Thanks for showing the ESP32. I have some esp32-cams sitting around I plucked one off from until I get some in from shyna express. Now I can use old 9v batteries to power the little PCB STS discovery shuttle thing I got off adafruit. Made by phyx
I prefer the ones with lower quiescent current, they work well to convert a lithium 18650 cell to be used to replace 2 AA cells, as the low quiescent current means months of operation between charges.
Just reading the datasheet and minimal math: SOT223 terminal resistance w/o any additional copper: 61deg/W 5V input to 3.3V output is 1.7V to turn into heat or 61*1.7deg/A as the junction temperature temperature above ambient. TI allows up to 125C on the die, so assuming 45C ambient (non-ideal situation) we have 80deg to spend. 500mA are within specs, but I would not go that far. Especially if it should work for a longer time.
Yes, to get these anywhere near 800mA you need to do a lot. Keep the input voltage down to below 2V over the output plus give up about 1 sq. in. of copper on both sides of the board with proper thermal vias between and ensure adequate airflow. It gets expensive fast.
Not sure if Dave would consider it jellybean already, but it's so popular you find one or a few almost everywhere. I like the separate sense line on your power supply cables there. And your Erem side cutters :).
Chenese products are more often AMS prefixed, usually can't withstand declared max input voltage, and their datasheets (in case they are bought from reliable international parts suppliers, which Alliexpress isn't) do not provide reliable data to properly calculate power dissipation over layout, and also provided electrical characteristics are possible only for TO-252 package on propperly designed layout, not mentioned in datasheets.
@AnalogueGround Have you seen manufacturers' datasheets download link on these sellers alli pages like on digikey, element14 subsidiaries, chinese lcsc for the sake of not being westerncentric, etc.?
@@deepblueskyshine No, but once I’ve established that a seller is selling genuine components I can download a data sheet from the manufacturer to check the spec. If a corporate customer requires certification of conformity then I’ll order from RS or Farnell and they pay more for parts. Aliexpress gets a bad rap as some of its sellers are passing on re-marked or even fake parts but it is only a few sellers and the price usually gives you clue. If you choose a seller wisely and even do a small test purchase to establish their integrity, there are some real bargains to be had especially when bulk purchasing.
I recently did consider using this part on a Commodore 64 cartridge that I am developing. I have rejected it, because C64 power supplies often deliver below 5V on the mainboard, 4.8V is quite typical and due to load on the mainboard, when your power arrives at your IC on the cartridge, it has dropped even more. I consider it desirable that my cartridge works correctly for inputs as low as 4.5V and then the dropout of the LM1117 becomes tight. Therefore I have decided not to use the LM1117 and am going for an MCP1826T-3.3, which has quite a bit better dropout voltage.
.... but the MCP Part is also more expensive. I think it does not care a lot for something for a C64. This is built more for a small group of customers who are happy to pay a few euros more if the components are also of premium quality. A product for the mass market is a different thing, every cent counts... And especially with such simple components, you look for components that are offered by several manufacturers so that you can go around delivery problems quickly...
Any alternative recommendations for a device which doesn’t regulate voltage by waste fully shedding heat? These LM’s seem to act as a programmed resistor. Maybe a miniaturized buck?
They do make modules that are drop in replacements for 78xx to-220 linear regulators. Linear regulators are good for low noise. When you put a switcher in there you get more ripple noise on the output.
@@misterhat5823 definitely! I've used them for that before. SMPS set at the most efficient Vi-Vo. If it needs to be as clean as possible, I throw a C L C filter on the output with a resistor for a minimum load. You can tune these for to intended load if it's going to be consistent. Wrap both sides in separate metal cans if you want to make the fcc happy
@@misterhat5823 In this use, you will want to provide the full 3V of headroom from the SMPS to the LDO to obtain the best ripple rejection from the LDO. Using only the minimum dropout headroom will allow more noise through the LDO. The LDO and LM78xx data sheets are tricky on this point; they only state the minimum dropout, for marketing purposes. But, if you check the ripple rejection graph, you will see that they use 3V of headroom for testing to achieve the lowest ripple. HTH.
Interesting electronic load... I love the Kikusui PLZ152w that I have on my bench, and that Miniware looks like something fun to tinker with.
Thanks for showing the ESP32. I have some esp32-cams sitting around I plucked one off from until I get some in from shyna express. Now I can use old 9v batteries to power the little PCB STS discovery shuttle thing I got off adafruit. Made by phyx
I prefer the ones with lower quiescent current, they work well to convert a lithium 18650 cell to be used to replace 2 AA cells, as the low quiescent current means months of operation between charges.
Just reading the datasheet and minimal math:
SOT223 terminal resistance w/o any additional copper: 61deg/W
5V input to 3.3V output is 1.7V to turn into heat or 61*1.7deg/A as the junction temperature temperature above ambient.
TI allows up to 125C on the die, so assuming 45C ambient (non-ideal situation) we have 80deg to spend.
500mA are within specs, but I would not go that far. Especially if it should work for a longer time.
Yes, to get these anywhere near 800mA you need to do a lot. Keep the input voltage down to below 2V over the output plus give up about 1 sq. in. of copper on both sides of the board with proper thermal vias between and ensure adequate airflow. It gets expensive fast.
Not sure if Dave would consider it jellybean already, but it's so popular you find one or a few almost everywhere.
I like the separate sense line on your power supply cables there. And your Erem side cutters :).
Chenese products are more often AMS prefixed, usually can't withstand declared max input voltage, and their datasheets (in case they are bought from reliable international parts suppliers, which Alliexpress isn't) do not provide reliable data to properly calculate power dissipation over layout, and also provided electrical characteristics are possible only for TO-252 package on propperly designed layout, not mentioned in datasheets.
Aliexpress are an umbrella for a multitude of sellers and some sell extremely high quality components, some specialising in industry overstock.
@AnalogueGround Have you seen manufacturers' datasheets download link on these sellers alli pages like on digikey, element14 subsidiaries, chinese lcsc for the sake of not being westerncentric, etc.?
@@deepblueskyshine No, but once I’ve established that a seller is selling genuine components I can download a data sheet from the manufacturer to check the spec. If a corporate customer requires certification of conformity then I’ll order from RS or Farnell and they pay more for parts. Aliexpress gets a bad rap as some of its sellers are passing on re-marked or even fake parts but it is only a few sellers and the price usually gives you clue. If you choose a seller wisely and even do a small test purchase to establish their integrity, there are some real bargains to be had especially when bulk purchasing.
A device that can take good advantage of the availability of Aluminium and Copper cored PCBs....cheers
These are in damn near everything!
I recently did consider using this part on a Commodore 64 cartridge that I am developing. I have rejected it, because C64 power supplies often deliver below 5V on the mainboard, 4.8V is quite typical and due to load on the mainboard, when your power arrives at your IC on the cartridge, it has dropped even more. I consider it desirable that my cartridge works correctly for inputs as low as 4.5V and then the dropout of the LM1117 becomes tight. Therefore I have decided not to use the LM1117 and am going for an MCP1826T-3.3, which has quite a bit better dropout voltage.
.... but the MCP Part is also more expensive. I think it does not care a lot for something for a C64. This is built more for a small group of customers who are happy to pay a few euros more if the components are also of premium quality. A product for the mass market is a different thing, every cent counts... And especially with such simple components, you look for components that are offered by several manufacturers so that you can go around delivery problems quickly...
Any alternative recommendations for a device which doesn’t regulate voltage by waste fully shedding heat? These LM’s seem to act as a programmed resistor. Maybe a miniaturized buck?
They do make modules that are drop in replacements for 78xx to-220 linear regulators. Linear regulators are good for low noise. When you put a switcher in there you get more ripple noise on the output.
@@xenoxaos1 You can use a switcher followed by an LDO to help with noise. It won't be drop in though.
@@misterhat5823 definitely! I've used them for that before. SMPS set at the most efficient Vi-Vo. If it needs to be as clean as possible, I throw a C L C filter on the output with a resistor for a minimum load. You can tune these for to intended load if it's going to be consistent. Wrap both sides in separate metal cans if you want to make the fcc happy
Check Monolithic Power Sytems (MPS) for
switching converters in a small module
with built-in inductor.
@@misterhat5823 In this use, you will want to provide the full 3V of headroom from the SMPS to the LDO to obtain the best ripple rejection from the LDO. Using only the minimum dropout headroom will allow more noise through the LDO. The LDO and LM78xx data sheets are tricky on this point; they only state the minimum dropout, for marketing purposes. But, if you check the ripple rejection graph, you will see that they use 3V of headroom for testing to achieve the lowest ripple. HTH.
What is the part number or info on the dummy load you used? Thanks.
ua-cam.com/video/SoDIe6WD77k/v-deo.htmlsi=xQoIe2XXIlbJ6qCX
Next, 723 with negative zero, please.
ua-cam.com/video/tTMFh5xvmAM/v-deo.htmlsi=2S3E8lWlkGHQ5Fay
A 723 uses one's complement arithmetic? I loved having both +0 and -0 on my UNIVAC 1108 back in the day!
@@IMSAIGuy totally wrong answere, i want a 723 that can regulate from 0 volt and up. 0-30 volt 🤘
@@sveinfarstad3897 Well you don't want a 723 then do you, check the google term 'Datasheet'
www.electroniq.net/power-supply/lm723-0-30v-adjustable-power-supply.html
XT60 connector.
To not use that connector is just senseless. 🤪
It's not, it's the tiny little XT30.
Nearly, half that rated current
no, current in the sense lines
@@IMSAIGuy Not in this application but that's what the XT range are marketed as unless something has changed