Hello Everyone,
is there an IBIS AMI model or something similar for the TUSB1104?
Thank you and best regards
Tobias
Hi Tobias,
Is this question in relation to the E2E thread here? I remember you had an active cable design in mind.
While looking for an s-parameter, our systems engineer pointed out that the 1104 by itself may violate the active cable spec for USB-C. Let me explain.
The USB-C Spec states active cables should have no more than 0dB of gain. You can see this in table 6-15:
This would not be a problem for high speed signals, as insertion loss will lower the gain below 0dB. The issue is for low speed signals, which are not as sensitive to insertion loss.
The TUSB1104 has a certain amount of "DC gain". This is gain applied to low speed signals that would boost them above the 0dB requirement in the type-C spec. There is no way to disable this gain, and it can be seen in the EQ graphs in the 1104 datasheet.
We are concerned that by having two 1104 re-drivers (one on each side), you would be applying this gain twice. For a low speed signal, the gain would rise above the 0dB max set by the USB spec.
From what I can gather, you would need some way to attenuate low speed signals enough to counter the gain from two 1104 re-drivers. A resistor may help to achieve this, but would also impact the signal integrity of high speed signals. Have you considered a solution for keeping low speed signals below 0dB of gain?
Best,
Shane
Hi Tobias,
You are correct that AC capacitors would block the DC component of the signal. The issue is that a capacitor would let through the AC component, even at (relatively) low frequencies. Keep in mind that when saying "low frequency" we are still talking about tens, to potentially hundreds of megahertz.
You would need to attenuate frequencies up to the point where insertion loss in your cable is enough to counter the gain from two 1104s. Judging from the datasheet graph, this looks to be about 1dB per 1104 at the lowest EQ setting.
Best,
Shane
