So the new precidip arrived and I did some comparisons...
- hc2.jpg (136.37 KiB) Viewed 5816 times
This is the HC CPU with the new sockets.. It looks like the spaces in between the IC pins go in a lot further.. Or they could simply just be longer than my other sockets..
- hcold.jpg (119.81 KiB) Viewed 5816 times
This is the HC CPU with my old sockets.. The plastic spaces do not really seem to go in between the IC pins.. The contact connections do look good though..
- ttl.jpg (129.11 KiB) Viewed 5816 times
This is the older TTL CPU with the new sockets. The spaces seem to fit nicely between the IC pins..
- hcold2.jpg (84.32 KiB) Viewed 5816 times
This is the HC CPU in the old sockets, it looks like the contacts are not connecting..
- hcnew.jpg (74.51 KiB) Viewed 5816 times
The HC CPU with the new sockets, it also looks like the contacts not meeting. But this could just be the angle of the pins it is difficult to say..
- ttlnew.jpg (82.79 KiB) Viewed 5816 times
This is the TTL CPU with the new sockets...
With the new sockets and the TTL CPU, looking all the way around the contacts I think they look okay overall..
With the old sockets and the TTL CPU, looking all the way around, does generally seem to look okay overall as well. Generally I have not had any problems with older TTL CPUs and sockets..
With the HC CPU and the new sockets, the gap between contacts looks larger, but it only looks like one side of the chip is making bad contact. But again I cannot say 100% as it could just be the angles of the pins making it look like there is a gap..
With the HC CPU and the whole sockets, pretty much looks like the new sockets, where there seems to be larger on some Pins but Not on Others.
I took the dimensions of both chips, was not really seem to be much, possible the HC pins overall have been squashed in some point.
HC dimentions
25.22mm x 25.03mm pins
24.32mm x 24.30mm case
TTL dimentions
25.10mm 25.14mm pins
24.33mm x 24.29mm case
Just for the record, during all testing sockets and chips, have gone round each pin on the socket on bent them outwards a lot, I have also been around the CPU is and bent the pins out as well and it did not make any difference. I also inserted small header pins which are used on my man upgrades to push the contacts out further in the sockets and this did not change anything either .So really it should be no reason for any bad contacts.
Overall, I am still no further forward with this I think. The HC CPU does not seem to fit correctly on new or old sockets. I have tried a couple of HC CPU is as well and they do look dodgy.
Going by the images and looking under the microscope, it just looks saying that the HC CPU for some reason does not make good contact in the sockets.
From what I can tell the precidip sockets do look like they fit more snug around the CPU. Though I cannot see any real differences in contacts between the precidip sockets and my original sockets.
One thing which is still in question, is that re-soldering IC sockets always seem to make things worse. It is possible that as the pins are heated up during soldering, maybe the plastic of the self distorts somehow causing the bad connections. It is even possible that the pins themselves could warp during soldering. Of course the sockets I photographed were both on unused sockets.
Only thing left is to physically try the new sockets in my booster board and see if they actually work or not...
EDIT:
- IMG_2145.JPG (59.14 KiB) Viewed 5811 times
Quickly made up a new test board with a precidip socket.. Similar as before, the TTL CPU works, but the HC one does not.. I know the HC CPU's work I tried a few of them is to be sure. Then put back the TTL CPU and it worked again.
When I had the plug-in STE booster looked into the socket ( see here
https://www.exxosforum.co.uk/forum/viewt ... 3164#p3128) , there was nothing on the board to interfere with anything (I even removed the GAL and wire linked over it to connect the CPU clock again .. ), it was just literally the CPU soldered onto the board going directly to the header pins which were pushed into the PLCC socket. So it proves the HC CPU can work with this booster board at least. They could possibly be like 1pF capacitance between the CPU pins due to the extra length, but I really do not think that will change anything. If that was the case, then the ST would surely try to at least boot up and just crash randomly. But the machine just simply acts like the CPU is not plugged in at all.
As said before, inserting the pins into the CPU socket alone with the HC CPU also did not work. I really have a hard time to believe the contacts were just bad during all these tests which I have done but it is the only conclusion I can come to.
Of course a sockets may not be directly to blame, it could just be some bad manufacturing tolerances on the CPU itself. But either way, the HC CPU and the sockets just do not work together..
I tried to track down the original sockets I use but they seem to not exist any more. I never had any issues with the sockets I purchase a long time ago. It was only on recent purchases of sockets I started to see this trouble.
I still basically am suspicious of the plastic spaces in between the pins, it was the only real difference I could see between my original sockets and the new sockets. I can only really assume the plastic parts are pushing the pins away from the contacts when the CPU is inserted. Then would think bending the pins upwards on the socket and inserting pins into the socket pins further would work.. Seems not.
Maybe there is something I am totally missing in all this, but I do not really see what
I think all future projects just not use the PLCC sockets. It is a little unavoidable for the ROM IC's currently, but so far I do not seem to have run into problems like this with those yet.
So overall I am still at a total loss really