logic level converter/Checking it/09. When nothing crosses
Checking it · 09 of 9

When nothing crosses

Four readings with a meter, black probe on GND, tell the usual faults apart: 5V on the header's 5V pin, 3.3 V on the socket's 3V3, 3.3 V on your board's L pin at rest and 5 V on the part's H pin at rest. A 3.3 V board's pin reading well over 3.3 V means it is on the wrong row. Readings that are all right and a signal that still does not cross mean the channel numbers do not match.

Four readings

Power the board with nothing driving the channel: no sketch running, or a sketch that leaves the pin as an input. Put the black probe on GND, and take four readings.

Four readings
Wiring
Readings off
0 of 4
Passes signals
yes
Five, three point three, three point three, five. The 5V pin is fed, the regulator is making 3V3, and each side of the channel rests at its own HIGH. If a sketch still sees nothing, the fault is past the TK97: the part, or a pin number in the sketch.
  1. The header's 5V pin. 5 V, give or take a few tenths on USB.
  2. The socket's 3V3 pin. 3.23 to 3.37 V, the ME6211's 2 % either side of 3.3 V.
  3. The L pin your 3.3 V board is on. 3.3 V, from its 10 kΩ.
  4. The H pin the 5 V part is on. 5 V, from its 10 kΩ.

What each wrong reading means

Pick each wiring in the figure.

  • All four at zero: no 5V. The regulator has nothing to work from, so the 3.3 V side is dead too. This is the most common one.
  • Your board's pin well over 3.3 V: it is on the header's row, not the socket's, and the H pin's 10 kΩ is pushing 5 V at it. That is the one reading that can damage something. Unplug it first.
  • All four right and nothing crosses: the two ends are on different channels. The meter cannot see it; the numbers printed on the board can.

Then the loopback

When the four readings are right, the loopback sketch checks the rest: two channels, both directions, with one jumper. If that passes and your part still does not answer, the TK97 is working and the fault is in the part, its library or a pin number in the sketch.

When it does not work

Every reading is zero.

5V is not reaching the header, or GND is not shared. Measure your board's own 5V pin against its GND: if that is 5 V, the wire between them and the TK97 is the fault. On a Pico, VBUS is 5 V only while it is on USB.

The 3V3 pin reads 5 V.

Something is feeding it from outside. The regulator cannot make more than about 3.37 V from 5 V. Unplug everything from the socket and measure again; a block or a wire that brought 5 V into the socket is the cause.

My 3.3 V board's pin reads well over 3.3 V.

It is on an H pin. Each H pin has 10 kΩ up to 5V, which pushes current into a 3.3 V input through its protection diodes; the reading lands between about 3.6 and 5 V depending on your board. Move it to the socket's row before anything else.

All four readings are right and the part still does not answer.

Check that your pin's L number matches the part's H number. A 3.3 V board on L1 and a part on H2 are on two different channels, and a meter at rest cannot show it. The loopback sketch reports it as always HIGH.

The 5V pin reads 4.6 V or so.

A USB supply under load, or a long cable. The regulator still gives 3.3 V from anything above about 3.42 V at 100 mA, so this matters only if the part on the 5 V side needs a full 5 V.

Where this goes next

Every block on the shelf, each with a handbook of its own.

The other TinkerBlock blocks →

Edit this page — content/books/logic-level-converter/when-nothing-crosses.mdx

Community

Questions about this product

See what other owners have asked, and read their solutions.

Ask a question ↗

Logic Level Converter

Loading discussions…

Discuss this article

Ask about this page. The answer stays here, on the page it belongs to, for whoever hits the same wall next.

Browse Modules and blocks on the forum →