power supply questions

[MrAl]

Hi,

I hate to say this, but it doesnt really look that big. You can still use it, but im not sure if you can pull 10 amps from it. Can you check the wire size of the secondary? That might tell us. You should probably test it at lower currents first and see if it overheats after several hours like overnight. A transformer that overheats melts the insulation and shorts out, blowing the fuse and then becomes useless except as a paper weight.

Also, if we measure the capacitance in ampere seconds per volt, at 50Hz the required capacitance is estimated to be C=0.01*I/V where V is the estimated desired ripple voltage. At 60Hz we could use a little less, 5/6 times that.

 

[audioguru]

The transformer with a 28.4VAC secondary (at full load?) produces a peak voltage of 40.2V. The loaded full-wave bridge rectifier drops 2V so the input to the regulator is 38.2V minus some ripple voltage. With a 5V/10A output then the power transistors must dissipate 332W which is a lot of heat to get rid of. You will need at least 5 power transistors each with an emitter resistor. You also need a fan to cool the huge heatsink.

 

[MrAl]

Hi audio,

Yeah haha, that's quite a bit of heat isnt it. I tend to believe that the intended transformer will not be able to handle this kind of current for long anyway, based on the picture posted previously. Looks like a 2 amp transformer or something like that.

 

[ChrisP58]

A 24VAC transformer for a furnace is probably only rated at 20 to 40 VA. After running it through a rectifier and filter cap, you'll be lucky to get even one amp DC.

What does your transformer weigh? A 60 Hz transformer big enough to handle the power you want will weigh about 8 pounds or more.

 

[bountyhunter]

here is the transformer i am using and i was looking for 5-30v ///
My grandpa got this for me from a furnace its sec is 28.4v and its pretty big so i suppose the current is over 10a

So you want a 300 Watt linear supply.

FYI, the RMS transformer current (before the FWB rectifier) is about 1.8X the DC load current. So you would need a 600 VA transformer to produce 300 Watts.

Wait until you see the size and cost of one.


Here is a brand new 30V/10A supply for $169:

**broken link removed**

 

[bountyhunter]

The transformer with a 28.4VAC secondary (at full load?) produces a peak voltage of 40.2V. The loaded full-wave bridge rectifier drops 2V so the input to the regulator is 38.2V minus some ripple voltage. With a 5V/10A output then the power transistors must dissipate 332W which is a lot of heat to get rid of. You will need at least 5 power transistors each with an emitter resistor. You also need a fan to cool the huge heatsink.

That's why you don't see 300W linear supplies. The first thing a designer learns is the distinction between possible and feasible.

 

[harold777]

I think I mentioned HEAT - and that transformer is too high a voltage (low-load). To replace it with something else is probably the best soultion.....
However, can OP tell us what this supply is for? What voltage range will do and why 10A - that's strikes me as being a "Spec" rather than an actual need.


To check a transformer, put it on a non-flam surface with fused input, then apply supply. Check temperture after 1hr, 2hr etc. It should settle about 50-60 Deg C.
Now build a load using a range of high wattage R's (these need to be mounted securly where they can keep cool, ideally wirewound). Check current by measuring the Resistance (when cold and later when hot). Measure V. Apply Ohm's Law.
Measure temperature every 15 min....this should give yu confidence to continue, if not STOP.


Given that OP can load the transformer at 10A this test will tell you
1) if it can supply the right OP voltage and 10A
2) that the temperature rise is withing reasonable limits.

Note this test is AC - no rectifier is required - that saves a lot of extra issues. The measurements are rms so apply factors for the actual supply required.

If the transformer falls short doing this test, there is no way it will be satisfactory in service and OP should look for a better altenative.

 

[bountyhunter]

Given that OP can load the transformer at 10A this test will tell you
1) if it can supply the right OP voltage and 10A
2) that the temperature rise is withing reasonable limits.

No, no, and no: the AC RMS current the transformer can supply is NOT the DC load current.... in fact, the DC load current the output will be able to supply is only about 55% of the AC RMS rating of the Xfmr secondary if you use a FWB and capacitor to make the unregulated DC. The OP wants 10A @ 30V. As I said before: if you want 300W of load power, you need close to 600 VA transformer. In this case, you would need a transformer rated at about 32VAC at 18 ARMS (or more).

 

[dr pepper]

My post removed, someone allready said what I was gonna say.

 

[MrAl]

No, no, and no: the AC RMS current the transformer can supply is NOT the DC load current.... in fact, the DC load current the output will be able to supply is only about 55% of the AC RMS rating of the Xfmr secondary if you use a FWB and capacitor to make the unregulated DC. The OP wants 10A @ 30V. As I said before: if you want 300W of load power, you need close to 600 VA transformer. In this case, you would need a transformer rated at about 32VAC at 18 ARMS (or more).

Hi,

Just curious, where are you getting the theory for this conclusion from? It sounds like you are saying that if we want a 300 watt DC output then we have to supply 600 watts input power to the transformer, which then that means somewhere we loose 300 watts.
Maybe i didnt understand your post correctly.

Just for reference, a typical 24vac 10 amp 60Hz transformer has the steel laminations core of these dimensions:
3.75 inches wide
3.125 inches high
1.625 inches stack
That's measuring the steel only, not any part of the windings.

 

[sparky!(Sean)]

There is a two amp power adapter and the transformer on it is tiny while I have a 700 watt transformer from a microwave that is bigger but not by much than the transformer from the furnace

 

[bountyhunter]

Hi,

Just curious, where are you getting the theory for this conclusion from?

30 years of power supply design experience. The ratio of AC RMS current to DC load current is well known for both FWB and FWCT.

It sounds like you are saying that if we want a 300 watt DC output then we have to supply 600 watts input power to the transformer, which then that means somewhere we loose 300 watts.
Maybe i didnt understand your post correctly.

You don't understand the theory: you are mixing VA (apparent power) and Watts (true power) which means you are ignoring the effect of power factor.

Just for reference, a typical 24vac 10 amp 60Hz transformer has the steel laminations core of these dimensions:
3.75 inches wide
3.125 inches high
1.625 inches stack
That's measuring the steel only, not any part of the windings.

IMHO, that size 60 Hz transformer could not possibly supply the power this design requires. Which as I said before, would be ballpark 32VAC/18A RMS.

 

[MrAl]

Hello again Bounty,


So then you are saying that it is a 600VA input and 300 WATT output?

I know that if the power is measured on the input side and the power is measured on the DC side, the ratio is more like 82 percent for a typical DC power supply that i've actually measured. There is loss in the diodes, the windings, or capacitor ESR, etc. 300 watt doesnt just disappear so im sure you are saying something and meaning something else here.

The transformer i gave the dimensions for was simply a reference to what a 10 amp 24v transformer would look like in size, which has nothing to do with what you are talking about. He said he had a 24v 10 amp transformer and i dont think it is really big enough to do that. Whether or not it can supply enough power to some DC power supply that has to meet some output requirement is another story altogether. It's just a transformer with known rating that has been measured by someone somewhere at some point in time.

 

[bountyhunter]

Hello again Bounty,


So then you are saying that it is a 600VA input and 300 WATT output?

correct for rough estimate

I know that if the power is measured on the input side and the power is measured on the DC side, the ratio is more like 82 percent for a typical DC power supply that i've actually measured. There is loss in the diodes, the windings, or capacitor ESR, etc. 300 watt doesnt just disappear so im sure you are saying something and meaning something else here.

If you have a true power efficiency of 82%, that is probably a switching power supply. You are building a linear supply here. Understand: the unregulated DC supply for a 30V output has to be about 40V to account for filter cap ripple, headroom for the regulator and rectifier diode drops. So, if the output is 20V, the efficiency is roughly 50% since about half the power is burned in the pass transistors (assuming we ignore diode power)..

In actuality, didoe power for this design would be high since each diode would have at least a volt forward drop and 18A through it at rated output. Diodes would be burning at least 40W.

The transformer i gave the dimensions for was simply a reference to what a 10 amp 24v transformer would look like in size, which has nothing to do with what you are talking about. He said he had a 24v 10 amp transformer and i dont think it is really big enough to do that. Whether or not it can supply enough power to some DC power supply that has to meet some output requirement is another story altogether. It's just a transformer with known rating that has been measured by someone somewhere at some point in time.

OK, just make sure if you buy a transformer for the design you understand what the specs mean. RMS current rating of the transformer is not the same as the DC load current.

 

[The Electrician]

Here's some pertinent information:

https://www.electro-tech-online.com/custompdfs/2012/06/Pg_006_7.pdf

 

[bountyhunter]

Good overview. It also says what I had posted earlier: RMS current in transformer secondary is about 1.8 times DC load current with FWB capacitive filter. Hence, you need at least 18 ARMS transfprmer to do a 10A DC power supply.

 

[ChrisP58]

The 1.8 multiplier is a commonly accepted scaling factor, but the true number needs to be determined on a circuit-by-circuit basis.

The attached LTSpice simulation is a typical 24VAC transformer circuit. It has a schottky fullwave bridge, 1000uF filter cap, and is loaded with a 1 amp current sink.

The waveforms to watch are: node A, node Vout, and the current of V1.

Even though the load current is one amp, the transformer RMS current is 2.2 amps. But the current peaks are over 6 amps.

The actual transformer current to load current ratio is largely a function of load current and filter capacitance. But other circuit parasitics will also play a part.

 

[MrAl]

The 1.8 multiplier is a commonly accepted scaling factor, but the true number needs to be determined on a circuit-by-circuit basis.

The attached LTSpice simulation is a typical 24VAC transformer circuit. It has a schottky fullwave bridge, 1000uF filter cap, and is loaded with a 1 amp current sink.

The waveforms to watch are: node A, node Vout, and the current of V1.

Even though the load current is one amp, the transformer RMS current is 2.2 amps. But the current peaks are over 6 amps.

The actual transformer current to load current ratio is largely a function of load current and filter capacitance. But other circuit parasitics will also play a part.

Hi Chris,


Now that i understand what this discussion is all about (little chuckle there if you dont mind) I have to agree completely. The 1.8 scaling factor should probably be taken as a "lower limit" de facto standard rather than a cut and dry calculation or measurement.

The main reason being that the current peak in the secondary is determined by the diode conduction angle, and the things that affect the conduction angle are many such as transformer resistance, capacitor value, capacitor ESR, line resistance, leakage inductance, etc.
For example, as the capacitor ESR goes down the diode peak current goes up, which means the power being delivered is being done over a shorter time, and with higher peak current the RMS current goes up even though the same amount of power is being delivered.
Likewise, for low transformer resistance we'll see the same thing. Thus, i wouldnt be surprised to see a factor of 2.4 in some cases.
However, with a transformer that is well matched to the load and some reasonable cap ESR and reasonable cap values i'd think a factor of 2 would be a good estimate to start with, taken along with the reservation that it could be somewhat higher.

 

[harold777]

This thread is getting very techy . . . but fundamentally it shows the difficulty of building a 10A PSU using a linear circuit.
Has OP come to any conclusions - or is this project in the bin, for now?

 

[sparky!(Sean)]

Ok and another question anyone know a reliable circuit for a 7 segment led volt meter