PSU with 48v,+/-25v, -22v, +/-18v, 12v or 5v. Harpo mods V0.3 48v updated

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

creal

Well-known member
Joined
Dec 19, 2008
Messages
498
Location
Paris,France
Hi All

I need to make a PSU for racking 2 channelStrips Neve 8108.
For this project, it need 48v, -22v, +/-18v for the channelstrips, 12v or 5v for the relays and leds and +/-25v for the output card (neve ba852).

I draw a schematic, but i like to know if someone can validate if it's do the trick.
I think i'll order a custom Xfrm like this

Primary 1 x 230v
Sec1 2 x 24v 1A
Sec2 1 x 22v 1A
Sec3 2 x 18v 1A
Sec4 1 x 12v 500mA

Thanks you for your advices.
Creal


 

Attachments

  • NEVE-8108_PSU.jpg
    NEVE-8108_PSU.jpg
    91.3 KB · Views: 192
Looks fine except the 48V rail!
the lm317 can only handle about 30 or 35 V on its input. you can use a 317 there but you should consider a "voltage-doubler" or "voltage tripler" setup with it. Look at the jlm-audio website for his psu. there you can see how it is implementet!

regards,
Wolfgang
 
Your 1k trimmers alone won't get you to the target voltages. You want an additional resistor in series to these trimmers. [Vout=1,25*(1+(Rtrimmer+Rx)/240R)] in your config.
As wolfgang already said, the 48V rail won't work as drawn (and the rectifier will blow when you join the gnds)
 
Hi,
Thx wolfgang, Harpo for the advices.
I make some changes this afternoon.

wolfgang said:
Looks fine except the 48V rail!
the lm317 can only handle about 30 or 35 V on its input. you can use a 317 there but you should consider a "voltage-doubler" or "voltage tripler" setup with it. Look at the jlm-audio website for his psu. there you can see how it is implementet!
regards,
Wolfgang

I redraw the 48v part with this changes:
1) replace the LM317 reg by a TL783,
2) add two caps for isolate the 18v part and the 48v, i saw this in the SSL9k PSU. (green frame)
3)add a voltage doubler circuit before the rectifier brige (red frame)

Harpo said:
Your 1k trimmers alone won't get you to the target voltages. You want an additional resistor in series to these trimmers. [Vout=1,25*(1+(Rtrimmer+Rx)/240R)] in your config.
As wolfgang already said, the 48V rail won't work as drawn (and the rectifier will blow when you join the gnds)

I add the resistor in serie to the trimmers.
I use this link to calculate the needed values.
http://www.whatcircuits.com/lm317-calculator-v2/

Here is a pdf version for a better view.
http://twin-x.com/groupdiy/displayimage.php?pos=-4224

I hope this changes fixe the issues.

Cyril



 

Attachments

  • NEVE-8108_PSU.jpg
    NEVE-8108_PSU.jpg
    97.5 KB · Views: 62
creal said:
I redraw the 48v part with this changes:
1) replace the LM317 reg by a TL783,
2) add two caps for isolate the 18v part and the 48v, i saw this in the SSL9k PSU. (green frame)
3)add a voltage doubler circuit before the rectifier brige (red frame)
The 'two caps for isolate the 18v part' with following diode network is a voltage doubler. No 100V caps rating needed here.
For 48VDC out you want a voltage tripler config with a 36VCT transformer in front.
Remove your arrangement in the red block. Connect the (aktually unconnected) both anode side diodes to unregulated +18V, giving you the voltage tripler.
The 10uF in front of the regulator is too small (min.100uF) and will need a 100V rating. The 100nF cap in parallel is located close to the regulator.
How many phantom mics are connected to this supply?

I add the resistor in serie to the trimmers.
In series to the trimmer, not to the adj.pin. (revisit the +18V and +25V config)

edit: something like this
 

Attachments

  • NEVE-8108_PSU.jpg
    NEVE-8108_PSU.jpg
    95.6 KB · Views: 29
creal said:
If i huderstand well, i need to replace the 2x18AC the third secondary of the xfrms by a 2x36AC ?
Can you explain to me?
From your schematic you don't have a 2x18V but a 36V center tapped secondary. No need to get a 72V CT transformer, but a 36VCT with voltage doubler will only get you to about 50.3V raw DC. Subtract the ~5V dropout voltage for the 783 regulator to maintain regulation and maybe 3.5V for 10% ripple giving max.42VDC. A voltage tripler (as drawn) will get you to your wanted 48VDC. You might leave out the -from datasheet not recommended- 10uF cap across the voltage setting resistor (dialed in for ~5k61) for the TL783.
 
Harpo said:
creal said:
If i huderstand well, i need to replace the 2x18AC the third secondary of the xfrms by a 2x36AC ?
Can you explain to me?
From your schematic you don't have a 2x18V but a 36V center tapped secondary. No need to get a 72V CT transformer, but a 36VCT with voltage doubler will only get you to about 50.3V raw DC. Subtract the ~5V dropout voltage for the 783 regulator to maintain regulation and maybe 3.5V for 10% ripple giving max.42VDC. A voltage tripler (as drawn) will get you to your wanted 48VDC. You might leave out the -from datasheet not recommended- 10uF cap across the voltage setting resistor (dialed in for ~5k61) for the TL783.

Ok, now i got a better understanding and why 36VCT it needed.
I made the mods on the 48v part and update the schematic.
http://twin-x.com/groupdiy/displayimage.php?pos=-4227
 
 

Attachments

  • NEVE-8108_PSU.jpg
    NEVE-8108_PSU.jpg
    95.1 KB · Views: 47
Thanks you Harpo for your time and advices.
Now it's time for me to learn to use Eagle and make the PCB for this PSU.
 
wolfgang said:
Looks fine except the 48V rail!
the lm317 can only handle about 30 or 35 V on its input. you can use a 317 there but you should consider a "voltage-doubler" or "voltage tripler" setup with it. Look at the jlm-audio website for his psu. there you can see how it is implementet!

regards,
Wolfgang

Not entirely accurate, you can use the LM317AHV, its like the special version, the output can be adjusted between 1.2 and 57V, and if im not mistaken the maximum input is 60V

http://www.fairchildsemi.com/ds/LM%2FLM317AHV.pdf
 
user 37518 said:
Not entirely accurate, you can use the LM317AHV, its like the special version, the output can be adjusted between 1.2 and 57V, and if im not mistaken the maximum input is 60V

http://www.fairchildsemi.com/ds/LM%2FLM317AHV.pdf
Not entirely accurate as well. It isn't the input voltage, but the differential voltage between input and output that you don't want to exceed.
Trouble starts, when your output is accidently shorted to 0V by a FI broken mic cable. After you replaced the broken mic cable, your mic might get fried from raw DC voltage that the already blown regulator passes thru. Maybe this thread can help a little on this subject.
 
Hi guys,

I have tried to do a similar version of the top psu and I get a 1000uf/50v blow on the circuit that is fed after the 100uf/50v (I'm actually using 220uf/50v). It is the cap which in on the negative side.
The AC measures as 19.2v coming into the bridge rectifier without the other connection connected to the negative side of the bridge, and 25v after, obviously when I connect the other feed of 25v DC it blows the 1000uf/50v.
Is my problem my 50v rating? As I think I'm just going over, something like 53v.
My toroidal measures 2 x 19.2v, which 19.2 x 1.4 = 26.88v x 2 for the voltage doubler = 53.76v
My aim is to get +30v, -10v, +18, -18v.
I'm basically connecting two of the attached picture circuits together.

Regards

Spence.

 

Attachments

  • image.jpg
    image.jpg
    324 KB · Views: 31

Latest posts

Back
Top