Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
02/16/04 16:01
Read: times


 
#64873 - RE: Massive Ground Plane - Kai
Responding to: ???'s previous message
I was just waiting for your expected comment !

I couldn't resist...

He always tells me it is difficult to realise massive ground planes in two sided PCBs with plenty of traces on both sides.

It is, it is, dear Raghunathan. But many circuits do allow it by carefully routing the lines. E.g. I layoutet a display circuit very similar to Oleg's one and it was able to provide a solid ground plane over the whole area of driver chips (74HC573), current limiting resistors and 7segment-LEDs. Also across the whole data and control bus. Only a small number of crossings were necessary, and when traces became too long, so that ground plane was interrupted too much, then I did not route a trace, but used a jumper, mounted on the ground plane side.
Of course, the 7segment-LEDs will introduce some radiation, due to internal created current loops, which is not screenable by ground plane of PCB. But nevertheless, total radiation level is much lower!

When it's needed to route circuitry on a double sided PCB with solid ground plane, it's mostly possible to remanage the traces, so that one side becomes as free as possible for solid ground plane. But it can be, that PCB becomes a bit larger. In most cases this does not make any problems. In any case, application profits of each cm^2 of added ground plane.

HCMOS do really profit from ground plane. LSTTL are less demanding, because they are a bit slowlier. But if you introduce ACMOS or similar technologies you really need solid ground plane, and if it's impossible to realize it with double sided PCB, then only a multilayer board will satisfy. With poorer solutions it's really hard to comply with ce standards.

Very many people totally underestimate the importance of a proper PCB!! To design the circuit on your desk is one thing. But to get it running on PCB is the real challenge, especially if high speed CMOS technologies are used. In critical applications I often change my schematic, when I see, that implementation on PCB is hard to achieve. It's a hand in hand process, all the time changing schematic and PCB layout. Circuit finally must 'look good' on PCB, not in schematic!

In former times one could see many layouts, which were routed by introducing lots of crossings. The upper side was used for vertical traces, the lower for horizontal ones. This really worked to some degree with LSTTL and older microprocessors. But with modern micros and fast digital chip technologies, this will end up with total desaster. Especially, if you want to connect some cables leaving the board to outer world. Without proper ground routing concept you will suffer from so much radiation, that you will never comply with ce standards.

I know, for many people it's difficult to become convinced, that they will profit from concept of solid ground plane. But if you ever worked with really fast stuff, then you will never go back to neat narrow ground wires again!

Tell your PCB guy, he shall try to put as many traces as possible on only one side. First he will heavily complain, but after one hour he will be surprised!

Bye,
Kai


List of 18 messages in thread
TopicAuthorDate
Memory IO issue            01/01/70 00:00      
   RE: Memory IO issue            01/01/70 00:00      
   RE: Memory IO issue            01/01/70 00:00      
      RE: Memory IO issue            01/01/70 00:00      
   RE: Memory IO issue            01/01/70 00:00      
      RE: Memory IO issue            01/01/70 00:00      
         RE: Memory IO issue            01/01/70 00:00      
            RE: Memory IO issue            01/01/70 00:00      
               RE: Memory IO issue            01/01/70 00:00      
                  RE: Memory IO issue            01/01/70 00:00      
                     RE: Memory IO issue            01/01/70 00:00      
                        RE: Memory IO issue            01/01/70 00:00      
                           RE: Massive Ground Plane - Kai            01/01/70 00:00      
                              RE: Massive Ground Plane - Kai            01/01/70 00:00      
                                 A lesson learnt - Kai            01/01/70 00:00      
      RE: Memory IO issue            01/01/70 00:00      
         Discuss of RC delay            01/01/70 00:00      
            Memory IO - resolved            01/01/70 00:00      

Back to Subject List