Building a simple oscillator

the

AnalogKid said:

For other circuits, probably. For the OP circuit, there is no (traditional) base current.

ak

Click to expand...

My comments were based on my suggestions with a regulated lower voltage not the video. However I recall the 2N2904 Vec will trigger between 8 and 9 V but will degrade rapidly with large cap (xx uF) discharge surge currents and temp rise. But if the OP just wants to make a super simple circuit work, get a plug-in breadboard and save time. If he wants to practice soldering, use magnet wire for neater appearance and tin the ends before connection. We have no idea on the purpose.

Be kind to your little 2N3904 and put some resistance in the discharge path.



Voltage across capacitor (yellow, top trace)
Voltage across LED and resistor (blue, Lowe trace)

Good suggestion Dick.

for the OP:
I see a trigger at 10.4V , a peak current limit of 26 mA with your parts and a holdoff at 6 mA resulting in an average brightness of around 10% of peak, which ought to be enough with a decent 5mm LED with high Iv candella. Audio inputs might be AC coupled already, but don't assume that all are.

Dick's RC values are suitable for low pulse rates but the OPS more for mid-band frequencies. Again we don't know what he wants.

Tony Stewart said:

When using BJT's keep in mind the base current must be at least 3% of the collector current as hFE drops to near 10% of it's value at rated saturation but actually rated at Ic/Ib=10 typically and 50 best case ($) Where as a 1 Ohm FET does not have that issue so smaller caps and much greater R values are possible for slow pulsing. Keep in mind duty cycle might not be 50% unless ratios are optimized.

https://tinyurl.com/22on3n6l

Here's a few different options.

Click to expand...

Tony, the 3% you quote rule of thumb, new to me so I would like to
understand what you are saying. Is it the knee turning into sat occurs
when Ib = 3% at beta 10% of its rated value ? Not sure I understand
the comment.


Rewgards, Dana.

danadak said:

Tony, the 3% you quote rule of thumb, new to me so I would like to
understand what you are saying. Is it the knee turning into sat occurs
when Ib = 3% at beta 10% of its rated value ? Not sure I understand
the comment.


Rewgards, Dana.

Click to expand...

3% current means Ic:/Ib= 33 instead of 10:1 or Ib=10% of Ic.
This means the Vce(sat) will be closer to 0.7 where Vcb ~ 0V rather than say Vce=0.2 when C-B is forward biased.

The actual characteristics of testing Vce(sat) is rounded off near 10% of max linear hFE , like 2%, 5% or 10% of Ic
So I suggested 3% as my rule of thumb because going to Ib= 10% of Ic only reduces the Vce(sat) by a fraction of the Vcb diode drop which is still low enough to satisfy most consumer applications and TTL low which is 0.8V max.

If you review a few hundred transistor datasheets for the test conditions of Vce(sat) @ x mA you will find ;

if hFE is < 200 Ic/Ib=10 ( 10% base current )
if hFE > 200 > Ic/Ib = 20 (5% base current
if hFE > 500 Ic/Ib = 50 ( 2% base current), some are rated 1%
These are Mfg choice, but later adapted by industry not laws of physics but rather production yield.

- then there are some parts with hFE >200 rated at multiple ratios 2%, 5% and 10% of Ic , many from Diodes Inc who have several hundred patents on their process with low Rce and high hFE up to 1500 I think. But some from Fairchild, now TI.


(some #'s from memory , you may verify *)


Rce is a parameter which, I have used for many decades as Rce=Vce/Ic in saturation which is approximately Rce< k/Pmax ( PN2222A is 2 to 1 ohms and 2N3904 is 6 to 4 ohms) where k depends on process design ranging from 0.3 to 1 for most but not all and also likewise for all diodes for Rs above 0.6V

Although more expensive to achieve 10 mohm in a BJT than a Enh-mode FET, the Cout is far lower in BJT's than Coss in FETS (ie. more speed or f_t or BW )

danadak said:

Is it the knee turning into sat occurs
when Ib = 3% at beta 10% of its rated value ?

Click to expand...

In short, it where I feel it is a tradeoff between drive efficiency and acceptable voltage drop for common small signal BJT's with hFE's >= 100 for "TTL acceptable levels" or 74HCT levels where the real threshold is two diode drops near 1.2V

It would not be suitable to use 3% if Ic when the linear DC gain, hFE is only 15 in a high power device.)

rjenkinsgb said:

Without trying it, I'm guessing that emitter and collector are reversed, so it's the VBE breakdown providing base current and it then turns hard on, until the cap discharges.

Click to expand...

This circuit is more like a voltage monitor than an oscillator.

shokjok said:

This circuit is more like a voltage monitor than an oscillator.

Click to expand...

NOT A VOLTAGE MONITOR but exactly a RELAXATION OSCILLATOR with AVALANCHE effect on reverse breakdown. This is not always destructive if the mJ energy is below some unspecified threshold. Beware of using large Caps and it is better to use < =1uF and larger R. But with more leakage than CMOS, there will be a lower frequency limit.

I believe this comment means you have yet to understand the operation. The Vbe is forward biased while Vbc is reversed operating like a Diac with negative incremental resistance or -dV/dI slope.


Those who misbelieve this works must also consider the leakage resistance // pull-up R and C combine to make T=RC for the asymptotic time constant to reach trigger threshold ( high tolerance near 2x Veb { absolute max } = 5 V )

This transition occurs very fast and produces the near 0 rise time limited by Ceb.*Re. During an Avalanche Effect breakdown.

The only uncertainty is how much physical damage affects reliability and depends on the energy discharged from the external cap.

The yellow trace in the teference plots from Coppel indicate the energy is reduced by the high holding current . Thus the difference in 1/2 CV^2 from 16.4V down to 15V is the energy dissipate in a very short duration at high power is what matters. One can try a high R up to maybe 100K and reduce C proportionally to improve MTBF in theory.

theyreinthehouse said:

Hi, hopefully this is the right place to post this question. I’m looking to build an oscillator. I have all the parts ready to order and I wanted to double check that I’m not making an obvious mistake. I’m completely new to this and I thought it would be worth a second opinion to know if the parts I found are correct based on the guide I’m following, if anyone has the time to take a look. Any pointers are appreciated.

Here’s a list of the parts based on the guide that I’m following:

1x 1k resistor,
1x 100k resistor,
1x 10k 16mm potentiometer,
1x 2N3904 transistor,
1x LED,
1X 10μF capacitor

And here’s the list of these parts that I’ve found and plan to order:

1K TruOhm 0.25 Watt Resistor - Irish Electronics.ie

irishelectronics.ie

1/2W Metal Film Resistor 15k Ohm -100k Ohm - Irish Electronics.ie

irishelectronics.ie

10K 16mm Linear Potentiometer - Irish Electronics.ie

irishelectronics.ie

2N3904 Audio Transistor NPN - Irish Electronics.ie

irishelectronics.ie

WHITE LED 5mm WATER-CLEAR - Irish Electronics.ie

irishelectronics.ie

TANTAL CAPACITOR 10µF / 16V - Irish Electronics.ie

irishelectronics.ie

I’m planning to solder it on this strip board - https://irishelectronics.ie/epages/...ath=/Shops/950018241/Products/2439&Locale=en_ IE

Here’s an illustration of the schematic from the video that I’m following - (http://imgur.com/a%2Fg1b0pOi)

Click to expand...

That circuit is useless. Use this one:

Lightium said:

That circuit is useless.

Click to expand...

Why?

ak

The schematic shows the transistor connected backwards. The top and parts view of the circuit has the transistor correct.

The maximum allowed reverse biased emitter-base of this 2N3904 transistor and most other NPN transistors is only 6V when it has avalanche breakdown and suddenly conducts. Then the capacitor discharges and the transistor turns off until the capacitor charges high enough for another avalanche breakdown, over and over.

Please note that the starter of this thread has not been back here since the 16th November.

So, he either has is simplified oscillator working, or, he gave up amid the avalanche of conflicting advice and went away.

JimB