Calling Power supply gurus

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, luketeaford, lisa, Joe.

User avatar
wahee
Common Wiggler
Posts: 212
Joined: Fri May 29, 2015 5:31 pm
Location: Belfast

Calling Power supply gurus

Post by wahee » Thu Aug 27, 2020 10:23 am

hi all

I was hoping to make a small desktop module which requires GND -12 & +12 rails.

I don't want to use a transformer (I have seen plenty of circuits for that.) due to mains issues but rather a 15VAC wallwart.

I have designed this for use with a DC barrel jack circuit and was wondering if those in the know could give it the once over to see if there is anything wrong with it.

I may add a fuse etc but the basic design I was more worried about

many thanks
PSU.JPG
You do not have the required permissions to view the files attached to this post.
Last edited by wahee on Sat Sep 05, 2020 6:59 pm, edited 1 time in total.
Crystalline Stricture now on
tonnrecordings.com

Elahrairah
Common Wiggler
Posts: 61
Joined: Sat May 02, 2020 12:25 pm

Re: Calling Power supply gurus

Post by Elahrairah » Thu Aug 27, 2020 11:17 am

MFOS and Frequency Central have a very very similar design, do read up on their schem for tips.

User avatar
mrand
Common Wiggler
Posts: 187
Joined: Thu May 23, 2013 11:26 pm
Location: Whitehorse, Yukon
Contact:

Re: Calling Power supply gurus

Post by mrand » Thu Aug 27, 2020 1:00 pm

I'm quite puzzled by this design. AFAIK, it's not the usual method.
Is this an innovative solution or just wrong?
Why are there caps before the bridge rectifiers?
I can't make sense of this. If I apply negative voltage at the input, I follow it through to both the ground and at the same time to the 7912 input,after 5 diode drops.
Hopefully some gurus can chime in on this one.
--ssdp--

Synthiq
Veteran Wiggler
Posts: 590
Joined: Mon Feb 06, 2017 3:10 pm
Location: California

Re: Calling Power supply gurus

Post by Synthiq » Thu Aug 27, 2020 1:29 pm

I have used the same circuit and it works. The two capacitors driving the lower bridge rectifier must have their positive side connected to the jack, though.

User avatar
wahee
Common Wiggler
Posts: 212
Joined: Fri May 29, 2015 5:31 pm
Location: Belfast

Re: Calling Power supply gurus

Post by wahee » Fri Aug 28, 2020 6:10 pm

hi all thanks for the replies. I did get the idea of the 2 bridge rectifiers from another circuit. The capacitor orientation was copied from that but it may be wrong.

I just need to generate the -12v and +12v so this seems like a way to do it.
Crystalline Stricture now on
tonnrecordings.com

User avatar
wahee
Common Wiggler
Posts: 212
Joined: Fri May 29, 2015 5:31 pm
Location: Belfast

Re: Calling Power supply gurus

Post by wahee » Sun Aug 30, 2020 4:15 pm

Yes the 2 caps should be + towards AC source, found this circuit.
bfb2a197acb9d3b895f0583608123aa1.jpg
You do not have the required permissions to view the files attached to this post.
Crystalline Stricture now on
tonnrecordings.com

User avatar
MikeDB
Common Wiggler
Posts: 129
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: Calling Power supply gurus

Post by MikeDB » Sun Aug 30, 2020 5:46 pm

Just connect the barrel (pin 1) of the AC input to GND.
Connect a single rectifier diode anode to pin2/3 input, cathode to +ve on C5.
Connect a single rectifier diode cathode to pin2/3 input, anode to -ve on C4.

As it's now a half wave rectifier you may need to increase values of C4 and C5 - see how much ripple there is on the outputs.
Personally I hate cats !

User avatar
mrand
Common Wiggler
Posts: 187
Joined: Thu May 23, 2013 11:26 pm
Location: Whitehorse, Yukon
Contact:

Re: Calling Power supply gurus

Post by mrand » Sun Aug 30, 2020 8:06 pm

MikeDB wrote:
Sun Aug 30, 2020 5:46 pm
As it's now a half wave rectifier you may need to increase values of C4 and C5 - see how much ripple there is on the outputs.
Isn't the full-wave rectified version discussed above a better solution?
wahee wrote:
Sun Aug 30, 2020 4:15 pm
Yes the 2 caps should be + towards AC source, found this circuit.
If someone has the patience to explain what C1 and C2 are doing... feel free! And how you would select those values.
--ssdp--

KSS
Super Deluxe Wiggler
Posts: 3413
Joined: Mon Jan 25, 2016 7:28 am

Re: Calling Power supply gurus

Post by KSS » Mon Aug 31, 2020 9:47 am

C1 and C2 are providing both a storage and a phase shift function. Remember -or refer to- ELI the ICE man. ICE in this case.
Treat utility modules as stars instead of backup singers.

Treat power supplies like Rockstars instead of roadies!

User avatar
revtor
Dialing it in..
Posts: 1784
Joined: Sat Nov 29, 2008 10:31 am
Location: North Jerz

Re: Calling Power supply gurus

Post by revtor » Mon Aug 31, 2020 10:26 am

Phase shift to reduce ripple?
Fellow North Jersey Synthaholic

User avatar
wahee
Common Wiggler
Posts: 212
Joined: Fri May 29, 2015 5:31 pm
Location: Belfast

Re: Calling Power supply gurus

Post by wahee » Tue Sep 01, 2020 9:56 am

KSS wrote:
Mon Aug 31, 2020 9:47 am
C1 and C2 are providing both a storage and a phase shift function. Remember -or refer to- ELI the ICE man. ICE in this case.
Very interesting, I always thought I needed a 3 tap transformer to easily get negative voltages, so this was a good if not perfect alternative
Crystalline Stricture now on
tonnrecordings.com

User avatar
MikeDB
Common Wiggler
Posts: 129
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: Calling Power supply gurus

Post by MikeDB » Tue Sep 01, 2020 10:48 am

wahee wrote:
Tue Sep 01, 2020 9:56 am
Very interesting, I always thought I needed a 3 tap transformer to easily get negative voltages, so this was a good if not perfect alternative
No need. See this example. Not sure it needs quite so much smoothing capacitance and it only really needs CR3 and CR5 or CR4 and CR6.

Image
Personally I hate cats !

User avatar
mrand
Common Wiggler
Posts: 187
Joined: Thu May 23, 2013 11:26 pm
Location: Whitehorse, Yukon
Contact:

Re: Calling Power supply gurus

Post by mrand » Tue Sep 01, 2020 1:27 pm

I'm interested to learn more about the merits of each of these two designs.

I'm still struggling to understand how the dual bridgerectifier design works, but if it gives us 120hz ripple instead of 60hz ripple, then it seems like the better choice... right?
--ssdp--

User avatar
MikeDB
Common Wiggler
Posts: 129
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: Calling Power supply gurus

Post by MikeDB » Tue Sep 01, 2020 2:05 pm

mrand wrote:
Tue Sep 01, 2020 1:27 pm
I'm interested to learn more about the merits of each of these two designs.

I'm still struggling to understand how the dual bridgerectifier design works, but if it gives us 120hz ripple instead of 60hz ripple, then it seems like the better choice... right?
The full wave design uses capacitors to effectively level shift the incoming AC below the GND line. The half wave design uses those same capacitors to provide the extra filtering needed for the 50/60Hz signal. So in many ways they are equal.

I personally wouldn't use either, instead prefering a simple rectification of the AC input followed by DC-DC convertor/regulators for the positive and negative lines. Far less crap on them this way.
Personally I hate cats !

KSS
Super Deluxe Wiggler
Posts: 3413
Joined: Mon Jan 25, 2016 7:28 am

Re: Calling Power supply gurus

Post by KSS » Tue Sep 01, 2020 5:26 pm

What MikeDB said.

There's also the additional diode drop of full vs half-wave rectification, and the added cost of all those diodes. Or integrated bridges.
C1 and C2 -refdes in the latest posted schematic- are being pushed very hard and had better be good quality caps rated for A/C rectification duty. Not all electrolytics are up to the task. This is *not* the place for cheap Tayda elcaps! While also true for the half-wave design, C1 and C2 are being subjected to the full A/C range of the input while those in the halfwave design are seeing only pulsed DC. Subjecting diodes CR1 and CR2 -refdes again in the latest posted WW schematic- to full A/C is much more in their 'wheelhouse' than doing the same with C1 and C2 in the full wave design.

This should explain why we see the half-wave version far more often than the dual bridge type. It's cheaper to make, can use lower rated parts of lower quality without the same 'hit' of doing so in the other design. If you have low voltage overhead then the additional diode drop factors in for the regulator headroom. Meaning the halfwave design can -ideally- use a lower transformer voltage. Functionally this isn't usually the case in a practical application, but we're drawing distinctions between the teo designes here.
---------------
@mikeDB
DC-DC converters bring their own set of 'crap'. But overall I'd agree that this is nowadays a better path, as long as one does their research and not simply copy some Buchla or Doepfer design. Doepfer exhibiting one of those 'crap' angles with the shutdown and noise factors some Euro users are seeing with them, and Buchla for the higher cost and reality that Don designed a closed system in the 200e and expecting that to work as well universally misses his expertise at aiming the solution toward a specific target.

My sig should also say Treat power supplies like Rockstars instead of roadies and your synth will work much better!
Treat utility modules as stars instead of backup singers.

Treat power supplies like Rockstars instead of roadies!

User avatar
MikeDB
Common Wiggler
Posts: 129
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: Calling Power supply gurus

Post by MikeDB » Tue Sep 01, 2020 6:05 pm

KSS wrote:
Tue Sep 01, 2020 5:26 pm
@mikeDB
DC-DC converters bring their own set of 'crap'. But overall I'd agree that this is nowadays a better path, as long as one does their research and not simply copy some Buchla or Doepfer design. Doepfer exhibiting one of those 'crap' angles with the shutdown and noise factors some Euro users are seeing with them, and Buchla for the higher cost and reality that Don designed a closed system in the 200e and expecting that to work as well universally misses his expertise at aiming the solution toward a specific target.
I've never even looked at Buchla or Doepfer designs so no idea how good/bad they are. Maybe I'll go and have a look. :twisted:

I usually try to sync my dc-dc convertors to a multiple of the master 96kHz clock in my designs so any noise problem is repetitive, although now that many of the recent new convertors are moving into the MHz region the bigger problem is making sure their switching frequency isn't leaking out the sockets and failing FCC Part 15.
Personally I hate cats !

User avatar
wahee
Common Wiggler
Posts: 212
Joined: Fri May 29, 2015 5:31 pm
Location: Belfast

Re: Calling Power supply gurus

Post by wahee » Sat Sep 05, 2020 7:23 pm

MikeDB wrote:
Tue Sep 01, 2020 10:48 am
wahee wrote:
Tue Sep 01, 2020 9:56 am
Very interesting, I always thought I needed a 3 tap transformer to easily get negative voltages, so this was a good if not perfect alternative
No need. See this example. Not sure it needs quite so much smoothing capacitance and it only really needs CR3 and CR5 or CR4 and CR6.

Image
this is great. I have learned so much in this post. The 2 rectifier design does seem a little OTT.
Essentially I need this design to run off eurorack or wallwart AC
I have created a very basic design with minimal smoothing, which I will add to, as it does seem to be a false economy to scrimp on the power supply
I was wondering about the purpose of the 2x 2.4k resistors. Are they for current limiting

thanks
minimal smoothing.JPG
You do not have the required permissions to view the files attached to this post.
Crystalline Stricture now on
tonnrecordings.com

KSS
Super Deluxe Wiggler
Posts: 3413
Joined: Mon Jan 25, 2016 7:28 am

Re: Calling Power supply gurus

Post by KSS » Sat Sep 05, 2020 8:56 pm

CR4 and 6 are not there for the same reason as CR 3 and 5. You can read about them in TI's Linear Circuits Databook, Vol.3. Pg2-436
4 and 6 are for output polarity reversal and 3 and 5 are for reverse bias protection. We do tend to see 3 and 5 more often in DIY published circuits.
But 4 and 6 are not redundant.

You need at least double the filter cap-acity of the full wave designs.

AIRC, Ray put those 2k4's in to provide a minimum draw on the suppy to ensure correct operation at low current draws. They become unnecessary at higher loads. I'm sure it's mentioned on his site.
Treat utility modules as stars instead of backup singers.

Treat power supplies like Rockstars instead of roadies!

User avatar
wahee
Common Wiggler
Posts: 212
Joined: Fri May 29, 2015 5:31 pm
Location: Belfast

Re: Calling Power supply gurus

Post by wahee » Sun Sep 06, 2020 8:19 am

KSS wrote:
Sat Sep 05, 2020 8:56 pm
CR4 and 6 are not there for the same reason as CR 3 and 5. You can read about them in TI's Linear Circuits Databook, Vol.3. Pg2-436
4 and 6 are for output polarity reversal and 3 and 5 are for reverse bias protection. We do tend to see 3 and 5 more often in DIY published circuits.
But 4 and 6 are not redundant.

You need at least double the filter cap-acity of the full wave designs.

AIRC, Ray put those 2k4's in to provide a minimum draw on the suppy to ensure correct operation at low current draws. They become unnecessary at higher loads. I'm sure it's mentioned on his site.
Thanks for the info on this, I checked out the Databook and it's a real goldmine of information. I see the reason for the extra diodes. The circuit is full of op-amps as mentioned in the book so I will add the extra diodes to the schematic. I'm not so sure I need the resistors as the circuit draws in excess of 250mA
Crystalline Stricture now on
tonnrecordings.com

User avatar
MikeDB
Common Wiggler
Posts: 129
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: Calling Power supply gurus

Post by MikeDB » Sun Sep 06, 2020 11:33 am

wahee wrote:
Sat Sep 05, 2020 7:23 pm

this is great. I have learned so much in this post. The 2 rectifier design does seem a little OTT.
Essentially I need this design to run off eurorack or wallwart AC
I have created a very basic design with minimal smoothing, which I will add to, as it does seem to be a false economy to scrimp on the power supply
I was wondering about the purpose of the 2x 2.4k resistors. Are they for current limiting

thanks

minimal smoothing.JPG
What you may like to consider is using more modern devices - the LM2940/LM2990. Gives better protection and can survive misuse far more - see the TI datasheets.

My point about the protection diodes was there are few applications where you are likely to get both positive and negative voltages coming into the supply from the loads, so you would normally consider what is likely, if indeed either. Usually it's just positive voltages stored on large capacitors downstream of the PSU during power down, which the LM29xx devices handle far better.
Personally I hate cats !

User avatar
wahee
Common Wiggler
Posts: 212
Joined: Fri May 29, 2015 5:31 pm
Location: Belfast

Re: Calling Power supply gurus

Post by wahee » Mon Sep 07, 2020 6:58 am

That's great I'll have a look at those devices. I think they are fixed output also.
The circuit is basic synth so nothing major in the capacitor department, just a few 220uf 16v
Thanks for your help with this.
Crystalline Stricture now on
tonnrecordings.com

User avatar
EATyourGUITAR
has no life
Posts: 5776
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: Calling Power supply gurus

Post by EATyourGUITAR » Mon Sep 07, 2020 8:40 pm

FYI the wallwart is a transformer. there is no way around it. switch mode supplies have transformers. there are designs that do not have transformers. a buck converter for example. I would not show you how to do it because those systems can be deadly if they are not used properly. something you will touch with your hands that has metal everywhere is a bad bad time without galvanic isolation. swap the Line and Neutral (US), then plug in a second piece of gear that is mains earth referenced, BANG. if your body was the conductor between these two live wires it could be lethal.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
wahee
Common Wiggler
Posts: 212
Joined: Fri May 29, 2015 5:31 pm
Location: Belfast

Re: Calling Power supply gurus

Post by wahee » Tue Sep 08, 2020 6:23 am

EATyourGUITAR wrote:
Mon Sep 07, 2020 8:40 pm
FYI the wallwart is a transformer. there is no way around it. switch mode supplies have transformers. there are designs that do not have transformers. a buck converter for example. I would not show you how to do it because those systems can be deadly if they are not used properly. something you will touch with your hands that has metal everywhere is a bad bad time without galvanic isolation. swap the Line and Neutral (US), then plug in a second piece of gear that is mains earth referenced, BANG. if your body was the conductor between these two live wires it could be lethal.
Thanks I am aware of that, I should have stated I wanted to avoid a 'PCB mounted transformer'
Crystalline Stricture now on
tonnrecordings.com

User avatar
Sandrine
Super Deluxe Wiggler
Posts: 2280
Joined: Tue Jun 09, 2015 10:28 pm
Location: BC Canada
Contact:

Re: Calling Power supply gurus

Post by Sandrine » Wed Sep 09, 2020 9:18 am

I would put in a second set of caps going the other way, electrolytics don't like being reversed, they tend to explode and stuff
Why not just rectify out to positive and use an isolated DC-DC module for the -12 (and post regulate both of course)?
I have done that several times

KSS
Super Deluxe Wiggler
Posts: 3413
Joined: Mon Jan 25, 2016 7:28 am

Re: Calling Power supply gurus

Post by KSS » Thu Sep 10, 2020 1:20 am

@Sandrine
There's no need for bipolar caps in the halfwave version. They are being 'filled' at a greater rate than being drained and therefore -combined with the pulsed DC of rectification- they don't see anything untoward to worry about.

It *is* still useful to employ capacitors *designed* for PSU filter use. The difference in life can be huge between a general use electrolytic and one designed for pulsing DC.

As I said originally, a PSU is not the place to try and save by using cheap capacitors. But neither should you overspend for some sort of audiophool capacitor magic.

------------
Most electrolytics can actually withstand up to about half their rated WV in the reverse direction. Best when this is an AC usage, but some caps have been discovered by techs in decades old gear which were soldered in backwards when mfd. <--manufactured, not capacitance. But the duality of language is appreciated and so it's left as first typed.
Treat utility modules as stars instead of backup singers.

Treat power supplies like Rockstars instead of roadies!

Post Reply

Return to “Music Tech DIY”