STEPS parts substitution
Jun 24, 2005 at 2:08 AM Post #16 of 30
Quote:

Originally Posted by tangent
2000uF is not fine with an LM108x family LDO, in my experience. The regulator will not power up reliably with that much capacitance on its output.

Oh, you could do something ugly like put a 1 ohm resistor inline with the regulator's output to allow it to work, but then remind me why you're using a regulated power supply in the first place?




umm... in fact, what i want to do is just reduce the heat generated by the LM317T, even it just drives my PPA with 0.11A, the heat sink on the regulator is too hot to touch, so i decide whether it can reduce the heat by replace the LM317T with LM108X series
 
Jun 24, 2005 at 2:45 AM Post #17 of 30
Quote:

Originally Posted by ceRon
umm... in fact, what i want to do is just reduce the heat generated by the LM317T, even it just drives my PPA with 0.11A, the heat sink on the regulator is too hot to touch, so i decide whether it can reduce the heat by replace the LM317T with LM108X series


Well then you've got a couple options that I can help you with:

a. turn up the juice, i.e. increase the output voltage. If your reg is getting too hot you're asking it to drop too much voltage (combined with current draw of course).

b. get a bigger heatsink.
 
Jun 24, 2005 at 2:54 AM Post #18 of 30
Quote:

Originally Posted by n_maher
Well then you've got a couple options that I can help you with:

a. turn up the juice, i.e. increase the output voltage. If your reg is getting too hot you're asking it to drop too much voltage (combined with current draw of course).


b. get a bigger heatsink.




for (a), in fact, the voltage drop is 47.6V - 24V = 23.4V, and the current drawn is 0.11A, so the power is 23.4V X 0.11A = 2.574W, is it too large for LM317T so that it make a lot of heat?
 
Jun 24, 2005 at 3:00 AM Post #19 of 30
Based on that calcuation, no it doesn't seem like too much. Mine has to deal with a lot more watts than that and I'm not worried about the heat. Why are you running 24V to a PPA? What opamps are you running? Have you actually measure the current draw at idle or just estimated it?

N
 
Jun 24, 2005 at 3:04 AM Post #20 of 30
I'd try a different transformer. 48V transformer (or is it a 36V transformer unloaded?) for a 24V 0.1A load is both larger and more expensive than needed, it's a part that might be more useful for some other project.
 
Jun 24, 2005 at 3:52 AM Post #21 of 30
Quote:

Originally Posted by n_maher
Based on that calcuation, no it doesn't seem like too much. Mine has to deal with a lot more watts than that and I'm not worried about the heat. Why are you running 24V to a PPA? What opamps are you running? Have you actually measure the current draw at idle or just estimated it?

N



i use AD8610 on my PPA and i think it is ok to run 24V on AD8610 as the max. voltage input of AD8610 is +/-13V

i do actually measure the current draw with the amp. switch on, and also i find the the current draw will rise gradually from 103mA to 110mA after i turned my PPA on and it stoped at 110mA
 
Jun 24, 2005 at 6:17 AM Post #22 of 30
Quote:

Originally Posted by ceRon
what i want to do is just reduce the heat generated by the LM317T


Changing out the regulator doesn't change the voltage drop across it. It will dissipate the same amount of heat in this situation. The only time LDOs help with heat is when you're operating it with a lower voltage drop across it than the equivalent standard regulator can tolerate.

The best solution is to use a more appropriate transformer to your required output voltage.
 
Jun 24, 2005 at 8:10 AM Post #23 of 30
Quote:

Originally Posted by tangent
Changing out the regulator doesn't change the voltage drop across it. It will dissipate the same amount of heat in this situation. The only time LDOs help with heat is when you're operating it with a lower voltage drop across it than the equivalent standard regulator can tolerate.

The best solution is to use a more appropriate transformer to your required output voltage.



i use the Amveco 70053 transformer, in fact, i just follow what the tangentsoft said, so 70053 is more than enough?!
 
Jun 24, 2005 at 8:24 AM Post #24 of 30
my aim is to reduce the heat generated by the regulator or to cool down the regulator, so which method is better for me?

1. replace the transformer with a lower voltage output
2. replace the regulator with a lower thermal resistance
3. replace the heat sink with a larger one
4. put a fan on top of the heat sink
 
Jun 24, 2005 at 9:00 AM Post #25 of 30
definitely 1. !
if you want to run your steps at 24V pick the 2x12V Amveco transformer tangent recommends

2. won't help at all, you have to decrease the voltage drop across the regulator

3. install the heatsink tangent recommends, it's available from Mouser and Digikey and cost only $1.30

a fan on top of the heatsink would be loud and the psu won't fit in the popular Hammond case anymore, so that's not really an alternative
 
Jun 24, 2005 at 9:04 AM Post #26 of 30
Quote:

Originally Posted by steinchen
definitely 1. !
if you want to run your steps at 24V pick the 2x12V Amveco transformer tangent recommends

2. won't help at all, you have to decrease the voltage drop across the regulator

3. install the heatsink tangent recommends, it's available from Mouser and Digikey and cost only $1.30

a fan on top of the heatsink would be loud and the psu won't fit in the popular Hammond case anymore, so that's not really an alternative




2 X 12V is no load voltage or sec-full load voltage? does 70052 suit me?
 
Jun 24, 2005 at 9:48 AM Post #27 of 30
yes, the 70052 suits fine

the 12V each secondary are AC at full load, after rectification and ripple rejection the DC voltage will be higher, at partial load even higher.
 
Jun 24, 2005 at 12:39 PM Post #28 of 30
Quote:

Originally Posted by ceRon
i use the Amveco 70053 transformer, in fact, i just follow what the tangentsoft said, so 70053 is more than enough?!


Let's be fair here, you followed some of the instructions, but not all of them. And believe me, I'm guilty of this as well. Just search this forum for "STEPS temp" and you'll find my thread about this. If you read through all the documentation on the site you'll find that the transformer you chose is better suited for higher voltage outputs, on the order of 30 to 35V.

Also, since you're using the 8610 you can go ahead and crank the STEPS output up to at least 26V as there will most likely be voltage losses in the system before it gets to the opamps. Double check this first, I know it to be true for the M3 but I'm not as familiar with the PPA's circuit.

HTH,

Nate
 
Jun 24, 2005 at 4:17 PM Post #29 of 30
Quote:

Originally Posted by n_maher
Let's be fair here, you followed some of the instructions, but not all of them. And believe me, I'm guilty of this as well. Just search this forum for "STEPS temp" and you'll find my thread about this. If you read through all the documentation on the site you'll find that the transformer you chose is better suited for higher voltage outputs, on the order of 30 to 35V.

Also, since you're using the 8610 you can go ahead and crank the STEPS output up to at least 26V as there will most likely be voltage losses in the system before it gets to the opamps. Double check this first, I know it to be true for the M3 but I'm not as familiar with the PPA's circuit.

HTH,

Nate



yes, you are right, the max. input voltage of AD8610 is +/-13V, but i am afraid there maybe some cases that the output of the STEPS will be slightly larger than 26V and burn the AD8610 out, that's why i use 24V, i think the sound quality would not be too large for that 2V difference
 
Jun 24, 2005 at 5:03 PM Post #30 of 30
Quote:

Originally Posted by ceRon
yes, you are right, the max. input voltage of AD8610 is +/-13V, but i am afraid there maybe some cases that the output of the STEPS will be slightly larger than 26V and burn the AD8610 out, that's why i use 24V, i think the sound quality would not be too large for that 2V difference


a. you're right, probably no sound difference, but it will decrease the voltage drop the regulator sees and every bit helps.

b. the chip will most likely never see the full output voltage, especially if you have things like a diode in the path to prevent reverse voltage problems. I think that part alone will drop the steps output by a full volt. Measure your power rails with the 24V out and I'd be downright shocked if you get 24V.

Nate
 

Users who are viewing this thread

Back
Top