For bidirectional transfer, definitely tristate. If you have only outputs to combine, then multiplexors (mux) may make sense. Another option is an open collector/open drain bus with resistor pull ups.

@zack1544

@W5VO Thanks for the reply. I only need to combine the outputs so I am thinking of using a combination of multiplexers or use tri state buffers.

@JonRB Looks more like just a polyfuse and TVS in a single box together, for premium money, from the overview/table info

@PlasmaHH Need anything from MiniCircuits?

Woohoo. Just spent buttloads of money!

@Asmyldof . . . on transistor wax?

:-D

On RF stuffs

hohum. Ansys installed

@Asmyldof Now you can be a Mückenficker too.

I'll stick with being a muggenzifter

And also, them is paying good money for a couple simulations and PCB improvements, for which I will be getting a training day from an ex-colleague

So, all is good, you can curse at me all you like

I'll take a few insults for a new car

@Asmyldof Considering you probably just spent a new car on your license...

Or at least a new-to-you car

@ThePhoton Yay for renting from another company

2 weekend days for training, 2 more weekend days for practise, 3 evenings for actual use

Just about the cost of a somewhat acceptable bicycle

:-)

Knowing people is for yays

@Asmyldof How much computer have you got?

Two late model i7's one with 980 graphics

And some more shit that's compatible to a $200 modern PC

About 3-5 minutes to simulate on a ~4-year-old i7, meshed at 5 GHz.

3D or 2.5D?

@Asmyldof 3D, we don't have a 2.5D license

Sucks for you

2.5D is a fraction of a fraction off and about 20 times faster

@Asmyldof Too bad you get paid by the hour

Me and another ex-colleague did a full run of 3D on a board, a couple of runs of 2.5D (+/-0.5dB on most params and points from the 3D) and reality was about 1:3 ratio between 3D and 2.5D

@ThePhoton Fixed rate

But my insane honesty did calculate for 2.5D times

@Asmyldof That would certainly make optimization runs less painful.

Yes

What we ended up doing was put 6 or 7 2.5D runs in a work day and then when we were somewhat confident we got it, run a full 3D at the end of the day and go home

Only once did it take another 3D run

We were running sim up to 30G, by the way and verifying IRL upto 20G

My ex colleague put the plots of everything on top of each other in a good structure and an intentionally bad one (smuggled that into the design for verification) and we were shocked how accurate the simulations were

@Asmyldof I was simulating up to 10 GHz, but Ansys recommends meshing at the middle of the band --- should probably go back and do one run meshing at 10 GHz to see how much difference it makes.

Unfortunately graphics card issues mean I have to reboot to switch between HFSS and Altium.

We had non-anti-pad oscillations neatly tracking each other in 2.5D and IRL

Guess I'm happy with two laptops of good enough quality

Even if that would be an issue, which it doesn't seem to be. But not using the license yet, so maybe at sim there will be problems

@Asmyldof My graphics card issues come as soon as I try to open an HFSS design.

You might avoid it with the 2.5D design capture tool, I dunno.

Or just by not having the same particular combination of graphics cards as I have.

Probably the latter