LM338 Current Limiting

[imix500]

Hello all, so I just finished building this seemingly simple circuit. It works well enough but it seems to be current limiting at just over 3 amps.
There is minimal current sag before the regulator, so it's got plenty of gas.
The device is not over heating, barely gets warm at 3 amps. All connections are ok, rechecked today. Any ideas?
Here's the schematic:

[BWRX]

What output voltage do you have the circuit set to?

[imix500]

Output flattens out at 27.7v
Under lesser load it will output a full 32.5v

[peranders]

How about your input voltage, how much and how much ripple?

[AndrewT]

Hi,
regulation seems to have stopped being effective long before you get to 27V out.

At what current does the output start to drop from 32V5 out?

What is the input voltage when the output voltage just starts to drop?

Can your DMM be set to measure AC to check the ripple on the input to the regulator (while still having the 32V5 bias on the ripple signal)?

[Nigel Goodwin]

Originally posted by imix500
Output flattens out at 27.7v
Under lesser load it will output a full 32.5v

The input voltage is too low for it to regulate, that's why it doesn't work. C1 is also far too small, although a higher voltage transformer is probably required anyway (and a larger C1).

[imix500]

Ok guys, here is some more info:
Unloaded rail voltage is about 36.1v
The rail voltage at 27v out and current limiting is 32v

I set my dmm to ac and measured 70v ac at the 2200uf cap next to the input of the reg, also at the output. But this must be a false reading of harmonics and such right? (it's not a true rms meter)

Forgot to mention, the main filter cap is 18000uf. The 2200uf is just a smaller one located next to the reg.
Thanks!

[peranders]

The dropout voltage is approx. 3 volts for the regulator but how stiff is your unregulated voltage?

Transformer?? + 18000 uF + 22000 uF + LM338 = True?

Notice also when you have so heavy smoothing the VA rating of the transformer will be 50% due to all reactive losses.

[Nigel Goodwin]

Originally posted by imix500

I set my dmm to ac and measured 70v ac at the 2200uf cap next to the input of the reg, also at the output. But this must be a false reading of harmonics and such right? (it's not a true rms meter)

Try putting a capacitor in series with the meter lead (0.1uF should be fine) and try reading again, although it won't read accurately anyway - but many meters read DC on the AC ranges as they don't have internal blocking capacitors.

[imix500]

Peranders, ah yes I always forget about that pesky derating of the transformer with full wave bridge and cap. It's something like 1.8x va is needed right?
The transformer is an industrial control type used for driving pilot lights and contactors and such, (freebie) I'll have to see if I can remeber the va rating but I think it's 150, which would be less than 1.8x.
Nigel, I'll do this when I get home from work tonight.
Thanks!