Skip to main content

General Discussion

DH parameter

Dear Chris Annin,

I'm writing to consult you about an issue with the DH parameters.

The DH parameter table in the left-side software doesn't match the one in the spreadsheet. For instance, the corresponding value of a₁ is 64.2, which is correctly filled in the "a" column corresponding to the J1 axis in the spreadsheet. However, in the host computer software, this value is filled in the next row.

Could you please help explain why this discrepancy occurs?

Thank you for your time and assistance.

Best regards,

[Zhongyuan.Duan]

DH parameter
  • 1 like
  • 3 comments

Chris Annin

Hello,

The difference you’re noticing is due to the convention used in each version of the model. The original spreadsheet was created using the standard Denavit-Hartenberg (DH) convention - the simplest geometric form, where the parameters are defined based on the relationship from one link’s Z-axis to the next joint’s Z-axis.

In contrast, the newer software uses the modified DH (MDH) convention, which defines parameters slightly differently — specifically, it shifts how link offsets and rotations are applied (from the previous joint’s frame instead of the current one). This change allows for cleaner matrix formulations and more advanced inverse kinematics calculations, which improve accuracy and numerical stability in the robot’s motion control.

That’s why a value like a₁ = 64.2 mm appears in a different row between the two tables — it’s not an error, just a difference in convention between the standard and modified DH systems.

Chris

Duan zhongyuan

Chris Annin 

Dear Chris Annin,

Thank you so much for your clear and detailed explanation! Your insights have been incredibly helpful in enhancing my understanding of the differences between the standard DH and modified DH conventions.

I truly appreciate you taking the time to address my question and clarify why parameters like 

 appear in different rows across the two systems. Your explanation about the shift in how link offsets and rotations are applied makes perfect sense.

Thanks again for your support and expertise!

Best regards,
[zhongyuan.Duan]

Chris Annin

Thank you very much - happy to help!