Arduino Forum

Using Arduino => General Electronics => Topic started by: bx3me on Jan 30, 2020, 07:11 pm

Title: +/-5v output?
Post by: bx3me on Jan 30, 2020, 07:11 pm
Hi, I have a industrial interface to which i am trying to send data. The device requires data to be sent on 8 wires: CLOCK+, CLOCK-, SYNC+, SYNC-, CHANNEL1+, CHANNEL1-, CHANNEL2+, CHANNEL2-.

The CLOCK is running at 2MHz and the SYNC needs to be low on 20:th bit sent.

How can i convert 3.3v logic level to +/- 5v?

Background:
I have used several platforms to generate the signal and have communicated sucessfully using the Teensy 3.6 Board. In this case all negative inputs where floating / NC. It worked but to much interference to be useful. Also the Teensy 3.6 was a bit to slow so i needed a better solution. I just bought the Teensy 4.0 and it is definitly fast enough but the 3.3v are not even close to give a readable output by the interface.

If i use long cables and pick up a lot of distortion on the negative (NC) wires random packets reach the interface so i only need to get the 3.3v signal to become -5v at 2MHz... I have loocked at some DC DC Buck converters but they all seem to max out at 400kHz.

Any idea on how i can produce 0v - (-5v) signal from 0 - (+3.3v) logic?
Title: Re: +/-5v output?
Post by: MarkT on Jan 30, 2020, 09:47 pm
Datasheet for the interface would be a good starting point.

I don't think you need negative voltages for differential signals.

The switching frequency of a buck converter is of no relevance to sending logic signals.  Buck converters produce DC power.
Title: Re: +/-5v output?
Post by: WattsThat on Jan 30, 2020, 10:09 pm
The LTC1045 can do that.

https://www.analog.com/media/en/technical-documentation/data-sheets/1045fc.pdf (https://www.analog.com/media/en/technical-documentation/data-sheets/1045fc.pdf)
Title: Re: +/-5v output?
Post by: herbschwarz on Jan 31, 2020, 02:16 am
I think maybe the - wires are just returns (grounds) for the
signal wires that are usually in twisted  pairs. If so, that may
help solve the interference problem.
Herb