I killed my LimeSDR... maybe

Hello!

So I’ve recently acquired my LimeSDR v1.4s USB-A, did all the testing as per the official information - all checked out well. I then performed the HF EasyFix1 and also a similar modification on the RX2_W following Marty’s experience.
It all worked well and the device checked out ok at this point also.

Then I went a started building my enclosure - I did some hardware work in the box itself, then I stripped the SMT mount LEDs from the device, tinned the through-holes for VCC_INT, Fan and LEDs and called it a day.

The next day I found out that my device does not enumerate in the USB devices list. After doing some preliminary measurements and consulting the schematic, I found out that my VCC3P3 power rail has failed and now measures about 0.95V. The VCC_INT rail also dropped from about 4.6V to 4.1V.

So at this point I suspect that something is loading down the 3.3 power rail, but I’ve visually inspected the board like 20 times and failed to find an obvious cause of this.

Next I’ll try to power the device externally and see what current it draws to see if it is too high or if something else has gone wrong in the voltage regulation.

In the meantime, if anyone has any advice or ideas - I’ll be most welcome. I’m prepared to fiddle around before calling the device a brick because I don’t think I’ve done anything obviously blatant.

Thanks.

ADDED: I tried supplying the power via the external connector. I tried voltages between 5 and 8V and it seems that the VCC_INT rail does not really change - it stays at about 4.1V. The current consumption my power supply reported was about 10mA from 7V and up, while it does not have the resolution to see what’s what on 5V.

I would like to bypass the power supply entirely and supply 4.8V directly to VCC_INT, but I’m not quite sure yet if this is a bad idea.

Ok so I mucked around a bit more and supplied 4.8V to VCC_INT directly from the lab supply. The board started consuming about 0.8A and when I hooked it up to the PC, the board got enumerated. I was able to run the LimeQuickTest, and the board passed all the tests.
So that’s good - now I need to find whats wrong with the power source selection circuitry.

P.S. Sorry if this has quickly turned into a monologue thread.

Glad to hear that you’ve made some progress and maybe @Zack might have some ideas as to what could have gone wrong.

1 Like

Thanks!

Now I’m trying to understand what should the inputs and outputs of the FPF3042 look like to try and assess if I damaged C367 or FR56 somehow. I have a scope, but I must understand what to look for first (like increased ripple on the output rail etc.).
Then I should probably check the VINSEL line just in case.

Alright, having read the FPF datasheet, I don’t quite understand where is VCC_INT voltage regulation done? The voltage on the USB jack is ok, but it is already 4.2V on C367, so it comes out of the FPF3042 switch that way. Both when I power from USB and external source.
Looking at the schematic, I cannot really see anything else connected to VCC_EXT though.

Is it a good idea to fit R167?

Thanks in advance for any hints.

Hi @OrsonMaxwell,

VCC_INT is not regulated. It should be close to VCC5P0_USB or VCC_EXT, while FPF3042 is power source selector where VCC_EXT takes priority if connected. What is external voltage and what voltage do you get on C367?
From your description it looks like FPF3042 is not working properly. You can supply the board by selecting VCC_EXT of VCC50_USB by soldering R166 or R167. But it is recommended to remove FPF3042.

1 Like

Hello, @Zack!

Thank you for taking the time to answer me. What baffled me earlier is the fact that I supplied 7.6V to VCC_EXT pin, and still got around 4.15V on VCC_INT and C367. The same happens both if I supply a different voltage to VCC_EXT (I tried adjusting from 6 to 9V on the PSU without any noticeable change on VCC_INT) and when I just plug in the USB cable. That was what I referred to as “regulating” in my previous post.
So if I understand this correctly now, the voltage does not get regulated between VCC_EXT and VCC_INT rails, and should be supplied as is to the individual regulators for different busses and rails.

Thanks for clarifying the NF resistors business - If I get pressed for time I’ll keep this option in mind. For now I’d like to understand what I broke.

Correct.

My guess is FPF3042. Would suggest to remove FR56 and check VCC_INT if it is close to VCC5P0_USB or VCC_EXT. If it is not, then FPF3042 is damaged.

1 Like

Yes, having understood the previous point, it makes much more sense to me, thanks. So the working hypothesis is that the FPF3042 chip is damaged and is loading down the power rail somehow similar to a zener diode.
From what I see through my magnifier, the chip looks like it has a silicon package as opposed to a plastic one - maybe I damaged it mechanically by accident when handling the board.

Alright, I’ll find a supplier and try to replace the selector, all the while having a viable workaround solution. Thank you very much for your support!

1 Like

This is what is used:
https://www.digikey.com/products/en?keywords=FPF3042UCXCT-ND
Here you can find BOM files where suppliers and product numbers are listed:
https://wiki.myriadrf.org/LimeSDR-USB#Board_Design_Files

1 Like

Yeah, I did clone the whole repo before doing the repairs, thanks!

Ok so I removed FPF3042 from the board, jumped R167 (as I only use the device over USB) and it works fine!

I ordered a couple new FPFs, so maybe later I’ll try to solder a new one on - would be a first time fitting a BGA package for me.

Cheers!

Glad to hear you that you managed to fix it. Keep us updated on BGA soldering.

1 Like

When I fried my power switch, I just bypassed it with a solder blob & run it off of a good quality powered hub (12V powered, not the cheap 5V powered hub). Works great. No problems with TX/RX at all.
Many use a cheap hub that will not supply enough amperage and fail.
I have since bypassed the power switch on my other Lime & it works flawlessly, too.
Actually, I have not had a single issue since doing so. No more reflashing the FX3, no more intermittent issues. Very stable.
Personaly, I would suggest not replacing the switch.

I can actually run both LimeSDR-USBs at the same time on the hub with no issues.

Ed

1 Like

I thought about that, it’s just that I’m curious to see if I can get the device back into the shape I bought it in. I don’t plan on using any hubs, and I didn’t have any issues prior to damaging the IC.
But since I damaged it, I thought about putting a small buck converter into the enclosure in case I need to power it via a car battery in the field.