Electronics (PCB) help

[choochoo_baloo]

I was happily using a 4 USB to ethernet adapter at my astrophotography rig for a while.

Until one night, for no apparent reason, the adapter would not power on upon plugging in.

The symptoms:

1) I was using a cheap 12->5V buck converter to power it. After plugging in, the buck makes a very faint whine. I’ve been told this means its inductor has failed (change in resonant frequency).

2) Having replaced the buck with a better and higher rated one, I notice its supply voltage change from 5.1 to 1.6V (according to built in voltage display) when plugged in. Must be a fault with the adapter...

3) whilst inspecting tye PCB while plugged, the back gets noticeably hot within few seconds. See photos circled red.

NB. smaller PCB sandwiches above larger PCB via the 10 header pins. That black chip thing circled would appear to be the source of the heat.

for the sake of a burnt component or so, I’m loath to throw out a £100 device!

Any help in diagnosis will be much appreciated.

Edit: I don't know where to start in identifying the faulty PCB component with multimeter etc.

Edited August 11, 2019 by choochoo_baloo

[pete_l]

1.) The component in question is labelled "D21". This implies it is a diode - by the looks of it a Schottky type.
2.) If it is getting hot, that is because it is passing an excessive current.
3.) If your power supply drops from 5V (nominal) to 1.6V when applied to this device, the drop is caused by the SMPS going into current-limiting mode due to a short circuit on its output.
4.) It is not unusual for inductors in SMPS's to "whine" when subjected to an overload condition - as your original one did / does.

That sheds some light on what is happening. I cannot say why it is happening. Except to say that the daughter board looks like it is a power supply, itself.

Edited August 12, 2019 by pete_l

[symmetal]

+1 for what pete_l said.

It's worth checking the diode D21 with your meter. It should measure higher resistance in one direction than the other (Reverse the meter leads). Your meter may have a diode check function anyway which can give useful info even though the diode is still connected to other components. Here's a Youtube video showing how to use the diode check function. As it's probably a schottky diode (low forward voltage drop) it would likely measure 0.2 to 0.3V.

Also check the three large electrolytic capacitors by it for shorts (or low resistance value in both directions) using the resistance function on your meter.

Alan

[choochoo_baloo]

4 hours ago, symmetal said:

+1 for what pete_l said.

It's worth checking the diode D21 with your meter. It should measure higher resistance in one direction than the other (Reverse the meter leads). Your meter may have a diode check function anyway which can give useful info even though the diode is still connected to other components. Here's a Youtube video showing how to use the diode check function. As it's probably a schottky diode (low forward voltage drop) it would likely measure 0.2 to 0.3V.

Also check the three large electrolytic capacitors by it for shorts (or low resistance value in both directions) using the resistance function on your meter.

Alan

@pete_l@symmetal

Thanks chaps for your input. After some reading up on remedial DC testing, I have confirmed that that diode seems fine, as with a nearby diode.

*However* the left most electrolytic capacitor in photo 2, is showing 0.3V in both directions - showing a short? I'll now remove it to confirm this.

Assuming this is the fault, can either of you confirm whether I need a special brand (other than obviously matching capcitance and Volt rating)?

[Dr_Ju_ju]

As long as the capacitance & voltage rating are the same, then any brand will do....  just make sure that the replacement is the same physical size, so it fits ok....

[symmetal]

Hi Matthew,

The electrolytic could be faulty, but as you say without removing it from the circuit it's hard to say. For high power SMPSs you want Electrolytics with a low ESR (equivalent series resistance, high temperature operation (105 degrees) and suitable ripple current rating. Here's the data sheet for KYA capacitors to give you some info.

However, as yours is a low power SMPS unit these features are not critical so as Julian says any 'recently made' Aluminium Electrolytic would be fine.

Alan

[boyandorion]

Hi. As for the caps in question, one of the failure modes of this kind is that they expand - as in u can see a bulge on top of them. They are made to do this. They don't have to be bulgy to be bad tho. As for testing them, there are a few methods of doing this depending on how you are set up with the tools. Easiest way if you aren't quite sure about how to do it, is to remove it from the circuit. There is good articles on the net on testing different ways. Like this:

https://www.electricaltechnology.org/2013/06/how-to-check-capacitor-with-digital.html

I'm not sure it was clear from the first post if you are testing the power supply plugged onto the main board or not. Just saying, the excessive current draw that causes this thing to more or less shut down, may be a failure in the load (the mainboard).

As for throwing it out (if the mainboard are ok), don't do that unless it's actually broken. It might very well be possible to drive the board with a completely other power supply. Not sure what's going on with this one tho, since it's so well eqipped at the pin header. Does it require other than a single +5v rail? A closeup photo of the silkscreen around the pinhead connector on the mainboard would maybe say something more about this.

Edit:
For cap replacement - same capacitance value, and same voltage rating or higher - as long as u can fit it in there, it doesn't matter much. Doesn't have to have the same lead spacing etc. as long as u can fit it in there and bend the legs to fit in it's holes. Only imagination is the limit hehe. You could even solder new capcitors on the other side of the board if needed.

Edited August 12, 2019 by boyandorion

[choochoo_baloo]

I should have said at outset: I am a physicist by training, not an electrical engineer, so although I know the fundamentals, but will need clarification with the jargon and methods

 

The plot thickens.

tested the lone capacitor, it is ~1M Ohm and outside range in the 2 directions -> can assume it is fine after all.

BUT

It was covering the circled Zener diode. As the picture shows, it measures 0.3V (diode test mode) *both* directions. It also appears cracked on the outer tube.

I measured a neighbouring Zener as shown for comparison. Is this the culprit?!

 

Edited August 12, 2019 by choochoo_baloo

[boyandorion]

The cap sounds like it's fine. They take some effort to ruin also, depending on the input voltage from the dc plug.

As for the zener, I know just a tiny bit too little about SMPSUs to take a glance at this board and say how it is supposed to work (give me a normal linear supply and I'll tell you all about it any day haha) but the zener diode should definitely not be broken .

Zener diodes is most likely part of the circuit that sets the voltage the regulator is outputting. I would say, a broken one will definitely make your supply not output what it's supposed to output, but what made it blow up... that's the question.

Is it possible to get hold of a schematic for this unit? It would be extremely helpful for "remote assistance" .

[choochoo_baloo]

5 hours ago, boyandorion said:

Is it possible to get hold of a schematic for this unit? It would be extremely helpful for "remote assistance" .

I'll phone StarTech in the slim hope they can send me a circuit diagram...

In the meantime, can anyone tell me how I can tell what spec' the aforementioned Zener diode is? I can then replace it with a new one, to see if that fixes it.

Thanks for your help.

[symmetal]

Hi Matthew,

The AX5510 8 pin IC below the zener is a boost converter, 5V in and up to 27V out. AX5510 datasheet. This typical application circuit from the data sheet may help but there is no zener shown connected to pins 6 and 7 as you have.

 

L1 above is L21 in your circuit and D1 above is a schottky rectifier which looks to be the SS34 schottky rectifier below your L21, and not the similar D21 rectifier which is getting hot. R3 above is R21 (100R) in your circuit. It looks like the zener is in parallel with R21 (that's how it looks but the apparent track layout may be deceiving) and may be a protection diode in case the 5V rises or something but isn't used as a reference voltage as that's internal to the IC. The output voltage of the AX5510 is set by R2 and R8 above.

The zener would have some numbers on it but they may be difficult to read and desoldering it will probably be necessary to find them. As it's cracked it has to come out anyway. The circuit would probably work without the zener but whether you want to power it with the zener removed is up to you. With the zener removed you can do a more valid diode test but at the moment it does look a suspect for your problem. Hopefully the cracked casing hasn't removed any letters/numbers printed on it. You'll need a magnifying glass to read them anyway.

Alan

Edited August 13, 2019 by symmetal

[boyandorion]

8 hours ago, symmetal said:

but isn't used as a reference voltage as that's internal to the IC.

You are of course right. I'm sorry for the misleading information on my part.

I totally misread the situatuion here tho... . When I read the first post over agian. I thought from reading the description in the post (very poorly read I might add), that the board was powered by 12V itself. I totally missed out the fact that there is also another power supply here. This combined with the fact that I did not have the datasheet for the IC lead me to believe that the device is something that it's not.

Sorry for that. I don't want my name associated with things like "that cocky new guy who just writes something something so it looks good" , as I am not that guy. I didn't read/check properly, and was wrong . . I will be more careful next time hehe.

[choochoo_baloo]

4 hours ago, boyandorion said:

Sorry for that. I don't want my name associated with things like "that cocky new guy who just writes something something so it looks good" , as I am not that guy. I didn't read/check properly, and was wrong . . I will be more careful next time hehe.

This made me smile, typical Norwegian good manners - like the Trafalgar Square Christmas tree. 🎄

[symmetal]

4 hours ago, boyandorion said:

You are of course right. I'm sorry for the misleading information on my part.

Hi boyandorion,

There's no need to apologize as there was nothing wrong with your post and it contained useful information.   Most of the time low power zener diodes are used as voltage references. That may be the case here too, perhaps for some other part of the circuit, but from trying to work out the circuit by following the visible board traces it's hard to see in this instance. 

If the zener is in parallel with R21   it could explain the odd readings across the zener as the 100 ohm resistor would upset the diode test. If Matthew confirms it is in parallel by using the continuity (buzzer) function of the multimeter (with power removed) then I believe it would work fine without the zener.

Alan

[boyandorion]

On 13/08/2019 at 15:14, symmetal said:

There's no need to apologize

All good. When I'm wrong about something that can send other people on the wrong track (trace ), I feel it's important to say so.

I have a few suggestions to how to proceed, but let's do one thing at the time. .

On 13/08/2019 at 14:50, choochoo_baloo said:

like the Trafalgar Square Christmas tree.

I am groot.