Amek Einstein PSU

[angelothewolf]

Hi everybody, im attempting to build the PSU for an Amek einstein console. i came across this schem, so i decided to work with it. i think this is from a smaller console model, but i think that changing some power ratings will do the job:



So i start to copying this schem on the ISIS, to draw a pcb later with the ARES.
What you think about this schem? gonna work? im not sure bout the tranny ratings for this 40 channel console im thinking on 10A for each 20V winding and 6A for the 40V one.

this is the Scchem i made with ISIS:

[angelothewolf]

nobody_?

[PA0PBZ]

One thing I noticed is that you have a LM317K in your -17V section, that thing only does like 1.5A.

[ciccio]

The 40 V section is used for the +48V dc phantom power for the microphones.
A maximum of 1.5 A will be needed. 6 A are excessive.
You can calculate this way:
a) each channel phantom power is supplied through a (semi-standard) 3K3 / 3K9 resistor on pin 2 and another one on pin  3 of the input XLR.
b) worst case: pin 2 and 3 shorted to GND (pin 1): each channel will sink 2 x (48/3.3) mA = about 30 mA.
c) 40 channels, worst case : 40 x 30 = 1.2 A.
Actually, most of the big console's phantom power supply is rated at less than 1 A.

I'm not sure that 1.5 A are enough on the  +18 / -18 bus. Maybe you could parallel two regulators. In any case 20 A are excessive.

Best regards

[tekfan]

I have an Amek Angela console and that is of similar size to the Einstein. It has two power supplies. One big one to power the channel strips (+17/10A, -17/10A, 12V relay supply, +48V PP)
And a smaller one (exactly the one you show in the pictures) just for the VU meters providing + and -17V (+48V PP is disabled).

The smaller power supply is actually originally meant for a Amek Tac console. So it is just too small to power an Einstein.
Maybe it would power the positive rail but the negative can't provide nearly as much current as needed to power all the channel strips.
I'll try to upload the images of the big supply on friday.

I also don't know if the Einstein has automation which also might require +5V.

[tekfan]

Here are the schematics and component layouts of the power supply. Beware of the wrongfully drawn bridge rectifiers before modules GL47 and GL48.
For the +17V, -17V and +10V module GL48 is used. For the -17V supply the positive output is tied to ground and the negative one goes to the output connector of the power supply. The +10V supply uses the same GL48 module but with different component values.

The filter capacitors after the bridge rectifiers for the +17V, -17V , +10V and +5V supply rails are somewhere around 20000uF with the appropriate voltage rating
(2000uF per ampere rule of thumb).

The +48V Phantom power supply is also fused in the same way all the other supplies are. The fuse is not drawn in the schematic!

It may look complicated but this is probably the best way to get a highly reliable power supply.

You may have to look at where exactly to apply the supply voltages to your console and the actual supplies it uses.
The pinout is almost certainly different for the Einstein.

[angelothewolf]

Here are the schematics and component layouts of the power supply. Beware of the wrongfully drawn bridge rectifiers before modules GL47 and GL48.
For the +17V, -17V and +10V module GL48 is used. For the -17V supply the positive output is tied to ground and the negative one goes to the output connector of the power supply. The +10V supply uses the same GL48 module but with different component values.

The filter capacitors after the bridge rectifiers for the +17V, -17V , +10V and +5V supply rails are somewhere around 20000uF with the appropriate voltage rating
(2000uF per ampere rule of thumb).

The +48V Phantom power supply is also fused in the same way all the other supplies are. The fuse is not drawn in the schematic!

It may look complicated but this is probably the best way to get a highly reliable power supply.

You may have to look at where exactly to apply the supply voltages to your console and the actual supplies it uses.
The pinout is almost certainly different for the Einstein.

many thanks for the info!! ...but, i dont see much difference with the design im working on besides the 5v & 10 v rails, and the -17v trik. this circuit deliver more current that the one i have_?

if i make two boards for each  +17 & -17 supply and parallel their outputs, it will deliver more current??

the idea is to build a reliable power supply for an day to day use on a recording studio.

thanks for the replies!

[tekfan]

many thanks for the info!! ...but, i dont see much difference with the design im working on besides the 5v & 10 v rails, and the -17v trik. this circuit deliver more current that the one i have_?

if i make two boards for each  +17 & -17 supply and parallel their outputs, it will deliver more current??

Yes, the power supplies are almost the same.
The schematics I posted also have SCR protected outputs. Since the series pass transistor (MJ2955) is under quite a bit of load it is succeptable to failure.
It can short out and apply all the voltage at the filter capacitors to the channel strips. There is an SCR protection circuit that fires and shorts the supply rail to ground if it exceeds a certain voltage, causing the fuse to blow. You wouldn't want to replace all the ICs in all the channel strips just because the power supply didn't have decent protection.

In theory it is possible to parallel two boards to get double the output current but in practice it is nearly impossible to have the current shared equally by both boards. The voltage would have to be exactly the same (within a few mV). It's much easier to add another series pass transistor. In this case even this isn't needed. The MJ2955 transistor is happy up to 15 amps so your circuit should work at 10 amps with no problems. Although at such high currents the SCR protection is a very smart idea to have.

As I said there are 3 exactly the same boards and one has two different components to make it work at a lower voltage. It can't be simpler than that.

[angelothewolf]

many thanks for the info!! ...but, i dont see much difference with the design im working on besides the 5v & 10 v rails, and the -17v trik. this circuit deliver more current that the one i have_?

if i make two boards for each  +17 & -17 supply and parallel their outputs, it will deliver more current??

Yes, the power supplies are almost the same.
The schematics I posted also have SCR protected outputs. Since the series pass transistor (MJ2955) is under quite a bit of load it is succeptable to failure.
It can short out and apply all the voltage at the filter capacitors to the channel strips. There is an SCR protection circuit that fires and shorts the supply rail to ground if it exceeds a certain voltage, causing the fuse to blow. You wouldn't want to replace all the ICs in all the channel strips just because the power supply didn't have decent protection.

In theory it is possible to parallel two boards to get double the output current but in practice it is nearly impossible to have the current shared equally by both boards. The voltage would have to be exactly the same (within a few mV). It's much easier to add another series pass transistor. In this case even this isn't needed. The MJ2955 transistor is happy up to 15 amps so your circuit should work at 10 amps with no problems. Although at such high currents the SCR protection is a very smart idea to have.

As I said there are 3 exactly the same boards and one has two different components to make it work at a lower voltage. It can't be simpler than that.

alright! now i get it, so apears that this will work for the amek einstein then. maybe its good idea to make an extra psu for the VUs, just for help a little bit.

i will try to draw that schematics with ISIS then, to make the PCB. i was thinking on mounting the regulators out of the board, directly to the heatsinks. and add some cooling fans...

I have a toroidal transformer manufacturer nearby. so what you think about the transformer ratings for this project_?

[Sm58]

I have a toroidal transformer manufacturer nearby. so what you think about the transformer ratings for this project_?

Hi
First, i have the same psu. Not shur from what a console it come from.
Anyway, other then yours, mine has the well known 2n3055 instead of a 3xx vr.
The ratings are the same. 10A for the +\- 18V and 3amps i guess for the 48V.

For a console a this size (-40 ch) i would not go well below that for +/- 18V.
At least 5amps per rail.
The phantom power seams to me as a bit of a overkill...

Transformer
As you do the math...

20V x 5 A x 1.8 = 180 VA per rail


 

[Sm58]

For a 10A version i would consider a slow/soft start circuit.

10A x 20V x 1.8 x 2 = 720VA !!!

[tekfan]

The toroidal transformer is very beefy indeed. As I recall it's about 20cm in diameter and 15cm high.

Check on the console around the plug for the power. There may be two plugs. One big one for the channel strips and a smaller one for the VU meters.
If there's just one plug then the console may only require one power supply.

And yes, big heatsinks and cooling fans are a must. My power supply has 4 fast fans inside just to keep it cool.
A soft start circuit also isn't a bad idea. It sometimes trips the breaker when I turn it on since all those big electrolytic capacitors have to charge up in one cycle.

[Sm58]

A soft start circuit also isn't a bad idea. It sometimes trips the breaker when I turn it on since all those big electrolytic capacitors have to charge up in one cycle. 

Exactly. The fuse on the primery side might blow quickley then

[angelothewolf]

For a 10A version i would consider a slow/soft start circuit.

10A x 20V x 1.8 x 2 = 720VA !!!

i think i dont get that calculation...

i need two 20V 10A windings for +17V & -17V rails , and one 40V 1.8A winding for phantom rail. isnt it_?

i already have the 338/317 regulators, so i will use them. i have four 10000uf/63V caps, can i use two of them after the rectifier as reservoir caps for each 17V rail?

I think the soft start circuit its a must too. maybe a temp-dependant resistor in series with me mains line its not a good idea for this project. do you know some active soft start circuit to recommend?

...and thanks for the replies, is being very helpful. and searching on the net i found lot of frustrated posts about similar large console PSUs. i think will be a good contribution for the forum!   

[Sm58]

i think i dont get that calculation...

Ok..
10 Amp per Rail
20 Volts per rail
2 rails

Due to the bridge rect. and the filter caps you must consider a bit more then that.
About 1.8 times more (its a constant, you find it in every basic electronic lecture)

So 10A x 20V x 2rails x 1.8 = 720VA


About the slowstart.
You can use a ntc on the primary with a relais that short the ntc out assoon as voltage is
Present on the secondary. Maybe delay the trigger for the relais and a fan which cool the ntc back down.

The ntc can get quite hot...

[angelothewolf]

i think i dont get that calculation...

Ok..
10 Amp per Rail
20 Volts per rail
2 rails

Due to the bridge rect. and the filter caps you must consider a bit more then that.
About 1.8 times more (its a constant, you find it in every basic electronic lecture)

So 10A x 20V x 2rails x 1.8 = 720VA


About the slowstart.
You can use a ntc on the primary with a relais that short the ntc out assoon as voltage is
Present on the secondary. Maybe delay the trigger for the relais and a fan which cool the ntc back down.

The ntc can get quite hot...

what you think about this...
http://electronics-diy.com/soft-start-for-power-supply.php

[tekfan]

Here's a detailed description of soft start circuits:

http://sound.westhost.com/project39.htm

[Sm58]

what you think about this...

Yes, very similar to what i talked about. Powerresistors instead of ntc.