Maker Pro
Maker Pro

Complete beginner help with finishing linear DC supply

Kyleag89

Oct 9, 2023
17
Joined
Oct 9, 2023
Messages
17
Hello everyone I've been learning more about circuitry and how to make diy projects using components I have laying around. I'm trying to utilize a center tapped transformer I salvaged out of a 1980s Kenwood hifi AV receiver. It's a 120v input with a 40 0 40 and 18v output(5 wires total). I removed the capacitors and diodes from the board and reassembled them into a bridge rectifier and now have 52vdc from that. What I would now like to do is turn this into an adjustable power supply to use to power various audio amplifiers, pre-amps, and tone controls for bench testing. What is the best way to go about achieving this? I have no idea what amount of current this is capable of either. I've tried looking up the transformer using the numbers on it but it seems impossible to find data on any of the transformers I've tried looking up from the random numbers on it.
 

kellys_eye

Jun 25, 2010
6,514
Joined
Jun 25, 2010
Messages
6,514
I removed the capacitors and diodes from the board and reassembled them into a bridge rectifier and now have 52vdc from that.
Can you show us how you did this?

The centre tapped transformer could deliver a dual-rail PSU solution which is often required for hi-fi pre-amp and output stage voltages. The difficulty is that you have a rather high DC level which makes the final design more complicated than it needs to be.

There are plenty of 3-pin regulators (variable) that can accept a 36V input (some can go slightly higher) and kits that adopt these devices are the simplest way to get to where you want to be. Buck-boost modules and even fully-made variable supplies are often cheaper than making your own.
 

Kyleag89

Oct 9, 2023
17
Joined
Oct 9, 2023
Messages
17
16968720806416634135328622303224.jpg

This transformer is massive. It's 4"x4"x3" tall and about 5 pounds.
 

Kyleag89

Oct 9, 2023
17
Joined
Oct 9, 2023
Messages
17
The capacitors are ELNA 56v 10,000uf and diodes are labeled S3V with another 3 after that has a horizontal line above it. I thought it might be useful in making a PSU that can power some 48vdc amplifier boards I want to test. I have 2 subwoofers that have blown plate amps I need to replace.
 

kellys_eye

Jun 25, 2010
6,514
Joined
Jun 25, 2010
Messages
6,514
Show how you connected the secondary of the transformer to the rectifiers/capacitors. The 40-0-40 secondary would be the red-black-red wires.

Note that the shortest wire between the two capacitors are likely unconnected and simply there to offer mechanical support to the soldering process.

Properly connected you would see +36V(ish) and -36V(ish) across each capacitor. The transformer looks to be capable of delivering a maximum of 2A or 3A so the original amplifier might have been 50 to 100 watts/channel.
 
Last edited:

Kyleag89

Oct 9, 2023
17
Joined
Oct 9, 2023
Messages
17
99% sure I had the black connected where the capacitors are soldered together neg to positive. Then the 2 red wires soldered where the ends of the diodes are soldered together(area I'm pointing to in pic). One red wire to each of the 2 "bridges". That would be in between the 2 diodes used for each wire.16968767934714648370420161945369.jpg
 

kellys_eye

Jun 25, 2010
6,514
Joined
Jun 25, 2010
Messages
6,514
That's about right but getting 52V seems rather high - it is, of course, not under any load but the capacitors are rated at only 56V and that seems too close for comfort. Adding a load shouldn't see a drop from 52V to the expected 36V(ish) so something seems amiss.

Make sure you measure from the black (common 0V) to the positive or negative of each capacitor.
 

Kyleag89

Oct 9, 2023
17
Joined
Oct 9, 2023
Messages
17
I just noticed that 56v rating while I took the pics for you and realized that they're not really high enough for 52 0 52. I don't know what I did wrong so I am going to temporarily hook it back up today and check my measurements again. You are spot on about the amplifier it came from, it was an old Kenwood 2 channel rated at 80 watts rms per channel! I did put that short wire between the 2 dummy contacts to give mechanical stability while working on them. I didn't have any solid copper wire so I cut the ends off of 45v scohtky diodes and used them. Hence the 2 pieces twisted together to make a long enough piece to connect - to + on the caps. If it would give me 36v ish out that would be much better because one of amplifiers I'm looking at requires a 30v to 50v supply with 48v being preferred for max output power. Giving it 36v however should be sufficient for my 8" subwoofer that plays well with between 80-120 watts rms.
 

Kyleag89

Oct 9, 2023
17
Joined
Oct 9, 2023
Messages
17
If the transformer is 120v 3a input does that mean I can only draw 3a or less from the rectified DC output?
 

Delta Prime

Jul 29, 2020
1,979
Joined
Jul 29, 2020
Messages
1,979
Your title: Finishing linear DC supply.
You have not even started!!!
Now you want variable?
You're all over the place! You have to relax, take a deep breath.
You have to focus; you should be asking what about short circuit protection?
What kind of fuse?
Transient protection (also known as) High voltage, high current spikes, from Mains connection. Thermal management am I not going to need a fan? heat sinks for components?

And what the hell is this crap!
1696888055349.png

Should be asking about safety!!
If you're not going to kill yourself you might kill someone else and it will hurt the whole time you are dying!!! :mad:
This should all be in a proper enclosure your an accident waiting to happen.
 

Attachments

  • LinearRegulators.pdf
    448.8 KB · Views: 2
Last edited:

kellys_eye

Jun 25, 2010
6,514
Joined
Jun 25, 2010
Messages
6,514
If the transformer is 120v 3a input does that mean I can only draw 3a or less from the rectified DC output?
Those parameters give some indication of the power rating of the transformer (VA) and equates to 360 watts but the 3A rating is usually overstated so I'd give that transformer a potential 250VA rating. On the secondary side, after losses, this might deliver 200-ish VA to a power supply such as yours so at +/-36V you might see a maximum of 3A - this fits nicely with the 80W per channel spec. Either way, you don't want to over load the PSU so don't try powering more than 50-75 watts per channel if you go that way.
 

Kyleag89

Oct 9, 2023
17
Joined
Oct 9, 2023
Messages
17
And what the hell is this crap!
View attachment 60969

Should be asking about safety!!
If you're not going to kill yourself you might kill someone else and it will hurt the whole time you are dying!!! :mad:
This should all be in a proper enclosure your an accident waiting to happen.

What part of temporary for testing didn't you understand? Been doing things like this for over 10yrs and haven't had anything bad happen yet so I guess I'm ok. If an accident does happen to kill me than oh well it would be doing me a favor. Not going to invest all the time buying and assembling it all into an enclosure until I know I can achieve what I want out of it. I assembled it tested it then even though not done testing disassembled it until I was ready to work on it again. What I'm learning from kelleys_eye is that it makes sense to toss the transformer and buy a smps.
 

AnalogKid

Jun 10, 2015
2,894
Joined
Jun 10, 2015
Messages
2,894
What I'm learning from kelleys_eye is that it makes sense to toss the transformer and buy a smps.
Depends on what your goal is. For signal circuit testing, the transformer you have is way overpowered, but that's not automatically a bad thing. There are several ways to limit the max current a power supply can deliver to a circuit, so you can throttle back the current to something like 500 mA. That way, if there is a circuit fault, you don't have amperes of current running around where it shouldn't be.

The split winding clearly is for the power amp section. An issue is the max voltage out of the transformer. It is so high that the usual chips people recommend for something like this can't handle it. Of course, there are ways around that, too.

A bench supply usually had adjustability down to 0 V. A 0 - 10 V voltage range and 0 - 1 A current range are common, and switching regulators do not do any of that well. Do you want to build a supply from scratch, wire up pre-assembled modules, or buy a complete supply?

For any of those options, start with what you want, then we can see if it is possible with what you have.

Bipolar (+/-) outputs?

Max output voltage?

Max output current?

Tracking?

ak
 

Delta Prime

Jul 29, 2020
1,979
Joined
Jul 29, 2020
Messages
1,979
No!
Linear power supplies have high voltages and currents in the transformer and the rectifier, which can cause electric shocks and burns. They also have large capacitors that can store lethal charges even after unplugging the device from the AC outlet. :mad:
I guess this one gets a Darwin award!
Presented by @AnalogKid
(Go Big Blue!)
:cool:
1696915520113.png
 
Last edited:

kellys_eye

Jun 25, 2010
6,514
Joined
Jun 25, 2010
Messages
6,514
That constructional layout seems perfectly acceptable to me and would even work in a 'finished' device when suitably boxed/housed.

If a PSU is derived from a mains AC source, as many SMPS devices also are, then the dangers are no different in terms of voltage and current. If it was down to preference I'd chose a linear supply every time - especially with audio applications where keeping noise to a minimum is essential and very difficult to achieve using SMPS designs.

You're taking this too seriously and for what reason?
 

Kyleag89

Oct 9, 2023
17
Joined
Oct 9, 2023
Messages
17
I hooked everything back up last night and I'm still alive to tell you the results! Lol I didn't stick my fingers or tongue on the diodes or capacitors either...

Anyway, I soldered one red to each pair of diodes in between them and the black to the -/+ connection from cap to cap. I measured 41 0 41 AC directly from transformer which was plugged into my 114v (at the time) outlet. Then measured 51vdc+ and 51vdc - on each terminal the diodes are connected to at the caps! What am I doing wrong?

BTW it I end up using this transformer I will desolder the mess I made and mount everything more neatly inside an enclosure. I got the idea of hooking it up this way by the few YouTube videos I watched with other people testing it exactly this way.
 

Delta Prime

Jul 29, 2020
1,979
Joined
Jul 29, 2020
Messages
1,979
You're taking this too seriously and for what reason?
I have witnessed the after effects first hand! I was powerless to help her as she convulsed on the floor before me.This was before the Advent of the portable defibrillators.
I Will stand down!
Also... I was wrong to mention
A well-respected member of multiple electronic community sites. @AnalogKid
My motivation was to bring attention to a very real dangerous careless design practice. We can't save them all can we...
 
Last edited:

AnalogKid

Jun 10, 2015
2,894
Joined
Jun 10, 2015
Messages
2,894
I measured 41 0 41 AC directly from transformer which was plugged into my 114v (at the time) outlet. Then measured 51vdc+ and 51vdc - on each terminal the diodes are connected to at the caps! What am I doing wrong?

BTW it I end up using this transformer I will desolder the mess I made and mount everything more neatly inside an enclosure. I got the idea of hooking it up this way by the few YouTube videos I watched with other people testing it exactly this way.

Nothing. A diode and capacitor form a peak detector. The transformer output has a low impedance, so it charges the cap about as fast as it rises. When the input is on the downside of the sinewave and there is nothing connected to the cap to discharge it, it holds the peak voltage and the diode disconnects it from the transformer because it now is reverse biased. The relationship between a sine wave's RMS value and peak value is sqrt 2. Your numbers are close enough.

And, I've seen some of those videos. That wire-frame wiring style is very popular in India diy circles.

ak
 
Top