DS15EA101: Simulation models for ADS

Hi Nick,

Thank you for the confirmation. Since DWDM and OUT_DA_SEL are set to 0 in all schematics, I will use them without changes.

I was checking the operation of IBIS-AMI in the cable schematic, but even when EQ_Bypass is set to 0, the CDR remained enabled. From the verification results below, it seems that if we want to bypass the CDR, we should use the ReDriver part from ADS. Is this the correct usage of the IBIS-AMI you provided?

Verification 1: In the cable schematic, I set DJ=0.15UI for Tx and switched between with and without CDR (EQ_Bypass=1/0) for simulation. Regardless of the CDR settings, DJ disappeared from the IBIS-AMI output.

Verification 2: When I conducted the same verification in the ReDriver schematic, DJ remained in the IBIS-AMI output regardless of the CDR bypass setting. In these schematics, the ADS parts for IBIS-AMI are different, with ReTimer part and ReDriver part being used respectively. It seems that to bypass the CDR, we should use the ReDriver part.

Best Regards,

Naoki

Hi Naoki,

I have not done this kind of validation myself and the part is not validated for the lower data rates. If you are seeing better results with the redriver part, that makes sense given your application. I have found that some of the configurable settings for the IBIS-AMI model are not used as much, so your work supports that understanding.

Can you explain your application to me one more time? With a block diagram to illustrate what you are planning to do.

Best Regards,

Nick

Hi Nick,

Nick said:
Yes, use the following sequence on the LMH0397:

Could you provide the details of the bits to change the equalizer characteristics using register 0x24?
Can I use the above sequence in actual applications?

The application I am referring to is shown below. I would like to transmit a 500Mbps signal through a high-loss transmission line.

Best Regards,

Naoki

Hi Nick,

I am currently verifying whether the desired functionality can be achieved with the 1297EVM.

I was able to manually set 0x25 on the ConfigIO page by the sequence you provided eariler(Sequence). For 0x24, I directly applied the results of EQ adaptation. However, after this sequence, the OUT0 output has ceased. The value for 0x02 indicates CB, suggesting that it is not in sleep mode. But the status is unclear because some register information has not been written in the programming guide.  Could you please advise on how to enable the OUT0 output after manually setting the EQ?

Additionally, I would like to disable the carrier detect to keep the output always active, even with a small signal. I configured the registers according to the 3.3 Force Carrier Detect Power Down as outlined in the programming guide, but this resulted in the loss of the OUT0 output. The value for 0x02 indicates 07, which means it is in sleep mode. What steps should I take to disable only the carrier detect function?

Thank you for your assistance.

Best regards,

Naoki

Hi Naoki-san,

I will get back to you shortly on this.

Best,

Nick

Hi Naoki-san,

Good to hear back from you. 

Since you are not going to be using the retimer, it makes the most sense for you to output raw data (EQ only) on OUT0. See 3.13.2.4 in LMH1297 programming guide.

It will not be necessary to use the signal detect because the output will not auto-disable. Try the fix above to disable the reclocker and let me know if that helps with OUT0 when it is powered up with pin-straps.

Best Regards,

Nick

Hi Nick,

I am attaching the verification results. OUT0 remains absent. I would appreciate your advice on the register sequence to achieve EQ manual setting, OUT0 output, and turning off Carrier Detect.
Just to inform you, the 1297EVM is configured to output raw data (EQ only) on OUT0 based on the pin configuration.

verification results.pdf
Page1
**Step 1:** After powering on the LMH1297EVM, I checked the CEI value. The OSC showed eye opening, and OUT0 was present.
**Step 2:** I executed the EQ manual setting sequence. The OSC indicated 0V, and OUT0 was absent.
**Step 3:** I executed the "OUT0 Raw Data" sequence. There was no change in the OSC, and OUT0 remained absent.

Page2
**Step 1:** I powered off the LMH1297EVM and then powered it on again.
**Step 2:** I confirmed the CEI value and that OUT0 was present.
**Step 3:** I executed the "Force Carrier Detect Power Down" sequence. The OSC indicated 0V, and OUT0 was absent. The LMH1297 entered sleep mode.

Page3
I checked the interrupt status before and after the EQ manual setting sequence.
0x54[7] was 1 before the sequence, but it changed to 0 afterward.
It seems that some settings related to Carrier Detect are being changed by the EQ manual setting sequence, which is causing OUT0 to be disabled.

I am changing the register values using a custom tool. Since this tool does not allow specifying a mask, I am calculating the 8-bit setting values. I am not very confident in my calculations, so if there is mistake please point out any mistakes in the register settings.

Best Regards,
Naoki

Hi Naoki,

Thank you for the information. I will take some time early next week to review this material and get back to you.

Bes Regards,

Nick

Hi, Nick

I found that the desired functionality with the LMH0324. I will send my verification results via private message. Could you please provide feedback on the validity of my verification results regarding the LMH0324 shown on pages 2 to 5 and any issues with my usage of the LMH0324?

I provided the LMH0324’s method as feedback for the LMH1297, but the OUT0 output did not recover. The LMH1297 has a more complex configuration, so it seems that some part is still not functioning as desired. I suspect the CDR and input/output MUX, but I cannot find any registers to check their statuses, which is hindering my investigation into the cause.

Best Regards,

Naoki

Hi Naoki-san,

I am glad that you have seen the desired functionality with the LMH0324. Thank you for verifying on your end.

I verified that I see the same functionality on the LMH1297. The CDR shuts off when we disable EQ adaptation on the LMH1297. Since you do not see the same issue on the LMH0324 and have reached the desired functionality using the register writes above, I do not see any issues.

What opportunity size is this project if you are able to use the LMH0324?

Best regards,

Nick

Hi Nick,

Thank you for confirming that there are no issues with the method I varidated for the usage of the LMH0324.

I am in a position to propose circuit configuration plans, and the decision on whether to adopt them is made by the customer. Therefore, it is difficult for me to estimate the opportunity size for the LMH0324. Moving forward, I plan to further evaluate the applicability of the devices that have been considered as candidates so far to the system.

Regarding LMH1297, is there a way to restart the CDR by changing the register settings and obtain the OUT0 output?

Best regards,

Naoki

Hi Naoki,

Thank you for the information. 

I experimented with restarting the CDR yesterday, however, there was still no output. I believe that by manually setting the rate this can be fixed.

Please keep me up to date on the results of your investigation.

Best regards,

Nick