Regulator heatsink

[Nick]

Yes, that was a typo, it should have been ESR, my bad. My thought was to look at the data sheet for the component in question. Its always usefull to know the spec's for many of the parts you need, they sheets are not hard to understand if you get into the hang of doing so. RS or Farnell web site is a good source of those sorts of references. And IMHO, getting used to checking such things, make the process of designing stuff more about design and less a painting by numbers sort of process.

Given the VA rating of the TX and the ESR of the cap, it should be possible to use PSD or paper to calculate the voltage drops and sort of peak current you will need the components to handle.

A 50VA TX, means that each secondary will be happy supplying 2.77 (25/9) amps into a resistive load. So the 1.5A load should be ok, (1.5*1.61 = 2.45A).

If we knew the regulaton of the TX we could get closer to the actual voltage. Assuming its 10% (And I can't remember which is the british or american way of calculating this Paul?), then 9*0.1*(2.77-2.54)/2.54 means the voltage into the bridge should be just under 9.1v so near enough to the rated value to not be worth worrying about.

[Andrew]

Andrew:

First time out I did the calcs

Oh, good, in that case, does my number make sense?

Yes, well I was with you, and agreed right up until the C/W calcs. Have a break whille I go through this. You have.....

Junction temperature will be equal to:

TJ = TA + PD (Q[HEAT SINK] + Q[CASE-TO-HEAT SINK] + Q[JC])

where Q[CASE-TO-HEAT SINK] is usually about 0.2 C/W

and you have Q[JC] as 4 C/W (seems high to me, is that the power section/control section or combined?)

Anyway,

TJ = 40°C + 6W (7.8°C/W + 0.2°C/W + 4°C/W) = 40°C + (6W * 12C/W) = 40°C + 72°C = 112°C

So yes, OK, assuming the chassis don't get above 50°C and the thermal bond to the angle iron is OK.

cheers,

-- Andrew

[Nick]

I took it from here

http://www.ortodoxism.ro/datasheets/nat ... 009060.PDF

Which gives it as 4 degC/W for the T package.

[Dave the bass]

Pah! Same results as mine except I have a cut finger

You Scientists

DTB

[pre65]

Just checked my parts box and i have 16 1N5820 which seem to be 20V 3A schottky diodes-will those be better ?

Hi-will these schottky diodes be man enough ?

[Nick]

Try simulating in PSD, in a bridge, maybe, the max reverse voltage they should see is 12v, full wave they would be too low voltage.

I use them in my 300b supply and they seem ok, so you should be ok, but the peak current will be far more than 3amp I suspect.

[pre65]

Hi-my PSU designer wont accept 1N5820 so i have got some others coming,

IR50SQ100 100V @5A.

Might as well try and get it done proper like.

[Dave the bass]

Yeah, I've been playing with Psud too, It doesn't have any UF4*** series either, is there an apdate or summink we don't know about or do we just use the 1N**** equiv's?

DTB

[Nick]

The INs should be close enough.

[simon]

What an excellent thread! I've been thinking about regulating the phono heaters but hadn't got my head round the maths fully. Thanks to Nick and Andrew for their input here. I've worked my way through it and knocked a spreadsheet up to model it - I find it easier to understand what's going on this way. Also has the added bonus that when I've forgotten it all this time next week when my head is full of work rubbish I can revisit the spreadsheet . The sense resistor bit is done, and most of the TX part. Just the heatsink calcs to do. I'll post it if anyone's interested, when it's finished.

So the 1.5A load should be ok, (1.5*1.61 = 2.45A).

I got a bit lost with the VA rating . Guessing wildly, should the 1.61 be 1.41 (root 2)?

then 9*0.1*(2.77-2.54)/2.54

Is 2.54 a typo - should it be 2.45?

[Nick]

I got a bit lost with the VA rating Confused. Guessing wildly, should the 1.61 be 1.41 (root 2)?

http://www.sowter.co.uk/htspec.htm

Is 2.54 a typo - should it be 2.45?

Yes, sorry.

[simon]

Ah, thanks Nick. I need to check another section of MJ out.

[Dave the bass]

Maths...Pah!

If you use my technique Simon wear gardening gloves to save cutting your finger, that way you wont need to file the burr off too

DTBodger

[simon]

I like your way Dave, there's a lot to be said for it. Unfortunately my instinct isn't that good, and my parts bin very small indeed. Using the spreadsheet I've just found out that the 50VA tx which I thought was plenty big enough is actually too small. Still, it'll be handy for something else.

[Nick]

I need to find where the 1.61 number comes from. I am not convinced its a exact value like 1.414 and 0.9 more a approximation depending on the amount of load current at the cap size.