TSH-06F IC Tester — Vol 3: Checking My KENBAK-1 Chips

3.1 Why test every chip

I am building a full-size KENBAK-1 reproduction with the original standard-TTL logic (see the KENBAK-1 unit in Iconic Machines, volumes 11 and 12). Its parts list calls for about 130 TTL chips of fourteen types. Texas Instruments still makes only three of them: the 7400, 7404 and 7416. The other eleven are new old stock: fifty-year-old parts from dealers’ drawers and eBay lots. Old stock brings three risks: a chip damaged in storage or handling, a chip that was a factory reject, and a chip that has been sanded and remarked as something more valuable.

Finding a bad chip in the bag takes ten seconds. Finding it on a populated board with a single lamp walking the wrong way takes an evening. The tester’s blind spots (Vol 2) don’t matter much for this machine. The KENBAK-1 runs a 1 MHz bit clock, slow by TTL standards, so the speed faults a static test misses are unlikely to bite.

3.2 Step 1: calibrate the tester on known-good chips

Before trusting a FAIL on a fifty-year-old part, find out what the tester does with a good one. Three of the types are in current TI production, so new reference chips are easy to buy:

Table 1 — Before trusting a FAIL on a fifty-year-old part, find out what the tester does with a good one. Three of the types are in current TI production, so new reference chips are easy to buy

Reference chip	Tests what	Expected
SN7400N (new)	Does standard TTL test correctly under the 74LS menu?	PASS as 7400/74LS00
SN7404N (new)	Same question, a second type	PASS as 7404
SN7416N (new)	Does the tester handle an open-collector output? Is “16” in the library at all?	Unknown: this is the question

For the 7416, try entering 16 in the 74LS menu first. There is no common 74LS16, so the library may not have it. If it doesn’t, try 06 and 05: the 7406 has the same pinout and function as the 7416 (hex open-collector inverter; the 7406 is rated to 30 V, the 7416 to 15 V), and the 7405 is the same with a lower voltage rating. Also run Search, and note what it says.

The result decides how to handle the open-collector parts (7403 and 7416):

• A new SN7416N passes: the tester pulls the outputs up, and open-collector chips can be tested normally.
• A new SN7416N fails: the tester has no pull-ups, and its result on an open-collector chip means nothing. Test the 7403s and 7416s on a breadboard instead (Step 4).

Keep the reference chips. They are the control every time a result looks odd.

3.3 Step 2: the chip-by-chip plan

Quantities are from my KENBAK-1 parts list (KENBAK-1 vol 12). “Menu” is where to look first.

Table 2 — Step 2: the chip-by-chip plan

Type	Qty	What it is	Menu / number	Watch for
7400	46	Quad 2-input NAND	74LS → 00	The bulk of the job. Test in batches of ten
7410	15	Triple 3-input NAND	74LS → 10	—
7495 / 7495A	13	4-bit shift register	74LS → 95	A sequential test; the 74LS95 shares the 7495’s pinout. Check against the data sheet before trusting a FAIL
7474	10	Dual D flip-flop	74LS → 74	Sequential test
7454	10	4-wide AND-OR-INVERT	74LS → 54	🔴 The 74LS54 has a different gate arrangement. A good 7454 can fail here. See Step 3
7430	10	8-input NAND	74LS → 30	—
7420	9	Dual 4-input NAND	74LS → 20	—
7442 / 7442A	4	BCD-to-decimal decoder	74LS → 42	—
7404	3	Hex inverter	74LS → 04	—
7416	3	Hex inverter, open collector, 15 V	74LS → 16, else 06 / 05	🔴 Open collector: trust only if the reference SN7416N passed
7486	3	Quad 2-input XOR	74LS → 86	—
7403	2	Quad 2-input NAND, open collector	74LS → 03	🔴 Open collector: same rule as the 7416
7408	1	Quad 2-input AND	74LS → 08	—
7451	1	Dual 2-wide 2-input AND-OR-INVERT	74LS → 51	🔴 The 74LS51 has a 3-input section on different pins (TI data sheet SDLS113). A good 7451 can fail here. See Step 3

That’s 130 chips. The two 1404A memories are not on this list: they are 1971 MOS dynamic shift registers that need two-phase clocks swinging about 12 V below ground, and no handheld logic tester drives them. They are tested in the machine, by the bring-up programs in KENBAK-1 vol 12.

For any part that reads “Not Found!”, the chip is fine as far as anyone knows. It’s just not in the library. Use Search as a second try, then the breadboard.

3.4 Step 3: the 7451 and 7454 traps

The tester’s library is built on 74LS and 74HC part numbers. For most types that does no harm, because the standard and LS chips of the same number share a pinout. Two of the KENBAK-1’s types do not:

• 7451 vs 74LS51. Texas Instruments’ data sheet covers both under one title, but its pinout drawings differ. The 7451 is two independent 2-wide, 2-input AND-OR-INVERT gates, with pins 11 and 12 unused. The 74LS51’s first gate has two 3-input AND sections, using pins 10 to 13 differently (TI SDLS113, 1988; Library copy).
• 7454 vs 74LS54. The KENBAK-1 builders’ notes flag the 74LS54 as “not the same gate arrangement” as the 7454, which is why the parts list insists on the standard part.

What to do:

1. Put a 7454 (or the 7451) in the socket and run Search. If the tester names it correctly and says it’s good, the library has the standard-TTL version, and the trap doesn’t apply.
2. If Search fails or names something else, test the same chip under the number. A FAIL here is expected and proves nothing.
3. In that case, test the chip on the breadboard (Step 4), using the data sheet’s function table.

A cheap way round both traps is a USB programmer with a logic-test mode. The XGecu T48’s open test library has separate 7451, 7454 and 74LS54 tests, and open-collector tests that expect an “off” output (Vol 4).

3.5 Step 4: the breadboard fallback

Any chip the tester can’t handle gets a five-minute bench test:

1. 5 V supply on pin 14, ground on pin 7 (for every 14-pin type on this list).
2. For open-collector parts, a 1 kΩ pull-up from each output to 5 V.
3. Drive each input HIGH (to 5 V through 1 kΩ) or LOW (to ground) by jumper, and read each output with a logic probe or meter. Work through the data sheet’s function table one row at a time.
4. A standard-TTL LOW output reads under 0.4 V. A HIGH from a totem-pole output reads over 2.4 V, and from a pulled-up open collector close to 5 V.

3.6 Step 5: log the results

The log template is 04-templates/kenbak-chip-log.csv: one row per chip, with type, marking and date code, maker, source, the menu used, and the result. A pencil mark on the tube, or a dot on each tested chip, keeps the passed chips apart from the untested ones.

Things worth noting while logging:

• Date codes and logos. A tube of identical TI chips from one week in 1974 is reassuring. A tube of mixed logos all claiming the same date code is a sign of remarking.
• Surface. Sanded tops, ink that rubs off with alcohol, or pins that are freshly tinned on old-looking packages are all remarking signs.
• Every FAIL gets a second test with fresh batteries, after re-seating, then with the reference chip, before it is binned.

3.7 After the tester

Every chip goes into a machined socket on the KENBAK-1 board, so a bad one that slipped through costs a swap, not a desoldering job. The board’s own bring-up (KENBAK-1 vol 12) is the final test: the machine runs at speed, under real load, with all 130 chips together.