idea... A data buffer to hold (up to) a fixed number of samples - but the number of samples and the size of the sample (byte, int, or long) - can be decided at run time - not at compile time. This allows a user selectable range - with more elements stored if using a smaller data size (byte (size ...

It does 'destroy' the data if you use GetValue. So if you need the last 5 values then GetIndexValue doesn't work - it does what you say - it writes to the end of the array and then starts again at the beginning. It would be possible to create a circular buffer that has the behaviour you require (pos...

If using the circular buffer - then if you don't mind destroying the contents of the buffer - then it's possible - it returns the earliest samples first (so - if you want the last 10 - then discard until 10 (or 10 minutes worth) remain - then average) If you need to keep all the values (user chooses...

Demo runs rather better if exported as a Flowcode app....


Martin