1. Field of the Invention
The present invention relates to the field of connectors, more specifically to connectors suitable for use in high data rate applications.
2. Description of Related Art
One known connector configuration is commonly referred to as a small form-factor pluggable (SFP) connector. SFP style connectors can be configured to provide two high data rate channels and a number of lower data rate channels. As can be appreciated, this configuration is sometimes referred to as a 1X connector as it provides for one channel of data communication for transmitting and one channel for receiving. Other connectors with similar form factors can provide more high data rate channels such as 4X connectors that provide four transmit and four receive channels. Because of the relatively small size, SFP-style connectors have proven useful for mounting in racks and other applications were space is at a premium and because of its performance, have also proven useful in relatively high performance applications. With ever increasing demands for more and more data, however, existing designs, even if potentially suitable for 10 Gbps data rates or greater, have begun to be less attractive for use in applications where it is generally desirable that the connector be somewhat future proof. Therefore, certain individuals would appreciate a SFP style connector that is suitable for applications where a higher data rates might be desired.
A connector is provided that includes a housing. The housing includes a mating face with a slot that has a width and a first and second side. The slot can include a plurality of terminals on the first and second side of the slot, the terminals respectively positioned in a first and second row. At least two pairs of terminals in the first row are configured to provide a differentially coupled signal pair. A ground terminal is positioned on each side of each signal pair. A terminal block can be supported by the housing and can support the first row of terminals in the housing and the terminal block can extend the length of the slot. The signal pairs can be configured to provide data rates of 16 Gbps or 20 Gbps or even 25 Gbps.
The present invention is illustrated by way of example and not limited in the accompanying figures in which like reference numerals indicate similar elements and in which:
The detailed description that follows describes exemplary embodiments and is not intended to be limited to the expressly disclosed combination(s). Therefore, unless otherwise noted, features disclosed herein may be combined together to form additional combinations that were not otherwise shown for purposes of brevity.
Connectors commonly use one or more sets of terminal supported by a housing. Depending on the application, the housing may be mounted on a circuit board by itself (e.g., for internal applications) and when there is a desire to control EMI interfering with and being emitted from the connector it can be surrounded by a cage (e.g., for external applications). The disclosure provided herein is directed toward a connector that in certain embodiment is suitable for both internal and external applications and could be used with any appropriate cage design.
To support and position the first set of terminals 70, the terminal block 80 can be inserted into the channel 52. As depicted, the terminal block in inserted from the rear side 50B toward the front side 50A, preferably in a manner parallel to a supporting circuit board. Unlike conventional waferized terminals (such as typically would be used for a stacked connector), however, the depicted embodiment allows the terminal block to be inserted into the housing in a first direction while provide a row 21 of contacts with the row of contacts being perpendicular to the insertion direction. In an embodiment, arms 82, 86 are mounted in notches 54, 56 in the channel 52 and the notches 54, 56 and the arms 82, 86 can be polarized so that the terminal block can only be inserted in a desired orientation.
As can be appreciated, signal terminals 70B are positioned so as to provide a signal pair 91 and 93 and both signal pairs are surrounded on both sides by ground terminals 70A. As can be further appreciated, a distance 102A (which is between terminals that form a signal pair) is less than a distance 103A. Similarly, as can be appreciated from
As can be appreciated, the terminals supported by the terminal block are at a first pitch at the contact and have a second pitch in the body section. As can be further appreciated, the terminals body has a free portion and a block portion, the block portion residing in the terminal block. To account for the change in dielectric constant caused by the use of the terminal block, the terminals can have one pitch between the free body and another pitch in the terminal block portion. In any event, as can be appreciated from
At the contact location it is difficult to vary the pitch due the desire to have a consistent and reliable connector with a series of contact pads on a mating card. It has been determined, however, that reducing the pitch between the differential pairs in the body section can provide a beneficial decrease cross talk. For example, in a connector with a 0.5 dB dip in insertion loss at about 8 GHz, by using preferential differential coupling it is possible to decrease the insertion loss to about 0.1 dB dip and to move the frequency of this dip loss out to frequencies greater than 11 GHz. Another measurement of the improvement can be determined by the crosstalk, which has a corresponding rise at the frequency of the insertion loss dip. An existing connector was tested and had crosstalk of about 20 dB at about 5 GHz. When the connector was configured in a manner similar to what is depicted in
In a typical first ground terminal 70A, first signal terminal 70B, second signal terminal 70B, second ground terminal 70A arrangement, the spacing between the grounds and signals terminals is kept constant. This is particularly true for stitch SMT style connectors, such as known SFP or QSFP connectors as it is difficult (and perhaps impossible) to vary the distance between stitched terminals if the contacts are going to be kept at a constant pitch. Thus, the distance might be 0.47 mm between each adjacent terminal (which could be 0.33 mm wide so as to provide a desired 0.8 mm pitch). This leads to situation where 33% of the energy is carried via signal pair coupling and 66% of the energy is carried via the signal-to-ground structure.
It has been determined that the energy carried via the multi-terminal ground structure can create resonances that cause dips in insertion loss (and corresponding increases in crosstalk,) such as noted above. Therefore, it can be beneficial to increase the % coupling on a differential pair 91, 93. It should be noted that while two differential pair are illustrated in
It has been determined that one beneficial way to increase the % coupling on the signal terminals is to change the distance between the terminals. The use of blanked terminals supported by a terminal block as illustrated helps allow the distance to be varied. Because of interactions between the terminals, assuming that the ground and signal terminals have a uniform cross-section and associated housing portions, it has been determined that for x (in mm) equal to the distance between the differential pair and y (in mm) equal to the distance between a differential terminal and a ground terminal, the following simple relationship of (1/x)/[(1/y)+(1/x)+(1/y)] provides the percent of energy carried via differential coupling for most symmetric terminal systems. In an embodiment where the distance between each body is 0.47 mm, for example, the formula for the % coupling via the signal pair is 1/0.47/[(1/0.47)+(1/0.47)+(1/0.47)] and this equals 0.33 or 33% coupling. By decreasing the distance between the terminals that make up the differential pair (and or increasing the distance between the signal terminals and the adjacent ground terminals), however, it is possible to provide solutions where the % coupling on the signal pair is increased by at least 10% compared to the symmetric case so as to reduce the energy carried via the ground structure, which tends to reduce potential resonant energy on the ground terminals. The reduction in energy on the ground terminals reduces amount of energy that is reflected and thus helps reduce crosstalk. As can be appreciated, further benefits can be obtained if a 20% increase in % coupling is obtained and even more benefits can be obtained if a 30% increase in % coupling is obtained. While the amount of increase in % coupling that is sufficient to ensure low crosstalk (e.g., less than 40 dB) due to energy reflections on the ground terminal will vary, it is expected that increasing the % coupling by about 30% will typically be sufficient.
As can be appreciated, further increases in % coupling versus the symmetric case can provide further benefit. For example, in an embodiment such as is depicted in
As there is generally a desire to provide a consistent impedance through the terminal, there are limits on how large of a percentage increase in % coupling is feasible. The terminal design illustrated in
In any event, it is beneficial to vary the ratio of the distance between the signal terminals that make up the differential pair and the distance between adjacent ground and signal terminals such that the % coupling is at least 36.5% (the 10% increase in coupling over the standard 33% coupling) and more beneficially is at least 39.6% (the 20 percent increase in coupling over the standard 33% coupling). Further benefits can be obtained by having at least a 30% increase in coupling (to about 43% coupling).
Because of the change in dielectric material, the terminals have a varied pitch and material thickness so as to reduce changes in impedance. The distance 102B is 0.45 mm and the distance 103B is 0.60 mm, which results in a % coupling of 40%. Thus, there is at least a 20% increase over the standard 33% coupling through the terminal body. It should be noted, therefore, while there are benefits to keeping the increase in % coupling consistent, in practice significant performance improvements can be obtained even if the increase in % coupling varies along the terminal. It should be further noted that as depicted, the distance the terminal is in the terminal block is about 2.7 mm and the total length of the terminal is slightly greater than 8 mm, thus terminal block occupies about a third of the total terminal length and based on a weighted average, the increase in % coupling is 0.33(7/33)+0.66(11.8/33), which equal about an average of about a 30.6 percent increase in % coupling. Generally speaking, using the weighted average allows the length of the terminal as well as other variations to be accounted for and is often beneficial.
The alignment bar 177 can be positioned near the contacts 74 and in an embodiment a front face 177a of the alignment bar 177 is positioned so that distance SD between the front face 177a and a center point 64a of the contact 64 is less than 20 mm. In another embodiment, the front face 177a can be positioned so that SD is less than 10 mm. If the distance SD is reduced, the alignment bar 177 can provide greater transverse support at the contact. To help ensure the alignment bar is retained in the desired position on the terminals, alignment notches 178 can be provided in the terminals. The alignment notch may also be beneficial in maintaining consistent impedance through the differential pair. However, if the alignment bar is small then the alignment notch it is expected to have only a minor impact on the impedance and can be omitted if it is not determined to be beneficial in maintaining the position of the alignment bar 178.
Another embodiment of a connector 200 is depicted in
Similarly to configuration of the connector 10, the terminals are arranged in rows. As depicted, the terminals in the lower side of the slot 215 have a row of tails 270a, a row of contacts 270b and a row of bodies 270c while the terminals in the upper side of the slot 215 have a row of bodies 240c, a row of contacts 240b and a row of tails 240a. The terminals are thus arranged in a first terminal set 239 and a second terminal set 270. The first terminal set 239 supports the terminals with a block 240 that is insert-molded onto the corresponding terminals. Similarly, the second terminal set 270 has a block 271 that is insert-molded onto the terminals. The blocks 240, 271 can be inserted into a channel 218 on the support face 210d and, as depicted, can be supported by cross-brace 217. Thus, the surface 218a and the cross-brace 217 support the block 240 and the surface 218b and the cross-brace 217 support the block 271. It should be noted that in an alternative embodiment, the block 240 and the block 271 could be configured so that they engage and support each other (thus removing the need for the cross-brace) and the cross-brace could be omitted. Thus, a number of possible variations exist for structures that could be used to support the terminal sets.
The depicted terminals of the first terminal set 239 are arranged so that there is a first signal pair 250a (which includes signal terminals 242a and 243a), a second signal pair 250b, a third signal pair 250c and a fourth signal pair 250d. As can be appreciated, ground terminals 241a-241f are positioned so that each signal pair is surrounded on two sides by a ground terminal. As noted above, in a convention connector such a configuration would tend to cause the differential coupling to carry about 33% of the energy. However, with the depicted arrangement (more of which will be discussed below) the differential coupling can care more than 40% of the energy, as was discussed above.
As shown, the signal pair 250a has terminals 242a and 243a that are separated by a distance D1 between the tails 240a and the block 240, are separated distance D6 in the block and are separated by a distance D4 between the block and the contacts 240b. In an embodiment, distance D2 and D3 are the same and D7 and D8 are also the same. Thus, each signal terminal is separated from an adjacent ground terminal by a distance D2 between the tail 240a and the block 240, by a distance D7 in the block 240, and by a distance D5 between the block 240 and the contact 240b. Thus, as depicted, the distance between the terminals in a signal pair (D1, D6 and D4 going from the tail to the contact) is less than the distance between a signal and adjacent ground (D2, D7 and D5 going from the tail to the contact). Or to put it another way, the pitch between the bodies of the terminal in a signal pair is less than the pitch between the bodies of an adjacent signal and ground terminal. However, the pitch between the tails and the contacts is substantially constant along the row of tails 240a and contacts 240c. Thus, tail mating interface 245 and contact mating interface 246 for each terminal in the first terminal set 239 can be on the same pitch. As signal terminals in the first set shift to a closer arrangement between point P1 and P2, which allows a substantial portion of the signal terminals to be preferentially coupled (thus providing the desired increase in amount of energy being carried on the signal terminals, as well as the reduction in crosstalk).
The second terminal set 270 also is depicted with four signal pairs 280a-280d and each signal pair is surrounded on two sides by a ground terminal, as discussed with respect to the first terminal set 239. For example, tail interface 275 and contact interface 276 are provided on a constant pitch while the bodies of the signal pairs are at a lesser pitch compared to a body of the terminals of the signal pair and the adjacent ground terminal. Thus, distance S2 is less than distance S1 and S3 (which may be the same) and distance S5 is less than distance S4 and S6 (which may be the same) and distance S8 is less than distance S7 and S9 (which may be the same). Similarly to the terminals discussed above, the reduction in pitch takes place between point P3 and P4 (thus along a majority of the length of the terminal).
Thus,
In the embodiment depicted, the first terminal set 239 is also supported by a tail frame 230 that includes a cross-bar 231. The tail frame 230 helps control alignment of the terminal tails prior to mounting the terminals on a circuit board. The tail frame 230 can be inserted into notches 222a, 222b so that the tail frame 230 is securely supported by the housing 210.
As can be appreciated, the housing 340 can include a combed edge 348. While not required, it has been determined that the combed edge 348 allows for a more gradual transition between the block portion and the free portion and thus can help further improve electrical performance. It should also be noted that while contact rows 340b, 370b are similar to above embodiments, rows of tails 340a, 370a are slightly different, specifically the tail portion of each terminal is substantially the same as an adjacent terminals. This optional configuration may be helpful in tuning an electrical response of the terminals.
The disclosure provided herein describes features in terms of preferred and exemplary embodiments thereof. Numerous other embodiments, modifications and variations within the scope and spirit of the appended claims will occur to persons of ordinary skill in the art from a review of this disclosure.
This application is a national phase of PCT Application No. PCT/US2011/024880, filed Feb. 15, 2011, which claims priority of U.S. Provisional Application No. 61/304,708, filed Feb. 15, 2010, which is incorporated herein by reference in its entirety.
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/US2011/024880 | 2/15/2011 | WO | 00 | 10/23/2012 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2011/100740 | 8/18/2011 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
5342220 | Kodama | Aug 1994 | A |
6478624 | Ramey et al. | Nov 2002 | B2 |
6565383 | Wu | May 2003 | B1 |
6663428 | Wu | Dec 2003 | B1 |
6827611 | Payne et al. | Dec 2004 | B1 |
6916209 | Casher et al. | Jul 2005 | B1 |
6969268 | Brunker et al. | Nov 2005 | B2 |
7070424 | Obikane et al. | Jul 2006 | B2 |
7160141 | Marshall et al. | Jan 2007 | B2 |
7166000 | Pharney | Jan 2007 | B2 |
7578705 | He et al. | Aug 2009 | B2 |
7591683 | Zhang et al. | Sep 2009 | B2 |
7591684 | Zhang et al. | Sep 2009 | B2 |
7670199 | Nagata et al. | Mar 2010 | B2 |
7674118 | He | Mar 2010 | B2 |
7699663 | Little et al. | Apr 2010 | B1 |
7794240 | Cohen et al. | Sep 2010 | B2 |
7794284 | He et al. | Sep 2010 | B1 |
7976344 | Brunker et al. | Jul 2011 | B2 |
8029319 | Kameyama et al. | Oct 2011 | B2 |
8038480 | Wei | Oct 2011 | B2 |
8157599 | Wei | Apr 2012 | B2 |
8226441 | Regnier et al. | Jul 2012 | B2 |
8251746 | Zhang et al. | Aug 2012 | B2 |
8323057 | Ho | Dec 2012 | B2 |
8371882 | Szczesny et al. | Feb 2013 | B1 |
8382519 | Lin et al. | Feb 2013 | B2 |
8419474 | Zhang et al. | Apr 2013 | B2 |
8535086 | Hunkins et al. | Sep 2013 | B2 |
8574011 | Shih | Nov 2013 | B2 |
8657631 | Lang et al. | Feb 2014 | B2 |
8702452 | Tai et al. | Apr 2014 | B2 |
20020048995 | Shindo | Apr 2002 | A1 |
20020102885 | Kline | Aug 2002 | A1 |
20030064626 | Kemmick et al. | Apr 2003 | A1 |
20050164561 | Casher et al. | Jul 2005 | A1 |
20060189212 | Avery et al. | Aug 2006 | A1 |
20080009147 | Avery et al. | Jan 2008 | A1 |
20080261448 | Yi et al. | Oct 2008 | A1 |
20080311801 | Chen et al. | Dec 2008 | A1 |
20090042450 | Zheng et al. | Feb 2009 | A1 |
20090042451 | He et al. | Feb 2009 | A1 |
20090149045 | Chen et al. | Jun 2009 | A1 |
20100267262 | Lin et al. | Oct 2010 | A1 |
20110269346 | Casher et al. | Nov 2011 | A1 |
20110300757 | Regnier et al. | Dec 2011 | A1 |
20120034820 | Lang et al. | Feb 2012 | A1 |
Number | Date | Country |
---|---|---|
2007-115707 | May 2007 | JP |
2009-117128 | May 2009 | JP |
422461 | Feb 2001 | TW |
461621 | Oct 2001 | TW |
M366779 | Oct 2009 | TW |
WO 03026078 | Mar 2003 | WO |
Entry |
---|
International Search Report for PCT/US2011/0248802. |
Number | Date | Country | |
---|---|---|---|
20130196550 A1 | Aug 2013 | US |
Number | Date | Country | |
---|---|---|---|
61304708 | Feb 2010 | US |