Level Translator Breakout - TXB0104 - CS Output Level Query

This is a novice question.

Pin high

Input is 1.78V (1.8V board)

Output is 3.35V

Pin low

Input is -7.6mV

Output is 0.94V

Why isn’t the output 0V when pin low?

I’m using it to shift to a LSM9DS1 using SPI without much success at the moment. Thanks.

Thanks for reaching out to us on this.

Luckily we have made a couple guides for getting started with these, check out https://learn.sparkfun.com/tutorials/logic-levels , which will give a very good overview of what signal levels you can expect in practice, and then https://learn.sparkfun.com/tutorials/tx … 1615791300 for this product specifically.

Hope this helps, and happy sparking!

Thanks. I’ve had a look at those. I still don’t understand why -7.6mV would output as 0.94V. Is the device operating correctly?

Real quick: Please provide a picture/video of your set-up/device wiring, it’ll help. Thanks

Photos attached. Spaghetti wiring but hopefully you can still see what you need to. Thanks.

Now with photos.

It’s really hard to follow the wiring in your photos, could you post a schematic?

Diagram attached.

Low doesn’t always mean 0 volts, it just means lower than what voltage is considered “low” or logic 0 for the device you’re using.

Inputting a voltage lower than zero for a logic low could be a problem though, some IC’s are not going to like that.

The gray board in your diagram, is that running at 1.8 volts?

What voltage are you feeding VCCA at?

What voltage are you feeding VCCB at?

Have you tried any other SPI boards with your setup?

DragonBoard running at 1.8V

VCCA = 1.8V

VCCB = 3.3V

I haven’t tried the setup with another board although I’m working on getting the LSM9DS1 up and running with an Artemis ATP. I’d still like to get it working with the DragonBoard though.

Many thanks

You might an Arduino Uno first since that’s a known working combination. Once you’re happy things work there, then move onto the Artemis and then to your Arrow board.

honestly… I do not understand where the negative voltage is coming from: GND should be GND. It must be a wiring/connection issue that a negative voltage is measured as it should be 0v everywhere.

Also, the 0.94V does not worry me looking at TXB0104 internal schematics. A PMOS or NMOS transistor/ MOSFET delivers a ‘weak 0’ as TS-CHRIS pointed out: low does not mean 0V, but low enough, 0.7V is what is very normal to expect.

However, looking at the Fritzing I do not see external pull-up resistors between the DragonBoard and TXB0104. They are available on the LSM9DS1 board and thus on the “VCCB” side, but you need those as well on the “VCCA” side. I2C physical layer is only pulling to 'GND ’ (SDA or SCL). A real pull-up (1.8V in this case) needs to be performed by resistors and thus make sure the level translator is working correctly as well. The right resistor values (like 2k2) will make sure enough current flowing to allow the TXB-0104 internal circuit to perform better.

with respect to my remarks about I2C, that was not correct as you use this in SPI mode. So please neglect that part.

Looking at the fritzing looking at the 78xxl , is the input really on the middle and GND connected on the left pin. (in most 78xx in TO220 packages it is the other way around).

You are not using CS_M is the by design? I wonder how the impact on communication if one part is set for I2C (mag) and the GYRO on SPI

There are a lot of wires, some half-way in the bread-board. Make sure they connect well. Also, you can try using SPI at a much lower speed and see whether there is impact.

Thanks Chris and Paul. I have decided to try to use an Artemis ATP and LSM9DS1 setup first - should be easier for a beginner.

In answer to a couple of your questions Paul:

The voltage regulator is LD1117VXX so middle is output which I think is what is shown on the Fritzing.

Had deliberatly not connected CS_M to keep things simple and because DragonBoard only has one CS pin. I can understand what you are saying though about the possible impact of not using connecting the pin.

You are possibly right about the breadboard connections. I’m using a 30 year old board which doesn’t take the connectors very well. Time for a change I think.

I’ll try the lower SPI at a lower speed when I get back to trying this setup.

In the meantime I’ll start a new topic for a query I have about this using the Artemis ATP.