Re: exxos blog - random goings on
Posted: 10 Jun 2019 11:53
I mashed up a "phase corrector" board...
so my plan is...
MMU 16MHz output > 7404 inverter > 5ns stepped delay line > 161...
..where the 161 generates the 8Mhz and 4MHz and 2MHz clocks... Which will be in phase with the MMU clock.
Delay wise, based on the 8MHz clock can vary up to 125ns. Which is basically what my scope was showing. So inverting it halfs that to about 62ns... the 7404 (SN74LVC1G04 ) will add about 2ns delay, so basically nothing.. the 161 will add about 5ns.... so 62ns - 7ns = 55ns stepped delay.. 5ns steps should be enough.
So basically what is happening is, we are taking the clock and delaying it until its back in phase to what its supposed to be. Because of propagation delays we can only add more delays until its back in proper phase again. So while a full clock may be 125ns, we run something like 100ns delay + propagation delays to bring it back to 125ns. This way I can take delays off in 5ns steps to adjust and compensate for propagation delays.
so my plan is...
MMU 16MHz output > 7404 inverter > 5ns stepped delay line > 161...
..where the 161 generates the 8Mhz and 4MHz and 2MHz clocks... Which will be in phase with the MMU clock.
Delay wise, based on the 8MHz clock can vary up to 125ns. Which is basically what my scope was showing. So inverting it halfs that to about 62ns... the 7404 (SN74LVC1G04 ) will add about 2ns delay, so basically nothing.. the 161 will add about 5ns.... so 62ns - 7ns = 55ns stepped delay.. 5ns steps should be enough.
So basically what is happening is, we are taking the clock and delaying it until its back in phase to what its supposed to be. Because of propagation delays we can only add more delays until its back in proper phase again. So while a full clock may be 125ns, we run something like 100ns delay + propagation delays to bring it back to 125ns. This way I can take delays off in 5ns steps to adjust and compensate for propagation delays.