Safe In-Wall AC to DC Transformers??

[rvendrame]

@petewill . the MOV goes in parallel with HLK input (and not in series like you did).

It was suggested to use .2A fuses (instead the .75A). And the Thermal fuse should be glued on HLK top (my guess).

[petewill]

@rvendrame said:

@petewill . the MOV goes in parallel with HLK input (and not in series like you did).

It was suggested to use .2A fuses (instead the .75A). And the Thermal fuse should be glued on HLK top (my guess).

Dang, totally forgot about all that in my rush to finish the diagram before lunch ended... That's why you guys design the electronics and I make the videos

How necessary are the .2A fuses? I couldn't find any on ebay at a reasonable price. Is the goal of this fuse to prevent large surges (like lightning) from hitting the MOV, or something else? If it's to prevent large surges would the .75A work ok?

Also, I was able to find cheaper varistors. I'm thinking 150VAC should work ok because the HLK is designed to handle 240VAC but maybe I'm off on that?
http://www.ebay.com/itm/5-x-Zinc-Oxide-Varistor-150VAC-15J-1200A-7mm-FREE-SHIPPING-/321024816822?hash=item4abe91f6b6

How does this look?

[Sparkman]

@petewill said:

Also, I was able to find cheaper varistors. I'm thinking 150VAC should work ok because the HLK is designed to handle 240VAC but maybe I'm off on that?

As long as you only feed it with 120VAC, there's no issue with that. For those that will feed it with 240VAC, they need to use it with a MOV rated above that.

Cheers
Al

PS Here's an option from Digikey: http://www.digikey.com/product-detail/en/MOV-10D241K/MOV-10D241K-ND/2407562

[Sparkman]

@petewill said:

How necessary are the .2A fuses? I couldn't find any on ebay at a reasonable price. Is the goal of this fuse to prevent large surges (like lightning) from hitting the MOV, or something else? If it's to prevent large surges would the .75A work ok?

The fuse serves two purposes, one to protect if the current draw of the power supply exceeds its rated capacity and the other to blow if the varistor starts conducting a large amount of current in a spike situation. A .75A will still provide protection, but will take a longer time to blow. You want to make sure the fuse doesn't blow because of the in-rush current at start-up, which a fast-blow fuse may do. Typically you want the fuse to be sized at about 150% max of the rated capacity so I would not exceed .3A.

Cheers
Al

[petewill]

@Sparkman

As long as you only feed it with 120VAC, there's no issue with that. For those that will feed it with 240VAC, they need to use it with a MOV rated above that.

I am in the USA so I will only be feeding it with 120VAC, but I will make sure to note that anyone using 240VAC will need a different value varistor.

the fuse should be after the varistor in the circuit

Ok, I'm still a little confused. Are you saying it should be 120VAC -> Varistor -> Fuse -> HLK? Or, 120VAC -> Fuse -> Varistor -> HLK? I thought it was the second one but maybe I misunderstood.

Also, I found some fuses that are rated at 300mA! Not a bad price either!
http://www.ebay.com/itm/Ceramic-Slow-Blow-Fuse-3-6-x-10mm-Axial-Leads-125V-250V-0-1A-6-3A-10-30pcs-/111433875797?var=&hash=item19f1fa0155

I think I am almost ready to start ordering parts. I am excited for this build! If anyone else has any feedback please let me know. Thanks!

[Sparkman]

@petewill said:

@Sparkman

the fuse should be after the varistor in the circuit

Ok, I'm still a little confused. Are you saying it should be 120VAC -> Varistor -> Fuse -> HLK? Or, 120VAC -> Fuse -> Varistor -> HLK? I thought it was the second one but maybe I misunderstood.

Sorry, the wording I used was unclear and was based on the 120VAC being on the right in your diagram :-). Yes, it is 120VAC -> Fuse -> Varistor -> HLK.

Cheers
Al

[petewill]

@Sparkman Great, thanks! The parts have been ordered. Can't wait!

[Fabien]

@ceech said:

This one is not "safe", but it is an In-Wall AC to DC converter. Transformerless. With a 3A Solid state relay:

The converter output is 3.3V at 100mA and the solid state relay is a Triac.

Do you have some shcematic of this board ? I'd like to adapt it for 2 relays

[Cliff Karlsson]

Can anyone give me ebay-links to the parts I need to order to use the HLK with 230V.

I am getting a little confused about all the links.

[rvendrame]

This is the one I sent to be reviewed:

http://www.aliexpress.com/item/Free-Shipping-1pcs-lot-HLK-PM01-AC-DC-220V-to-5V-mini-power-supply-module-intelligent/32408565688.html

[Cliff Karlsson]

Ok thanks, but I meant the fuses, varistors and stuff.

[petewill]

@Cliff-Karlsson
I haven't tested any of these parts yet but this is what I ordered. I am in the USA so this is spec'd for 120 VAC. If you're using 240 you will need to change the size of the Varistor but everything else should be fine for 240.

Varistor for 120VAC - http://www.ebay.com/itm/321024816822?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

73°C Thermal Fuse - http://www.ebay.com/itm/221560426284?_trksid=p2057872.m2749.l2649&var=520415979885&ssPageName=STRK%3AMEBIDX%3AIT

250V 300mA Slow Blow Fuse - http://www.ebay.com/itm/111433875797?_trksid=p2057872.m2749.l2649&var=410420838583&ssPageName=STRK%3AMEBIDX%3AIT

HLK-PM01 - http://www.ebay.com/itm/351418782712?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

[ceech]

@Fabien @DrJeff Here is a link to board and schematic files for the transformerless AC-DC converter:
https://github.com/ceech/AC_SR087

[MartinP]

Would this varistor work with 240v?
http://www.ebay.com/itm/10-x-Zinc-Oxide-Varistor-250VAC-60J-2500A-10mm-/260848704608?hash=item3cbbcb5c60

[DrJeff]

@ceech can't look at the files right now is this a dimmer switch that is transformerless?

[ceech]

@DrJeff Yes

[DrJeff]

@ceech Thanks yet again!

[petewill]

@MartinP Based on my limited knowledge, yes, that should work. Maybe someone else can chime in to confirm?

[Fabien]

thank you @ceech

[Sparkman]

@petewill said:

@MartinP Based on my limited knowledge, yes, that should work. Maybe someone else can chime in to confirm?

I would personally go with a higher voltage rating. Something like 15-20% higher than your normal voltage otherwise you may get regular nuisance trips. So for a 240 VAC system, I would use something close to 290V.

Cheers
Al

[petewill]

@Sparkman Thanks!

[RJ_Make]

[Yveaux]

@ServiceXp long evening ahead

[Steve Caster]

Hi all,

I was wondering, I want to be able to adress my blinds using mysensors. Using the relay, radio etc works fine, yet I am still wondering on this power supply. So using the above component I basically prevent fire by using the thermal fuse and an overload by using a regular fuse, the varistor and the HLK.

Now, I am using a Nano (I have still 15 of those) which appear to be using 230mA, the radio itself and a relay. The relay (from the spec sheet) uses 180mA. This should be ok to be used with the HLK shown above. Now, my main concern is, how do you make a all sit together nicely? Is there a PCB type of board you are using? Any pointers welcome

Kind regards

Steve

[filipq]

Hi Steve

this may be useful http://forum.mysensors.org/topic/1540/110v-230v-ac-to-mysensors-pcb-board

Regrads
Filip

[petewill]

@ServiceXp Nice! Let us know how it goes.

[filipq]

Another one http://forum.mysensors.org/topic/1586/roller-shutter

[cygnus]

I'm planning on using the HLK-PM01 with the extra precautions discussed here; gluing a thermal fuse on top of the HLK-PM01 and adding a fuse and varistor.

My only problem right now is that I'm unable to find a good source for 0.3A slow blow fuses with axial leads on eBay. The only variants I can find are without leads and it feels a bit unnecessary to add a fuse holder to the circuit since the fuse is meant to be non-replaceable. The ones that petewill posted a link to are out-of-stock. Does anyone have any good suggestion where to find good fuses to use?

[Cliff Karlsson]

@cygnus said:

HLK-PM01

+1

[Sparkman]

@cygnus said:

My only problem right now is that I'm unable to find a good source for 0.3A slow blow fuses with axial leads on eBay.

Looks like the .4A ones are in stock. I would just go with those.

Cheers
Al

[GertSanders]

found these:

http://www.aliexpress.com/item/Free-Shipping-25pcs-0-25A-250mA-250V-3-6x10mm-Fuse-Slow-Blow-Glass-Axial-Leaded/32452469177.html?spm=2114.01020208.3.138.0Cb2oJ&ws_ab_test=201556_3,201527_3_71_72_73_74_75,0_0

[Cliff Karlsson]

If I want to use a relay for in-wall use. Could I use the same Slow Blow fuse, thermal fuse and varistor to get it to be more safe?

[novicit]

This has been an excellent thread! I am left with one practical question though. How does one safely mount the thermal fuse on the HLK-PM01? Even if the thermal fuse thru hole on the PCB was exactly next to the middle of the HLK-PM01, the lead will still need to run 2+cm to the top of the HLP-PM01, then another 2+cm back down to the PCB. Only the thermal glue holds it in place. It seems it could easily come loose with minimal pressure - and remember it is 110 or 220V! Of course the leads would be protected by electrical tape or shrink tube, but it is still not firmly secured. Plus it is more likely that the leads on both sides would be closer to 3cm, plus the 1 cm for the fuse for a total of ~7cm of high voltage 'leads' not securely tied down to the PCB. I am only a hobbyist, so I am curious to hear others thoughts on this. Also, note that this board (http://forum.mysensors.org/topic/1540/110v-230v-ac-to-mysensors-pcb-board) has great protection, even fuse & varistor on the low voltage side. How would one mount a thermal fuse on the HLK-PM01 without having a 'loose' lead running to the top of the HLK? Or am I over concerned.

[GertSanders]

I would suggest that the PCB have a cutout UNDER the HLK-PM01, so that you could stick the termal fuse under the powermodule against the bottom.
If you look at the images of an opened powermodule, then you will see that the internal PCB is on top. I would expect the heat from any problem to be noticable on the bottom as well.
Could be a safer option (less long wires) ?

[TD22057]

This has been a great thread with a lot of good information. I'm still a little worried about putting something like this in my wall though. What are peoples thoughts on benefits of this approach versus using something like a Samsung cube charger which is about a 28 mm cube (without the prongs), handles 120/240 V, and is probably very safe?

[GertSanders]

My guess is that the HLK-PM01 is of the same quality, but it only gives 700mA @ 5V. On my repeater it never feels warm at all (and this is in a completely closed box).
You could add a temp sensor to monitor and set an alarm in your home control system, maybe a solution to feel more secure.

[lrtsenar]

What do you think about this ? 1.79$ 5V @ 700mA dim: 3cm x 2cm x 1.8cm :
http://www.ebay.fr/itm/321846469504?_trksid=p2060353.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

[GertSanders]

That is similar to the HLK-PM01, but without any of the certifications the HLK-PM01 received. And it is completely open, so unsafe in my view.

[pdey]

I use these in Australia:
https://www.clipsal.com/Trade/Products/Electrical-Accessories/Electronic-Accessories/USB-Charger

AS/NZ safety certified, etc.

[Chester]

@pdey What is the per unit cost of those? And do they fit easily inside the wall box say sideways to allow for arduino etc to also fit?

[Cliff Karlsson]

@Chester said:

@pdey What is the per unit cost of those? And do they fit easily inside the wall box say sideways to allow for arduino etc to also fit?

I looked around a little and they where pretty hard to find but the prices I found where 30-35 $/unit + shipping. To expensive for me.

[pdey]

@Chester: They're usually about $30-35 per unit. On the expensive side, but I'd prefer to have something buried in my wall that I know I won't have to pull out in 12 months time - and also that is unlikely to start a fire...

They're quite small. Designed to fit into the Clipsal faceplates for Cat5/Cat6 jacks.

[sundberg84]

http://www.ebay.com/itm/Ceramic-Slow-Blow-Fuse-3-6-x-10mm-Axial-Leads-125V-250V-0-1A-6-3A-10-30pcs-/111433875797?var=&hash=item19f1fa0155

Hi all! I tested the Slow BLow fuse after @m26872 was a bit suspissions and it didnt blow. Looks like its some sort of bi-directional TVS and not a fuse. I could "run" a 1000Watt vaccum on this "fuse" which at 240v/200mA should blow at 48 Watt.

Good find by m26872

[petewill]

@sundberg84 Thanks for posting this. I'll have to take a look at mine to see if I got the same thing. That sucks that you didn't get what you ordered. Can you post a link when you find some new ones that you plan on using?

[nunver]

@petewill Thank you very much for this excellent thread. It's great to see the community is in agreement on how to implement a safe power supply.

Just one comment; for 220V, shouldn't we opt for a lower amp rating fuse? If we stay with 300 mA, I am thinking we will not protect the device?

[dakky]

hmm have some questions too:

1. is this setup only needed/recommended for in wall setups? or do you use this for all ac/dc converters?
2. does someone has a picture for me to share, how this looks in reallive?
   I'm not shure how to build my node in this case
3. what kind of cables (profile) do you use for the "high voltage" parts?

[sundberg84]

@nunver - A question i have also asked myself doing my PCB - i dont know if it makes any difference but HLK has a maximum imput of 200mA (1A spikes) according to its datasheet. What is the logic here and why - anyone that can explain?

@dakky I think this would be just as great for a outside ad/dc converter, the main idea is to protect us but in-wall you dont have the ability to remove the heat made from the HLK that easy. Pictures from my PCB here (Not completed) but based on this thread.

[petewill]

@nunver I am by no means an expert in this (which is why I started the post :)) but it is my understanding that the fuses are there to protect your house (not sensor) from damage if something goes wrong. I can't remember where I read it but it was advised if the 300ma fuse blows the HLK should be replaced as a safety precaution.

@dakky

is this setup only needed/recommended for in wall setups? or do you use this for all ac/dc converters?

So far I have always used phone chargers for all my sensors which haven't been in the wall. I have never had an issue. I use the phone chargers because I can usually get them free from work when they are recycled. The HLK is much smaller though so it's a great option.

I'm not shure how to build my node in this case what kind of cables (profile) do you use for the "high voltage" parts?

Good question. I'd be curious to hear what others are doing. I haven't had time to build an in-wall sensor yet but I was planning to use 14 gauge wire because that's what is in most of my house.

[nunver]

@petewill I am no expert either, which is why I am reading this thread . It is true that we protect the home, not the device. But fire starts at the device if not properly protected. The rating of the device says absolute maximum 0.2 A. This translates to 24W for 120 V and 44 W for 220 V, if my math is correct. Since this is way above the supply rating of the device, we must consider 0.2A as really absolute maximum. So, if we want to protect, in my opinion, we need 0.2A fast acting fuse.

Again, not an expert, so please disregard this post if I make no sense

[sundberg84]

What about that the HLK can manage spikes up to 1A (according to datasheet). If so, this will blow the fuse even it its not necessary, atleast if you have a fast one?

[HenryWhite]

Here's some more valuable information about the HLK: http://lygte-info.dk/review/Power Mains to 5V 0.6A Hi-Link HLK-PM01 UK.html

[nunver]

@sundberg84 Correct, need to move the fuses out of varistor loop, between varistor and HLK.

[sundberg84]

@nunver - Its not a good idea to have the varistor before the fuses i think... we have discussed this above and here
Since varistors only can handle a short period with high load you need the fuses first.

[nunver]

@sundberg84 Yes you are correct there as well. Maybe two fuses like you have but different locations?

[TD22057]

FYI: Here's an emissions test of the HLK-PM01. I don't have any experience with this sort of thing so I'm not sure what to make of the results:
https://skippy.org.uk/quick-look-at-the-hlk-pm01/

[punter9]

this doesn't mention specifically but a good read

http://lygte-info.dk/review/Power Mains to 5V 0.6A Hi-Link HLK-PM01 UK.html

[petewill]

@TD22057 Thanks for posting! It's great to get more info on this. Maybe someone else can weigh in who has more experience but this is comforting "While it is not great, I would consider using it on my own projects..."

[TD22057]

One more note: There is also an HLK PM03 which outputs 3.3V.

[Soloam]

Hello, did any one attempted to make this project? what is the feedback? Is it safe? I was looking for something like this, but I must confess that this is a big step! I never liked using mobile chargers, I always asked my self if they are safe to use... If this is a step up to make my project safer, I would give it a go!

[sundberg84]

I did: https://www.openhardware.io/view/13/In-Wall-ACDC-Pcb-for-MySensors
But still there are many questions - I dont know what happens if it fails for example and if the components meant to protect it is good enough.
I have some temp test but not been able to test it with overload and stuff like that.

Also a relay board with HLK PM01: http://forum.mysensors.org/topic/1540/110v-230v-ac-to-mysensors-pcb-board

[Pierre P]

Do you want some tips to make "safe" crash tests ? Like putting it in a closed electrical box and switch on and off an old vacuum with no electronic ? Or make a shortcut make the relay ko... Because you cannot test that if you don't have a good electrical installation at home...

[sundberg84]

@Pierre-P That i have tested (the vacuum test) and the fuses blew... i was thinking about a long but not high overload.

[alexsh1]

@punter9 said:

http://lygte-info.dk/review/Power Mains to 5V 0.6A Hi-Link HLK-PM01 UK.html

Please see this:

https://skippy.org.uk/5v-acdc-converter-switch-power-supply-module-3w-700ma-industrial-voltage-regulators/

1. It does not comply with the UK standards (I live in the UK)
2. It is dangerous and not safe.

[sundberg84]

Yes, if you read the threads about this - we have alof of those examples... thats why we are using the HLK-PM01 and are having this discussion. That link is not about the HLK-PM01 @punter9

Edit: that guy you linked to tested the HLK as well:

"While it is not great, I would consider using it on my own projects, however it would need filtering on the mains side if I was going to use it in a project I was going to place on the market (read as sell), as under my control I can keep an eye on it causing unwanted interference with other equipment, this is why we have the limits set in the standards."

and if you read our conclusions from this thread its the same.

[alexsh1]

@sundberg84 Did you see it? The guy is using fuse (Farnell part number 2464451) and varistor (Farnell part number 1004390) for HLK-PM01. It is 0.8A fuse

[sundberg84]

@alexsh1 Yea i saw it - and the text say: "I got this generic even cheaper board" - its not a HLK he is testing...

[alexsh1]

@sundberg84 Yes it is - sorry I am now talking about this link https://skippy.org.uk/quick-look-at-the-hlk-pm01/

This is the fuse he is using
http://uk.farnell.com/bussmann-by-eaton/c310fc-800-r-tr1/fuse-pcb-0-8a-250vac-fast-act/dp/2464451?selectedCategoryId=&exaMfpn=true&categoryId=&searchRef=SearchLookAhead

This is a fast act 250V 0.8A fuse

[sundberg84]

@alexsh1 - Yea, ok - we have discussed that test above in this thread as well.
As far as I know the HLK is the best so far.

We also have another test where we have got different advice about fuses, (normal/temp) and varistor (see above).
Im accually using a 0,2A fuse for 240V. Read through this post if you are interested - its a great discussion.

[alexsh1]

@sundberg84 I have gone almost through the whole topic. Excellent discussion!

I think the difference is that you are using the slow blow fuse 0.2A and he is using fast acting 0.8A one. I really wonder which one would be more safe. I know slow blow fuses are used with inductive load (like motors), but not sure about this application.

I know it has been mentioned here as well, but did you make any provision for a thermal SEFUSE (say, 73C) or do you think it would be too much?

[m26872]

Let's hope skippy.org.uk also will recommend some good EMC line filters for the HLK-PM01 soon.

[m26872]

And about that cheap converter test. Since is an open type one with high res pictures from seller - too bad he didn't test an arbitrary other where the failed "visual inspection" could have passed simply by a look at the pictures. (I even think those blue caps could be small ceramic safety ones).

Btw, found a version of EN 60335 here.

[alexsh1]

@m26872 said:

Btw, found a version of EN 60335 [here].

I've used this link

[m26872]

@alexsh1 Thanks, it was better in every sense.

[sundberg84]

@alexsh1 said:

I think the difference is that you are using the slow blow fuse 0.2A and he is using fast acting 0.8A one. I really wonder which one would be more safe. I know slow blow fuses are used with inductive load (like motors), but not sure about this application.

I know it has been mentioned here as well, but did you make any provision for a thermal SEFUSE (say, 73C) or do you think it would be too much?

Its a good a releveant question... im not sure about slow or fast, but when i read the slow blow can tolerate high levels of voltage for a short period and a fast (or quick) blow fuse is a fuse that bursts instantly when high power voltage is passed through it.

How i see it, we have the protection against high quick power voltages allready in the Varistor? (Im not educated in this - this is my brain/logic trying to add everything up). Allthough what i read is what you are saying - slow blow are often used in motors and fast blow on home appliaces.

In my PCB I use a 73dgr therma fuse as well, im using the exact same circut that petewill has in its first post in this thread.

[alexsh1]

@sundberg84

The varistor is against high voltage transients and surges. The fuse is for overcurrent.

I have just ordered 0.8A fast blow fuses - they will be installed on my boards.

[sundberg84]

@alexsh1 - yes offcourse, you are right - mixed that up.
I have some 0.5 fast blow - sounds like thats a better idea than slow blow @petewill.

[alexsh1]

@sundberg84 Given that HLK-PM01 can withstand current spikes to 1.0A, I decided to go for a 0.8A fast blow fuse. Additionally, having a thermal fuse would be an advantage, but not a must. Unless you want to push the PSU to around 600mA then it is going to operate close to its limit and you need some extra protection

[sundberg84]

@alexsh1 - what do you base it on, its not a must? Im trying to be as safe as possible but i really wants to learn as much as possible about safety regarding ac/dc converting.

[alexsh1]

@sundberg84 Obviously, there is no harm from having it, but if your consumption is, say, 150ma and the PSU is handling 600ma, this is just an extra in my view.

Did you see the IR images taken after about 45 minutes of operation over here?

[sundberg84]

@alexsh1 said:

150mA

Ok, sounds logic - tnx. If I want to run a relay and some sensors it might be higher than 150mA so I like the extra safety Yea, have read that review about 100 times now, and emailed the dude as well for questions, haha. Much to learn!

[petewill]

@sundberg84 said:

@alexsh1 - yes offcourse, you are right - mixed that up.
I have some 0.5 fast blow - sounds like thats a better idea than slow blow @petewill.

Are you guys sure you're not confusing the input current with the output? In the above mentioned review he says there is a "Maximum input current: <0.2A" which is the reason for the slow blow fuses (I think). Or, maybe I'm misunderstanding and you're talking about adding an additional fuse to the output of the HLK?

[sundberg84]

Yes, max input is 0.2A with spikes ok upp to 1.0A but
@petewill we were discussing about using slow or fast blow fuse on the input.

Slow blow seems to be more for motors and allowing equipment making spikes when they start up. Then the fuse wont blow due to a higher current during start. In home appliances it seems its a fast blow fuse we should use.

[sundberg84]

Found what seems to be a interesting article: https://www.ieee.li/pdf/essay/safety_considerations_in_power_supply_design.pdf (Texas instrument).
Not read through it or know if the source is reliable but Texas instrument seems like a good author/source. No references inside so I guess its "how they see it".

[petewill]

@sundberg84 Ok, I understand. I will defer to the experts here and update the first post when a consensus is reached. I do have one more thing I'd like clarified though. It still seems to me like a 200/300mA slow fuse would be better as it would survive small over-current spikes but it wouldn't allow sustained draws over the HLK's rating of 200mA. With a 600mA fast blow fuse the HLK could draw 500mA as long as it wanted without blowing. Maybe this doesn't matter though?

On a side note, when I tested one of my slow blow fuses it still blew pretty fast with a moderate load above it's specs. I tested with a 33 watt floresent bulb and it blew in less than a second (not sure of the exact time/wattage as it was really just a test to see if it was a fuse and not something else).

[Sparkman]

Transformers typically have an in-rush current as well, so a fast-blow fuse could pop because of this if it's not sized for that.

Cheers
Al

[sundberg84]

Interesting @Sparkman!
https://en.wikipedia.org/wiki/Inrush_current "When a transformer is first energized, a transient current up to 10 to 15 times larger than the rated transformer current can flow for several cycles."

[RJ_Make]

This in-rush 'effect' can be seen on any induction load.

[m26872]

I´m no expert but I thought and researched some. Could be wrong, please correct me then.

I think you're confusing AC inductive load inrush current with what we're interested in, which is the inrush to our switched power supplies. In our case the inrush current is to the primary side rectifier filter capacitor, not to the transformer. The transformer inside SMPS is for high frequency and would probably give a small inrush even if it was subjected to normal AC. Now it's DC and behind the switch controller and FET. And the controllers I've seen usually have some start-up time.

General about fast/slow types of small fuses (from here

The fuses must not open in less than one hour at 125% of rated current and open within two minutes at 200% of rated current. The 1000% overload is used to determine the fuse characteristic. The opening time for each rating is listed below.
Type FF: Less than 0.001 sec.
Type F: From 0.001 - 0.01 sec.
Type T: From 0.01 - 0.1 sec.
Type TT: From 0.1 - 1.00 sec.
These characteristics correlate to the terminology used in IEC 60127-1.

So in theory a type FF fuse could open "instantly" for 10*In (rated current) ? (in reality you check its specs of course). But let's see how a more normal type F fuse would work given the setup from a "normal" SMPS. I use this, but from what I've seen, primary side looks very similar to the HLK-PM01.

Rectifier filter cap value of my SMPS: 4.7uF
A normal 5x20mm fuse resistance value: 5 ohm (usually more resistances than just the fuse, but let's be conservative)
Max voltage peak-peak 700V ( ~250Vrms )
Regular capacitor calculation:
I(t0) = U/R = 140 A
T = RC = 54.7u = 23,5 us
I(t) = U/Re^-(t/T) = 140*e^-(t/0.0000235)

How much inrush current do we have at for e.g. the fastest F-type fuse characteristic:
I(tmin=0.001) = 140e^-(0.001/0.0000235) = 140e^-(1/0.0235) = 3.3*10^-19 (i.e. zero current)

How much time until I=0.1A (which would correspond to fuse value 0.01A):
I=U/R*e^-(.) <=> e^-(.)=IR/U <=> (.)=-ln(IR/U) <=> t = -ln(IR/U)0.0000235
t = -ln(0.15/700)*0.0000235 = 0.00017 = 170us

So even a 0.01A type F fuse should be far from blowing in a worst case and with conservative assumptions. For extra fast fuses, type FF, the characteristic must be checked.

Edit: This is only regarding the inrush, of course the load must be added to this eventually.

[sundberg84]

@m26872 So, to sum up - there is no need to use a slow fuse, since the in rush current isnt "long" enough to blow a fast fuse. So considering safety (and if you have the same mA value for both fuses) its better and should work to use a fast rather than slow ??

[RJ_Make]

@m26872 I'm no expert either. We may be saying the same thing however... I don't think it matters (AC or DC), if there is a transformer in the circuit, there is going to be a current peak on startup. If you have a filter cap before your transformer, you will see 2 peaks. The first one will be caused by the capacitor charging up, and the second from the inductive load caused by the transformer powering up. I'm thinking the 2nd peak should be much less due to the filter cap.

Of course I could be all wrong.

[RJ_Make]

Here is an interesting article http://www.vptpower.com/wp-content/uploads/downloads/2012/01/info_inrushCurrent.pdf on the subject. Looks like there is a LOT more to it (not that I thought there wasn't) than my over simplified post above.

[petewill]

@m26872 Thanks! So, am I understanding this correctly that we would want a 200mA or 300mA fast blow fuse? Since I know some of us have already purchased the 300mA slow blow fuses do you think it is safe to use them? If I'm understanding correctly they should still blow relatively quickly if they are subjected to high current right?

[m26872]

@sundberg84 Yes, a fast fuse should give a higher level of protection without issues.

@RJ_Make That second "inrush" is what I called the load. It's completely governed by the controller and should not stress the fuse more than running overloads. If there's a peak it´s because of capactive load on the output and has nothing to do with the transformer.
Thx for the link. It confirmes that these turn-on-times are a few milliseconds just like the HLK-review said (around 3ms). No chance to superimpose the first peak.

@petewill Regarding fast fuse, yes. But the value I don't know. The HLK review shows overload capabilties, but I don't know about the input currents at that point. Honestly I think it's very unlikely that you'll be saved by using a fast fuse instead of a slow. It's more a matter of e.g. better varistor (overvoltage) protection and the personal feeling of maximum safety etc.

[sundberg84]

So there is another safety issue we should discuss for the HLK.
Creepage (separation between two solder eyes measured along the surface) and clearance (shortest distance between as measured through the air) on the AC input.

According to datasheet there is 5mm center-center, but real distance with solder and pads its more like 3, maybe 3.5mm.

I found this and this which was a great summary of some IEC standards states that if your are going to use 240v and put the HLK into a environment with pollution degree III higher you need more creppage than 3.2mm. What am i saying? If i understand this right its not safe to use the HLK in any environment where moisture condensation may occur (outside, bathroom). My guess when i try to read the contamination groups is that if you put the HLK inside a sealed case you get the contamination level down to 2 and then we are safe above needed 2,5mm.

To be on the safe side I have in my latest pcb added holes between AC input to increase creepage:

Am i right? Dont know - its hard to read all these articles and Im new to this so please bring anything to the table... corrections, questions or thoughts.

[TD22057]

@sundberg84 I think that looks like an excellent idea. I was looking at the tear down of the HLK in the test that was run and it doesn't look like there was any reason for them to do this - at least as far as I could see, there was plenty of room to put those connectors at the corners like the DC side.

[alexsh1]

@sundberg84 How about a watertight enclosure?
In the UK it is IP67 (this is what it means -> http://aceeca.com/handhelds/ip67)

http://www.rapidonline.com/mechanical-fastenings-fixings/hammond-ip67-watertight-enclosure-with-flanged-lid-1555cf22gy-120-x-66-x-42mm-30-4338

[sundberg84]

@TD22057 Yes, you can see in the teardown they have thought about the creepage in the optical feedback unit (cut a hole in PCB) between primary and secondary. I dont know why they so close, but 5mm is enough - maybe they forgot about solder/pads would take some space?

@alexsh1 If i understand Contamination levels right the difference between III and II is that in III occurs conductive contamination so if you seal it up this should get it down to II.

Then there is a difference between materials - and if compare material I and III its a big difference in creepage - does anyone knows what this means? What material is a normal PCB?

[martinhjelmare]

Maybe they put the pins closer on the AC side than on the DC, to make sure people would not put the unit the wrong way, which could be dangerous? But as you say, they should have done a better job, though.

[sundberg84]

@martinhjelmare - thats a great thought - must be so, but if you think safety they should do the other way around, ac side wide apart and dc closer.

[martinhjelmare]

@sundberg84

Yeah, , they really screwed that up. Unfortunate, when the rest of the design, seems so good.