PSU Problems

[Sebo]

Hi all:
I built the Elektor Formant PSU, based on the Bergfotron PCB (http://hem.bredband.net/bersyn/psu.htm).
I used two 18V - 1A transformers. Until now it worked OK, but with the recent add of some more modules
to my system, I noticed that I have hum in some modules (specificaly the VCFs). If I disconnect
some modules (specially the Quantix-8 that draws 150mA) the hum goes away.
I measured the total current draw of my system at the positive rail, and it is only 550mA, in the Bergfotron site it's stated that this PSU should give 930mA and it's recommended not to go over 800mA for
best regulation.
At this point I was about to build another little PSU (300mA) to feed the Quantix-8 alone, but I'm curious about
why the PSU is not working as it should.
Could it be the transformers that aren't really 1A as they should be?
That's the only apparent reason I could think.
I dind't measure the current draw of the negative rail, but I'm sure that it is much less. Could I just replace the transformer
of the positive rail?
Thank you.

[blue hell]

It could also be that the filter cap after the rectifier is too small, what value is used there?

And what voltage is the transformer? If that's a bit on the low side that could give hum too, which then probably could be fixed by adding a filter cap as well.

Oops .. I now see .. 18 V .. that should do.

[Sebo]

Hi Blue Hell:
The caps I used are 2200uF x 35V as stated both in the Formant schematics and in the Bergfotron PCB layout.

[blue hell]

ok, still 2200 uF is not too much ... if you have some big ones laying around try to add one in parallel - on the positive supply only should do, likely your problem will be fixed then.

[Sebo]

I have an extra 2200uF x 35V here, I will try it over the weekend.
Thank you.

[Sound]

Sebo,

So the problem is the Hum?

How is your power distribution done? Could you try adding section to the ground conductor? try also connect the Quantix with a cable directly to any point close to the power suply, I mean without sharing bus with other modules.

[Sebo]

Hi:
Today I tried connecting a 2200uF x 35V in parallel with the existing one, and
the problem is still the same.
Also tried to reeplace the 18V - 1A transformer with a 18-0-18V - 1A that I
had at hand (using only one half of it), and the problem gets worse. Maybe this
transformer has a slighty lower otput voltage.

@ Sound: The problem is not the Quantix-8, because all worked OK until I
added more modules. I disconnected it because it is the module that draws
most of the current (150mA), and it's easier to disconnect one module than 3 or 4.
Anyway I tried to connect it at other points of the distro board (close to
the power inputs, and far from it), but the problem still the same.

At this point, I don't know if a bigger transformer (1.5A) will make any
difference. I don't want to spend money in it, and later discover that the
problem is still there. If nothing comes to my mind (or yours) in the next
few days I will build the second PSU.

Thanks.

[blue hell]

[Sebo]

No problem, your help is alway great!!!
Thank you.

[Sound]

I would like insist in the power distribution,
Once I have that hum and it disappeared increasing the section of the ground conductor.
The cable that fed the modules to the bus was three cables 0.25mm2 ,one for positive, one for negative and other for ground. I added other 0.25mm2 cable for ground and the problem disappeared.
Because the cable resistance was reduced avoiding ground loops.
I think is good oversize ground section from module to power supply.

I pointed to Quantix because it has digital ground but more I did intuitive than knowing exactly why.
One thing that I don't understand is why are you pointing to the positive rail.
Other thing that I'm wondering if two transformers could work same that a centre tapped.
A question: are the two transformers exactly the same?

Regards.

[yusynth]

Try to check the grounding of your modules.
Do you use a star distribution or a buss distribution ?
This kind of problem can occur if you use long busses, in such a case try to reorganise the connector on the buss as a function of module current draws.

[blue hell]

[Sebo]

Hi all:
I do all my power wiring with 0.75mm2 (something between 19 and 18 AWG).
I don't think that wire is the problem. The wires are a little oversised, I think.
I think the grounding of the modules are OK, all have good ground, I don't
know what more to check about it.
My distribution scheme is like this: from the PSU are 3 x 0.75mm2 wires
going to a PCB that have big copper stripes, there are mounted all connectors
for power, also there is a second PCB almost identical to the other that take
its power from 3 x 0.5mm2 wires connected directly to where the wires from the PSU
connects to the first PCB. From this connectors I run wires (3 x 0.75mm2 again)
to the modules. These PCBs are about 15 cm long the first one, and about
20 cm the second, and have 12 and 18 connectors each (the connectors
are aligned in 2 rows in every PCB).
I think power distribution is OK too.
Also, as I posted before, I moved the connection of the Quantix-8 to different
positions in the distro board with no change.
Thanks for the help.

[Sebo]

Oh, I forget:
I refer to the positive rail because is the one that more current draws.
And I'm pretty sure there is a problem of current consumption.
I used 2 transformers because the PSU is designed to work with a transformer
with 2 secondaries (not the same as a center tap one) or with 2 transformers.
Using a center tap transformer doesn't work.
When I said I used a center tap transformer, I was only replacing one
transformer alone.

[Sound]

hello Sebo, how is it going? found any trial? any solution?

[magman]

I'm not sure if this will help, but it may be worth looking at.

I've been trying to hunt down an article I read on the web recently about the phase of primary windings on transformers (without success so far) that may apply to your problem. If the primary windings aren't wired up in phase, you can cause all kinds of strange problems, especially as you load up the transformers.

You have effectively got two primary windings, even though they are in separate transformers, so you may have to be careful about how the mains is connected to the primary windings.

You may also have some issues with the magnetic fields of the two transformers interfering with each other. Are both transformers mounted on a metal case of a wooden case? They may need to be mounted on a metal chassis for magnetic coupling. Things to try may include rotating one transformer 90 degrees to the other, or at the opposite extreme making sure they are in line and magnetically linked (using the central mounting screw to link the two mounting brackets together, to increase the coupling).

I'm far from sure that these things will help, as I still have a lot to learn about transformers, but it may be something less obvious like this that is causing your problems.

Off course, the simple solution is to get a properly sized dual secondary transformer, to completely remove any transformer related issues that could be causing your problems.

Hope that helps.

Magman

[Sebo]

Hi guys:
I didn't do anything yet, as I'm busy doing some music
I will do some experiments next week.

> Magman: Thanks for the tips! The transformers are mounted on a wood
case, they aren't in contact with each other, and are side by side at about
4 or 5 cm (2 inches). The transformers aren't toroidal ones. Getting a
transformer with a dual secondary is very hard here, and costs 3 or 4 times
more than 2 simple secondary or a middle tap one. That's because I chose
to use 2 single secondary transformers.
I will try some of your tips, but I don't think I will move one transformer to
the other side of the cabinet, it will take a lot of effort because of the current layout of the PSU, distro board, and some other circuits.

I will report what I find next week.