# DIY Power supply kit suggestion (Fan & pull-up resistor, cap)

I bought this 0-30v power supply DIY kit... There is three suggestion I need...

1. I want to use a 12V fan, so I'm changing the 7824 to 7812. So, I'm thinking to change the pull-up resistor (2.2K) to 1K/1W. Is it right? (as $$\V =I * R\$$)

3. For C1, can I use 2200uF instead of 3300uF? The provided capacitor is okay but came dented, and I have some 2200uF capacitors handy. (Is there any equation to get capacitor value?)

• can you tell us which purpose R1 fulfills? It's really not very clear what it does. Oct 6, 2019 at 10:53
• The R1 is attached to U4 (7824) pin1 & pin2... This is for if the fan is not connected...
– BLUE
Oct 6, 2019 at 10:55
• uh, ok, not quite sure why that's necessary! Do you happen to have a full schematic? I just noticed that D4 also seems to be connected to fan negative, and that makes preciously little sense like this. Oct 6, 2019 at 11:00
• I could not find any proper schematic... Just found this manual, netzener.net/images/portapower/operation-and-maintenance.pdf, and this schem image, qsl.net/z33t/0-30V_dc_regulated_power_supply/shema_small.jpg
– BLUE
Oct 6, 2019 at 11:35

# R1

You name this a pull-up, but if it's really between a 7824's Pin 1 (input) and Pin 2 (ground), it's an input-loading pull-down resistor. Since you don't change the input voltage, don't change the resistor.

# Replacing 7824 with 7812

That means you'll have 12 more volts that the resistor has to drop. Also, a 12 V fan of same power as the original 24V fan needs to draw about twice the current – in sum, you will have more than twice the heat produced in U4.

It's unlikely that the heatsink was sufficiently oversized (these things are relatively expensive) to support that.

So, don't do that, unless you want to toast your 7812. Instead, you'd connect a switch-mode voltage regulator specified for the input voltage, 12 V output and the current you need. It's easy when you have such through-hole components – you can simply solder in wires to a sufficiently beefy buck converter module (don't go ebay/aliexpress on that – only buy things with proper datasheets).

# heat sink "plate"

That's an isolating barrier that still lets thermal energy pass, typically made from mica. Yes, you'll need that if you want to attach your heatsink to a 78xx-style regulator.

# 2200 µF instead of 3300 µF

Obviously, someone designed in a 3300 µF capacitor because they decided that using a smaller value wasn't appropriate. That alone should tell you that, no, just reducing capacitance by 50% isn't a good idea.

Then again, this is just a buffering capacitor to smooth us the irregular and potentially very noise current draw of the fan. It is kind of usual to slightly overdimension that to allow for degradation and variability and noise-sensitive environments, so you might be fine. However, again, when going from a 24 V fan to a 12 V fan with the same power, you're doubling currents and hence probably also significantly increase current ripple – and that might be so much that it makes the life of your regulator hard.

So, I'll go with a "no, unless you know very well through measurements that the ripple you see is negligible and you don't run into EMI problems, you can't just replace a component with one 2/3 of the values".

• For R1, I just read this article, learn.sparkfun.com/tutorials/pull-up-resistors/all, anyway, I'll never going to use this circuit without fan, so, I'm not changing that resistor as per as you suggested...
– BLUE
Oct 6, 2019 at 11:43
• That article has nothing to do with the circuit here. Oct 6, 2019 at 11:51
• About the capacitor, I bought two new 3300uF cap from local, but my GM328 tester showing, they are about 2100uF...!!! Total waste...!!! The provided cap is reading 3100uF... Would you suggest to use the dented capacitor... It's just irritates me...!!!
– BLUE
Oct 6, 2019 at 11:52
• @BLUE measuring large capacitances is actually pretty hard, I don't really have any recommendation Oct 6, 2019 at 12:32

Assuming it's a linear regulator...

the effect of a 2200uF cap will be to increase ripple on the UNregulated supply, especially at full load current. The regulated output will probably be fine up to about 27V but may drop out (start showing ripple) somewhere before 30V. (Assuming all else is OK ... correct transformer, nominal mains voltage, etc.).

This would be a serious defect in a piece of lab equipment for general use (which is supposed to be trustworthy anywhere within its ratings).

But if this is for your own use, you may decide you're OK with that.

Just check the supply voltage looks clean on a scope whenever you're using it above 25V. (Actually, do that anyway, whatever the voltage setting, as a confidence check).

• I don't have a scope, that's the real problem, anyway I'm still waiting for a good capacitor... I just tensed about, not to damage the circuit or other parts... But, after your reply, I assume it only effect the output, so I can at least complete the circuit and use it below 27v without any problem or damage the circuit...
– BLUE
Oct 6, 2019 at 12:14