FAQ
TL;DR: Goal: light 6 LEDs in sequence with ~1 s steps, then blink all together; "Much easier with logic components or even better, an MCU." [Elektroda, Steve Lawson, post #21664606]
Why it matters: This FAQ translates a constraints-only school brief (op-amps + MOSFETs/BJTs) into workable sequencing options without 555s or decoders.
Quick-Facts
- Target pattern: 6-LED ramp (1 s per step), then all blink in unison. [Elektroda, elsie li, post #21664604]
- Allowed parts: op-amps (e.g., LM358), MOSFETs, BJTs; no 555/decoder ICs per brief. [Elektroda, elsie li, post #21664607]
- Clock available: square wave from a bistable multivibrator is already on hand. [Elektroda, elsie li, post #21664604]
- Typical alternative if rules were relaxed: CD4017 + 555 or MCU for simpler build. [Elektroda, Mark Harrington, post #21664608]
- Shift-register idea (rule-dependent): 74HC595 can step LEDs with minimal wiring. [Elektroda, Julio Flores, post #21664605]
Quick Facts
- Target pattern: 6-LED ramp (1 s per step), then all blink in unison. [Elektroda, elsie li, post #21664604]
- Allowed parts: op-amps (e.g., LM358), MOSFETs, BJTs; no 555/decoder ICs per brief. [Elektroda, elsie li, post #21664607]
- Clock available: square wave from a bistable multivibrator is already on hand. [Elektroda, elsie li, post #21664604]
- Typical alternative if rules were relaxed: CD4017 + 555 or MCU for simpler build. [Elektroda, Mark Harrington, post #21664608]
- Shift-register idea (rule-dependent): 74HC595 can step LEDs with minimal wiring. [Elektroda, Julio Flores, post #21664605]
What does “sequentially” lighting LEDs mean in this project?
Here, sequentially means one LED turns on after another until all are on, then all blink together. Timing is even, about 1 second per step. The pattern repeats until power off. This matches the described 6-LED ramp followed by synchronized blinking. [Elektroda, elsie li, post #21664604]
Why avoid 555 timers or decoders like CD4017 here?
This is a constraints-only exercise: you were given op-amps, MOSFETs, and BJTs to use. The usual approach would be a 555 clock driving a CD4017 decade counter, but the assignment forbids that to focus on discrete or op-amp design. [Elektroda, Mark Harrington, post #21664608]
How hard is it to do this with only op-amps and transistors?
It’s doable but tougher than with logic ICs. An expert noted it’s “much easier with logic components or even better, an MCU.” Expect more parts and tuning when using only LM358, MOSFETs, and BJTs. [Elektroda, Steve Lawson, post #21664606]
I already have a square wave. Can I use it to step the LEDs?
Yes. Feed your bistable multivibrator’s square wave into a chain that progressively latches LEDs on, then switch to a global blink mode. The 1-second period suits human-visible sequencing well. [Elektroda, elsie li, post #21664604]
Can I prototype with just 3 LEDs first and then scale to 6?
Yes. Build a 3-stage version to validate timing and gating, then expand to 6. Starting small makes debugging current paths and thresholds easier before duplicating stages. [Elektroda, elsie li, post #21664607]
What’s a practical discrete approach without 555/CD4017?
Use a ring-counter-like chain with transistor flip‑flops or FET flip‑flops. Clock it with your square wave, add feedback, and gate the final stage to switch into the all-blink mode. “Think ring counter and think feedback.” [Elektroda, Mark Harrington, post #21664612]
Is there a reference for FET flip‑flops as building blocks?
Yes. A contributor pointed to a FET flip‑flop reference for inspiration. You can adapt the concept for LED-driving gates using your available devices and thresholds. [Elektroda, Mark Harrington, post #21664611]
How do I switch from “ramp-up” to “all blink together”?
One suggestion is to let the last stage toggle a bistable that globally gates all LED drivers. When that bistable is active, the same clock makes all channels blink in unison. [Elektroda, Steve Lawson, post #21664614]
Can I emulate a 555 timer using op-amps and transistors?
Yes. You can replicate the 555’s comparator, SR latch, and discharge transistor with an op‑amp and a BJT/MOSFET. That satisfies constraints while yielding a clean clock or one‑shot. [Elektroda, Jiajun Chen, post #21664610]
What is a bistable multivibrator in this context?
It’s a two‑state circuit that produces a square wave used as your clock. You already have one, so you can drive the sequential logic without adding a 555. [Elektroda, elsie li, post #21664604]
Why do many recommend a CD4017/555 or MCU instead?
Because it’s the standard, simpler solution for LED chasers. As one expert put it, the normal path is a decade counter with a 555, or nowadays a microcontroller. [Elektroda, Mark Harrington, post #21664608]
What’s a shift-register option if rules allow it later?
A 74HC595 can shift in bits with each clock and present 8 outputs. You’d load a pattern that grows lit LEDs, then a full-on pattern for the blink mode. [Elektroda, Julio Flores, post #21664605]
Any edge cases I should watch for when using only linear parts?
Threshold drift and unequal LED currents can stall or skip steps. The expert caution was that achieving this only with linear components is difficult, so design margins matter. [Elektroda, Steve Lawson, post #21664606]
Can you outline a simple concept to start from? (3-step How‑To)
- Build a three-stage FET/BJT flip‑flop chain and clock it from your square wave.
- Add gating so each stage latches on in sequence, then routes the final stage to a global gate.
- Use the same clock to blink all channels when the global gate is active.
[Elektroda, Steve Lawson, post #21664615]
What timing should I target for visibility?
About 1 second per transition is readable while debugging and demos. You can scale later, but start with 1 s to verify logic and current handling. [Elektroda, elsie li, post #21664604]
Can LEDs and LDRs form a ring counter variant for this?
Yes. A creative suggestion was a ring counter employing LDRs, LEDs, and FETs. It’s unconventional but satisfies the “no 555/decoder” constraint creatively. [Elektroda, Mark Harrington, post #21664611]
