Years ago constant current diodes (a Jfet with a resistor) were available and maybe still today.

 

One possible disadvantage of that circuit is that there can be a significant difference in the constant-current from unit to unit, due to the large variation in the JFET's Vgs control voltage (see below for the 2N3819 values, which shows a 15:1 difference).
This would require adjusting the source-resistor value to get the desired current.

View attachment 248931

I wonder what the manufacturing process is that causes such large variations?

 

No one has given an explanation as to why the initial spike appears. The transistor has a collector-base capacity. Due to the fact that the supply voltage rises from zero to 36 V in 10 µs, there is a charging current of this capacity. This current enters the base of the transistor and opens it. Replace the transistor with a perfect transistor (no capacitance by default) and there will be no overshoot. Adding a resistor to ground prevents the harmful effect. You can increase the power on edge with a PWL type source. For example Vcc 36V 0 0s 0V 1ms 36V. The harmful effect will be less!

 

I wonder what the manufacturing process is that causes such large variations?

It seems to inherent in the process used to make JFETs (which likely hasn't changed much in years).
Apparently the channel pinch-off voltage is very sensitive to small process or doping variations during its fabrication.

 

Hi,

Again commenting with some finding.
Did anyone observ any failure about senet diode diode?

What is the minimum current needed for zener to be in breakdown mode ?

Per current circuit zener current is 550uA and further increase in R1 value may reduce the zener current. This may stop zener to be in breakdown mode.

Could anyone help me theoretically ?
As I see in datasheet of zener , the Izt is in mA e.g. 5mA.

Thanks in Advance ?

 

The datasheet for the UMZ12K zener diode says that it is not recommended for new designs. It is 12V +/-10% at a current of 5mA and its datasheet shows a "typical" curve of it only a little less than 12V (about 11.9V) at a current of 1uA.

 

The datasheet for the UMZ12K zener diode says that it is not recommended for new designs. It is 12V +/-10% at a current of 5mA and its datasheet shows a "typical" curve of it only a little less than 12V (about 11.9V) at a current of 1uA.

Thanks for your reply !
I tired to look the graph but I did not find it.
Could you please share the snip here ?
Thanks !

 

hi,
I always to try to use the same circuit the TS has posted, unless he asks for a design modification.
They are usually interested in how their circuit performs, not my version of it.
E


Update:
Look at these two sims, with and without R5
View attachment 248918

Hello,
Could you please suggest any other solutions here ?
Just to have provision of RC delay upto 1sec.

 

hi m,
A simple option.
E

 

Could you please suggest any other solutions here ?
Just to have provision of RC delay upto 1sec.

Can you use an IC or does it need to be built from discrete parts?

 

Can you use an IC or does it need to be built from discrete parts?

Hello,
Cost effective solution.
Discrete circuit will less cost.
Which IC are you talking about ?

 

hi m,
A simple option.
E
View attachment 249031

Hi C,
By tuning the RC we can tune the delay time, this I understand.
Here JFET part selection is critical. Since Vds is 36V so we have to choose the part accordingly.
Then value of Vgs(off)= ? And Id = ?

To be very frank I was not able to calculate the value of IR1=?
I tried using Id = idss(1-Vgs/Vgs_off)^2 but this formula did not give Ir1=1.3mA.
How do we decide Vgs voltage here ?

For me simulation shows everything but theoretically I was not able to calculate and decide those parameters.

If some one help me to do the on paper analysis.

Thanks !

 

Hi,
For this particular 2N3819 a –2.1Vgs will enable an Ids of 1.27mA,

So 2.1v/1.27mA = ~1k6.

It will depend upon the production spread of the jFET parameters.

E

 

UMZ12K datasheet details:

 

Discrete circuit will less cost.
Which IC are you talking about ?

A 555 timer which is dirt cheap, likely less than the cost of the Zener and JFET.

Do you specifically need that transistor to turn on, or do you just need a signal?

 

A 555 timer which is dirt cheap, likely less than the cost of the Zener and JFET.

Do you specifically need that transistor to turn on, or do you just need a signal?

I need transistor to be turned on after some delay.
555 timer in monostable state need inverter circuit to turn ON the Transistor permanently. Also it need additional external trigger circuit.

Actually I am trying to design low cost soft start circuit.

 

See

 

See

That circuit won't tolerate 36V.
Needs an added regulator of some sort.

 

Here's a delay circuit using an LM339 or LM393 comparator.

The comparator has an open-collector transistor output, so I don't think an added transistor is needed.

Note that the delay is independent of supply voltage.

The pulse delay after power application is ≈R1*C1 as determined by the resistor and capacitor tolerances.

The LED serves two purposes; it gives an indication that the power is applied, and it reduces the supply voltage by about 2V since the LM339/393 is rated for a maximum of 36V.

 

Here's a delay circuit using an LM339 or LM393 comparator.

The comparator has an open-collector transistor output, so I don't think an added transistor is needed.

Note that the delay is independent of supply voltage.

The pulse delay after power application is ≈R1*C1 as determined by the resistor and capacitor tolerances.

The LED serves two purposes; it gives an indication that the power is applied, and it reduces the supply voltage by about 2V since the LM339/393 is rated for a maximum of 36V.

View attachment 249114

Here's a delay circuit using an LM339 or LM393 comparator.

The comparator has an open-collector transistor output, so I don't think an added transistor is needed.

Note that the delay is independent of supply voltage.

The pulse delay after power application is ≈R1*C1 as determined by the resistor and capacitor tolerances.

The LED serves two purposes; it gives an indication that the power is applied, and it reduces the supply voltage by about 2V since the LM339/393 is rated for a maximum of 36V.

View attachment 249114

Hello C,

Thanks for your support,

This looks quite good for me.
The only problem i see here, the logic is reverse of mine circuit posted in post #1.
Open collector outputs means opamp has V1 at output before delay arises and comparators works.

I need output high after delay.

Regards,