Final tweaks to the 1130 core memory replacement (MRAM) board

FINAL VALUE CAPACITORS INSTALLED IN THE TIMING CHAINS

The updated capacitor values for the timing chains arrived, but very late. They were shipped Fedex ground with a scheduled delivery of Aug 14. Due to a string of missteps and poor execution, such as sitting 36 hours in some depots before moving, it finally was out for delivery on Monday August 18th.

I did receive the package by 1PM Monday, in time to still get over to my shop for a short time. I put these new timing capacitors on the board. I then installed all 17 of the bypass capacitors at each integrated circuit location. 

BEGAN ANOTHER ROUND OF TESTING. 

As I was verifying that the replaced chips output a negative going sense pulse when a bit being read was 1, I realized I had installed 74LS03 chips which have a minimum VCC requirement of 4.75V but my board operates on 3.3V. They were working properly however I ordered 74HC03 chips which work fine with 3.3V VCC and are the same footprint. I will replace the five output chips with the new 74HC03.

I was able to verify that the MRAM chip does store data and return it properly, at least for a default storage address. I used a bench tool to generate a positive pulse for storage read and for storage write, setting up the input data and observing the sense output of the board. 

I am going to set up a bit more elaborate test where I set various storage addresses, store different data patterns in them, and then read the addresses to check that these are discrete locations giving the associated data patterns. It still uses a debounced pushbutton to emit the read and write signals. All the timing on the oscilloscope shows it should operate correctly at 1130 speed. 

CABLED UP FINAL POWER CONNECTIONS

I put together the wiring between the power regulator module and my board, along with wires that will attach to the 1130 power supply terminal strip at the bottom of the gate where the memory compartment sits. They have ring terminals on them so that I can quickly hook up the memory when I am ready to test on the IBM 1130. 

SHIELDING BOX TO BE BUILT AND INSTALLED

I do plan to build a metal box to fit over the MRAM chip since it is sensitive to magnetic fields. I set up four ground pads on the PCB where I expect the box will be soldered down. The ground plane underneath the MRAM chip is unbroken as well. 

I found a good shielding box on Amazon that would fit reasonably well - a bit larger footprint but will clear other parts on the board. It is made of Mu-Metal, a material with high magnetic permeability that would strongly attenuate any changing magnetic fields outside the box. However, once it was ordered, the vendor confessed that it is not in stock and cancelled the order. The product listing remains on Amazon with the same unrealistic delivery estimates to lure the unwary. 

I may have to form my own box, which may challenge my skills with metal work given the limited tools I own for the purpose. However, if I can form the part up so that I can solder overlapped seams and it is not too misshapen or ugly, I could use that. I bought some heavy copper sheets, since MuMetal prices were near $100 in the minimum quantities they offered. 

FINAL STEP WILL BE CONFORMAL COATING OVER THE BOARD

I will coat the board with silicone to insulate and protect it while it lies inside the gate of the IBM 1130. However, I will wait until I have completed testing and won't need to change the board or any parts on it.