Ethernet for the QL

Nagging hardware related question? Post here!
User avatar
Dave
SandySuperQDave
Posts: 2621
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Ethernet for the QL

Postby Dave » Tue May 25, 2021 4:59 pm

Surprisingly true.

Summer here is mostly days in the mid 40s, with humidity up over 80% during the day and 100% at night. At 3am it sometimes gets down to 32C/ It's not conducive to projects, so I just go on road trips to Colorado or Arizona. The whole COVID situation means I haven't had that element in my life for 16 months, so the second I was vaccinated and every state was open that I wanted to pass through, I jumped on the change for a bit of shoulder season tourism before EVERYONE ELSE does the same thing.

Peter: The 2200 can only handle the magnetics being up to about 40mm away and then it all goes to hell. I have the RJ45 with the magnetics built in. Are you aware of any separate magnetics that have the same characteristics? If I can identify some, I can use those and a standard RJ45, allowing it to be much further from the board. The dimensions of the ones I have are really restrictive as they're larger than ordinary RJ45s.


User avatar
Peter
QL Wafer Drive
Posts: 1295
Joined: Sat Jan 22, 2011 8:47 am

Re: Ethernet for the QL

Postby Peter » Thu May 27, 2021 11:22 am

Suitable separate magnetics certainly exist. But I never selected any, since I prefer them integrated.
Not sure what you mean by "up to about 40mm away and then it all goes to hell".
The signals are differential, so if you have a groundplane and keep them in pairs, I don't see why the CP2200 should be so restrictive.


User avatar
Dave
SandySuperQDave
Posts: 2621
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Ethernet for the QL

Postby Dave » Thu Jun 03, 2021 4:41 pm

If the magnetics are close to the IC it works fine. I was trying to put them on a short cable so the socket could be located somewhere with more room (can't close the case) but the signal is destroyed. The datasheet also says the magnetics need to be within 3-4cm of the IC.

I've ordered a revised PCB that solves this conflict. Also fixes a missing JTAG line.


User avatar
Dave
SandySuperQDave
Posts: 2621
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Ethernet for the QL

Postby Dave » Tue Jun 08, 2021 4:56 pm

Expanding on my previous comment:

I placed the jack facing left towards the expansion port. Even very gentle wiggling of the cable makes the board jump out of the socket. The CPU sockets are UTTER CRAP. I will try with it facing backwards (exiting the ROM port) to see if the reduction of leverage helps. If it doesn't, it will have to go elsewhere.

I used turned pins, as I do not want to de-spring the leafs. I could use square pins but they would ruin the socket for anything else that follows. The SPEM internal 512K used to do that too. So obviously that's not an option.

I'll drop the schematic onto a small expansion bus card just to get the testing out of the way. A couple more weeks?

I hope that explains a bit better what's going on.


User avatar
Peter
QL Wafer Drive
Posts: 1295
Joined: Sat Jan 22, 2011 8:47 am

Re: Ethernet for the QL

Postby Peter » Tue Jun 15, 2021 7:02 am

Dave wrote:I was trying to put them on a short cable so the socket could be located somewhere with more room (can't close the case) but the signal is destroyed.

I was not aware you were using a cable. That is of course problematic.

Dave wrote:The datasheet also says the magnetics need to be within 3-4cm of the IC.

From what I read, it's not that strict. They say it "should" be for "optimal performance". With decent differential tracks and ground plane I'm sure at least twice of that is no problem. But since you didn't mean to stay on PCB, that does not help of course.

Dave wrote:I placed the jack facing left towards the expansion port. Even very gentle wiggling of the cable makes the board jump out of the socket. The CPU sockets are UTTER CRAP.

Sorry to read you have so much trouble.

I also have no good ideas for a compact mechanical solution with (S)GC systems.



Who is online

Users browsing this forum: No registered users and 6 guests