Engine Simulation Custom Device Feedback

[StephenB]

Hi Jamesy,

Changing the location of the missing teeth unfortunatley does not simply shift the entire pattern by that many degrees. As discussed before it just moves the mask placed on the pulse train... in the post 150 you referenced.

So like you said the only realy solution here is to shift the "TDC configuration". Unfortunatley we do not support shifting by an offset in degrees but rather relative to the tooth. Beware, the mask stays in the same place even when you shift the teeth by changing the "TDC configuration". I recommend playing with the missing tooth location and the TDC configuration to see if you can remove the error. Make sure you don't see any half teeth though.

The other option is to not use the N teeth M missing generation but instead the fully custom generation where you can define all 60 teeth's angles explicitly.

And the last item to discuss... are you sure you haven't found a bug in the unit under test? Perhaps it is the one off by one degree?

Take care,

[jamesy]

Hey Stevey,

Much as it would please me to find a bug in the UUT unfortunately with a scope i can see that moving the TDC 1 degree has no effect on the signals produced. So the bug is the "old functionality" provided by the custom device.

When will you release the fully custom config ?

Sounds like exactly what we need.

James.

[StephenB]

Fully custom configuration already exists. See the maing page for the custom device:

https://decibel.ni.com/content/docs/DOC-19122

Its usually used for Cams but ive seen many people use it for cranks as well.

Just make sure your bitfile has the fully custom teeth component in there, then when you load the bitfile with the custom device it will show up. When you drop the component into the FPGA VI, change the angle reference from the default cycle angle to crank angle. Then in NIVS uncheck cycle angle and change it to crank angle.

[StephenB]

Also, if your generation does not change when you change the TDC Configuration.... you might be using a very old revision of the N Teeth M Missing FPGA code that did not support TDC configuration. If you look at your bitfile with a text editor or look at it with LabVIEW FPGA, do you see a U64 control on your N Teeth M Missing code named NTeeth.<Component Name>.TDC Offset?

[jamesy]

Yes its there : NTeeth.<Component Name>.TDC Offset

The TDC does move just not at 1 degree increments.

I just removed 1 degree during post processing of the HILAPI captures so that will work ok for now and get things up and running.

James.

[Sarci1]

I would make a change to the custom device page which states that the AES Library for HIL must be installed after the NI VeriStand Based FPGA I/O Interface Tools.

[StephenB]

Done! Thank you    

[StephenB]

This Add-On is deprecated and will no longer be updated with new features or bug fixes. Therefore, it is not recommended for use in new projects. A next-generation engine simulation toolkit will be posted this month.

[jamesy]

Stephen , Looking forward to that, any teasers as to the content? Plant model perhaps?

[StephenB]

Hey Jamesy,

Glad to hear that

No plant model unfortunatley. I'll post up full details soon, but here is a space/performance teaser:

7854R PXIe 8130 PXIe 1082	AES template implemented with AES IP	AES template implemented with new EST IP
Slices	6,703 of 17,280 (38.8%)	3,159 of 17,280 (18.3%)
Registers	14,882 of 69,120 (21.5%)	5,912 of 69,120 (8.6%)
LUTs	19,702 of 69,120 (28.5%)	8,239 of 69,120 (11.9%)
DSP	24 of 64 (37.5%)	2 of 64 (3.1%)
BRAM	21 of 128 (37.5%)	13 of 128 (10.2%)
40 Mhz Max	42.69	65.35
Compile time	37 minutes	21.7 minutes
RT Loop Duration	305 to 335 uS	138 to 186 uS w/ RIO 13.1 132 to 165 uS w/ RIO 14.0