Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

RF amplifier can be damaged #974

Closed
mossmann opened this issue Oct 20, 2021 · 10 comments
Closed

RF amplifier can be damaged #974

mossmann opened this issue Oct 20, 2021 · 10 comments
Assignees
Labels
help wanted help from community wanted and appreciated

Comments

@mossmann
Copy link
Member

mossmann commented Oct 20, 2021

The RF receive amplifier (U13, part number MGA-81563-TR1G) on HackRF One seems to be damaged in the field more than any other component. This has been known for a long time; some tips for avoiding damage may be found in this video:
https://greatscottgadgets.com/sdr/5/

As far as we know, damage does not occur without exceeding the -5 dBm limit documented here:
https://hackrf.readthedocs.io/en/latest/hackrf_one.html#receive-power

However, it is possible to exceed that limit by operating a HackRF One in close proximity to a higher power radio transmitter, something that many users would like to be able to do safely. Regardless of whether or not damaged units have been subjected to power levels exceeding the maximum specification, we would like to characterize the conditions under which the RF amplifier may be damaged and reproduce the failure in the lab.

Less frequently we have seen damage to the TX RF amplifier (U25, same part number), and we are interested in characterizing that failure as well.

Diagnosis of either failure is straightforward: If you enable a working amplifier, the power level should rise. If you enable a damaged amplifier, the power level drops instead.

If you have experienced this failure and have some idea of how it happened, please comment with any information that could help us reproduce the condition.

@supersat
Copy link

I wish I could tell you exactly what went wrong with my HackRF, but both of my amps are blown. There doesn't appear to be any change in gain with the RX amp turned off or on, but when I turn the TX amp on, the signal disappears.

I suspect it might have died at Toorcamp 2016 when I tried to hook it up to a sketchy 1.5W PA from eBay, although I'm not exactly sure what might have caused that. It looks like the input to the amp is AC-coupled through a cap. I could send you both the amp and the HackRF for analysis if you'd like. (There are also some people around here who can decap and image chips with a SEM, so that might be interesting?)

My friend also damaged his HackRF transmit capabilities when testing it in the lab with a DVB-S2 receiver. In that case, I think the receiver put out a DC bias voltage for the LNB. It was connected through some attenuators, but I don't think that protects you against DC bias voltages. I'm not sure if the HackRF's bias was turned on or not, but at any rate both U14 (an RF switch) and U25 (the TX amp) were blown and had to be bypassed.

@straithe straithe added user experience help wanted help from community wanted and appreciated and removed user experience labels Oct 21, 2021
@mautas
Copy link

mautas commented Oct 30, 2021

After receiving some MGA-81563 spares, today I replaced blown U25 (Tx LNA) on my HackRF One (Original Rx LNA is still good).

I don't know how and when original U25 was blown.

After replacing the part I connected HackRF One to this RF Power Meter with a -30 dB attenuator. (https://www.passion-radio.com/swr-power-meter/rf-power8000-926.html) and started testing ouput at various amateur radio frequencies:

7000 khz
28000 khz
146 MHz
440 Mhz

(yes, all in the amateur bands)

These tests where performed with Hack RF One and a Portapack H2+ with Mayhem FW powered from my notebook USB Port. After a while I wanted to make some tests using SDRAngel and set the device to HackRF "native" mode from the portapack.

Regretably I cannot tell exactly when, but while doing some more tests switching on and off the Tx LNA in 28 MHz and 146 MHz things started to go weird with HackRF One kind of resetting... As everything was working OK before it took me some minutes to figure out this happened when I turned on Tx with the LNA on... as you may be guessing, the new LNA just replaced a couple of hours ago was blown and it seems to have ended up in some kind of short circuit that makes the whole board reset.

All I can say is that all the time everything was safely tied to my wattmeter. Even though this did not prevent the LNA blowing up while doing "legitimate" stuff (ie. not doing nothing normal device configuration prevents from doing). May be it was just bad luck and the new LNA was not a good part. I replaced it again, just checked out it was working on 7000 kHz and 146 MHz, turned it off and I don't think I will be enabling it anytime soon.

As a ham radio enthusiast I absolutely love HackRF One but this whole MGA-81563 issue looks like a really really weak point. I regret not being able to offer much more details about what happened today besides saying that this happened without any operation error or misconsideration from me, just my HackRF One tied to my wattmeter and operating it within its intended limits. I will follow this thread and I hope something good comes out of it, like useful knowledge on how to avoid blowing this IC and maybe a fix for future hardware releases of HackRF One.

@mautas
Copy link

mautas commented Nov 4, 2021

After replacing the part I connected HackRF One to this RF Power Meter with a -30 dB attenuator. (https://www.passion-radio.com/swr-power-meter/rf-power8000-926.html) and started testing ouput at various amateur radio frequencies:

7000 khz 28000 khz 146 MHz 440 Mhz

(yes, all in the amateur bands)

Tonight a friend of mine that gave a look to the MGA-81563 datasheet directed my attention to the fact that this IC range of operation starts at 100 MHz (0.1 GHz). May be this is not critical for Rx operation where power magnitudes might be low enough not to cause damage but using it for Tx is a whole different situation.

I think HackRF Firmware should prevent enabling the Tx MGA-81563 below 100 MHz since in my experiencie this might be a situation where you will probably end up with a burned Tx amplifier, like me on saturday testing the new IC. Even though I am not very sure if the chip went bad in the 10 meter band or the 2 meter band, now I am prone to think this happened while doing tests at 28 MHz. Perhaps people at GSG could do some tests to reproduce this situation and check out if operating this IC for Tx below 100 MHz is safe.

Meanwhile I would recommend not enabling this stage for Tx below 100 MHz minimum specified frequency for the IC.

@mossmann
Copy link
Member Author

I suspect it might have died at Toorcamp 2016 when I tried to hook it up to a sketchy 1.5W PA from eBay, although I'm not exactly sure what might have caused that. It looks like the input to the amp is AC-coupled through a cap. I could send you both the amp and the HackRF for analysis if you'd like.

Thank you for the offer! We will follow up with you out of band about shipping.

@mossmann
Copy link
Member Author

Meanwhile I would recommend not enabling this stage for Tx below 100 MHz minimum specified frequency for the IC.

That's an interesting suggestion. Thank you! The data sheet does not specify a frequency limit in its absolute maximum ratings, but we will do some testing.

@mautas
Copy link

mautas commented Nov 15, 2021

Meanwhile I would recommend not enabling this stage for Tx below 100 MHz minimum specified frequency for the IC.

That's an interesting suggestion. Thank you! The data sheet does not specify a frequency limit in its absolute maximum ratings, but we will do some testing.

The datasheet title goes

MGA-81563
0.1– 6 GHz 3 V, 14 dBm Amplifier
Data Sheet

and
Introduction
This high performance GaAs
MMIC amplifier was developed for
commercial wireless applications
from 100 MHz to 6 GHz.

So, I guess operating this below 100 MHz puts us in uncharted territory

@eegan
Copy link

eegan commented Jan 22, 2022

Reviewing the data sheet, indeed the amplifier may not have been characterized below 100MHz, but there is no reason to think it will suddenly stop working below this point.

It does look as if its gain might be a bit higher sub-100MHz, as well as its noise figure, if you extrapolate from the provided curves. And of course the S11 also seems to be getting worse at lower frequencies.

@straithe
Copy link
Member

Thank you all for this feedback. We haven't gotten as many specific examples and experiences as we were hoping for, but your comments have helped. We will continue to look into this issue and will be happy to hear about more examples of U13 damage from anyone who would like to post here.

@electrical-pro
Copy link

Today my RX MGA-81563 died (short circuit). I don't know at which point it failed, prior to this I was using Hack RF with LNB, but MGA-81563 was off, so maybe it is unrelated. Already replaced with a new one, so everything is back to normal.
A long time ago my TX MGA-81563 died too, also exact moment is unknown (prior to that I was experimenting with an external amplifier connected to Hack RF).
After the first failure, I bought 10pcs MGA-81563, I still have 8 of them left, I hope they not gonna die often :)

@otpisani
Copy link

otpisani commented Jun 28, 2022

I can also confirm that my tx MGA-81563 burned, I wad doing some testing with gnuradio-companion and transmit with attached exteranl amp, but I didn't get even close to magic number of -5dbm. I also have couple of spare MGA-81563, but did someone try suggested part like LVA-123+ or similar, did hackrf one operates or not, also when you are using mayhem and transmit something @mautas with replaced MGA-81563, hackrf one operate like it should or not???

@greatscottgadgets greatscottgadgets locked and limited conversation to collaborators Oct 24, 2023
@greatscottgadgets greatscottgadgets deleted a comment from otpisani Oct 24, 2023
@greatscottgadgets greatscottgadgets deleted a comment from surpscare Oct 24, 2023
@greatscottgadgets greatscottgadgets deleted a comment from LilaQ Oct 24, 2023
@greatscottgadgets greatscottgadgets deleted a comment from otpisani Oct 24, 2023
@greatscottgadgets greatscottgadgets deleted a comment from bm16ton Oct 24, 2023
@greatscottgadgets greatscottgadgets deleted a comment from LilaQ Oct 24, 2023
@greatscottgadgets greatscottgadgets deleted a comment from Mclute0 Oct 24, 2023
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
help wanted help from community wanted and appreciated
Projects
None yet
Development

No branches or pull requests

7 participants