DipTrace 4.1 beta
Re: DipTrace 4.1 beta
An interesting b4.1 (v4.0.9.1) Layers Panel "resize cursor" bug
When moving the mouse cursor downward (do not click) from the Layers panel towards the Design Manager panel in the area highlighted here... ...the cursor changes from a "selection arrow" to a "resize symbol" as expected.... However, further downward movement of the now-changed cursor (do not click) onto one of the three selection icons in the Design Manager panel results in the "resize symbol" not changing back to the "selection arrow".
When moving the mouse cursor downward (do not click) from the Layers panel towards the Design Manager panel in the area highlighted here... ...the cursor changes from a "selection arrow" to a "resize symbol" as expected.... However, further downward movement of the now-changed cursor (do not click) onto one of the three selection icons in the Design Manager panel results in the "resize symbol" not changing back to the "selection arrow".
Tom
Re: DipTrace 4.1 beta
I am not able to use the IPC-7351 pattern generator to generate a pad layout with the pad numbers in the correct order for the MEMS sensors made by Bosch/ST/TDK (Invensense).
https://www.st.com/en/mems-and-sensors/lsm6dso32.html
https://www.st.com/en/mems-and-sensors/lsm6dsrx.html
https://www.bosch-sensortec.com/product ... #documents
https://invensense.tdk.com/products/mot ... icm-42605/
https://invensense.tdk.com/products/mot ... m-42688-p/
They all use the same package with the same pinout (minor connection differences) and same pin numbering scheme.
They use what they call an LGA-14L package, but the LGA generator in diptrace creates patterns with pins on a grid, not with pins around the edge. The PQFN generator in diptrace can generate landing pattern with the correct dimensions of pads and 3D model, but the pin order is incorrect. I tried using the "Pin 1 Orientation" option and reversing the D/E dimensions and pin counts but that didn't help.
Their landing pattern:
Generated pattern and settings:
Recovery code:
```
[Y;PQFN;1;Y;Y;1;0;0;Y;Y;Y;Y;101;Default;|3;4;0.5;0;;;0.87;;;0.2;0.25;0.3;0.05;-0.05;2.95;
;3.05;0.05;-0.05;2.45;2.5;2.55;0.05;-0.05;0.425;0.475;0.525;0.05;-0.05;0.074;0.1;;;;1.35;
.475;0.25;1.85|0;|;;;;;;;;;0.25;0.25;0.25|||||3;2;;|||||||||4934475;15461355;16119285;Y]
```
Actual pattern that I've used on previous PCBs using DipTrace 3.x
Note the pad number ordering.
https://www.st.com/en/mems-and-sensors/lsm6dso32.html
https://www.st.com/en/mems-and-sensors/lsm6dsrx.html
https://www.bosch-sensortec.com/product ... #documents
https://invensense.tdk.com/products/mot ... icm-42605/
https://invensense.tdk.com/products/mot ... m-42688-p/
They all use the same package with the same pinout (minor connection differences) and same pin numbering scheme.
They use what they call an LGA-14L package, but the LGA generator in diptrace creates patterns with pins on a grid, not with pins around the edge. The PQFN generator in diptrace can generate landing pattern with the correct dimensions of pads and 3D model, but the pin order is incorrect. I tried using the "Pin 1 Orientation" option and reversing the D/E dimensions and pin counts but that didn't help.
Their landing pattern:
Generated pattern and settings:
Recovery code:
```
[Y;PQFN;1;Y;Y;1;0;0;Y;Y;Y;Y;101;Default;|3;4;0.5;0;;;0.87;;;0.2;0.25;0.3;0.05;-0.05;2.95;
;3.05;0.05;-0.05;2.45;2.5;2.55;0.05;-0.05;0.425;0.475;0.525;0.05;-0.05;0.074;0.1;;;;1.35;
.475;0.25;1.85|0;|;;;;;;;;;0.25;0.25;0.25|||||3;2;;|||||||||4934475;15461355;16119285;Y]
```
Actual pattern that I've used on previous PCBs using DipTrace 3.x
Note the pad number ordering.
Re: DipTrace 4.1 beta
Swap "D" and "E" for Pin Count and table values, then rotate everything using the hotkey combination Ctrl + Alt + r until pin 1 winds up on the upper-left side...
Tom
Re: DipTrace 4.1 beta
https://www.snapeda.com/parts/ICM-42688 ... ef=digikey If you open this, click on the 3D model tab, zoom in until you pass through the model, the dimensions feature is nice.
I fought that pattern awhile ago before the IPC thing existed. Now either digikey or mouser generally has CAD files.
The IPC generator got me with the manufacture calling a part a LGA (Land Grid Array as in ARRAY) when it was a RECESSED QUAD FLAT No-lead or RQFN or RPGFN I spent some time learning about SON and PSON instead of DFN since the pattern library only has 4 lead DFN devices!
There is a lot of information here https://diptrace.com/books/Pattern_names_help.pdf
Re: DipTrace 4.1 beta
Hi Tom,
You mean close the IPC pattern generator, then rotate manually? What I was looking for was a solution that was purely IPC pattern generator UI based so that all changes to the pattern are made using the IPC UI. It feels like there's simply some missing 'pin ordering' options missing from the UI to me. The danger is that you make the part and 3d model using the UI, then you rotate it outside the UI, then you go away for months, come back, use the pattern generator to adjust some parameters like lead length, apply it and forget that you also needed to do the rotation... Trying to avoid accidental errors as much as possible.
Also if you rotate the pattern outside the IPC UI when won't the model be mismatched from the land pattern (e.g. pin one location)
-
- Posts: 63
- Joined: 22 Jun 2013, 22:20
Re: DipTrace 4.1 beta
So, still no push & shove? *Sigh*
Re: DipTrace 4.1 beta
One thing that I hoped would make it into DipTrace 4.x is a DRC rule for minimum solder mask width.
My PCB suppliers are able to specify the smallest width that solder mask should be, e.g. so you get solder mask between the legs of a 0.5mm pitch LQFP without it flaking off or failing to print.
For suppliers like PCBWay the value is about 0.15mm, currently I have no way to check this in DipTrace and rely on Gerber review or manually measuring distances in other tools.
For small parts, the solder mask is CRITICAL to ensure high yield and low re-work volume.
My PCB suppliers are able to specify the smallest width that solder mask should be, e.g. so you get solder mask between the legs of a 0.5mm pitch LQFP without it flaking off or failing to print.
For suppliers like PCBWay the value is about 0.15mm, currently I have no way to check this in DipTrace and rely on Gerber review or manually measuring distances in other tools.
For small parts, the solder mask is CRITICAL to ensure high yield and low re-work volume.
Re: DipTrace 4.1 beta
Agreed.
Re: DipTrace 4.1 beta
Yes.dominicc wrote: "...You mean close the IPC pattern generator, then rotate manually?..."
It does look that way.dominicc wrote: "...It feels like there's simply some missing 'pin ordering' options missing from the UI to me..."
From what I have seen so far it looks like the rotation of the pattern is retained even after changing some of the parameters inside the IPC pattern generator; this is coming back to the Pattern Editor the next day. DipTrace automatically rotates the 3D model when you rotate the pattern. I checked the pick-and-place rotation angle in the PCB and, without rotating it on the PCB, it is 0 degrees as expected. Of course this is only one sample in a brief trial so I guess there are no guarantees. Anyway, just thought I would throw that out there.dominicc wrote: "...The danger is that you make the part and 3d model using the UI, then you rotate it outside the UI, then you go away for months, come back, use the pattern generator to adjust some parameters like lead length, apply it and forget that you also needed to do the rotation... Trying to avoid accidental errors as much as possible.
Also if you rotate the pattern outside the IPC UI when won't the model be mismatched from the land pattern (e.g. pin one location)"
Tom
Re: DipTrace 4.1 beta
b4.1 (v4.0.9.1) Route Net bug
1) Connect a ratline from C1(2) to C2(1).
2) Complete a trace from C1(2) only half way to C2(1) by terminating it with the [Enter] key.
3) Right-click on C1(2), select "Route Net" in the context menu and observe what appears to become one completed trace connecting C1(2) to C2(1).
4) Drag the left half of the trace upward and the right half of the trace downward. Observe one connected path and one disconnected path. This problem does not exist in v4.0.0.5.
1) Connect a ratline from C1(2) to C2(1).
2) Complete a trace from C1(2) only half way to C2(1) by terminating it with the [Enter] key.
3) Right-click on C1(2), select "Route Net" in the context menu and observe what appears to become one completed trace connecting C1(2) to C2(1).
4) Drag the left half of the trace upward and the right half of the trace downward. Observe one connected path and one disconnected path. This problem does not exist in v4.0.0.5.
Tom