Hi all!

The board is this one, 12MHz variant:
https://theretroweb.com/motherboards/s/leadma … inc-lm-103fs-16
(probably should end -12)

Cpu is running, TURBO ON, 25% slower than rated.
9.45 MHz instead of 12.5 MHz
NPU follows suit.

TURBO OFF, 6.45MHz

Landmark tests indicated Cpu frequency as 6.249 MHz (remeber this one, experts already know what it is by this value alone)

I have 3 frequency sources:

1.) 25 MHz oscillator (cpu) ----> 25 MHz
2.) 14.31... MHz crystal -------> 14.3.. MHz
3.) 32 KHz crystal -----------> 32 KHz

They all check. 100% spot on.

I probed all frequency sources and it was all checking.
Except for this one: pin#123 Hispeed (it determines CLK*2 - i.e. 25MHz speed - or CLK*2 /4 - i.e. 25/4=6.25 or 6.249)
Traced chipset frequency inputs to sources and found that a component wasn't issuing the correct signal, oin 123 should have been high and it was 0.2V (LOW).

Some more tracing and I got to a HD74LS74A (Dual D-type Positive Edge-triggered Flip-Flops), with a LOW output of 0.2V.
More tracing and went to another component, 74F00 (Quad 2-Input NAND Gate), and one of the 2 required inputs that output the 0.2V is connected to a Jumper that can, in real time, set "Hispeed" to CLK*2 or CLK*2 /4.
The jumper must be set shorting to ground to allow a "high" singal to output.

Now all checks perfectly and CPU works at 12.5 MHz as intended.
It was a missing jumper adventure, TRW doesn't have a jumper table, I'll add that later.

- Fin -