I think we should also give the original designers some credit. I thought the typical topology was a pretty clever means of maintaining the proper impedance for a uni-directional bus, while remaining "friendly" to practical PCB design constraints. Here are 3 topologies (pardon the crude ascii-art attempt, "+" being an intersection): 1) 37.5_ohm_driver/pkg -> 37.5_ohm_t-line -> + filter -> 75_ohm_vga_cable -> 75_ohm_pulldown | 75_ohm_pulldown The 37.5_ohm_driver first sees the 37.5_ohm_t-line, and then the 75_ohm_pulldown in parallel with a 75_ohm_vga_cable (for a 37.5 ohm net impedance, the filter impedance being left out of consideration), and then the final 75_ohm_pulldown at the receiver. No impedance discontinuities from driver to receiver. Unfortunately, this requires 37.5 ohm routing on the motherboard: not practical. 2) 37.5_ohm_driver/pkg -> + 75_ohm_t-line -> filter -> 75_ohm_vga_cable -> 75_ohm_pulldown | 75_ohm_pulldown The 37.5_ohm_driver now sees the 75_ohm_t-line in parallel with the 75_ohm_pulldown (for a 37.5 ohm net impedance), followed by the the filter and 75 ohm vga cable, and then the final 75 ohm pulldown at the receiver. Unfortunately, this requires 75 ohm routing on the motherboard: not practical. 3) 37.5_ohm_driver/pkg -> + 50_ohm_t-line -> + filter -> 75_ohm_vga_cable -> 75_ohm_pulldown | | 150_ohm_pulldown 150_ohm_pulldown The 37.5_ohm_driver now sees the 50_ohm_t-line in parallel with the 150_ohm_pulldown for a net impedance of 37.5 ohms. The 50_ohm_t-line then sees a 75_ohm_vga_cable in parallel with a 150_ohm_pulldown for a net impedance of 50 ohms (again, the filter impedance being left out of consideration). Then, the final 75_ohm_pulldown at the receiver. No discontinuities from driver to receiver, and allows routing at 50 ohms, a common impedance. This last topology is typically what's called for. When I came across this and reverse-engineered the reasoning behind it, I was impressed at its elegance. True, it doesn't deal with reflections from an imperfect receiver termination well, but it does a nice job with what it has available. I would agree that, as video resolutions go up to incredible values (and pixel times drop), it may not be adequate. But, for the timeframe in which it was conceived, it's worthy of respect. And certainly is worthy of study by folks in this field. This represents a best-guess about the original engineering involved (since I wasn't there); I'd love to hear from anyone with more insight if this was indeed the thinking behind it. P.S. If you're an SI-geek and want to have a GREAT time, start looking at the RGB (aka VGA analog) signals in-depth (with an O'scope, with different patterns on the screen). My dealings with them have been about the funnest thing I've come across. They're also the most confounding, as different video patterns can perfectly mimick overshoot, slow risetime, ringback, etc. Cheers, Jeff Loyer -----Original Message----- From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of olaney@xxxxxxxx Sent: Thursday, October 25, 2007 4:23 PM To: joel@xxxxxxxxxx Cc: si-list@xxxxxxxxxxxxx Subject: [SI-LIST] Re: Question about VGA termination Distributing the near end termination resistance is an attempt to reduce reflections at the VGA connector. However, it's sloppy. Looking from the VGA output, the nearby resistor looks like nothing more than a resistor, and can be ignored. Then you see 50 ohm line, then 150 ohms in parallel with the 75 ohm coax and far end termination at the monitor. 150 ohms at the connector in parallel with 75 ohms =3D 50 ohms. The 50 ohm line is matched and does not "see" the transition into the 75 ohm line.=20 The nominal video level for a VGA monitor is .7 VPP, with black at ground. Peak current is therefore .7/75 =3D 9.33 mA. The chip is = doubly terminated, so must source twice this, or 18.7 mA per your statement. All is well if the monitor itself is well terminated. This is not always the case. Reflections coming back on the 75 ohm line encounter first, 150 ohms paralleled by the 50 ohm line, or 37.5 ohms, yielding a nasty negative re-reflection. However, at the far end of the 50 ohm line, the signal will meet the 150 termination at the chip, which makes another nasty but positive reflection. Keeping the distance between these mismatches short helps to overlap the reflections to achieve some cancellation. The time skew of several nonoseconds between them leads to spikes of noncancellation traveling back to the monitor that are hopefully beyond the monitor bandwidth, and will be low in amplitude if the pixel rise/fall times are longer than the spike widths. The 12 inch rule is to maintain adequate overlap. Beyond that the spikes become wide enough to be increasingly visible. In other words, the recommendation is a clever way of being stupid. The preferred method is 75 ohms at the chip, a 75ohm line to the connector, and only a few inches at most between the connector and the chip! The fact that the connector is not itself a good approximation of 75 ohms (usually higher) might be the driver behind all the rigmarole. It might indicate that the design was optimized with a network analyzer (not a good idea for video). Only a TDR knows for sure. Orin Laney On Thu, 25 Oct 2007 15:20:56 -0700 "Joel Brown" <joel@xxxxxxxxxx> writes: > I am working on a design with VGA analog video interface. The mfg of=20 > the graphics chip recommends 150 ohm termination at the graphics chip=20 > and another 150 ohm termination at the VGA connector. > They also recommend 50 ohm trace impedance between the two 150 ohm=20 > resistors, the resistors may be up to 12 inches apart. There is a=20 > passive filter between the second 150 ohm resistor and VGA connector.=20 > A monitor which has 75 termination resistors plugs into the VGA=20 > connector via a 75 ohm video cable. So this brings up some questions: >=20 > =20 >=20 > Why not use a single 75 ohm termination at the graphics chip? >=20 > =20 >=20 > What is the purpose of having a 150 ohm termination resistor close to=20 > the VGA connector and / or filter? >=20 > =20 >=20 > Why 50 ohm trace impedance (and not 75 ohm) between the 150 ohm=20 > resistors? >=20 > =20 >=20 > Why is there a limitation of 12 inches between the 150 ohm termination > resistors? >=20 > =20 >=20 > I have been running some spice simulations on a filter circuit to be=20 > used on this interface, however there is no spice model for the analog > video outputs of the graphics chip. >=20 > So I am simply using a current source with 0.0187 Amp full scale, is=20 > this a reasonable approach? >=20 > =20 >=20 > Thanks - Joel >=20 > =20 >=20 > =20 >=20 > =20 >=20 > =20 >=20 > =20 >=20 >=20 >=20 > ------------------------------------------------------------------ > To unsubscribe from si-list: > si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field >=20 > or to administer your membership from a web page, go to: > //www.freelists.org/webpage/si-list >=20 > For help: > si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field >=20 >=20 > List technical documents are available at: > http://www.si-list.net >=20 > List archives are viewable at: =20 > //www.freelists.org/archives/si-list > or at our remote archives: > http://groups.yahoo.com/group/si-list/messages > Old (prior to June 6, 2001) list archives are viewable at: > http://www.qsl.net/wb6tpu > =20 >=20 >=20 >=20 ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: =20 //www.freelists.org/archives/si-list or at our remote archives: http://groups.yahoo.com/group/si-list/messages Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu =20 ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list or at our remote archives: http://groups.yahoo.com/group/si-list/messages Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu