The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> High-Speed I/O Design >> Reflection
https://designers-guide.org/forum/YaBB.pl?num=1174225916

Message started by SATurn on Mar 18^th, 2007, 6:51am

Title: Reflection
Post by SATurn on Mar 18^th, 2007, 6:51am

Hi,

I am currently trying to design an LVDS driver for 200-MHz applicatoins. My problem is that reflections due to the bonding inductors and also PCB makes the shape of final signal so ugly. My question is that is there any way to improve the performance of this circuit? Is it a good idea to use internal termination resistor?

Thanks in advance,
Saturn

Title: Re: Reflection
Post by ywguo on Mar 20^th, 2007, 2:38am

Hi,

Have you used any termination resistor until now? Would you please describe your circuit detail so that others can give you helpful advices?

Best regards,
Yawei

Title: Re: Reflection
Post by SATurn on Mar 21^st, 2007, 10:26am

Hi,

What I am using is just a simple LVDS driver without any termination resistor inside the chip. Of course there is a 100-Ohm (float) termination resistor in receiver side. I have not used internal termination resistor just to keep the power consumption of LVDS driver low. What I see is some ringing (caused by series inductance of bonding pad as well as PCB).

So, my question is that is there any way to compensate the effect of these parasitic elements? Like equalization?

Thanks,
SATurn

Title: Re: Reflection
Post by SRF Tech on Apr 12^th, 2007, 9:07am

Equalization is a possible route to improving your signaling as a result of losses in the channel, but I doubt it will solve your problem as this is more an issue of impedance matching. You could use an RF approach and try to impedance match the bondwire to the external transmission line, from the driver, by use of an on-chip capacitor in a L. I personally have never done this, but I have wanted to try. ;D

This would help transmit cleaner signals as well as reduce/terminate returning reflections, without the use of a resistor. (note that most of these techniques are narrowband, so you would need to optimize for the most fundamental frequency of your signal patterns) To be honest, I stated, I would like to try this out myself sometime, but for now, I see a lot of obstacles towards this approach for digital signaling.

Title: Re: Reflection
Post by mg777 on Apr 12^th, 2007, 9:01pm

1. Is the ringing present in the differential mode?

2. Try using a (small) differential resistive attenuator on the source side.

M.G.Rajan
www.eecalc.com

Title: Re: Reflection
Post by SATurn on May 5^th, 2007, 3:32am

Hi mg777,

The ringing is on differential signal.

I just didn't get your comment on adding a differential resistor on source side. Do you mean adding a differential resistor at the output of LVDS driver?

Thanks for your comments.
SATurn