MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Calex (or Power one or similar) PSU & panel ground
MUFF WIGGLER Forum Index -> Oakley Sound Systems  
Author Calex (or Power one or similar) PSU & panel ground
catatemycaps
It was time to replace my Oakley PSU and PS-20 with something more powerful so I bought this Calex +/- 12 - 15V, 1.7A open frame PSU.

http://uk.rs-online.com/web/p/embedded-linear-power-supplies/7127155/

I've done all the earthing and other wiring but I don't quite know what to do with the panel ground which was naturally straight forward with the Oakley PSU. hihi

Should I just connect the panel and 0V grounds as in MOTM? Of course this will nullify the idea of separate 0V and panel grounds but is it even possible with this PSU?


Cheers.
Synthbuilder
catatemycaps wrote:
Should I just connect the panel and 0V grounds as in MOTM?


The panel ground and module ground (pin 3 and 2 respectively on the Oakley power headers) should be connected together at the power supply. If you have Dizzies then simply take both ground wires back to the new power supply and join them onto the 0V output pin. This is the same as it is on the Oakley PSUs - I've just got two output pins connected together on the PCB.

The Oakley PSU will quite happily work with higher currents. The limit is essentially set by the PA-20 and any heatsinking you have. If you are happy with wiring mains power then replacing the PA-20 with a more substantial internal power transformer can be cost effective.

Bear in mind that the Calex also needs heatsinking or put somewhere where it can dissipate all that heat.

One other thing about the Calex and other Power One type power supplies - beware of latch up on power on. If you find occasionally that only one rail powers up this is latch up. It is caused by excessive start up currents taken by the modular. The best way to avoid this is to use a PSU that doesn't latch up on power up, but if you do come across it, then a timed power up system like the PCB project sold by CGS (Ken Stone) can help.

Tony
catatemycaps
Quote:

The panel ground and module ground (pin 3 and 2 respectively on the Oakley power headers) should be connected together at the power supply.


Ok. Great. smile
This is actually how I do it when I use my benchtop PSU for testing modules. I just was not 100% sure if this was good for a permanent solution.

Quote:

The Oakley PSU will quite happily work with higher currents.


Actually my plan was to use 80VA transformer and replace the two limit resistors with 0.56R values but I then realized that 1 - 1.2A per rail will not be able to handle my backlog. smile
If I want more than that I will have to replace the smoothing caps as well.
I got the Calex PSU at a good price so I went for it.

I have got a couple of big heatsinks that I can slab on it if necessary.
I don't think I will draw more than 1.4A from it so convection cooling should suffice. The unit is mounted on sturdy stand offs so air can travel around it. I will watch out for the heat, though.

Quote:

One other thing about the Calex and other Power One type power supplies - beware of latch up on power on


Hmm...according to the specs this model should have protection against it.

"COMMON-MODE LATCH UP
All 32000 SERIES dual power supplies have
incorporated a unique anti-latch circuit to minimise
common mode latch up"

Hopefully it works. smile


Thanks for the help, Tony! Guinness ftw!
Synthbuilder
catatemycaps wrote:
"All 32000 SERIES dual power supplies have
incorporated a unique anti-latch circuit to minimise common mode latch up"


That is good to know. It's right ruddy PITA to sort latch up problems. Especially as it can be a very intermittent thing.

Tony
catatemycaps
Everything was going fine until now.
I installed my 12th module to the modular and then everything just locked up.
I'm using lock up in lack of a better term.

All the leds light up and nothing works.

When I removed the 12th module everything was back to normal again.

All the power connectors are the the right way (I installed one module at a time and tested).
I have tested with different power cables so it is not that either. Also, it does not matter which module I plug in after the 11th.
Can't be a broken Dizzy. There are other modules plugged in the same Dizzy.

I have not measured voltage because I dare not keep the modular powered on when it locks up.
In normal operation all dizzy boards are providing +/-15V.

I seriously doubt that I am drawing more amps than the psu can provide as it should be capable of 1.7A per rail according to the specs.

hmmm.....
Synthbuilder
Sounds like one rail is going down if the LEDs are lighting up.

Does sound like power on latch up... 12 modules is a bit naff if it is though.

Try plugging in the last module into the Dizzy when the unit is powered up. Use a less fussy module like a Multimix or Fourmix and make sure the power plug goes in fast and correct.

If the modular still powers up just fine you have a power latch up problem. It's caused by the sudden surge in current when the modules are powered up. Each module has a small 2.2uF capacitor on the power supply rail to ground. These need to be charged up at the moment of switch on and the only thing that restricts the current at this point is the resistance in the wiring, the internal resistance of the capacitor and the power supply's current limiting.

The thing is, the PSU's current limiting depends on the output voltage of the power supply. It's more sensitive on power up, hence the latch up.

If it is power on latch up, then one of those timed start up modules that Ken Stone does will do the trick. These power up banks of modules one at a time reducing the initial current surge.

Tony
catatemycaps
Quote:

Does sound like power on latch up... 12 modules is a bit naff if it is though.


Well, so much for the latch up protection they have built in this unit.

Quote:

If it is power on latch up, then one of those timed start up modules that Ken Stone does will do the trick.


Well...bugger. This is straight from Ken's site.
CGS63 Power Supply Delay. discontinued


Hmm...time to get back to the drawing board and perhaps build two Oakley PPSUs + two 80VA toroidal transformers.

I'll give the hot plug in you suggested a go once I get home from work and see how that goes. Hopefully nothing will go Dead Banana

Thanks, Tony. Guinness ftw!
catatemycaps
I tried plugging in one module while the unit was powered up. However, aligning the molex connector and pushing it quickly in was harder than I thought it would be so I decided not to do it.

I managed to power up with twelve modules once. Overdrive II being the 12th module. Subsequent boot ups failed again, though.

I think your assumption of latch up is correct.

What a shame. I trusted Calex would serve as a good PSU since they advertised it was protected against latching up. sad banana
Synthbuilder
A couple of things to try:

1. Reduce all the 2.2uF caps on the power supply inputs on each module to 1uF. It shouldn't affect module performance.

2. Use multiple Dizzies with a double pole switch on each one. This will allow you to manually power each bank up one at a time.

You could modify the Calex but we really need to see the schematic for it. I've modified various Power One PSU to stop them latching up.

Tony
catatemycaps
Quote:

2. Use multiple Dizzies with a double pole switch on each one. This will allow you to manually power each bank up one at a time.


This is actually a great idea. I've got four dizzies. Only using two at the moment.
I just need to shop for some DPST switches.

Having all four dizzies behind switches will work as an extra safety measure as well. If that Calex goes up in
smoke some day on boot up it will not take all the modules with it to modular heaven. smile
And if I don't need the top row or the bottom row of the unit for a patch I can keep that row unpowered.

This PSU came only with a diagram of how to wire up remote sensing, but no schematics.
They might have them on the manufacturer's site.
jvq
Interesting thread! I am using an Oakley RPSU to power a CrowBX (which is not yet finished). If I understand all this correctly the CrowBX will be giving a bit of a kick when it powers up... it has a total of 10 100uF capacitors per rail at +19V and -19V that all have to charge up when the power is connected.

This makes me think twice about at least putting a power switch on my build so that I can time this better; I normally power up my synth gear 5-8 units at a time with power strips.

The RPSU and the PA-30 (and the electricity outlet at my work bench) have not protested yet; my build is at 16 out of the 20 100uF input filter capacitors now.
terjewinther
Synthbuilder wrote:
The Oakley PSU will quite happily work with higher currents. The limit is essentially set by the PA-20 and any heatsinking you have. If you are happy with wiring mains power then replacing the PA-20 with a more substantial internal power transformer can be cost effective.
Tony


This is very interesting, because I have just aquired a rather large cabinet, that can probably hold up to 80 modules, and obviously I need a power supply for this. I thought about using the Oakley PSU, but since you state "small to medium modular" I didn´t go any further.

My idea was to get a substantial ring-core power transformer and build something myself, but if I can use the Oakley PSU, I would be happy. I have done a lot of mains wiring, so that is OK.

I now see that you have added some info regarding mains powering of the Oakley PSU. Do you think it will be possible to get 4A out of it?
Synthbuilder
terjewinther wrote:
Do you think it will be possible to get 4A out of it?


No. 4A is huge. I reckon 1.5A should be the top limit of the Oakley PSU. There are many reasons why there is this limit but chief among them is that the series pass transistors require a controlling base current to drive them and the amount of base current is dependant on the current taken from the power supply. 4A would require more base current than the controlling circuitry can provide.

There is another thing to consider. Multiple small power supplies are always better than one big one. The main reason is to do with distribution. It's really hard to distribute accurate stable voltage over a large system. The various resistances in the connecting leads will produce voltage drops across those leads. While constant voltage drops across the conductors are generally acceptable within reason, the varying currents in modular synthesiser will produce varying voltage drops creating unwanted interference between modules.

It is therefore far better to use smaller power supplies around your system. Preferably one per case. The various 0V (module ground) points from each case should then be connected together with a nice thick cable.

And finally, consider what would happen should a power supply fail to regulate in the event of a fault. It could take out your whole modular system. Better to restrict any such fault damage to one case.

Tony
terjewinther
Synthbuilder wrote:
Multiple small power supplies are always better than one big one.


That is of course a very good point. For this large cabinet it might be a good idea to actually have two power systems inside one cabinet. I have some calculations to do. Thanks.
MUFF WIGGLER Forum Index -> Oakley Sound Systems  
Page 1 of 1
Powered by phpBB © phpBB Group