Voltage drop with just 40mA...why? (schematic inside)

[Auke Haarsma]

More troubles with my DIY powersupply...

I've built one with this schematic:



With no current draw the output voltage sits nicely on 8.9V. When adding some pedals which draw about 5 mA each (totalling 20mA), I still have 8.8V output.

Now, when I add more pedals, so the total current draw goes over ~40mA the output voltages drops to 7.2V.

Why? I really don't understand it!

Transformer -> 220V AC primary, 12V AC secondary. About 15V DC after rectifying.
Transformer capable of ~500mA
voltage regulators capable of 1A.
Big Caps before output to prevent voltage peaks.

What am I missing?

[PerroGrande]

A couple of comments/questions...

1) You should include a protection diode as per the 7809 datasheet.  You have a fairly large capacitor on the output side of the regulator which can discharge through the regulator (damaging it) when the power is removed. 

2) Are you *sure* that the current draw is only 40mA?  Is it possible that one of those pedals is drawing a lot more than you thought?

3) Is there any heating of the 7809? 

[Auke Haarsma]

John, thnx for your reply!

1) Thanks for pointing that out, I didn't know that.

2) Well, I measured it with my DMM. Some more data:
-0mA -> 8.9V
-18mA ->8.9V
-23mA ->8.88V
-34mA -> 8.31V
-48mA -> drops to 6.3V

3) the 7809 gets a bit warm. Not hot, just warm. I can easily touch it.

NB. I have a Rangemaster, an Omnidrive, a SHO and a Korg DT-10 tuner connected. The DT-10 adds the most mA. Especially when it is trying to tune I get the voltage drop and increased current draw.

[amz-fx]

If the transformer secondary is 12v AC for both windings in series, then that is your problem.  Not enough voltage on each to overcome the inherent drop in the 7809.

You should be using a secondary that is 20v to 24v for both windings in series.

Get rid of the 100R resistors too. They are creating an unnecessary voltage drop. You might be losing 5v across them at 50ma.

regards, Jack

[Gus]

Do you have heatsinks on the regs?

What is the secondary winding(s) of the transformer? 
12VAC CT?
Two independent 12VAC
Two independent 6VAC?

[Auke Haarsma]

@Jack. I measured it all over again, just to be sure. Voltage after rectifying is about 18V DC. I understood a 78XX regulator needs about 3V above what it will put out. That's why I thought 12V DC would be enough.

I removed the 100Ohm resistor. And the voltage drop is gone! THANKS! I've seen that resistor in other PS-designs for some filtering, that's why I put it in. Right now I have 110mA at 8.68V and with 200mA it drops to 8.5V.  That seems more normal to me.

@Gus: secondaries are two independent 12VAC, rated at 500mA each.
I don't have heatsinks (yet). I do have them lying around, but the regulators are not getting hot at all.. Your remark got me thinking... I tested again now with over 100mA current draw....the regulators do get hot. I'll attach the heatsinks (that why I ordered them in the first place).

This place is so full of cool knowledgeable people! Thanks for the help everyone!

[R.G.]

Welcome to power supply design.

Go look at the first illustration in "Power Supplies Basics" at GEO -

http://geofex.com/Article_Folders/Power-supplies/powersup.htm

The current in a full wave rectifier circuit comes in very high pulses near the top of the AC voltage wave. The bigger the filter cap compared to the current being drawn, the higher the multiple of the DC current output that the rectifier pulses are because they are drawn for a shorter period of the AC wave. With under 40ma being drawn and a 1000uF cap, the pulses are very large compared to 40ma. How high is not easy to calculate without measuring all the components in the chain, but many times larger.

We need to know what we're working with, so it's back to the data sheet. The data sheet for the 7809 shows a couple of interesting numbers.
1. The 7809 needs typically 2.0V more than its output to regulate. By my experience, when it gets less input voltage than that, the output follows the input down with about that same differential voltage.
2. The output of a 7809 can be between 8.55 and 9.45V and still be "correct" as these are the stated limits on the chip. So when you get:
> 0mA -> 8.9V - this is a valid regulated output.
> 18mA ->8.9V - this is a valid regulated output.
> 23mA ->8.88V - this is a valid regulated output.
> 34mA -> 8.31V - At this point, the regulation has failed - Vin is less than the regulator chip needs
> 48mA -> drops to 6.3V  - and here it's fallen further.

Is there any forensic information we can use? Yep - The raw voltage fell about (8.31 - 6.3) = 2.01V while the current went up 14 ma. That means your power supply looks like it has an internal impedance of 2.01V/0.014A = 143 ohms. It's pretty obvious where 100 of that is coming from. So yes, you got good advice - get rid of the 100 ohm resistor. Is that enough?

Maybe.  It's hard to tell. What would really help is to measure the voltage at the 1000uF cap with 0ma load and with 48ma load. That would tell us how the average DC voltage declines from as load increases and give us some idea how it will decrease as load increases. Even better would be to measure the DC resistance of both primary and secondary of your transformer (unplugged!) as well as measuring the AC voltage on the secondary with 0ma load and then with a 100R resistor and nothing else across the 12V secondary. From those measurements we could compute the internal resistance of the transformer and predict how much it will sag under load.

Your numbers do make sense; it's the peculiar nature of rectification and voltage regulator chips that are giving you trouble.

[R.G.]

oops - you posted while I was typing.

[PerroGrande]

lol -- that'll teach me to look at the schem. a little more carefully before posting...  I didn't even see that silly 100R resistor sitting there.   

I think I read something somewhere about old-age and its effects on observation and memory, but I forget the details...

Nice call Jack and RG -- great, informative posts as always.

[Auke Haarsma]

R.G. thanks for your reply. As always it's a good and informative read!