#1879

Episode 1879
programmable current source
Be a Patron: / imsaiguy

[@craigs5212]

If you get a chance before sending it out, add a contact pad area for testing SMD diodes. Version 2 : USB-C for power with possibly a boost converter to get a higher voltage. Version 3: USB-C charging for Li cell power.

[@alexloktionoff6833]

I'm afraid 30v power supply just will burn MOSFET gate while no zener is connected. I think you need a 10-15v zener between gate and ground.

[@peterjones2411]

you should have minimum of 1ma at all times passing through tl431 to maintain regulation. specified in data sheet

[@ramighazzawi3907]

Yes, and 5mA is ideal

[@johnlucas2445]

Really bad design. Well below what you are capable of. Use a current limited dc power supply. Set the voltage and current limit and read the value on the power supply voltmeter meter. You could use any test voltage with no limit as long as you could set the current limit. Tl431 is a bad choice for a current source. A LML317 or equal as a current source would be better with a less part count. Why buy expensive 1% resistors or such when I am sure you have 10 turn pots in your junk parts bin. Dial that current right in.

[@ChrisSmith-tc4df]

One must just take care to connect/disconnect the DUT with the output channel off so that the output capacitor doesn’t dump into the DUT. Zeners make poor flash lamps.

[@andymouse]

As you don't pay you might as well have ENIG (Electroless Nickel Immersion Gold ) having a surface that won't oxidise for the contact pads is a perfect example of why such a surface treatment exists !...cheers.

[@uni-byte]

Good idea Andy. I'll have to remember that for a future PCBWay project.

[@_nemo171]

Nickel/phosphorus is always a good idea. You can do it at home also.

[@andymouse]

@@uni-byte :)

[@uni-byte]

I did a diode checker a while back. Handy tool. Mine is self contained and uses a completely different approach though.

[@ramighazzawi3907]

I stopped ordering resistors ans capacitors off digikey.. LCSC is legit 1/30th of the price.. digikey shipping is also unreal..

[@VEC7ORlt]

May I ask, why even bother? Lab supply will do all the above and why even bother measuring zeners - ballpark figure is more than enough. Also you might want to use dip switches instead of jumpers and you can replace 10k resistor with a 2 transistor current source and ensure +1mA current through the 431 regardless of supply voltage.

[@ChrisSmith-tc4df]

One must take care to have the lab power supply channel output off while connecting the DUT, or the output filtering capacitor inside that’s wired directly on the output posts will briefly dump a bunch of current into the DUT upon first contact - potentially blowing it out. Care taken, it is a viable method that I’ve used myself.

[@gkdresden]

A nice-to-have would be a current adjustment which is inversly proportional to the Zener voltage. I think it is easy to implement it into the circuit so that the measuring of the Zener voltage goes more semi-automatically.

[@kumarayushmittal2099]

which software are u using?

[@alaricsnellpym]

Could you make a better current source than the datasheet one by using the TL431 current sink circuit to feed a current mirror? Two more transistors, of course...

[@andymouse]

Occam's Razor. he's not making a metrology grade analytical instrument here ! But of course you're free to. The obvious answer to your question is 'Yes' but you knew that already.

[@ghlscitel6714]

An output current DAC may also be a choice like e.g. the AD DAC08.

[@andymouse]

@@ghlscitel6714 And what do you do with your analogue output ? Its got clips for a DMM already to use, that's the design

[@ghlscitel6714]

@@andymouse Drive the Zener diode with the DAC output current and measure the voltage with a DMM

[@andymouse]

@@ghlscitel6714 It dosent need this level of complexity :) my final point.

[@ghlscitel6714]

Cool circuit! Which EDA tool is this? Maybe another MOSFET like the Onsemi FDN86501LZ is advantageous with its more than 10 times higher transconductance.

[@alexloktionoff6833]

Why not to make the PCB totally on SMD components? let's say SOT23 1206 are very easy to solder manually.

[@bobdoritique7347]

clever scheme!

[@AnalogDude_]

You only got 3 holes. Someone is making musical instruments using zeners to create white noise.

[@_bogbean]

For those Vz and Vg test points try the Kicad test point with ‘bridge’ in the title, it shows the wire loop on the 3D model.

[@markcummings150]

I wouldn’t have put the current flow direction on the pcb because we don’t all use conventional current. I would have put the zener symbol somewhere in parallel to the cutout somehow. Personally I was taught (a few decades ago) and still use electron current flow as it makes more sense, especially when it comes to valves.

[@VEC7ORlt]

Literally everyone uses conventional current flow, its you who is the weirdo here.

[@timgraham8975]

@@VEC7ORlt I must be a weirdo also! I was also taught electron flow.

[@markcummings150]

@@VEC7ORltwhat’s weird is putting a current flow direction on a pcb. All PCBs I’ve ever seen only have component orientation on them.

[@markcummings150]

@@timgraham8975my point exactly. So @VEC7ORIt is ‘literally’ wrong.

[@VEC7ORlt]

@@timgraham8975 so was I, in 7 or 8th grade, and it proves jack SH.