EECALC
Design tools for electrical engineers
1. Where is EYECALC online help?
2. What is an eye diagram?
3. Who can use EYECALC?
4. What is a Dual Dirac?
5. Seriously, what's a bathtub?
6. What is the .NET framework?
7. Do you have other products?
8. Do you offer discounts?
9. Tell me more about EECALC
2. What is an eye diagram?
3. Who can use EYECALC?
4. What is a Dual Dirac?
5. Seriously, what's a bathtub?
6. What is the .NET framework?
7. Do you have other products?
8. Do you offer discounts?
9. Tell me more about EECALC
1. Where is EYECALC online help?
The online pdf help for EYECALC can be found here.
2. What is an eye diagram?
An eye diagram is used to study the quality of digital signal waveforms, especially after their passage through communication channels that cause intersymbol interference (ISI). The majority of eye diagrams relate to serial data transmission using Non-return to Zero (NRZ) symbols, where a logic LOW is represented by 0 V and a logic HIGH by (say) 1 V. The eye diagram is formed by triggering an oscilloscope with the clock signal corresponding to the data and selecting a high display persistence. If the binary input data is random or pseudorandom the resulting waveform displays all possible histories of the signal. You can think of an eye diagram as representative of the 'bad karma' of a signal.
The online pdf help for EYECALC can be found here.
2. What is an eye diagram?
An eye diagram is used to study the quality of digital signal waveforms, especially after their passage through communication channels that cause intersymbol interference (ISI). The majority of eye diagrams relate to serial data transmission using Non-return to Zero (NRZ) symbols, where a logic LOW is represented by 0 V and a logic HIGH by (say) 1 V. The eye diagram is formed by triggering an oscilloscope with the clock signal corresponding to the data and selecting a high display persistence. If the binary input data is random or pseudorandom the resulting waveform displays all possible histories of the signal. You can think of an eye diagram as representative of the 'bad karma' of a signal.
In particular, an eye diagram shows 'rails' at the top and bottom corresponding to continuous 1's and continuous 0's, apart from the various transitions between the 0's and 1's. It is customary to display half a bit period on either side of the central eye. Note that each 'eye' is identical to the others; if not, it usually means that the data pattern used to generate the eye is not sufficiently random. For an ideal ISI-free channel the eye diagram is a rectangle, or a 'box'. While the eye in the figure above is not a 'box', it is still 'open' enough to be considered excellent. Furthermore, note that all the bit transitions 0-1 and 1-0 fall right on top of each other. This knife-edge repetition in time means that the midway crossings of the waveform can be used to extract or 'recover' an accurate clock signal at the receiver end.
In the presence of ISI each symbol persists for a time that can be much longer than bit period (hence the term intersymbol interference). The figure below shows the same signal after passage through a channel causing a substantial amount of ISI.
In the presence of ISI each symbol persists for a time that can be much longer than bit period (hence the term intersymbol interference). The figure below shows the same signal after passage through a channel causing a substantial amount of ISI.
Note that in the presence of ISI, each 0-1 transition is different from another owing to different bit histories, and likewise with the 1-0 transitions. The ISI manifests in the eye diagram as horizontal and vertical 'closure' of the eye. The vertical closure decreases the immunity of the transmission to additive noise, where it should be noted that the bit error rate (BER) of the link for a given noise level varies exponentially as the square of the vertical eye opening. Hence even a 25% vertical eye closure represents a severe penalty on the signal integrity of the link. The horizontal closure (or timing jitter) decreases the margin available for receiver decision and corrupts the accuracy of any clock that is recovered from the signal.
3. Who can use EYECALC?
EYECALC can be used by anyone who designs or works with serial communication links such as PCI Express, SATA, XAUI, and SONET. Its strength lies in working off a behavioral model for the channel, as opposed to getting caught up with the detailed physical modeling of the channel. As a result, 'what if' scenarios can be easily posed and resolved.
4. What is a Dual Dirac?
First, the Dirac refers to a Dirac delta function d(t), also called an impulse. It is so named after its inventor, the physicist P.A.M. Dirac. You can think of an impulse as a sharp, narrow pulse or blip in time. We would expect such a blip to have plenty of high frequency content. What makes an impulse special is that it contains all frequency components with equal weight. Such an impulse would have infinite energy and hence be unphysical, but it is still a useful abstraction. The response of a linear system to an impulse carries the signature of all frequencies, so with the system's impulse response in hand we can obtain the response to any other signal via Fourier transformation. Clearly an ideal channel will pass an impulse perfectly, so its impulse response is also an impulse.
A Dual Dirac is an impulse response containing two delta's separated by a time interval. It is primarily used to model systems with echos (such as printed circuit board traces interrupted by connectors) so the time interval between the impulses is typically the round trip time of an echo. It is easy to see that a 50-50 Dual Dirac (a=0.5, equal impulse heights) has a spectral null at a frequency equal to half the reciprocal of the split time (the time interval separating the two implses).
3. Who can use EYECALC?
EYECALC can be used by anyone who designs or works with serial communication links such as PCI Express, SATA, XAUI, and SONET. Its strength lies in working off a behavioral model for the channel, as opposed to getting caught up with the detailed physical modeling of the channel. As a result, 'what if' scenarios can be easily posed and resolved.
4. What is a Dual Dirac?
First, the Dirac refers to a Dirac delta function d(t), also called an impulse. It is so named after its inventor, the physicist P.A.M. Dirac. You can think of an impulse as a sharp, narrow pulse or blip in time. We would expect such a blip to have plenty of high frequency content. What makes an impulse special is that it contains all frequency components with equal weight. Such an impulse would have infinite energy and hence be unphysical, but it is still a useful abstraction. The response of a linear system to an impulse carries the signature of all frequencies, so with the system's impulse response in hand we can obtain the response to any other signal via Fourier transformation. Clearly an ideal channel will pass an impulse perfectly, so its impulse response is also an impulse.
A Dual Dirac is an impulse response containing two delta's separated by a time interval. It is primarily used to model systems with echos (such as printed circuit board traces interrupted by connectors) so the time interval between the impulses is typically the round trip time of an echo. It is easy to see that a 50-50 Dual Dirac (a=0.5, equal impulse heights) has a spectral null at a frequency equal to half the reciprocal of the split time (the time interval separating the two implses).
In EYECALC the Dual Dirac impulses are themselves broadened by a channel's impulse response, which is more representative of the loss and dispersion of PCB channels. Figures below show a 50-50 Dual Dirac impulse response and the corresponding eye diagram when the bit period equals the Dual Dirac split time. Note that this is not two 'open' eye diagrams sitting on top of each other - the middle of the eye is actually completely closed!
5. Seriously, what's a bathtub?
A bathtub is a plot of the bit error rate (BER) of a serial communication link as the receiver decision instant (or phase) is swept across the pulse period or unit interval (UI). Clearly the BER will be high at the edges of the pulse owing to the reduced vertical eye opening there. Likewise the BER will generally be smallest in the middle of the eye (the 50-50 Dual Dirac above being a notable exception). Owing to the exponential dependence of the BER on eye opening, the log(BER) versus sampling phase curve is shaped like a bath-tub.
A bathtub is a plot of the bit error rate (BER) of a serial communication link as the receiver decision instant (or phase) is swept across the pulse period or unit interval (UI). Clearly the BER will be high at the edges of the pulse owing to the reduced vertical eye opening there. Likewise the BER will generally be smallest in the middle of the eye (the 50-50 Dual Dirac above being a notable exception). Owing to the exponential dependence of the BER on eye opening, the log(BER) versus sampling phase curve is shaped like a bath-tub.
6. What is the .NET framework?
The .NET framework is a platform-independent managed software execution paradigm developed by Microsoft. The EYECALC Help has a small note on the .NET framework, which you can find here.
7. Do you have other products?
Yes, two other products are currently under development. In addition, work is underway on the next release of EYECALC. Note that licensed users of EYECALC 1.3.3 (the current release) will get a free upgrade to future releases for a period of 1 year.
8. Do you offer discounts?
Yes, we offer substantial bulk discounts. We also offer a special license for corporate use. Please contact us for more information.
9. Tell me more about EECALC
EECALC is a small software company located in Bangalore, India. It is led by a PhD in EE with many years experience in the design and test of high-speed electronic hardware. We understand the special needs of electronic designers and focus exclusively on this market segment. We are responsive to customer issues, please contact us if you have any questions.
EECALC products utilize a number of third party software components licensed from companies in the US and Europe, so we are truly representative of the new global economy.
The .NET framework is a platform-independent managed software execution paradigm developed by Microsoft. The EYECALC Help has a small note on the .NET framework, which you can find here.
7. Do you have other products?
Yes, two other products are currently under development. In addition, work is underway on the next release of EYECALC. Note that licensed users of EYECALC 1.3.3 (the current release) will get a free upgrade to future releases for a period of 1 year.
8. Do you offer discounts?
Yes, we offer substantial bulk discounts. We also offer a special license for corporate use. Please contact us for more information.
9. Tell me more about EECALC
EECALC is a small software company located in Bangalore, India. It is led by a PhD in EE with many years experience in the design and test of high-speed electronic hardware. We understand the special needs of electronic designers and focus exclusively on this market segment. We are responsive to customer issues, please contact us if you have any questions.
EECALC products utilize a number of third party software components licensed from companies in the US and Europe, so we are truly representative of the new global economy.
FAQ
Bathtub
© 2007 EECALC, All Rights Reserved