logic generator

Thanks to all. Now I have to find which solution is for lazy dumbasses, ie, me. That rules out the build-your-own...

Which PIC can generate signals with 10 ns resolution?

On a sunny day (Thu, 22 Oct 2009 05:59:36 +0200) it happened Frank Buss wrote in :

LOL you have to overclock a bit ;-) Did not know that 10 ns was a requirement.

If 10 ns is a requirement, then just use an FPGA evalulation board. My Digilab 2 would do, but is no longer produced it seems, but later models are available from here:

Agreed, the hp stuff can be quite heavy, but it is tough and lasts. The early valve tek scopes were elegant, right down to the little reel of silver loaded solder for the ceramic tagstrips, but some of the later stuff was much more lightweight and the plastics could become brittle with age. The old 500 series with separate psu's would keep a lab good and warm in the winter. Bought an HP175a eventually, which I kept for over ten years and only sold in the late 80's. It was a very good scope, but just took up too much space and the performance was no longer good enough. You would have to prise the current 2465 out of my dead hands, even though the tube is getting a bit dim and had to fix the psu (switching mosfet and diode) a couple of weeks ago. Digital scopes are very usefull, but a good analog scope is still hard to beat for some tasks.

I never owned a 547 and they are very thin on the ground in the uk. That's the dual trace 100 Mhz scope that looks a bit like a 561a from the front, right ?. Really solid bit of kit from what I remember and istr was a specially ruggedised model for the military. I do have a 468 in storage though, which is an early digital storage scope based around the 465 chassis.

Have also kept a rack mount 7000 series mainframe just for the plugin ability. 7s11/7t11, etc, but it hasn't been used in anger for several years. Used 7000 series for years and the only thing that ever went wrong was a psu tant decoupling cap, in a really inaccessable spot behind the plugin area. Took me hours to find it...

Regards,

Chris

PIC from Ubicom would do. Quite a few microcontrollers are capable of 100+ MHz.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Do you have some examples for generating signals with 10 ns resolution? Usually they can't output signals at this speed, even if the internal clock can be much higher. Maybe bigger microcontrollers can do it, like AT91 chips from Atmel, maybe with DMA transfer and external RAM.

I agree with some other posters, that the easiest solution would be to use some FPGA board, if you want to build the generator yourself, but don't want to buy a commercial generator. Any small FPGA board should do it, e.g.:

A slow 40 MHz external clock is ok, because of the internal PLL of most FPGAs.

I do that routinely; it is not that difficult to generate waveforms to one clock resolution. However, I use BlackFin or TMS28xx.

Provided the waveform edges are spaced sufficiently far (few microseconds or so), the arbitrary signal can be generated by MCU timers reloaded at every edge. If a higher rate is desired, you can generate a piece of code which does bit-banging and execute this code from RAM.

Jitter is the problem with MCU generated signals as well as FPGA generated signals. It is hard to get below 50...100ps.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant