Author Topic: Wiring two RD Tech DPS power supplies in series.  (Read 6338 times)

0 Members and 1 Guest are viewing this topic.

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Wiring two RD Tech DPS power supplies in series.
« on: February 13, 2018, 06:34:53 pm »
I would like to wire my two supplies in series so I can take the middle point as ground.

DPS5020 + DPH5005

I'm just a little wary though.  One is running from a battery, so is obviously floating.  The other does not appear to have a direct reference to ground.

If I connect the + of one PSU to the - of the other how will it take it?
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #1 on: February 13, 2018, 06:43:47 pm »
I did a quick test at 1V on each.  Both show 0.99V, 0.000A and measuring across them I have 1.99V on the DMM.

Should I start upping the voltage and hoping or should I use two tranformer based 12V wall warts to be safer?
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Online rstofer

  • Super Contributor
  • ***
  • Posts: 9890
  • Country: us
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #2 on: February 13, 2018, 07:08:18 pm »
Any decent power supply will have the outputs floating so that they may be connected in series of parallel.

Since you want a common center point ground, use two 470 Ohm resistors between the common side of the PS and the 'ground' point.  You should be able to vary the voltages on both sides but you will have limited the current just in case they supply outputs are grounded internally.

Finally, remove the resistors and you should be good to go.
« Last Edit: February 13, 2018, 07:10:05 pm by rstofer »
 

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #3 on: February 13, 2018, 07:21:58 pm »
Well, they haven't gone bang.  I have 24V across them and +12V GND -12V on the breadboard.  Zero current is flowing from either.

Fingers crossed!

Hopefully when I load them nothing dramatic happens.  I have both set to 100mA limit.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #4 on: February 13, 2018, 10:40:22 pm »
Well they performed fine, but I was only drawing about 10mA through them.

Possibly odd was that the 0 +12V supply was always showing more current than the lower -12V supply.

My circuit was just one op amp and an LED.  Both across both rails.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline not1xor1

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #5 on: February 14, 2018, 07:01:01 am »
I would like to wire my two supplies in series so I can take the middle point as ground.

DPS5020 + DPH5005

I'm just a little wary though.  One is running from a battery, so is obviously floating.  The other does not appear to have a direct reference to ground.

If I connect the + of one PSU to the - of the other how will it take it?

AFAIK to serialize 2 or more power supplies you must ensure they have diodes of appropriate power connected with reverse polarity in parallel with the outputs.

As you can easily see with a spice simulation, in some cases, the outputs of one of the PSUs might be reversed. The diodes prevent damage.
 

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #6 on: February 14, 2018, 09:16:51 am »
That makes no sense to me.  Maybe if there were in parallel and one current limited before the other, then the non-limiting PSU could back feed the limiting one due to higher voltage.

I'm not sure how two supplies putting out positive (from their perspective) voltage can reverse polarity.

Batteries are put in series all the time and they are much more sensitive to reverse polarity.

Finally both supplies have reverse polarity protection.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Kalvin

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #7 on: February 14, 2018, 09:49:21 am »
Interesting thread. I am thinking about similar solution, but I intend to use a transformer with two separate secondaries which allows series connection (similar to paulca's arrangement) or common ground connection (two DPS3005's with different output voltages and common ground).
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #8 on: February 14, 2018, 10:58:43 am »
Batteries are put in series all the time and they are much more sensitive to reverse polarity.

And the weakest cell often becomes reverse charged, - bad news if they are were rechargeable.
« Last Edit: February 14, 2018, 11:00:21 am by StillTrying »
.  That took much longer than I thought it would.
 

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #9 on: February 14, 2018, 11:14:08 am »
Batteries are put in series all the time and they are much more sensitive to reverse polarity.

And the weakest cell often becomes reverse charged, - bad news if they are were rechargeable.

And why you always use similar cells and in the case of NiMH make an effort to use similarly charged cells or in the case of Lithium use a balancer and cell monitor.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline glarsson

  • Frequent Contributor
  • **
  • Posts: 814
  • Country: se
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #10 on: February 14, 2018, 11:31:50 am »
It is just as important to make the two serially connected power supplies equal. The first supply to go into current limit will see a reverse current driven into it by the other supply. That is one reason for a reverse diode on the supply output.
 
The following users thanked this post: not1xor1

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #11 on: February 14, 2018, 11:40:46 am »
Do you have a drawing or schematic of what you mean?
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Maxlor

  • Frequent Contributor
  • **
  • Posts: 565
  • Country: ch
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #12 on: February 14, 2018, 11:50:35 am »
AFAIK to serialize 2 or more power supplies you must ensure they have diodes of appropriate power connected with reverse polarity in parallel with the outputs.
And the weakest cell often becomes reverse charged, - bad news if they are were rechargeable.
Neither of those apply to what paulca is doing, so he is quite safe without any diodes.

The diode protection on a power supply is needed when you have two supplies working against each other, say, when you're changing a battery. If the PSU is inadvertently switched off without being disconnected from the battery, the battery will try to feed power into the power supply, which might damage it. The problem only arises when you have the negative rail of one PSU connected to the negative rail of the other PSU (or battery), and the positive of one to the positive of the other.

The battery voltage inversion only happens when you have a good amount of current running through an empty battery. It happens if the voltage drop over the internal resistance is larger than the voltage created by the chemistry. This typically happens when you have several batteries in series, and most of them are still good enough to supply a decent amount of current; the current is simply pushed through the bad battery, and its internal resistance creates a voltage potential that is inverse to its normal voltage. This always happens with batteries of course, but as long as they're good the internal resistance is small enough for it not to overcome the voltage potential created by the chemistry.

Now paulca is using power supplies, not batteries, with internal resistance that is negligible and unchanging.

Or is it? What happens when we change paulca's setup (and to be clear paulca: this is not what you're doing, so you don't need to worry about it) and say, we set a current limit on the weaker PSU of 1A, but not the other, and then short the output of the combined serialized PSUs? The current-limited PSU will reduce the voltage of its output as much as it can, ideally to 0V, against the other supply, who'll try to force some voltage onto there until it hits its maximum current of 20A. Whether the current-limited PSU survives this depends on its design I reckon. I'd expect it to be ok as long as you're under the limited PSU's max current capability (so under 5A), but above that, damage will occur eventually.

This is where having OCP (Over-Current Protection) will save you. If the PSU sees more current than the OCP level, it won't just reduce its output voltage (as it does when you hit the current limit), it will disable the output thus breaking the circuit. OCP in the weaker PSU should be able to do that on its own, but to be on the safe side, I'd lower the OCP in the stronger PSU to match the capabilities of the weaker one.
 

Offline glarsson

  • Frequent Contributor
  • **
  • Posts: 814
  • Country: se
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #13 on: February 14, 2018, 11:52:07 am »
Like the attached image. Many supplies have the diodes built in.
 

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #14 on: February 14, 2018, 01:05:14 pm »
AFAIK to serialize 2 or more power supplies you must ensure they have diodes of appropriate power connected with reverse polarity in parallel with the outputs.
And the weakest cell often becomes reverse charged, - bad news if they are were rechargeable.
Neither of those apply to what paulca is doing, so he is quite safe without any diodes.

The diode protection on a power supply is needed when you have two supplies working against each other, say, when you're changing a battery. If the PSU is inadvertently switched off without being disconnected from the battery, the battery will try to feed power into the power supply, which might damage it. The problem only arises when you have the negative rail of one PSU connected to the negative rail of the other PSU (or battery), and the positive of one to the positive of the other.

I have wondered about this for charging batteries with a bench PSU.  Luckily all of mine seem to have survived having fairly large current capable batteries connected to their outputs without protection.  That would include a 150A capable LiPo and a 800A capable SLA.

I probably 'should' have diodes on them, but I only use the bench PSU when I want to exceed the 6A or so the Accucel puts out to fast charge a LiPo and a 10A diode is big, ugly and the voltage drop will mean I have to measure the battery locally and adjust the PSU.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #15 on: February 14, 2018, 03:19:09 pm »
Without the diodes, if the -3V side can't supply enough current because of its current limit, or it hasn't powered up yet, it gets up to 12V across it in reverse, and it might not like that.
.  That took much longer than I thought it would.
 

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #16 on: February 14, 2018, 03:21:34 pm »
So make sure the settings on the PSUs are identical and don't short or overload it. :)
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Kalvin

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #17 on: February 14, 2018, 03:32:07 pm »
So make sure the settings on the PSUs are identical and don't short or overload it. :)
Just add two diodes as glarsson suggested, and the outputs are protected from the potential polarity inversion StillTrying was describing.  :-+
 

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #18 on: February 14, 2018, 04:04:31 pm »
But diodes are irritating.  They drop voltage which varies in bizarre ways with current and temperature and for high current ones they are expensive and large. Not to mention having to solder up leads that include them etc.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Kalvin

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #19 on: February 14, 2018, 04:32:40 pm »
But diodes are irritating.  They drop voltage which varies in bizarre ways with current and temperature and for high current ones they are expensive and large. Not to mention having to solder up leads that include them etc.

No, those protection diodes are connected in reverse direction, directly across the module's output (anode to 0V output and cathode to positive output). The diodes are there only to protect the modules from voltage reversal as described by StillTrying above. See glarsson's diagram above for actual wiring. Those diodes are not conducting normally when things are okay, so there won't be any extra voltage drop or any other current /temperature dependent behavior as would happen with a series protection diode which you are probably referring to.
 
The following users thanked this post: analogo

Online paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4051
  • Country: gb
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #20 on: February 14, 2018, 04:57:34 pm »
I see.  I believe my power supply already has those internally.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline glarsson

  • Frequent Contributor
  • **
  • Posts: 814
  • Country: se
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #21 on: February 14, 2018, 07:06:02 pm »
I see.  I believe my power supply already has those internally.
Believe is not enough. The magic smoke might escape.

When attempting to connect power supplies in parallell or series it is important to think through all conditions that might cause problems and verify that the power supplies are properly protected. It is not enough to "adjust the power supplies to the same voltage and current" as conditions might change, i.e. by shorting one of the power supplies on the breadboard. What does the manufacturer specify? Note that thinking through the setup is also valid if one of the power supplies is a battery, e.g. what happens when charging a big lead acid battery and the mains fails? Does the power supply burn?

Btw, sorry for the drawing posted earlier. It was drawn using a finger on a phone ...
 

Offline glarsson

  • Frequent Contributor
  • **
  • Posts: 814
  • Country: se
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #22 on: February 14, 2018, 09:22:24 pm »
Also, when connecting power supplies in series all power supplies must still be within its allowed voltage range with respect to ground. Taking a 12V supply and connecting it in series with a 300V supply (to get 0, 300V and 312V) might cause serious security issues.
 

Offline not1xor1

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #23 on: February 15, 2018, 06:26:02 am »
That makes no sense to me.  Maybe if there were in parallel and one current limited before the other, then the non-limiting PSU could back feed the limiting one due to higher voltage.

I'm not sure how two supplies putting out positive (from their perspective) voltage can reverse polarity.

Batteries are put in series all the time and they are much more sensitive to reverse polarity.

Finally both supplies have reverse polarity protection.

Sorry for my late reply. I've been offline since yesterday morning.

When you connect two PSU in series, even if they are built to track each-other, most of times, there will be a small difference in their current limit. So once you hit the limit of the first one the other will see a reverse voltage.



I'm attaching a real basic LTspice schematic. Just play with it, change current limit values, remove the diodes and safely see (no magic smoke in the virtual world  :)) what happens at the output terminals of the two PSUs.

« Last Edit: February 15, 2018, 06:28:21 am by not1xor1 »
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6987
  • Country: ca
Re: Wiring two RD Tech DPS power supplies in series.
« Reply #24 on: February 15, 2018, 06:51:41 am »
I see.  I believe my power supply already has those internally.

The DPS power supplies each have an N-ch MOSFET in-series with the negative output terminal.
So the two reverse-diodes are absolutely needed here, they are not present.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf