I add power to my latest TTL design, and it don't work. What 5 volt power supplies work the best.
I hope to beef up the power and ground when I get some thick wire in soon.Ben.

A bench PSU with current-limiting is a good one for these purposes, to avoid a melt-down.

Some while ago we chose this dual channel one, but it hasn't been used enough to be a strong recommendation:
https://cpc.farnell.com/tenma/72-10495/ ... dp/IN06823
Tenma Dual Output Linear DC Bench Power Supply - 30V, 5A

That's £150. Much much cheaper would be an ordinary PSU for a PC, maybe £30 to £40. Mean Well is a brand I hear recommended. You might need a dummy load or a sense wire connected up to convince it that you want some power.

I built my own... some 35 years ago... And a couple of years back I upgraded it from 7805s to modern buck/boost converters off ebay with LCD display, current limiting, display of voltage, amps, power and input voltage. I think they were under £5 each. 3A max, output volts from 3 to 30.

I do need to upgrade the old transformer though - it's only rated 1 amp at 9v but for now, it's OK.



Zero points for prettiness ... 100% functional though.

-Gordon

All I could find was this one, as I hope is linear power supply.
I was kind of hoping for a simple design to build off the web, like this one, from 1980.

https://electronics.stackexchange.com/q ... 110#335110

So get a nice 9v, 30va toroidal transformer, 3 amp bridge rectifier, a big 16v, 1000µF capacitor - that circuit above, put it in a box and off you go...

-Gordon

I use an old ATX power supply for TTL.

There are pinout/wiring diagrams on the web. It it fairly simple though, you just need to add a switch that connects the on signal wire to ground.

The problem is I have a almost working computer, with some sort of dynamic bug. I have the mico code patched so it just incriments
the ACC. It does nothing, just leds displaying something. Front panel switches all off (0's). I turn switch #33 on (1) and it starts counting.
Set the switch to 0, it stops counting.
The front panel is not used at all.
I want to try a linear power, to check if is not switching power supply glitch.

Well I found 1 bug with a 22v10, fixed that, restored other logic to normal from single step clock.
I was now getting garbage with the upper 16 bits. Tried replacing the upper 16 bit slice, powrerd it
up. I think I killed my ALU bit slices.
Going back to a very primitive, 20 bit cpu design,with software emulation first, and hardware when my budget
can afford it, and more testable PCB's.

I am going to small PCB modules from now on,as they may be more easy to test.
The first pcb is a 4 bit slice, using cheap 2901's. Ths has the MAR. IO reg, input REG
and input multiplexer. Card is 4" x 2". 4 cards make 16 bit ALU.,5 cards 20 bit ALU,
8 and 9 cards for a 32 or 36 bit alu.
PS: cost for 10 PCB's was $50 US with shipping.

The PCB's are in. Made a off by 1 wiring error on a little used output pin. Used my last spare pin to fix that.
Decided to make the data bus active low,so for that I need new parts. 74ALS640 vs 74ALS643. OCTAL BUFFERS.
Started working on the tester for the bit-slive, just some leds and switches. Hope to test a alu module tommorow, too
sleepy today.ZZzz

new alu for a 24 bit cpu

Is that an LED I see before me?

YUP! Good old power led.
Starting a clean 24 design. Microcode is 256 x 9 x 2 that fits in 22v10's
Short address, immedate and register memory formats.
PSH and POP. byte (12) and int (24) data types.

Revised the pcb for better layout, with a slight change in clocking.

and now a tester to test it. 3 til311's display input,output and status.