Skip to content
Buy Spikeling units

Network with two units

This experiment builds a minimal spiking network using two Spikeling units (or a two-neuron setup in Emulator mode). You will create a connection from one unit’s axon output to the other unit’s synapse input, then explore how coupling sign, strength, and presynaptic firing pattern shape the postsynaptic output.

The point is to teach the simplest network truth in neuroscience:

Neurons influence each other through spikes, and synapses determine whether those spikes drive or suppress activity.

Background reading: - Concepts → Synapses and networks - Synapses and inputs

Learning goals

By the end of this experiment, students should be able to:

  • wire a two-unit network (axon → synapse) safely and consistently
  • demonstrate excitatory vs inhibitory coupling
  • show how presynaptic firing rate affects postsynaptic Vm/spiking (temporal summation)
  • create basic motifs: feedforward excitation, feedforward inhibition, and E/I balance (optional)
  • measure a simple “network transfer function” (presynaptic rate → postsynaptic rate)

What you need

Hardware route

  • Two Spikeling units (Unit A presynaptic, Unit B postsynaptic)
  • TRS cable for: Axon output (A) → Synapse input (B)
  • Spikeling GUI

Emulator route

  • Emulator mode with Auxiliary Neuron 1
  • Optional: Auxiliary Neuron 2 (for E/I balance motifs)

Recommended signals to display (postsynaptic): - Vm - Synapse input current (Synapse 1 / Synapse 2, if available) - Optional: Total input current (Itot)

Part A — Build the connection

Hardware wiring

  1. Power and connect both units to the GUI as appropriate for your setup.
  2. Patch the TRS cable:
  3. Unit A Axon output → Unit B Synapse input 1

Cable sanity check

If the connection behaves oddly, confirm you have not swapped stimulus and axon ports. Use your labelled cables and the mapping in Controls and I/O.

Emulator wiring

  1. Start emulator.
  2. Configure the main neuron as postsynaptic.
  3. Open Auxiliary Neuron 1.
  4. Enable coupling using Synapse main neuron.

Part B — Establish stable presynaptic spiking (Unit A)

  1. Choose a neuron mode that produces reliable tonic spiking.
  2. Use injected current (hardware current injection pot / patch clamp slider) to set a steady spike train.
  3. Keep noise off for the first pass.

Goal: a stable presynaptic spike train makes synaptic effects easier to interpret.

Part C — Excitatory coupling (positive synapse gain)

Setup (Unit B)

  1. Set Unit B to a stable baseline below threshold (silent).
  2. Set Synapse 1 gain to a small positive value.

Sweep

  • Increase synapse gain until you observe:
  • clear depolarising postsynaptic responses, and then
  • postsynaptic spiking (if strong enough)

What to look for

  • postsynaptic Vm deflections time-locked to presynaptic spiking
  • summation when presynaptic rate increases
  • recruitment into spiking when net drive crosses threshold

Part D — Inhibitory coupling (negative synapse gain)

Setup

  1. Keep the same presynaptic spike train.
  2. Set Synapse 1 gain to a negative value.

What to look for

  • postsynaptic Vm becomes more hyperpolarised or less excitable
  • spiking is delayed, reduced, or abolished
  • inhibition can create spike skipping if the postsynaptic neuron is already spiking

Part E — Network transfer function: presynaptic rate → postsynaptic rate

This is a clean quantitative network measurement.

Steps

  1. Fix synapse sign and gain (e.g., moderate excitatory).
  2. Run several presynaptic firing rates by changing Unit A injected current:
  3. low rate
  4. medium rate
  5. high rate
  6. For each presynaptic rate, record 10–20 seconds and measure:
  7. presynaptic firing rate (Hz)
  8. postsynaptic firing rate (Hz)

Expected observations

  • excitatory coupling: postsynaptic rate increases with presynaptic rate (often non-linear)
  • inhibitory coupling: postsynaptic rate decreases as presynaptic rate increases

Part F — Timing and coincidence (optional)

If you have three units (or emulator with two auxiliaries), you can show coincidence detection:

  • Unit A → Synapse 1 on Unit B (excitatory)
  • Unit C → Synapse 2 on Unit B (excitatory)

Adjust presynaptic timing so spikes sometimes coincide.

Observation: coincident input can recruit a postsynaptic spike when either input alone cannot.

Part G — E/I balance (optional but powerful)

Use two inputs with opposite sign:

  • Synapse 1: excitatory (positive)
  • Synapse 2: inhibitory (negative)

Explore how inhibition shapes excitation:

  • vetoing spikes
  • delaying spikes
  • sharpening spike timing
  • controlling output rate without changing mean input current directly

Measurements (choose one, depending on level)

Simple, robust metrics:

  • Postsynaptic firing rate vs synapse gain
  • Postsynaptic spike probability per presynaptic burst
  • Latency between presynaptic spike and postsynaptic spike
  • Summation index: peak Vm deflection after a burst / after a single spike

Common issues

  • No postsynaptic effect
  • presynaptic unit may not be spiking
  • synapse gain may be near zero

  • wrong ports/cable routing

  • Postsynaptic neuron saturates

  • synapse gain too strong
  • postsynaptic bias current too depolarising
  • reduce gain and re-centre the postsynaptic baseline

Discussion prompts

  • Why does synaptic decay enable temporal integration?
  • How can inhibition control spike timing without necessarily changing mean Vm much?
  • Why is the network transfer function non-linear near threshold?
  • What would happen if both neurons were adapting strongly?