The Godfather talking
This is god damn my place! Capisci?
Sonsivri
 
*
Welcome, Guest. Please login or register.
Did you miss your activation email?
December 11, 2016, 08:59:08 08:59


Login with username, password and session length


Pages: [1]
Print
Author Topic: Enhance Parallel Data Transmit with Long Distance Design  (Read 1757 times)
0 Members and 1 Guest are viewing this topic.
Help
Junior Member
**
Offline Offline

Posts: 37

Thank You
-Given: 33
-Receive: 0


« on: November 03, 2007, 05:35:06 05:35 »

Hi,

I'm trying to transmit 8-bits data from A location to B location which is 20 meters long. Do i need to do any enhancement design on both A and B location so the data can able to transmit more stable condition?

Please advice.

Thank You.
« Last Edit: November 03, 2007, 05:45:51 05:45 by Help » Logged
FriskyFerret
Hero Member
*****
Offline Offline

Posts: 564

Thank You
-Given: 513
-Receive: 358


Put it in, take it out.


WWW
« Reply #1 on: November 03, 2007, 06:02:22 06:02 »

Without more information only a broad answer will you get: Yes

You'll need to consider grounding, transmission line effects, switching rate, settling time, driver impedance, passive or active termination, cable impedance, connectors, etc. All depends what data rate you're looking for and what error rate is acceptable, if any. For 1 Hz, use as is.

Is this your homework you're wanting us to help with?
Logged

Dancing pants and leotards, that's what I'm talkin' about!
Help
Junior Member
**
Offline Offline

Posts: 37

Thank You
-Given: 33
-Receive: 0


« Reply #2 on: November 03, 2007, 07:57:18 07:57 »

Hi,

Sorry, i wonder why my attach file couldn't show on the board? Please can you teach me how to do attachment? I'm still new on this forum.

I'm using 8052 MCU. The MCU port will connected it to 74HC573 to pull out the 8-bits data at location-A then transmint the data to location-B (location-B most of that is TTL device, the location-A 8-bits data will link it to 74HC259 and some other logic device. The HC259 output will be connected to x8 LEDs). Between location A and B will be very long distance, 20 Meter. I just plan to use parallel method to transmit the data. I know serial method can solve this problem. I just want to take experience on it. Hope can use parallel method to solve it.

I already did the design but is on testing board. But haven't test it in long distance. 
1) How to solve the transmission line effects?
2) the switching rate is not so frequent. If can turn ON the output LED within 1sec will be nice. Is it possible to do that?
3) driver impedance - using normal TTL driver HC573 (data output driver), HC259(input device), 4077, HC153, HC00 and so on..
4) I think this design should be passive termination.
5) Normal parallel port 25-pin connector and cable.

What other issue i have to consider?

Thank you.

 
Logged
jasmine
Guest
« Reply #3 on: November 03, 2007, 08:30:04 08:30 »

You will have to think from grounding point of you as well.
(1) Your circuit ground at point A and other circuitry on point B should have the  common
ground.
(2) Consider Power Supply filtering as the hight priority. As the return path noise
and noise riding on supply can mesh up entire system.
(3) PCB layout point of view

Above said are the basics... but the most important

If need more feel free to ask me
Logged
Help
Junior Member
**
Offline Offline

Posts: 37

Thank You
-Given: 33
-Receive: 0


« Reply #4 on: November 03, 2007, 09:18:29 09:18 »

Hi,

1) Be sure, i will concentrate the ground. I'm using 25-pin connecter but not all the pin i'm using so i will connected it all to ground. Include the DB25 connector casing.
2) How to filtering the power supply?
- I'm using one 12V/5Amp power supply to supply for location A and B, each location would use 5V regulator to get 5V.
- The supply cable is saparated with the 25-pin connector.
- I would like to put 4700uF/50V and 1uF/50V capacitor on each location at supply input for stablelize the voltage.
- each TTL IC would put one 100n cap beside the Vcc pin.
any other points to reduce the noise?
3) what's the PCB layout point of view i have to concern?

On the HC573 output will use 10kOhm to pull it high. Do we need 27pF cap or RC design on HC573 output?

I heard some people say:
They don't think so can really work.. because for long distance transmission of low level signals, loading occurs which cause the signal to distort in the middle.. also when we make changes on the line like from 0 to 5 or 5 to 0 then we might not get the desired results.
And if any frequent change occur in the signal then it may be possible that we don't get it properly on the other side.

 
Thank for your kindly help and advice. I'm very appreciate.

Thank You.
« Last Edit: November 03, 2007, 11:27:53 11:27 by Help » Logged
smainj
Active Member
***
Offline Offline

Posts: 125

Thank You
-Given: 69
-Receive: 75



WWW
« Reply #5 on: November 08, 2007, 11:59:53 11:59 »

tray to send it in serie and use max232
Logged

Silent_Thunder
Junior Member
**
Offline Offline

Posts: 57

Thank You
-Given: 28
-Receive: 17



« Reply #6 on: November 14, 2007, 09:51:33 09:51 »

Hi;
I think the only problem you gonna face is current drop down over this distance, and of course the data rate, in slower data rate there will be no problem, consider poosting your current ( for example using ULN2803 taking in your consideration the data will be inverted), or use a RS485 but that will require adding extra hardware and another MCU to handle the recived data
Logged
Parmin
Hero Member
*****
Offline Offline

Posts: 544

Thank You
-Given: 357
-Receive: 126


Very Wise (and grouchy) Old Man


« Reply #7 on: November 15, 2007, 11:04:45 23:04 »

IMHO, you better use some converter from Parallel to Serial and them back to parallel.

It should not be too hard by using some PIC with built in USART, it could be quite fast too.
Or if you must, you could use some shift register too.

Either way, they are more reliable than just sending the data out with amplifiers in parallel.

Par.
Logged

If I have said something that offends you, please let me know, so I can say it again later.
kolin
Junior Member
**
Offline Offline

Posts: 40

Thank You
-Given: 25
-Receive: 19


« Reply #8 on: November 20, 2007, 05:39:21 17:39 »

Use 2 micros with UART and MAX485 chip or equivalent. Wire it up this way
http://aquaticus.info/system/files/pictures/atmega8_max485_0.preview.png
and write a very simple program.
I have done a lot of this stuff in work, ask me by PM if you want
You can do galvanic insulation fairly simple if needed
Logged
Help
Junior Member
**
Offline Offline

Posts: 37

Thank You
-Given: 33
-Receive: 0


« Reply #9 on: December 08, 2007, 04:32:49 04:32 »

Hi Dear All,

Thank for your all advice and shared the knowledge.
 Kiss
Last week i just done the testing. It working fine.
I'm using ribbon cable to carry my data along 24 feets and i'm using low-level logic signal as my data signal.
The transmision data rate is around 1Kbps.

I'm not sure that is the long term solution but i'm be observed one week it working very well. Alot of people don't advice to use parallel data for long distance transmission. I will try to switch the design to serial 485.

Thank again... Smiley

« Last Edit: December 08, 2007, 05:23:06 05:23 by Help » Logged
bogdantk
Junior Member
**
Offline Offline

Posts: 96

Thank You
-Given: 13
-Receive: 18


« Reply #10 on: December 10, 2007, 09:10:40 09:10 »

Hello
I tested a link full duplex over 1000 m long in an industrial environment
(very noise !) with very good results using  75176 RS485 drivers and
CAT 5 twisted pairs cable. One cable for 4 bits might be expensive but I was
able to work at speeds over 115200 bps over 1000m and with no errors
(tested 2 hours with to PCs 0% error rate).
All the best
Bogdan
Logged
yoda
V.I.P
Junior Member
*****
Offline Offline

Posts: 58

Thank You
-Given: 611
-Receive: 17


« Reply #11 on: December 13, 2007, 11:44:25 23:44 »


On the HC573 output will use 10kOhm to pull it high. Do we need 27pF cap or RC design on HC573 output?

Hello, you said you where interested in trying out the parallel possibilities of this 20m connection, so no advice for going serial (which is better of course). But your 10k pull-up's are a bit weak then, for such a long cable.
They should be about 120 ohms (cable termination resistance) but that is to much for this driver. Try to find out how low you can go (from 10k to 120 ohm) and still get valid zero levels.
I would at least try 1k or so, (I dont know what the 537 can sink). You will notice a higher possible transmission rate then.
Don't use cap's without checking with a scope here. You are definitely having reflections on the lines now, but they just don't hurt probably at the moment.
On the receiving end you also should terminate (for minimising reflections), you can experiment at that point (try what 1k to +vb does for you).
Bye.
Logged
orange
Active Member
***
 Muted
Offline Offline

Posts: 126

Thank You
-Given: 703
-Receive: 199


« Reply #12 on: December 14, 2007, 08:21:23 08:21 »

Normally, TTL level could communicate reliably within 5 feet.
Logged
yoda
V.I.P
Junior Member
*****
Offline Offline

Posts: 58

Thank You
-Given: 611
-Receive: 17


« Reply #13 on: December 14, 2007, 03:58:49 15:58 »

Normally, TTL level could communicate reliably within 5 feet.

Yes, but the OP said he already obtained a "transmision data rate around 1Kbps" which is possible if you make sure you sample  (clock-in) the data somewhat later than you enabled it. So not on the leading edge but after a slight delay.You can generate this delay bij monitoring the data or by using a separate strobe pulse.
Bye.
Logged
Help
Junior Member
**
Offline Offline

Posts: 37

Thank You
-Given: 33
-Receive: 0


« Reply #14 on: December 15, 2007, 08:06:23 08:06 »

Hi yoda,

Thank for your advice... Smiley
Now i already change the design, change the driver HC573 to HC07 buffer with 10k it working fine. Is it ok for 10k pull-up's? for HC573 internel IC which have build in 120 ohms i think.
Is it we put lower resistor for pull up's the data signal on transmition would be more stable? Because enough current to flow through the cable?
On the receiving end...how can i terminate the reflections?
I had do some study on LVDS   http://www.interfacebus.com/Design_Connector_LVDS.html  the LVDS Bus mothed is quite similar what i had did

Thank you.
Logged
yoda
V.I.P
Junior Member
*****
Offline Offline

Posts: 58

Thank You
-Given: 611
-Receive: 17


« Reply #15 on: December 16, 2007, 01:06:15 01:06 »

Is it we put lower resistor for pull up's the data signal on transmition would be more stable? Because enough current to flow through the cable?
On the receiving end...how can i terminate the reflections?
Hello Help,
Yes, a lower resistance causes more current to flow and thereby charges and discharges the (line) capacities faster, so that the rising and falling edges of the datawaveform stay steep enough, otherwise you get timeshift problems. That is one reason, the other one is to prevent reflections, and they become less in amplitude if your resistance (on both ends of the line) get nearer the characteristics of the cable.
Your cable will have a characteristic resistance between 100 and 200 ohms I think, so your 10k should go down as far as possible to that value. But the driver must be able to sink that low resistance (from the pull-up) to ground, otherwise you dont cross the threshold of the detection on the receiving end. So you must find a compromise.
I would try 1k as pullup on the transmitter and 1k (to +vb) on the receiving end.
See what happens to your throughput (speed).
I'll see what the 573 and the 07 can drive and take a look on that interfacebus site.
Good luck, Yoda.


Posted on: December 15, 2007, 11:04:41 23:04 - Automerged

.how can i terminate the reflections?
I had do some study on LVDS   http://www.interfacebus.com/Design_Connector_LVDS.html 

Hello Help,
I look'd at the interfacebus site, the LVDS interface is not wat you want, it is a balanced line design, like RS485 or 422.
I also saw that the HC573 is a good choice (you made earlier), it can sink 25 ma, so your termination resistors may go as low as 200 ohms.
So you may change your pull-up's to 470 ohm on the 573 side, and also 470 ohms on the receiver side (in parallel they deliver less than 25 ma then).
Because you want such a low data rate (1 second? per byte) you should not worry too much for the receiver side (1 to 10 K) will still work there.
Only if you want faster throughput of data the reflections on the line must be taken care of. Also decouple your powerline at the driverpoint with 10uF or more, because 8 or 16 x a zero in the data, means around 300mA current. Take care that your oVolt or ground wire can supply that current (extra wire from driver to psu connection, and the elco on the 573 supply pins).
A secure way to deliver data this way is to send a separate strobe signal to the receiver, a short time (1 or more milliseconds) after the data is enabled, (and the lines came to rest, (reflection has gone then)) if you do that, the whole resistor termination story can be forgotten, important than remains: a low impedance circuit is less sensitive to external inductions then a high imp. one.
Good luck again.
Logged
Help
Junior Member
**
Offline Offline

Posts: 37

Thank You
-Given: 33
-Receive: 0


« Reply #16 on: January 05, 2008, 04:01:13 04:01 »

Hi yoda,

Sorry for so long to reply you.

What does you mean "Also decouple your powerline at the driverpoint with 10uF or more, because 8 or 16 x a zero in the data, means around 300mA current"?

Now i change all the transmiter drive to 74HC07.
Actually how to connection look like? If we connect the 470ohms resistor to ground at the receiver and the resistor pull-up at transmiter is 470ohms we would get very low voltage because we already divider the voltage. I already draw out the diagram. Please give some advice.

Thank you.
Logged
yoda
V.I.P
Junior Member
*****
Offline Offline

Posts: 58

Thank You
-Given: 611
-Receive: 17


« Reply #17 on: June 14, 2008, 01:47:49 01:47 »

Hi yoda,  Sorry for so long to reply you.
What do you mean "Also decouple your powerline at the driverpoint with 10uF or more, because 8 or 16 x a zero in the data, means around 300mA current"?
//
Actually how the connection looks like? If we connect the 470ohms resistor to ground at the receiver and the resistor pull-up at transmiter is 470ohms we would get very low voltage //
- end quote -


Hello, also a long time ago ..

All the (470 ohm) pull up's go to plus Vb, (on the tx and the receive end) so no division will occur.
Decouple the powerline here means an elco over the power-connections from the chip that delivers the drive current (to the cable).

You will have this problem solved by now I think, because I forgot to check this conversation in januari, sorry for that.

Greetings. yoda.
« Last Edit: June 14, 2008, 01:51:11 01:51 by yoda » Logged
Pages: [1]
Print
Jump to:  


DISCLAIMER
WE DONT HOST ANY ILLEGAL FILES ON THE SERVER
USE CONTACT US TO REPORT ILLEGAL FILES
ADMINISTRATORS CANNOT BE HELD RESPONSIBLE FOR USERS POSTS AND LINKS

... Copyright 2003-2999 Sonsivri.to ...
Powered by SMF 1.1.18 | SMF © 2006-2009, Simple Machines LLC | HarzeM Dilber MC