transformers, current, voltage?

[Volter]

Hello Bob, no, it is not homework, at my age I could care less about education, waste of time and money! It is a bit of my idea that I would like to build (design and build) and see if that could work and maybe turn it into something … !?
Very likely I might not succeed in what I am trying to accomplish but on the other side I am relearning what I should keep going in. Hopefully something else might come out of it.
It is interesting how one gets lost in those spec sheets and not find the right one that would give the right results.
dyarker bought me into that LM317HVT even I already ordered the LM5118, now dyarker has to convince me that the 36Vdc (rms) is better then the 48Vdc (rms) transformer. The output requires 40Vdc regulated with 700mA! dyarker is the genius here!
Thanks.

[CeaSaR]

I do believe that Bigglez has hit the proverbial "Nail on the head". It would be helpful to
all involved in helping you if more of the project were revealed other than just '40 Volts
at 700 ma'.

No need to reveal whatever secret design, just a bit of info on what is going to need that
'40V @ 700ma'. It may be that the needs are over / under estimated.

CeaSaR

[Volter]

You guys you're experts who knows all the tricks and ideas. The black box, 40V, 700mA, just does not fall into that category especially if I can manage the design of the load and cannot make up the supply part, just toooo simple! The complex part is this power management for the black box, well, complex for me not for you. The project IS the Power Supply!

[Bigglez]

The black box, 40V, 700mA
...The project IS the Power Supply!

So you have a 57.2 ohms, 28 Watts resistor?

[dyarker]

This is getting silly, so I'll try to make it simple for you.

Biglez(sic) the max dif is the same but you circled 117, 317 is further down the page

Really? Are you actually looking at the LM317 datasheet or the LM317AHV

This time READ what is inside the circle YOU drew.
Vin minus Vout less than or equal to 40V
For example:
80V in minus 40V out equals 40V, meets the spec.
67V in minus 40V out equals 27V, 27V is less than 40V, meets the spec.

[dyarker]

Anyway 40V out plus 40V differential is 80V, still enough.

The abs max from input to output, or input to adjust is 40V.
This is a process limit and can't be exceeded.

If the LM317 is used as a current source, with a load side
sense resistor feeding back a voltage proportional to current,
the output to adjust voltage will be Vref (1.2V typ).



In that special case the input can be 40V higher than the
output (of the LM317).

If we only knew what the OP is trying to do with this
project, it would have been completed by now...

Your circuit works too, but it's not special, nor is it what I discribed. I put the load (a series of LEDs) between the positive supply and Vin of the regulator IC. The supply voltage can be higher than the max of the IC (41.2V in this case) because of the forward drops of the LEDs. I admit my version would not be a good idea for loads that do not have a relatively fixed drop; for LEDs it is fine as long as there are enough LEDs.

[dyarker]

Volter,

Please excuse the two reading lessons.

Not trying to "sell" my idea, especially if you already ordered LM5118. In fact it looked interesting to me. In fact from the spec sheet, it looks like you can put less than 40V in and still get 40V out (boost mode). This would be useful for driving a 40V series string of LEDs in the car from 12V.

As for 36V transformer -

48VRMS * 1.414 = 67V peak (like you already calculated)
67V - 4V for diodes = 63V
63V - 40V load = 23V across LM317
23V * 0.7A = 16W to be dissapated by LM317

36VRMS * 1.414 = 50V
50V - 4V = 46V
46C - 40V = 6V across LM317
6V * 0.7A = 4.2W to be dissapated by LM317

result = a much smaller heatsink needed for LM317, and electric meter runs a tiny bit slower.
Both 48V and 36V calculations assume a filter capacitor after the diodes for less than 1V ripple.

Cheers,

[CeaSaR]

Well done Dyarker.

Now that's what I call a concise answer. I'm sure others on here would have posted an
extremely similar answer. Hopefully this explains why you can use a 36 V transformer to
obtain 40 V DC.

Let us know how things work out.

CeaSaR

[Volter]

edited
Sorry, I got carried away on this one. If it says 36V how it can support 40V? I guess only expert can know!
Now, where do you get these type of Transformers.
Thanks.

[Bigglez]

This time READ what is inside the circle YOU drew.

I drew that circle to point that the maximum
voltage allowed between any terminal to any other terminal on a 40V
process part is, well, forty volts.

The part can be floated above ground, which is not what the OP asked
for in his earlier posts.

[Bigglez]

If it says 36V how it can support 40V?

36V is RMS for a sinewave. 40V is DC after rectification.
RMS has the same energy as as DC, but root-2 times the
peak voltage. i.e 36 * 1.414 = 50.91 Volts.

The rectifier converts the positive and negative symmetrical
sine wave to positive only half-wave voltage. The smoothing
or reservoir capacitor charges to the peak and holds the
average voltage near the peak during the valleys of the
rectified AC voltage. This produces almost pure DC output.

The transformer current during the AC peaks must support the
load and the capacitor charging, so the RMS current in a
transformer is 1.8 times the DC load current.

Many hobby power supplies fail due to oversight on this
vital point. The secondary current causes heating in the
transformer, and the wire resistance causes sags in the
output DC.

A linear regulator bridges the gap from the desired DC
output and the lowest practical AC voltage from the
transformer. (Under low line input, max load current, and
worst case transformer losses).

A rule of thumb is to make the transformer RMS voltage
under load equal to the DC output voltage required and
use the regulator to 'waste' the rest as heat, while keeping
the output steady.

Now, where do you get these type of Transformers.

Any electronics component supplier. Digikey and Mouser both
offer excellent on-line service and competitive pricing.

[Volter]

I came up with 16700uF for the Cap!? Is that correct? Vr = 1 / C f That's: C = 1 / 1V x 60Hz = 16.7mF? Something is not right!
Another formula:
Vr = (0.0024s) x Iload / Cf That's: Cf = (0.0024 x .7) / 1V = 1,700uF? This is more like it should be but is it correct?
Thanks.

[Bigglez]

I came up with 16700uF for the Cap!? Is that correct?

You (and therefore us) are missing an important spec. What is the maximum
ripple voltage that you allow on the output.
Hint: The greater the ripple voltage the lower the capacitance,
the smaller the physical capacitor size, and the lower the cost.

[Volter]

Well, is it not the task of the Voltage Regulator to eliminate these farther ripples?
Bigglez, you say that the first formula is correct but there is no such Cap available in that size!?
As it was stated earlier it is to drive LEDs and they are capable of handling min and max voltage and so the 1V ripple is just fine.
Thanks.

[dyarker]

This time READ what is inside the circle YOU drew.

I drew that circle to point that the maximum
voltage allowed between any terminal to any other terminal on a 40V
process part is, well, forty volts.

The part can be floated above ground, which is not what the OP asked
for in his earlier posts.

An LM317 is ALWAYS floating, and spec sheet says DIFFERENCE BETWEEN Vin and Vout it does NOT say any pin.

From top section of spec sheet YOU used:
"Since the regulator is floating and sees only the input-to-output differential voltage, supplies of several hundred volts can be regulated as long as the maximum input to output differential is not exceeded, i.e. avoid short circuiting the output."

[end reading leason number three, if it still doesn't sink in it's hopeless]

Volter said he was rectifying with a diode bridge, and also said drop was over 4V, therefore he's using a full wave bridge. That means the frequency when calculating the capacitor is 120HZ, not 60Hz.

Volter, have fun with your project. Tried to straighten this out gracefully on page 2, but he wouldn't let go, sorry.