Hello everyone, I am working on a project on the Izhikevich circuit in 32nm technology, I am having difficulty simulating the circuit..

phatdatgl

Newbie
Joined
Jun 17, 2024
Messages
4
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
32
Hello everyone, I am working on a project on the Izhikevich circuit in 32nm technology, I am having difficulty simulating the circuit, I wonder if anyone here has simulated through this circuit?
 

The original article is evidently written by Izhikevich. It appears to be available by subscription at the IEEE site. Normally Edaboard is not allowed to display those articles. A neuron's response seems to be adjusted close to a borderline of touching off other neurons, or else being triggered to fire several times repeatedly. A few scope diagrams look like a phenomenon known as 'squegging': Peaks occur at a lower frequency (discharges or spikes), while each spike consists of several spikes in rapid sequence at a much high frequency. It's hard to build deliberately into an oscillating circuit and when it happens accidentally it's usually unwanted.

There's a neon bulb oscillator which triggers a row of neon bulbs to flash one after another.

You can try a Forum search on neuron and neural, sinse it's been an occasional topic in discussions.
--- Updated ---

homemade-circuits.com/neon-lamps-working-and-application-circuits/
 
Last edited:



Thanks for helping me in the previous questions, now I have another circuit but I have not made the desired waveform, but have not done the desired result, it is the Izhikevich circuit
 

VLSI (Very Large scale integration) implies the schematic is intended to be etched on the same substrate. Probably in large numbers if it's imitating a cortical network. Evidently it sends and receives signals from other sections (c, d, Vin, Vbias, Spike out, Post Synaptic Input etc.). Certain signals probably require careful adjustment in order to create desired current flows. I imagine it requires repeated adjustments.

Perhaps it can succeed since you mention using 32nm process. It will require an expert to tell whether the original schematic can work, and whether your own simulation can work. It appears to make use of current mirrors and long-tail pairs. It would help if functions could be distilled into a simpler flow of steps. An expert certainly can do that better than I can.
 

Cookies are required to use this site. You must accept them to continue using the site. Learn more…