Fuses and transformer, again...

[scaesic]

Before i test my power supply i want to check that i've got the right fuse sizes and in the right places.

At the moment i have a 300mA slo-blo fuse in position shown by fs1.

Now the supply is a bi-polar +/-20V. And the circuit shouldnt pull out anymore than 6A (the main juice is drawn by a tda2050 (2-5A ), the rest is drawn by a 4558 bax circuit and amz mosfet boost, so shouldnt be taking more than 60mA). So thats 240W of power?

I cant find the resitance of the primary coil but guess it has a dc resistance of around 20 ohms. So the continuous current at the primary should be about 1A, but could see inrush currents on switch on of about 17A - {(root(2)*230)/20} - is this right?

heres the trafo im using - (15+15 120VA http://www.rapidonline.com/productinfo.aspx?kw=88-2582&tier1=Electronic+Components&tier2=Transformers&tier3=Toroidal&tier4=Toroidal+transformers&moduleno=35506)

Re-working i can get 1.725A steady state on the secondary side if i use a 300mA fuse, which is what someone reccomended i use.

Will the tda2050 really use 5A?

in reality i think il stick with the 300mA sloblo and move it to 500mA if it keeps blowing, but i wanted to check.

[R.G.]

OK, you're making +/-20Vdc out of a (centertapped?? ) transformer to run a TDA2050. The datasheet for the 2050 says it will do something like 66% efficiency with something like 30W out at +/-22V supplies. So if 30W out is 66% of the total power from the power supply, the 2050 must be dissipating about 15W, and the DC being taken from the supply is an average of about 45W. Call it 50W for conservatism.

The power supply is 40V, and it's supplying 50W, so on average it's providing 1.25Adc at this high power level.

It's a quirk of full wave rectifiers that all the power comes in as big spikes of current near the peak of the AC power wave, so the RMS value of the current flowing in the transformer secondary is much higher than the DC average out. In most cases, it's 1.6 to 1.8 times as much, so the RMS current in the transformer secondary could be as much as 1.25A*1.8 = 2.25Arms. Fuse 2 will likely pop if it's less than 2.5A and you play loudly for long periods of time or test it into a dummy load at full power.

The mains current is the inverse of the voltage ratios. For 30Vct that means that for 120Vac, the primary RMS is at least (30/120)*2.25 = 0.563Arms. For 240Vac, it will be half that 0.281A. That's a continuous current rating, so if you put a 0.5A fuse in for 120vac or 0.3A for 240Vac primaries, the fuse will pop on sustained full power runs.

You need a primary fuse of at least 0.75A for a 120V primary and 0.5A for 240Vac. I personally would use 1A and 0.75A respectively, both slow blow because toroidal transformers pull BIG startup transients.

Remember: F1 is to prevent FIRES, not to protect circuit parts. F2 is to protect the transformer, not anything else. F2 probably should be 2.5A.

[scaesic]

OK, you're making +/-20Vdc out of a (centertapped?? ) transformer to run a TDA2050. The datasheet for the 2050 says it will do something like 66% efficiency with something like 30W out at +/-22V supplies. So if 30W out is 66% of the total power from the power supply, the 2050 must be dissipating about 15W, and the DC being taken from the supply is an average of about 45W. Call it 50W for conservatism.

The power supply is 40V, and it's supplying 50W, so on average it's providing 1.25Adc at this high power level.

It's a quirk of full wave rectifiers that all the power comes in as big spikes of current near the peak of the AC power wave, so the RMS value of the current flowing in the transformer secondary is much higher than the DC average out. In most cases, it's 1.6 to 1.8 times as much, so the RMS current in the transformer secondary could be as much as 1.25A*1.8 = 2.25Arms. Fuse 2 will likely pop if it's less than 2.5A and you play loudly for long periods of time or test it into a dummy load at full power.

The mains current is the inverse of the voltage ratios. For 30Vct that means that for 120Vac, the primary RMS is at least (30/120)*2.25 = 0.563Arms. For 240Vac, it will be half that 0.281A. That's a continuous current rating, so if you put a 0.5A fuse in for 120vac or 0.3A for 240Vac primaries, the fuse will pop on sustained full power runs.

You need a primary fuse of at least 0.75A for a 120V primary and 0.5A for 240Vac. I personally would use 1A and 0.75A respectively, both slow blow because toroidal transformers pull BIG startup transients.

Remember: F1 is to prevent FIRES, not to protect circuit parts. F2 is to protect the transformer, not anything else. F2 probably should be 2.5A.

AHHHH! took me ages to figure out where your 120Vac was coming from!

i'm in the UK, i think its nearer 230Vac, so 0.5A fuse could have been enough? I think i'll start low and move up if popping becomes a nuisance because of the large inrush currents, better safe than sorry. Originally i wasnt going to include f2. but i think i might considering the hassle and money it'd be to replace the toroidal.

This was really useful btw. Maybe you should add it to the geofx power supplies section under a fusing heading?

So tentatively 0.5A slow blow for f1, and move up to 0.75A if it becomes a problem. and 2.5A for f2 (i had no idea about that 1.8 factor...). Does the transformer act as a kind of buffer to prevent the large transients getting to the bridge rect? i.e f2 doesnt need to be a slo blow?

OK, you're making +/-20Vdc out of a (centertapped?? )

yep, secondary ct'd

Im using the raw power supply from the caps to power tjhe tda, and i have a couple of 15V rectifiers (+/-) to power the pre-amp and tonestack. Just hope the tda doesnt get noisey as a result, else il have to buy a bigger transformer and a couple more rectifiers.

[R.G.]

I think the 2050 is OK. These things are designed to ignore power supply ripple; that's the "power supply ripple rejection" number in the datasheet.