🔴 Build A Constant Current Source / Current Sink - Sponsored by PCBWay - No.1005

[@TheDefpom]

I have just gone through and renamed things, on reflection it is a sink not a source... i have updated the project on PCBWay too (just a naming change, nothing else.)

[@daveodessa]

Mounting the TL431 on the bottom side of the PCB may help with effect of the heat from resistors. That pinout It just goes to show that there are standards and standards.

[@t1d100]

Fun! Thanks. IIRC, the OEM 431 comes in several pinout models. Yep, the project is worthy of an OEM 431. Comes as LM431 and TL431, too. Don't forget to correct the cap hole spacings, before posting the project files.

[@TheDefpom]

I’ve done a rev1.1 which is what’s posted on the PCBway site, it corrected the cap spacing, I also refined the design a bit and made the board smaller and added more accurate information on the resistances for different supply voltages following the testing I did with this board.

[@andymouse]

Cheers.

[@SuperBrainAK]

Just apply a constant voltage into the base of the transistor, regulated voltage on the input, then a potentiometer as a voltage divider, resistor as a current limit, then into the base of the NPN transistor. The Emitter is still going through a low value resistor to give it a voltage drop. whatever the base voltage is equals the current through the collector of the transistor (minus the base current and voltage drop of the transistor junction). I have actually covered that kind of super simple ciruit in my videos for driving LED's (expecially LED backlights). That way you can wind the current down to nothing and have very good control of the backlight brightness. I have been putting them on all of my displays so when I record a KZbin video the camera can pick up the readout better. I wish you the best of luck on your project and hope you can use my idea so you never complain that an LCD is washed out on camera! Cheers!

[@TheDefpom]

I had considered making a variable version, but this is just going to be part of a larger project which wants a fixed value.

[@SuperBrainAK]

@@TheDefpom yea the circuit I described would be great for other applications. But they are very similar in principle

[@helmuthschultes9243]

Noted the linked pcb design has current set resistors vertical, while your assembly has horizontal placement. But in general looking at that pcb as displayed on the link the layout is OK as far as TL431 pinout, matching to my data on hand and supply of these devices (have over 1000 in stock).

[@jrnandreassen3338]

This is the very first video with DPE I've ever seen, and I hit subscribe almost immedeatly. You explain things in a way even a layman can understand, and belive me: That is highly appreciated. Now where is that "recommend to rest of the world button? Hm!

[@TheDefpom]

Welcome to the channel and thanks for your feedback!

[@simonbaxter8001]

A note on your comment about holding the component legs that you snip off ... I've seen someone take one in the eye! and it stuck in. It wasn't pretty and he was lucky it didn't cause long term damage. The shorting other gear on you bench is top notch advice too. As for the TL431, check the Eagle footprint, some datasheets show TO-92 pinouts as viewed from the pins upwards. I've seen many a footprint reversed because it's been allocated the pin order looking down on the device. I always double check every library component I use and mark it as 'verified'. Been caught out too many times in the past.

[@TheDefpom]

The foot print is OK, I already checked the datasheet etc. which is why I struggled to figure out why it didn't work like it should, and resorted to experimenting, it is definitely the device in this case, oddly!

[@Rob_III]

@@TheDefpom I got a bunch of these TL431A's, mine are labeled "WS TL431A 633SB." but I haven't gotten around to test them. What are yours labeled?

[@TheDefpom]

@@Rob_III These are labelled "TL431 1308C TL"

[@Rob_III]

@@TheDefpom Hmm, thanks! I'll try to test mine soon and see if they're 'swapped' too - I'll let you know.

[@bitsurfer0101]

What is the accuracy of your multimeter? Did you get it calibrated?

[@bitsurfer0101]

What is the max current you could generate if you had a decent heatsink on the transistor?

[@briansauk6837]

Curious as to why not use an LM317 - fewer parts and very straightforward.

[@TheDefpom]

I almost did, but wasn’t sure if it would have the same precision as a proper voltage reference, plus I wanted to build something using the TL431.

[@briansauk6837]

@@TheDefpom Yeah, temperature spec for 317 voltage is pretty mediocre, and it requires 3V min input / output differential, which would not work well with your proposed li ion battery, since that is 3V on top of the 1.25V reference. So your design is definitely better given the constraints. I should have checked the specs first…

[@TheEmbeddedHobbyist]

Technically his is a constant current sink, as it defines the current being sunk to ground. A constant current source would provide a constant current to your circuit from your positive power rail. Nice to see it work in the end, would have been better to show a circuit diagram.

[@TheDefpom]

I debated the source vs sink definition, but as power is routed through the board I went with the source definition.

[@TheDefpom]

The diagram is on the PCBway project page, of my rev1.1 which has some minor improvements, electrically it’s identical.

[@TheEmbeddedHobbyist]

@@TheDefpom I just based the comment on where the control element was. it was between your external circuit and the ground so i went for sink. also your circuit will still work if the external circuit was powered from a different source as long as the grounds were tied and the voltage did not exceed the VCE of the transistor.

[@TheEmbeddedHobbyist]

@@TheDefpom Sorry if I missed that in the video, running on 1 hr sleep today. had to do a 230mile round trip during the night to drop somebody off at the airport.

[@TheDefpom]

@@TheEmbeddedHobbyist yeah thinking about it that way you would be right, damnit.

[@radiotvrepair1059]

Thanks for this video. I downloaded the gierbers file but the files won't open

[@andymouse]

Interesting stuff so what was wrong, the silk screen, the circuit or the device please ?....cheers.

[@TheDefpom]

The device!

[@andymouse]

@@TheDefpom I see below it comes in a couple of pin out variants thanks to this I am aware now...cheers.

[@stevenbliss989]

It is a current SINK, not a source, ...sorry to nitpick, but I needed a controllable CS, and was hoping you might have one I could adapt. Nice board, but the TL431 pinout is a TRAP, ...WOW!!!! I better check the ones I have in stock, both to-92 & SMDs.

[@TheDefpom]

I have renamed it to be a sink now, as that is more correct on reflection. I actually think this could be modified to be a source quite easily, if the +current and -current connections are linked together and the 0V connection is then used as the -current connection, this could be used as a + side current limiter, I haven't tried it but I think it could work, it would also mean it would then only require 2 connections instead of 4, where the + supply passes through it via the +3.5-24V and the 0V (sudo -current) connections, the down side is then the supply is powering the circuit too, so its own current consumption would effect the current, so a fixed voltage would be required to be used to keep that contant.

[@stevenbliss989]

@@TheDefpom yes, re-arange a bit, use a PNP and your done :)

[@helmuthschultes9243]

67 ohm? Measured? Standard series value is 68 ohm. Or using precision (0.1%) very low temp coeftype, does not look like it. Yep just reviewed video, on the assembly looks Blue Grey Black Silver (x0.1) so 68 ohm! Not really of any issue as easily covered by normal tolerance range anyhow. The tolerance band does appear dark, so not sure about that.

[@TheDefpom]

Yeah it measured 67… at least it did when I setup my parts drawers many years ago! It was a work around for not having the 51 ohm I wanted… those have now been ordered !

[@helmuthschultes9243]

That reversed pinout was unexpected and pure bad luck. I have not checked if there are standard variants. One possibility is some Asian supplier/manufacturer who is not conforming to correct TL431 pinout.

[@TheDefpom]

There is a TL432 which would match this pinout, so maybe it is a 432 and not a 431!

[@IanScottJohnston]

There is a mirrored pinout of the TL431 T092 package available from some manufacturers, but not all. The TL432 is smd only if i remember correctly.......could be wrong.

[@helmuthschultes9243]

@@TheDefpom Had a look at TL432data from Texas Instruments, the TL432 is not even available as TO92 style. Interesting is that they have a new type with " LI " suffix with improved characteristics. Other suppliers may have had a TL432 in TO92, but that should be clear by label on the device if it is a TL432 that you have. Most probable still is a sloppy asian source with non standard or bad batch bondout. Just the kind of parts sold via cheap Ebay type retailers.