Author Topic: Drawing too much current from AC to AC wallwart?  (Read 1581 times)

0 Members and 1 Guest are viewing this topic.

Offline K3mHtHTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: ca
Drawing too much current from AC to AC wallwart?
« on: January 17, 2022, 12:56:51 am »
Hey everyone,
Ayy.. so... I recently shipped about a dozen little widgets that are powered by a 16VAC adapter. The voltage inside gets rectified and turned into a +/-15VDC rail (LM317/LM337). I was measuring the current draw of the actual rails and never bothered to measure the current draw out of the 16VAC adapter until one of the first recipients of my widget said that their wallwart died after 20 minutes.

During testing, I left it on for weeks and my wallwart was fine so I never thought I was approaching the limit of the 16VAC adapter. Sure enough after testing it, I noticed I'm drawing 950mA from the wallwart that is rated for 1000mA.

Wallwart data sheet: https://www.jameco.com/Jameco/Products/ProdDS/100108.pdf

The wallwart shows a thermal fuse, so I assumed if there ever was a problem that would trip and then return back to useable state after it cooled down, but this was not the case. I'm waiting for the blown wallwart to be returned so I can do an autopsy, but in the meantime I have a few questions:

How bad is it that I'm drawing 950mA from a 1000mA rated AC wallwart? I know I'm not leaving a smart margin of error, and the widget definitely needs to be redesigned to lower the current - but I'm curious about the fate of the dozen widgets already out in the field.. maybe the one that blew was just a runt / weakest link? Or maybe I should expect all 12 to eventually give up the ghost?

I know, I'm definitely stupid for not measuring this before shipping.

 

Offline Kim Christensen

  • Super Contributor
  • ***
  • Posts: 1258
  • Country: ca
Re: Drawing too much current from AC to AC wallwart?
« Reply #1 on: January 17, 2022, 04:21:01 am »
Since a rectifier doesn't have a very nice power factor, you could be exceeding the VA rating of the wallwart by a lot more than you think. Can you post a schematic of the rectifier / filter cap portion of the circuit and how much current you are drawing from each +/-15VDC rail?
« Last Edit: January 17, 2022, 05:20:21 am by Kim Christensen »
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 8972
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: Drawing too much current from AC to AC wallwart?
« Reply #2 on: January 17, 2022, 04:41:49 am »
If it's a very unbalanced load, it could be DC biasing the transformer which would cause it to heat up more than usual. Is the adapter getting hot to the touch?
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline Kim Christensen

  • Super Contributor
  • ***
  • Posts: 1258
  • Country: ca
Re: Drawing too much current from AC to AC wallwart?
« Reply #3 on: January 17, 2022, 05:21:25 am »
I would guess you are using a fullwave voltage doubler type circuit to get +22Vdc and -22Vdc to feed your LM317 & LM337. If it's something like the circuit attached (ignore cap values), all you can really pull would be around 280mA off each rail.

« Last Edit: January 17, 2022, 05:30:41 am by Kim Christensen »
 

Offline TheMG

  • Frequent Contributor
  • **
  • Posts: 866
  • Country: ca
Re: Drawing too much current from AC to AC wallwart?
« Reply #4 on: January 17, 2022, 05:29:23 am »
The thermal fuse in those things is non-resettable. Once it's blown that's it, wall wart dead.

Also keep in mind high ambient temperature can cause the transformer to overheat sooner. Same thing if the wall wart is buried underneath a pile of junk and the heat has nowhere to go.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19342
  • Country: gb
  • 0999
Re: Drawing too much current from AC to AC wallwart?
« Reply #5 on: January 17, 2022, 10:19:47 am »
The thermal fuse in those things is non-resettable. Once it's blown that's it, wall wart dead.

Also keep in mind high ambient temperature can cause the transformer to overheat sooner. Same thing if the wall wart is buried underneath a pile of junk and the heat has nowhere to go.
Going from the ones I've opened: some have bimetal strips, which are resettable, others are just one-shot thermal fuses, but one should always assume the latter.

Hey everyone,
Ayy.. so... I recently shipped about a dozen little widgets that are powered by a 16VAC adapter. The voltage inside gets rectified and turned into a +/-15VDC rail (LM317/LM337). I was measuring the current draw of the actual rails and never bothered to measure the current draw out of the 16VAC adapter until one of the first recipients of my widget said that their wallwart died after 20 minutes.

During testing, I left it on for weeks and my wallwart was fine so I never thought I was approaching the limit of the 16VAC adapter. Sure enough after testing it, I noticed I'm drawing 950mA from the wallwart that is rated for 1000mA.

Wallwart data sheet: https://www.jameco.com/Jameco/Products/ProdDS/100108.pdf

The wallwart shows a thermal fuse, so I assumed if there ever was a problem that would trip and then return back to useable state after it cooled down, but this was not the case. I'm waiting for the blown wallwart to be returned so I can do an autopsy, but in the meantime I have a few questions:

How bad is it that I'm drawing 950mA from a 1000mA rated AC wallwart? I know I'm not leaving a smart margin of error, and the widget definitely needs to be redesigned to lower the current - but I'm curious about the fate of the dozen widgets already out in the field.. maybe the one that blew was just a runt / weakest link? Or maybe I should expect all 12 to eventually give up the ghost?

I know, I'm definitely stupid for not measuring this before shipping.
How much current does the DC side use?

Is it balanced?

You might be better off using a switched mode 15VDC power supply and a buck-boost to get -15V.
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 14067
  • Country: de
Re: Drawing too much current from AC to AC wallwart?
« Reply #6 on: January 17, 2022, 11:46:30 am »
For the current the RMS current is relevant. With the usually poor powerfactor it makes a difference if the meter to measure the AC current is RMS or average reading.

With a current drawn quite close to the maximum the supply would run quite hot and will have a higher failure rate.
 

Offline K3mHtHTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: ca
Re: Drawing too much current from AC to AC wallwart?
« Reply #7 on: January 18, 2022, 05:39:29 am »
Thanks for all the replies so far! Some replies and details:

For the current the RMS current is relevant. With the usually poor powerfactor it makes a difference if the meter to measure the AC current is RMS or average reading.

I'm using a Fluke 177 true RMS meter, so the AC current measurement should be accurate right?

How much current does the DC side use?

Is it balanced?

You might be better off using a switched mode 15VDC power supply and a buck-boost to get -15V.

The +15V rail uses 275mA
The -15V rail uses 236mA

I would guess you are using a fullwave voltage doubler type circuit to get +22Vdc and -22Vdc to feed your LM317 & LM337. If it's something like the circuit attached (ignore cap values), all you can really pull would be around 280mA off each rail.

Hi Kim, yup that is the circuit exactly.. but I think it's called a half-wave rectifier. And yup! I'm pulling dangerously close to 280mA from each rail (275mA on the positive side).


Since I'm pulling 95% of the recommended current from these wallwarts and would love to be below 80% as a safety factor. I can't really see how that would be possible with the circuit.. I've tried a few SMPS solutions but have had noise issues - I will revisit those. But honestly at this point, I might just switch to a 16VAC 2000mA wallwart - which I found for North America, and will look into MOQ's for getting European versions made as well.




 

Offline Kim Christensen

  • Super Contributor
  • ***
  • Posts: 1258
  • Country: ca
Re: Drawing too much current from AC to AC wallwart?
« Reply #8 on: January 18, 2022, 06:25:37 am »
Quote
Hi Kim, yup that is the circuit exactly.. but I think it's called a half-wave rectifier.

Yes, it is a halfwave for the positive supply and another half wave for the negative supply. Together, they load the transformer just like a full wave would.

Quote
But honestly at this point, I might just switch to a 16VAC 2000mA wallwart

That's the simplest to implement and probably the best solution at this point.

« Last Edit: January 18, 2022, 06:27:40 am by Kim Christensen »
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19342
  • Country: gb
  • 0999
Re: Drawing too much current from AC to AC wallwart?
« Reply #9 on: January 18, 2022, 11:20:27 am »
Thanks for all the replies so far! Some replies and details:

For the current the RMS current is relevant. With the usually poor powerfactor it makes a difference if the meter to measure the AC current is RMS or average reading.

I'm using a Fluke 177 true RMS meter, so the AC current measurement should be accurate right?

How much current does the DC side use?

Is it balanced?

You might be better off using a switched mode 15VDC power supply and a buck-boost to get -15V.

The +15V rail uses 275mA
The -15V rail uses 236mA

I would guess you are using a fullwave voltage doubler type circuit to get +22Vdc and -22Vdc to feed your LM317 & LM337. If it's something like the circuit attached (ignore cap values), all you can really pull would be around 280mA off each rail.

Hi Kim, yup that is the circuit exactly.. but I think it's called a half-wave rectifier. And yup! I'm pulling dangerously close to 280mA from each rail (275mA on the positive side).


Since I'm pulling 95% of the recommended current from these wallwarts and would love to be below 80% as a safety factor. I can't really see how that would be possible with the circuit.. I've tried a few SMPS solutions but have had noise issues - I will revisit those. But honestly at this point, I might just switch to a 16VAC 2000mA wallwart - which I found for North America, and will look into MOQ's for getting European versions made as well.
I wouldn't worry about the current being 95% of the PSU's rating, which will result in 90% of the copper losses, compared to full load, as P = I2R. 41mA of DC in the transformer, is more of a problem.

I'd advise using a 30VDC PSU and a rail splitter, such as the TLE2426.
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 6732
  • Country: pl
Re: Drawing too much current from AC to AC wallwart?
« Reply #10 on: January 18, 2022, 02:13:22 pm »
Why would you do that?

Dual half wave rectification is efficient when load isn't equal on the rails. Not sure if TLE2426 has thermal shutdown, it could probably cook itself on ±15V if something goes wrong and it could run hot even if everything is right.

The only problem is that the rails don't come up simultaneously.
 

Offline K3mHtHTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: ca
Re: Drawing too much current from AC to AC wallwart?
« Reply #11 on: January 18, 2022, 05:44:10 pm »
Here's a scope shot of the 16VAC going into the instrument under full load. Interesting that there's some spikes on the direction changes.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19342
  • Country: gb
  • 0999
Re: Drawing too much current from AC to AC wallwart?
« Reply #12 on: January 18, 2022, 07:40:33 pm »
Why would you do that?

Dual half wave rectification is efficient when load isn't equal on the rails. Not sure if TLE2426 has thermal shutdown, it could probably cook itself on ±15V if something goes wrong and it could run hot even if everything is right.

The only problem is that the rails don't come up simultaneously.
Adding a rail splitter and using a 30V PSU, would avoid DC in the transformer, which is the downside of dual half-wave rectification, when the output currents aren't balanced.

The data sheet for the TLE2426 says it's short circuit proof, but it's only specified to 20mA, so I agree, it's not beefy enough. I suppose a couple of pass transistors could be added, to boost the output current a bit.
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 6732
  • Country: pl
Re: Drawing too much current from AC to AC wallwart?
« Reply #13 on: January 18, 2022, 08:16:30 pm »
Adding a rail splitter and using a 30V PSU, would avoid DC in the transformer, which is the downside of dual half-wave rectification, when the output currents aren't balanced.
Is it an issue though?

DC voltage on the primary is a known problem as it affects the magnetizing current.
But is there any problem with asymmetric load current?
I suppose mains output impedance is low enough that it doesn't cause much DC voltage to develop at power levels we are talking about here.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19342
  • Country: gb
  • 0999
Re: Drawing too much current from AC to AC wallwart?
« Reply #14 on: January 18, 2022, 09:34:08 pm »
Adding a rail splitter and using a 30V PSU, would avoid DC in the transformer, which is the downside of dual half-wave rectification, when the output currents aren't balanced.
Is it an issue though?

DC voltage on the primary is a known problem as it affects the magnetizing current.
But is there any problem with asymmetric load current?
I suppose mains output impedance is low enough that it doesn't cause much DC voltage to develop at power levels we are talking about here.
Yes, DC load current has a similar effect as DC voltage on the primary. It biases the core at DC, making saturation more likely, especially if it's run close to the maximum current, as is the case in the original poster's application.
 

Offline K3mHtHTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: ca
Re: Drawing too much current from AC to AC wallwart?
« Reply #15 on: January 19, 2022, 06:35:15 pm »
Yes, DC load current has a similar effect as DC voltage on the primary. It biases the core at DC, making saturation more likely, especially if it's run close to the maximum current, as is the case in the original poster's application.

Would I see this on the scope shot I posted? It seems very symmetrical..

Regardless, I have found some 2000mA wallwarts to replace the 1000mA one I've been using.. so I think I will switch to that and be happily at 50% of the rated current.


 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19342
  • Country: gb
  • 0999
Re: Drawing too much current from AC to AC wallwart?
« Reply #16 on: January 20, 2022, 09:46:45 am »
Yes, DC load current has a similar effect as DC voltage on the primary. It biases the core at DC, making saturation more likely, especially if it's run close to the maximum current, as is the case in the original poster's application.

Would I see this on the scope shot I posted? It seems very symmetrical..
That doesn't show the magnetic flux inside the core, just the secondary voltage.
Quote
Regardless, I have found some 2000mA wallwarts to replace the 1000mA one I've been using.. so I think I will switch to that and be happily at 50% of the rated current.
That's much safer.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf