# Power supply series connection

Discussion in 'Power Electronics' started by Viju, Dec 15, 2017.

1. ### Viju

12
0
Dec 15, 2017
Hi,
Can I connect a 19 v and a 12v supply in series. What should I take care about if it can be done. Also please advice on the type of diodes and diode name I have to use. A connection diagram would be appreciated.

2. ### Audioguru

3,440
738
Sep 24, 2016
If the outputs are DC and are completely separate (not each one grounded to earth) then simply connect the + of one of them to the - of the other to get 31V. Why do you need diodes?

3. ### Viju

12
0
Dec 15, 2017
Thx for the reply. The diodes came into question as when I googled series connections, everyone was taking about diodes needed to protect one power supply from the other if one starts giving output early. They were talking about some reverse biased diodes I think. Also if it will a problem if one power supply is 1A and the other is 2A.

4. ### kellys_eye

4,282
1,152
Jun 25, 2010
If one PSU is lower amperage then the total MAXIMUM amperage will be equal to the lowest supply value. In your example you could get the 31V at 1A maximum.

Diodes are usually specified when you parallel PSUs - in which case the total amperage is additive (i.e. the sum of all the PSU amperage maximums).

5. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,500
2,840
Jan 21, 2010
Placing diodes across the power supplies is also a very good idea when placing them in series.

If you short out the combined supplies, it prevents one power supply from driving the other with reverse polarity.

6. ### AnalogKid

2,543
725
Jun 10, 2015
Every time the supplies are powered up, one of might reach its output voltage before the other one is powered up and stable with a low output impedance. Whichever one comes up first will appear as a reverse voltage on the output of the slower one, through the load. Depending on how much current the load can pass backwards when underpowered, this might be a total non-issue. OR, it might kill one of the supplies after a small number of power-on cycles. Reverse-polarity diodes prevent the voltage impressed backwards on the slow supply from exceeding one diode Vf.

Again, the current through the protection diode *might* be the full rated output current of the faster supply, so the diode might have to be large.

ak

Last edited: Dec 15, 2017
(*steve*) likes this.
7. ### Viju

12
0
Dec 15, 2017
Thank you all for chipping in. I think I should go ahead and use diodes for protection. Exactly what type of diodes should I use, any model number etc. I need to power a mono amp with 32v and all I could find from scavenging through scrap merchants is a 19v 2a and a 12v 1a switched power supply. The amp is a TDA7498 Of rated output power of 150w.

8. ### AnalogKid

2,543
725
Jun 10, 2015
If you want the full output of 150 W, you will need power supplies that can deliver 4.5 A minimum, and preferably more like 6 A each. But with 1 A supplies you can determine if the rest of your circuits work correctly and get some volume. for what you have now, something in the 1N5401 - 1N5404 range will work well. It is an older part, but very reliable.

https://www.diodes.com/assets/Datasheets/ds28007.pdf

ak

9. ### Audioguru

3,440
738
Sep 24, 2016
You ain't getting 150W.
The datasheet of the TDA7498 shows that it is overloaded with 4 ohm speakers but with your 31V supply its output per channel into 8 ohm speakers is 70W when it is clipping like mad with a horrible sounding 10% distortion. A graph shows 35W per channel at a more reasonable 0.2% distortion. It has class-D efficiency so its max current when barely clipping is 2.3A. Your 1A power supply will curl up and die when the output power is more than only 16W per channel.

10. ### Viju

12
0
Dec 15, 2017
Ok. The TDA7498 is bridged mono and spec shows a rated power output of 130w. I have a 10 inch 100w rated subwoofer single voice coil which I intend to use with the amp. Please suggest a power supply spec from which I can get the max out of the amp and sub. The sub is 4 ohm.

11. ### Audioguru

3,440
738
Sep 24, 2016
No. The TDA7498 is a stereo amplifier that already has both of its two channels separately bridged. You cannot bridge them again into mono. In its datasheet I see nothing about 130W. I also see nothing about it driving a 4 ohm load, it shows 6 ohms and 8 ohms.

Oh, the document for their "demonstration board" is different to their datasheet. It shows stereo outputs of 130W per channel into 4 ohm loads when the supply is 31V and is clipping like crazy at about 15% distortion. Each channel produces 41W into 4 ohms at 0.2% distortion with your 31V supply. I do not know how you can parallel their outputs for a little more power.

12. ### Viju

12
0
Dec 15, 2017
I brought this from AliExpress.
TDA7498 150W Subwoofer amplifier board Mono Audio Bass Digital power amplifier
http://s.aliexpress.com/faUvmYrY
(from AliExpress Android)

13. ### Audioguru

3,440
738
Sep 24, 2016
Ali is a Chinese company. They do not know how to speeky zee Engrish! They exaggerate!
One place on their ad says, "Maximum power per channel: 130W", another place says, "100W + 100W" and another place says, "Channels: 1". They do not say one word about how much clipping distortion. They do not say if they connected the two channels in parallel.

14. ### ramussons

385
73
Jun 10, 2014
Hmmmmmmmm......... how about an example ?

15. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,500
2,840
Jan 21, 2010
If you connect two power supplies in series, then short the output, it is like connecting two power supplies together positive to negative and negative to positive.

This is worse than simply shorting the power supplies. In this case, rather than the positive rail being forced to ground, it is forced below ground. For a very simple series regulator this can defeat a current limit.

Placing reverse biased diodes across the outputs means that in the worst case, the positive output can only go a diode drop below ground. This will typically allow even a simple shunt regulator to limit the output current.

16. ### ramussons

385
73
Jun 10, 2014
Ah.. YES. I get it. Thanks a ton.

17. ### Viju

12
0
Dec 15, 2017
If the power supplies are exactly the same type (same voltage, amperage, etc.) and make does it need diode protection. What if both are the same model laptop chargers?

18. ### ramussons

385
73
Jun 10, 2014
I think it does call for a protection. 2 similar power units are not likely to behave exactly the same under a short circuit / overload condition.

19. ### Viju

12
0
Dec 15, 2017
Thanks a lot. I found something on eBay which will suit my need. It's a 15v 5a laptop charger. I am planning to get 2 of these and use in series. I am just waiting to hear what u guys think about the diode I should use (model number, type). Any circuit diagram would be appreciated.

20. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,500
2,840
Jan 21, 2010
Diodes rated at moire than 30V and more than 5A would be best.

Even some 1N5408's would probably be fine. They will be called on very rarely (if at all) and their short-circuit failure mode will probably protect the power supply (as long as it is protected against short circuit) if s large reverse voltage appears at the output.