Short circuit

[Rolf Andersson]

Hi All

I just assembled my brand new HL2. 

I connected a small 12V/5A linear supply and leds started to blink

I also connected ethernet and a dummy load as antenna. Even the Lan leds started to blink.

After spending one our or two installing quisk I tried to connect withouth success.

I noticed that the HL2 leds are off and it is more or less short circuiting my supply.   Disconnecting the HL2 and supply is recovering. I tried another PSU with adjustable current limiting.  I get about 2V at 3Amps and I don't want to increase current.

Any idea about where I should go from here?

Thanks

Rolf

  

[Steve Haynal]

Hi Rolf,

I'd remove the HL2 from the enclosure, remove the small aluminum heat shim, remove any thermal paste you applied, remove the N2ADR board and try to power up the HL2 standalone. It should draw less than 230 mA in this configuration. If the HL2 comes up, I'd double check that the aluminum heat shim or enclosure is not shorting something. I'd also check that the thermal paste you are using does not conduct electricity. 

If the HL2 does not come up in this configuration, I'd do a visual inspection of the HL2 for any component which looks like it passed high current. Also, with power supply connected (ideally briefly) check if any area of the HL2 board is becoming hot. Please report  back with any findings.

73,

Steve

kf7o

[Rolf Andersson]

Hi Steve and thanks

It is the same result with the bare board.  I can't see anything burned. U16 the 1,2V regulator for the FPGA gets hot however it seems to work and delivers 1.35V on pin 4 on DB9.

The output in the fuse F1 seems stuck a litte above 2V. I increases slightly when when I allow current to go from 2 to 3 amps.

Cheers and 73

Rolf/SM7LMY

[Steve Haynal]

Hi Rolf,

With the power supply disconnected, can you measure the resistance between the +12V input line and ground? I measure very high resistance >100MOhm with the Ohm meter leads in one configuration, and then ~30MOhm when the leads are switched.

Also, please post high resolution photos of the front and back of your HL2. A camera phone is usually sufficient.

73,

Steve

kf7o

[Rolf Andersson]

Hi Steve

In one direction it is higher than I can measure. In the other direction it is 10MOhm but sometimes about 35KOhm.  It is like a pattern depending on the direction of the probes.

a) High resistance

b) change direction then  Low resistance

c) disconnect, connect again with same direction 10MOhm.

This pattern is consistent.

I have attached a couple of pictures.

73/Rolf

[Peter Johnson]

Hi Rolf

I have just looked at your pictures of the board.

On backside of the board where the heat-sink fits if you zoom in on that
area it looks as if there are burn marks in the fixing hole plus a burr
as if the hole has been drilled to make it bigger for a larger fixing screw.

I maybe mistaken but it look like that is where the problem lies.

73 Peter m1bcv

> --
> You received this message because you are subscribed to the Google
> Groups "Hermes-Lite" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to hermes-lite...@googlegroups.com
> <mailto:hermes-lite...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/hermes-lite/612c11ca-054e-466b-9a07-20e4977eed43n%40googlegroups.com
> <https://groups.google.com/d/msgid/hermes-lite/612c11ca-054e-466b-9a07-20e4977eed43n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Graeme Jury]

Hello Rolf,

Has the hole for pulling the PCB onto the case runners been drilled out
a bit? It may be creating a short on the internal layers.

73, Graeme ZL2APV

[SNIP]

[Rolf Andersson]

Hi Graeme and Peter

The hole has not been drilled however it might have been slightly deformed by the scew I used to hold it to the "shim" and case.

If I look inside the while there are marks from the screw on the plating. Is there no margin in the PCB for this?

Thanks and 73/SM7LMY Rolf

[Graeme Jury]

A bit difficult to see properly on the photo so I wondered if it might have been the case. I don't use the shim and only put gentle pressure on the screw, enough to make sure it is in good contact with the case rails so there are no marks around the PCB hole. I will make a shim before summer as it gets right on 50C with sustained TX on a hot day.

73, Graeme

--

You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.

To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/7d8cbb4e-33b5-4929-97ca-2e03522af65en%40googlegroups.com.

[Rolf Andersson]

Hi Graeme 

I git the shim with the kit. 

Here is a closeup of the hole. It might look a little burned. But why is it so sensitive?

[J P Watters]

In my opinion, the HL2 Circuit Board hole between the final transistors is not large enough to support installation of a 4mm machine screw. It almost is, but not quite. One can thread the 4mm machine screw into the board and allow the threads to expand the hole as the narrow dimension (3.75mm) of the end of the machine screw can be started into the hole with moderate pressure. I cannot imagine what happens to internal layers as a screw pushes on the board layers. 

From the photo, it would appear that a 4mm or 5/32" machine screw was used. I also used a 4mm machine screw. Additionally I tapped the shim with a 4mm tap. Further I used a sharpened machinist's hole reamer to ever so slightly cut away the hole edge to expand the hole in the circuit board to allow the screw to penetrate the board without resistance. I think the preferred screw size should be 3.5mm

Hindsight, had me replace the 4mm screw with a nylon one or use a nylon 3.5mm screw and nut. I can only think of all the ways introducing a metal nut atop the screw would make for unexpected consequences if it were somehow where it didn't belong upon powerup.

..jpw J P Watters

KC9KKO

Morris, IL USA

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/0a0c7521-d663-4d14-92c7-37231d88f0cbn%40googlegroups.com.

[Alan Hopper]

Hi jpw,

where did you get the 4mm from, if it is written down anywhere it is wrong and should be removed. The correct size is 3mm.

73 Alan M0NNB

[Chris Gerber]

Please read all my .doc about mounting the Shim. There was always mention

to use only an M3 (3mm) screw !!!

That is a fact !!

Chris HB9BDM (Creator of the shim)

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/CAGCXWQO8VctBTrAsY1inf3Ay4yF0PKFEQVRukRSBZsdLayORjA%40mail.gmail.com.

[Steve Haynal]

Hi Rolf and Group,

With or without the heat shim, a M3 screw is what is recommended on the assembly wiki page. But I don't think Rolf's problem is a larger screw. All 4 PCB layers in the vicinity of this hole are ground with quite a bit of margin. The four layers are already shorted via the plating in that hole. 

Please measure for shorts between the various power supplies as described here. This is done with an Ohm meter and no power connected. Also, please report what resistance you measure across for F1, again with no power applied. 

How interested and/or comfortable are you with soldering/desoldering components? Depending on your measurements, a likely next step would be to remove D12. Another option would be to send it to me or someone in Europe willing to make repairs.

73,

Steve

kf7o

To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite+unsubscribe@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/0a0c7521-d663-4d14-92c7-37231d88f0cbn%40googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.

To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite+unsubscribe@googlegroups.com.

["Christoph v. Wüllen"]

It would be interesting to know whether the current „flows“ through
the PA FETS. If you see 3A at 2V then it is not a short circuit.

A „bogus“ bias, as well as a „blown“ FET my very well show
such „short cuts“.

You could de-solder connection „2“ of the output transformer T2
(possibly not too difficult) to this end.


Just my $0.02,

Christoph DL1YCF.

[Rolf Andersson]

Hi Steve and others.

I want to say thank you for all your support.

FYI I'm using a 3mm screw but my hole in the case is perhaps not 100% aligned with the PCB.

I measured the resistances between the supplies. Everything looked ok ( > 500 Ohm) except the 1.2v supply. There the resistance is only 5 (five!)Ohm to ground.

Over F1 the resistance is about 0.6 Ohm.

I comfortable enough to solder the passive components, but not ICs.

Thanks and 73

Rolf/SM7LMY

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/0a0c7521-d663-4d14-92c7-37231d88f0cbn%40googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.

To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/CAGCXWQO8VctBTrAsY1inf3Ay4yF0PKFEQVRukRSBZsdLayORjA%40mail.gmail.com.

[Steve Haynal]

Hi Rolf,

Your 5 Ohm measurement between 1.2V and GND is definitely low and a problem. I remeasured a board here and still see just over 500 Ohms between 1.2V and GND no matter which way the leads are used.

In the picture you sent, there looks like a short involving pin 46 (1.2V) of the FPGA. Please see the attached picture with red arrow identifying the pin. Please check this.

If this is not the short, the 1.2V supplies portions of the FPGA and KSZ9031 Ethernet Phy. There are lots of capacitors between 1.2V and GND. Hopefully it is one of these capacitors which has become a resistor.

While measuring the ~5 Ohm resistance, try applying finger pressure to the various capacitors (most are underneath or near the 2 ICs) and regulator to see if you can find any sensitive area which makes the resistance change. Also try applying finger pressure to the FPGA U2 and KSZ9031 U4. Be sure you are statically discharged by also touching a good ground or using an ESD strap. All the capacitors and ICs connected to 1.2V are shown on pages 2 and 3 of the schematic:

https://github.com/softerhardware/Hermes-Lite2/raw/master/hardware/hl/hermeslite.pdf

Also, if you have a microOhm meter, you can try measuring resistance across several of the capacitors to see if you can zero-in on an area with least resistance. 

Finally, with your current limited power supply on and connected briefly, check with your finger if any of the many capacitors or components on the 1.2V supply are heating.

73,

Steve

kf7o

[Rolf Andersson]

Hi Steve

I couldn't find anything particulary hot except u16 and the area around.

However I measerured the resistance between VSUP (not Vsupply) and GND it is also 4.5 Ohm. Then I measured the resistance between VSUP and 1.2v it is ZERO.

So I guess U16 is shot? What worries me is that maybe everything else on 1.2v might be shot?  I might have gotten 2v here when I played around with the current limiting supply. 

Thanks and 73 

Rolf

[Steve Haynal]

Hi Rolf,

So the picture with a red arrow did not point to a short?

Your 4.5Ohm between VSUP and 1.2V is definitely part of the problem. I measure over 1K here. A next step would be to remove U16 or at least desolder and bend up pins 6 and 7 to disconnect it from VSUP and 1.2V, and check if the short still exists.  

73,

Steve

kf7o

[Rolf Andersson]

Hi Steve

No it was not a short just some dirt. I will try to do some soldering. But that has to wait a couple of days.

Thanks and 73 Rolf/SM7LMY

[Rolf Andersson]

Hi Again Steve

I have lifted pins 6 and 7 on U16 from the PCB. It is a short between them (zero Ohm) and about 5 Ohm to ground.

Now it is 380 Ohm to ground from 1.2V on DB9

It is 20 Ohm from VSUP to ground on DB9.

Hmmm?

[Steve Haynal]

Hi Rolf,

The 1.2V to GND is looking closer. I measure ~600 Ohms so maybe it is okay for you.

The 20 Ohm from VSUP to GND is too low. I measure at least 1K.

Pin 1 of U16 also connects to VSUP. Maybe that is still part of a short. Can you try lifting that pin?

You could remove U16 and U8 and then put U8 on the U16 footprint. U8 is only on and used during TX. At least you would know if the unit is still alive for receive. I can then send you some pulled parts I have for U8.

How old is this board? Do you want to investigate a repair or exchange with Makerfabs?

73,

Steve

kf7o

[Rolf Andersson]

Hi Steve

My board is very new. I ordered it less than two weeks ago.  U16 died within 2 hours of operation.

I have now removed U16 and VSUP to ground is now a couple of KOhms. 

I could try to get a new regulator locally and see if it helps.

But what is the real problem? Bad regulator? Something else?

Im a little suprised that I have this problem. Nowadays electronics seems to be very robust.

Thanks and 73 Rolf/SM7LMY

[Steve Haynal]

Hi Rolf,

I can send you some pulled regulators if you like. The problem could be more than the regulator, but I think it is worth trying with a replacement regulator. U8 can always be pulled and used to replace U16 as a test.

To keep costs down, we only do screening tests of the HL2 boards. This makes us more likely to hit problems with the bathtub curve. Overall I think our failure rate after reaching a customer is in the 1 to 2 out of a 100 range. I am sorry that you are one of the unlucky ones.

Your options are to try and replace U16 yourself, or send it to me via regular air mail. I will either repair it and ship it back using the repair funds pool we have, or request Makerfabs to send you a replacement. Please let me know what you prefer.

73,

Steve

kf7o

[Rolf Andersson]

Hi Steve

At work we design FPGA boards (I write the VHDL/Verilog stuff).  Our PCB engineer removed U16. The device is cheap so he will get one for me and replace it.

That will take a couple of days I guess.  It that doesn't help I would like to ship the board to you.

Is that Ok?

Thanks and 73 Rolf/SM7LMY

[Steve Haynal]

Hi Rolf,

Sounds good. I do have one new or a coupled of pulled regulators which I can send to you if you have any difficulties procuring the part. Again, sorry for the problems.

73,

Steve

kf7o

[Steve Haynal]

Hi Rolf,

Any update on your issue? Was the replacement of U16 successful?

73,

Steve

kf7o

[Rolf Andersson]

Hi Steve

I got the replacement regulator, but due to the covid situation I haven't been to work since last week. So it is not mounted yet.

73 SM7LMY/Rolf

[Rolf Andersson]

Hi Steve

I have replaced the regulator myself, since our engineer is working from home now.

It might not be perfect since I didn't solder the back of the package to the pcb only the leads. It seems to work now and doesn't get

hot at all. Current stabilizes to about 170mA at 12V. DHCP works.

Tonight I will try to get it to run with quisk.

73 and cheers

Rolf/SM7LMY

[Steve Haynal]

Hi Rolf,

I am glad to hear you fixed your HL2! There is a hole to allow for hand soldering of the hidden pad if you like.

73,

Steve

kf7o