Other Parts Discussed in Thread: TUSB216-Q1, TUSB217-Q1, TPS25830-Q1, TPS2583XAQ1EVM-147
Hello,
I need the SPICE model for TPS25830A-Q1
Thanks,
Mani M
Hello,
I need the SPICE model for TPS25830A-Q1
Thanks,
Mani M
Hello Seong,
In the absence of SPICE model, can i use RDS_ON (4.3ohm) value for simulation? Along with RDS_ON, CIO_ON also include?
Provide a circuitry between DP/DM_IN to DP/DM_OUT.
This may helpful for our simulation.
Thanks,
Mani M
Hello Manikandan,
Yes, 4.3Ω is RDS_on for the DP and DM switch on-resistance. CIO_on, the DP/DM switch on-state capacitance is 10pF.
See the Functional Block Diagram for the DP/DM circuitry below.
Hope this helps,
Seong
Hello Seong,
Ideally i used 4.3ohm resistance in USB lines, anyway i will use Cio capacitance also.
Additionally, in datasheet i saw the insertion loss is -3dB near 800MHz, i hope this large loss chances of degrade the eye diagram, when using USB cable (roughly 1.6m). Is there anyway to improve the performance of eye diagram or any other TI parts recommended with lesser insertion loss?
Thanks,
Mani M
Hello Manikandan,
Please see our application report, "How to Improve an Eye Diagram Using Long USB Cables" here. You have two options:
BR,
Seong
Hello Seong,
Thanks for your updates.
Regarding LC network to DP/DM lines, how to choose those values? Whether the LC value depends on the USB cable?
For Eg., if i use 1.5m, 2m & 5m cable, the LC value constant for all USB cable length?
Thanks,
Mani M
Hello Manikandan,
USB cables create an equivalent capacitance and resistance making the USB port bandwidth more narrow and impede the signal integrity. So signal integrity will become worse with longer cables.
The eye diagram pattern should be measured first to be used as reference when selecting the LC values. To improve the the eye opening, a series inductor between 5nH to 15nH should be added first and tuned to extend the bandwidth at the expense of less linearity. Then add and tune a shunt capacitor to compensate the inductor and improve the linearity. 4.7pF is a good value to start with.
Hope this helps,
Seong
Would you be able to generate the Spice model or S-parameter for this IC ?
Thanks!
Hello Seong,
In datasheet, the insertion loss mentioned as -3dB near 800MHz. How to include this data transmission loss in Simulation. Since i considered only RDS_ON value only.
How the insertion loss plot derived? Can you share more information about this.
Thanks,
Mani M
Hello Aptiv team,
Unfortunately, I won't be able to help when it comes to generating a Spice model for this device.
I believe a network analyzer was used to measure the insertion loss. I will see if I can reproduce the same measurement for DP and DM, and if successful, I will try to generate an SxP file to be used for simulation. However, it may take up to about a week before I can deliver this. Will two S2P files (one for DP and one for DM) or an S4P file work for you?
BR,
Seong
Noted. I will keep you posted.
This is 3years old request, where TI has commented S-parameters model is available.
We believe both parts have similar insertion loss characteristics. Can you help to get S-Parameter model for TPS25830-Q1?
Hello Akhilesh,
Yes, based on the network analyzer shots in their datasheets, I would expect their insertion loss to be more or less the same.
If I can generate the TPS25830A-Q1's touchstone file for you, I will also generate the TPS25830-Q1's touchstone file for you as well.
BR,
Seong
Hello Manikandan,
Yes, based on the datasheet, 4.3ohm is the typical RDS_ON for the data switches.
BR,
Seong
Hello Seong,
In the datasheet, the data transmission loss of -3dB measured @800MHz. In this measurement regards, can you clarify below queries.
Thanks,
Mani M
Manikandan,
Figure 8-29 is when the device is in Active Mode. Figure 8-28 shows the eye diagram when data is passed through the TPS25830A-Q1 data switch. This signal integrity is good and is a reflection of how a -3dB loss at 800MHz does not affect the Eye Diagram test, but this is with a 10cm cable.
The problem occurs when a 1+ meter cable is connected from the USB port to a front-end head unit. The parasitics of the cable will affect the impedance, thereby the signal integrity.
I am currently in the process of sourcing two 50ohm semi rigid cables. I will let you know once I have these on hand so I can begin using a network analyzer to capture the S-Parameter measurements and generate the S4P file.
My plan is to measure each data line separately and create two S2P files for each data line. Then I believe I will need to use MATLAB or a similar software to convert the two single-ended measurements to a single S4P. Considering what must all be done to deliver this simulation file, I'm not certain how quick the turnaround will be and may take another week or so.
BR,
Seong
Hello Manikandan,
I have reserved time to use a network analyzer next Monday. I will provide the S4P COB on the 24th.
BR,
Seong
Hello Manikandan,
I've tried measuring the S-parameters with the TPS2583XAQ1EVM-147 using a network analyzer recently, but the results were not accurate. I believe it is because of the cables I was using. The conductor side that is soldered on the USB traces are exposed without much of the GND shield (see images below).
I will try again this week by cutting the conductor shorter, but I cannot guarantee that I will be able to deliver what I've committed to. For accurate measurements, we may need to design and build a board with on-board SMA connectors connected to the data pins for this very purpose. I will keep you updated.
BR,
Seong
Hello Manikandan,
I plan to take a 2nd try at the measurement tomorrow.
I'll let you know how the measurement turns out.
BR,
Seong
Hi Manikandan,
After making my modifications, I took a second try at measuring the Sdd21 and generating the simulation file you requested. However, I was unable to improve the accuracy of the measurements today (the results were the same as my previous measurement), so I won't be able to provide the s4p file. My apologies.
BR,
Seong