As I mentioned in my previous post, I was unhappy with the standard 0.1″ headers that the Xprotolab Plain used for measuring analog signals, so I created an adapter board that would connect standard BNC connectors to the oscilloscope header. BNC connectors tend to be the standard for analog signals due to their compatibility with coax cables which reduce the noise on analog signals.  In addition to routing the two oscilloscope channels to BNCs I added a third one to route the AWG waveform out of.  However, due to the voltage range only being +/- 2V, I decided to add some signal amplification to increase the range to +/- 8V.

Below I’ll break the schematic down into pieces and explain the purpose of each.

 

This is the connector to mate with the Xprotolab Plain board.  Both oscilloscope channels, the oscilloscope trigger pin, and the AWG are broken out.

 

Rather than adding a step up regulator to the board to boost the Xprotolab’s 3V to the 5V that is necessary for the Max680 circuit, I just left a connector on the board to feed the 5V supply.  Due to the large amount of USB components that will be in the tablet, I expect it will be easy to connect the USB power to this input.

 

I’ve routed the external trigger for the oscilloscope to a 0.1″ header.  Because it’s a binary signal I don’t need to worry as much about high speed signalling and noise.

 

 

These are the three BNC connectors for the two oscilloscope channels as well as the AWG output.

 

This is a boost converter that takes in 5V and generates +/- 10V to use for the amplifier rails.

 

This is just a basic inverting amplifier using one of the op-amps in the LM358.  It takes the AWG signal as an input and amplifies it up to a +/- 8V signal.  Due to the inner workings of op-amps, it’s not possible to use the full +/- 10V, so I rounded down to the next even number.  The resister ratio (R2/R1) needed to be 4x to achieve the proper gain to amplify the 2V signal to 8V.  However, because most resistors have a 5% tolerance, I used a 10kOhm potentiometer for R2 rather than a 4kOhm resistor so I could adjust the gain to be exactly 4x.

Here’s the final layout for the board.  Notice I didn’t use a powered plane on the top board layer like I usually do.  This is to prevent any power bus noise from interfering with the oscilloscope signals.  I also tried to keep the traces as thin as possible because thicker traces cause more signal degradation.  For anyone interested in ordering this board for themselves, it’s available on OSH Park here.  I kept all of the components as through hole parts to make soldering easier.