TL750M08 - regulated power circuit...HELP

[Robert Reed]

OK, Dale brought my attention to the reason for a 12 volt supply. Funny how one can read thru these things. So the circuit should work OK as explained. Whether 250 ma or 500 ma-a dropping resistor will help and If the load is >250 ma you may be able to get by with just a clip-on fin style of heat sink. If it is <250 ma probably no heat sink will be needed. Lower load current will require a higher value dropping resistor and the value is pretty linear from that 12volt/4 ohm value as stated. I guess its your choice as to what components you have at hand. However low power resistors (2-5 W) are cheaper than heat sinks if you have to purchase them
As far as cap voltage rating vs. applied voltage, 'Lytics have a voltage coefficient that will give slightly reduced capacitance with lower applied voltage. But for your circuit the difference is insignificant.

[bob332]

thanks again for all the assistance. as far as heatsink vs 2-5W resistor, it really doesn't matter to me as i need to acquire either one which is no problem. is one version 'better' than the other? would the resistor have any issue if it is 140F ambient temps? should the voltage regulator have a heatsink regardless seeing it will be in up to 140F ambient temps? whichever you guys think would be the best i will go with - hoping after this is project is put in place it will just run and be forgotten about, so it will be a 24/7/365 item, at least that is the plan

[Robert Reed]

Any components wattage rating will have to be derated after a certain ambient temperature is reached. Wow-140 degrees ambient! That will take one huge heat sink to keep the regs internal junctions down even inspite of eating up some wattage with a preceding power resistor. If you want to scrub off a couple of watts with a preceding resistor, you may want to go with a 10 watt rating at that ambient. But why do these devices have to operate in that kind of heat. Could they not be in a remote location?

[bob332]

Any components wattage rating will have to be derated after a certain ambient temperature is reached. Wow-140 degrees ambient! That will take one huge heat sink to keep the regs internal junctions down even inspite of eating up some wattage with a preceding power resistor. If you want to scrub off a couple of watts with a preceding resistor, you may want to go with a 10 watt rating at that ambient. But why do these devices have to operate in that kind of heat. Could they not be in a remote location?

these are actually in a remote location, but i live in phx, az and this will be outside, so there is no way around it. 140F may be just a bit high but not much as i can guarantee 120F ambient no problem - my front door faces west and in the hottest parts of the summer it will hit 160F surface temp when the official temp is 118F or so. since this will be encased in box that will be either in direct sunlight or a possibly double encasement, it is still going to be very hot. i am trying to make it for the worst it will see and i know 120F+ ambient is very possible as i have lived through that. i should note that the max temps will only be for a couple hours a day for probably a month straight, don't know if that matters or not, so it will not be in 140F temps 24/7/365, the extreme temps will probably be more like 5/7/30 or so. i just don't want this project to fail during the heat of the summer.

i had thought about some type of fan,but figured it would be just another point of failure as i don't know how well they deal w/ that kind of heat.

[Bob Scott]

The 1K8 crap started in either the UK or Europe and seems to be slowly spreading our way.

Robert,

Yes that "1K8 crap" is a European convention, just as their schematics have boxes instead of zig-zags for resistors. I first saw this used by Philips on resistors when I was a kid back in '62. I'm sure the convention is older. I think it is a good way to get rid of one character, the decimal point, which can be hard to see on copied documents. Yup, I think it's cool.

Everybody,

Nobody has mentioned that a variable regulator can be made adjustable quite easily. See this document: Figure 2 on Page 8:

http://focus.ti.com/lit/ds/symlink/ua7805.pdf

That and the LM317 (?) series.
All those are so old and have high dropout. I like those newer low dropout types (LM750) for low voltage applications. Less power will be wasted.

[Einar M]

For enclosed spaces reducing total power is desirable. I would recomend a high efficiency 5V switcher and bypass the on board reg. As for Al caps, higher voltage usually means better performance: lower ESR and leakage. Your situation calls for the high temp versions.

[Einar M]

One further point; you should check the specs on the arduinlo. This type of board is rarely designed for high temps; it's not likely to last long.

[bob332]

would i be better off going w/ a different type of cap all together for the drop from 12->8V w/ the voltage regulator? or do you mean the caps that are rated for 105C? any suggestions, i am open to them all. if staying w/ Al, 25V? 50V? 100V? all 105C rated?

as far as the arduino, i have looked into it and haven't found concrete temp #s, but i figure after the initial unit is done and all bugs are worked out w/ the programming, if need be i could always make my own w/ higher quality parts for the temps i am needing.

i think on this first one i need to incorporate a temp log into it just so i can see exactly what temps the unit sees. it wouldn't even need time of day, just grab the temp every 10mins or so, that shouldn't be too hard to incorporate.

fwiw, here is the schematic i was planning to build this off of - don't know if makes a difference or not:



thanks for all the advice

edit 1: glad i checked here as i was just getting ready to order some parts...will hold off a bit.

edit 2: is there any issue ordering Al type for the prototyping/breadboarding, getting everything running and then just changing to a different type of cap be an issue? to be honest, i would rather just make up a board in eagle using 0805 size components as i enjoy smd soldering anyway and can usually get the parts cheaper going that route too. plus i could probably make the board just a couple postal stamps size and have plenty of room unless there is some issue w/ this.

[Robert Reed]

Should not matter what type of 'Lytic you use in prototype, only as Eimar mentioned go with the 105 degrees in finished version. You don't need a 'Lytic to the input terminal as long it's supply has even moderate smoothing, but you do need a ceramic of about 0.33mf right at the input terminal (when using long leads as you mentioned). 10 mf will be sufficient on the output terminal and keep the 0.1 mf in place there to. These regulators can react super quick for load changes and therefore require only minimal filtering on their output terminal.Also, I see no benefit in >25 volt ratings on your 'Lytics.One other thing that may help in the finished version is a shiny aluminum heat reflector on the sunny side of your enclosure.

[bob332]

thanks for the info, went ahead and ordered earlier today - was hoping to get them sat since i chose priority but ordered a bit past the cut off times. ended up going w/ 50V 105C electroylytics and did a 50V 85C .1uF on accident. figured this is for prototyping so it will be ok. need to order a temp sensor to wire up the the arduino so it can keep track of the temp.

other than speed of assembly and ease of board design, do the smd versions of these items offer anything else? in theory should they put off less heat themselves just do to surface area of the components? i figure i can use part of the pcb copper for the heatsink for the tl750m08, both part of the top and maybe some vias to connect it to the back or is that wishful thinking to use a via to transmit heat?

thanks again for all the assistance you guys have given, it has helped tremendously,
bob