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

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
Kai Klaas
05/31/05 05:18
Read: 1108 times

Msg Score: +1
 +1 Good Answer/Helpful
#94136 - Answers...
Responding to: Farshid Jafari Harandi's previous message
Farshid said:
i have a relay board consist of 24 relays which is fired this way :
Main board (MCU AT89S52) -> prepherial board (74HC541->74LS138->74HC164) -> relay driver board (74LS541->ULN2803->relays)
it's a long way ! but i think these should buffer the switching from the main board enough ?

Not really, if you neglect filtering. Really high frequency interference isn't stopped by just putting electronics on different board. Only adequate filtering can help.

- cables going to the prepherial board is a 28AWG flat cable connected to diffrent port pins of the MCU.with no ground in between of each wire, just a common ground signal.

Do always use ground wires between signal wires, whenever digital signals are transmitted.
By the way, '51 micros are not designed to directly drive cables or longer copper traces. Due to the special bidirectional port configuration, they do not show low output impedances all the time, which is extremely necessary when directly driving longer connections at port outputs. You should use 74HCMOS ships to drive the signals to the peripheral board, with filters at their according outputs (common mode ferrite choke preferred).

- cables going to the relay driver board is a 40 pin 28AWG flat cable with a connector same as IDE with ground between each signal ,grounded at one end in the prepherial board (near the HC168 outputs).

Ground wires must always be connected to ground at BOTH ends. Otherwise they do not really work as a shield.

do i have to use a toroidal ferrite on these cables ? how should i wound the cable around the ferrite ? a sample winding(the flat cable around the toriod) i seen in my Camera's USB cable is seen in the below picture.

Compress the flat cable in such a way, that the part of cable which enters the torroid and the part of cable which leaves the torroid does not touch each other! If they touch, then there is again increased stray capacitance between input and output, which heavily decreases the dampening of choke.
By the way, there are also special ferrite chokes for flat cables, where you only need to feed through the cable, without any turns.

is this right,how big the toroid or ferrite have to be ,or what should be the specification of the ferrite ?

The actual specifications isn't really interesting. They do not differ so much. Main thing is, you use them!

do you mean Toroidal Cores by 'triac choke' ?

Yes. These toroid chokes are often used with triac applications.

so what about the amperage rating of the inductor ?

Isn't critical either, as long as you use them for signals. If DC currents have to be filtered, then look, that no more than about 25% of maximum current rating is applied. As with all magnetic metrials, the more current means the more magnetic flux, and somewhere there's a saturation then, where furtherly increasing of H will not show a further increase in B.

question is how to calulate the inductance ?

You worry too much about inductivity of EMI filters. Again, it can be much better to choose a choke of very low inductivity but superior high frequency performance, than to choose a big one, but with bad high frequenc performance. If you use digital chips, like 74HCMOS or micros like AT89S52, then you need dampening at frequencies of up to 1GHz. Choose a choke, which shows high impedance even at 1GHz, and do not worry about actual inductance.
Soft ferrite beads do not provide dampening by their inductance, but by their ohmic core losses, which develop at frequencies higher than about 1...10MHz. So, the inductance only tells you, what is going on at frequencies below 1MHz.
You can roughly estimate the inductance of a soft ferrite choke (in the frequency range below 1MHz) if you just take the impedance Z1 at 1MHz and calculate L = Z1 / (2 x pi x 1MHz). But again, it's the ohmic losses of core at frequencies between 10...1000MHz which makes the ferrite choke being an amazingly good filter, not so much the inductance!

just in case , i'm interested to know what happens if i use this filter before the 5V regulator ?

I just meant to use this filter directly at input of 5V regulator. That's the proper place for this kind of filter, because DC voltage drops caused by the filter will have the least relevant effects here.


List of 22 messages in thread
Info on Murata Inductors for LC filter ?      Farshid Jafari Harandi      05/27/05 03:55      
   Ground routing and filtering        Kai Klaas      05/27/05 04:57      
      Examples of soft ferrites      Kai Klaas      05/27/05 05:07      
         good link      Farshid Jafari Harandi      05/27/05 17:13      
      Ground routing      Mehdi      05/27/05 07:24      
         Ground routing techniques      Kai Klaas      05/31/05 04:34      
      Ferrite on cables and filtering      Farshid Jafari Harandi      05/27/05 17:05      
         Answers...        Kai Klaas      05/31/05 05:18      
            many of my questions are answered ! :)      Farshid Jafari Harandi      05/31/05 12:16      
               Long lines      Kai Klaas      06/01/05 08:50      
                  the other reason      Erik Malund      06/01/05 09:24      
                     Buffers good !      Farshid Jafari Harandi      06/01/05 17:18      
                  my next PCB would be way much better !      Farshid Jafari Harandi      06/01/05 17:18      
   BL02RN2 in Tehran      Payam Soltany      05/27/05 20:10      
      thanks a thousand dear payam      Farshid Jafari Harandi      05/28/05 09:28      
         On-Board Type EMI Suppression Filters      Payam Soltany      05/28/05 18:37      
         Basics of EMI Filters      Payam Soltany      05/28/05 18:38      
         More...      Payam Soltany      05/28/05 18:40      
         Tell me when it's enough        Payam Soltany      05/28/05 18:43      
            you're doing a favor for me!      Farshid Jafari Harandi      05/28/05 19:49      
               Not all available in Tehran probabely//      Payam Soltany      05/28/05 20:40      
               BL02RN2 Datasheet      Payam Soltany      05/28/05 21:13      

Back to Subject List