EXTTR failure

[Richard Wyatt]

I have had my HL2 for a couple of months and been using it with a 60dbm neptune 100w amp for the last month. Tonight I have noticed that the amp is not going into TX when the HL2 is.... upon further investigation its not the amplifier. My HL2 is not grounding the EXTTR when in transmit anymore. Other than that my HL2 transmits and receives absolutely fine.

From what I can gather, this is a hardware issue as the EXTTR port is not software switchable.

I have uninstalled/reinstalled thetis twice and deleted all the extra folders for a fresh install, reset the database.....just in case, but no change.

I have checked the port with my multimeter and its a constant open circuit, even when HL2 is transmitting.

Is this a component failure?

Can anyone shed any light please. 

Thanks in advance.

[Graeme Jury]

Hi Richard,

I have had this happen to me twice and others have posted with the same issue. The switching FET is pretty light for the job in many cases and fails. I finally replaced mine and followed it with an npn transistor driving a high voltage power FET with a pulse swallower diode bridged across it. Problem solved. I had to search around a bit to find an FET that switched reliably with 3 volts but there are several out there - most need about 5 volts.

73, Graeme ZL2TE

--
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/f616bfb7-37b2-46bd-90b2-c87f237dd76fn%40googlegroups.com.

[Richard Wyatt]

Thanks for the info Graeme. I thought this may be the case. Q5?? What is frustrating is that I was in the process of building an amp buffer for this exact reason, and to switch 2 amplifiers. Although the 60dbm Neptune quotes less than 1mW to switch the PTT. So absolutely well within the specs for the driver transistor. That's what I cant understand. Anyway I've ordered a few replacement NUD3124 and hopefully be able to use my soldering skills to replace the offending transistor. Any tips on clean removal as you've done it a couple of times already. I will be able to resolder the new one in place with little issue. Its just the removal as I don't have a hot air gun.

[Graeme Jury]

It comes off pretty easily. I just went round the pins a couple of times with the iron with a small amount of sideways pressure and slid it off the pads which I cleaned up with solder wick and swabbed with isopropyl and a cotton bud. If you are worried about lifting the pads, try sliding a piece of 35 ish awg wire between the body of the NUD3124 and the pin in a loop then apply heat and slide the wire along the pin and pad which wipes off the solder as you pull it out and leaves the pin floating. I have removed a 120 pin microprocessor the same way without a hitch.

73, Graeme ZL2TE

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/c945d638-db61-4f9e-8bc6-ee7ed7cfb2f8n%40googlegroups.com.

[Richard Wyatt]

Thanks for the tip Graeme. A little knowledge helps a lot. I just have to wait for the part to arrive and ill get cracking. I inspected the transistor with a magnifying glass but it looks perfect, nothing obvious like cracks or a hole in the housing etc. Only time will tell.

[Richard Wyatt]

OK so I replaced Q5 and EXTTR is short to ground. So I powered HL2 up and measured TP4  to ground with no Q5 fitted. Its a constant 3.041V........so what next??? 

Regards 

Rich

[Richard Wyatt]

Anybody???

[Mike Lewis]

Have you verified it is not a Thetis software related issue and try using another SDR app or the PTT jack on the HL2 itself?  

Reflash the firmware on the HL2?

 

• Mike K7MDL

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/c0a60a27-b43e-435b-8879-1a2935d31920n%40googlegroups.com.

[Richard Wyatt]

Hi Mike, thanks for the reply.  This issue is present outside of thetis, as soon as power is applied 3.041v is present at tp4 causing q5 to switch on.

Everything was working well until it just decided to mess up while in the middle of a QSO. The 60dbm.com neptune 100w requires less that 1ma to switch to tx, so well within the 80ma maximum rating of Q5. This failure should not have occurred?

From what I have read on this forum, this appears to be a failure of q5 which has damaged the FPGA??  I'm not the only person that has suffered this failure??

As it stands I have been without a fully functioning radio for over a week and I am £20 out of pocket by ordering the parts from mouser that did not fix the problem.

My HL2 was 59 days old at the time of failure. Do I need to RMA? 

Please advise.

[Steve Haynal]

Hi Richard,

Just to confirm, without Q5 installed, with the N2ADR filter board removed, with your AMP disconnected, with no ethernet (obviously no software), but with the HL2 powered on, do you still measure 3V at tp4? Is this a clean 3V when measured with an oscilloscope or is there a signal?

There have been a few problems with Q5 but I don't think it is the device. I think it is operator misuse like not having a common ground, switching unsupported voltages (the 60dbm amp does seem to be well within spec), or RF entering the EXTTR port from a poor amplifier.

A constant voltage at TP4 can be a symptom of a blown FPGA pin or FPGA IO bank. This signal is on bank 2 of the FPGA. Are other pins on bank 2 still working? For example, do you see accurate temperature reporting or is it pegged to some untrue high or low value? When you switch bands in RX, do the filters switch on the N2ADR filter board?

If it is just the pin which is bad, I can build you a custom gateware which uses another pin (must be wired manually) for EXTTR.

The repair and warranty process is described here:

https://github.com/softerhardware/Hermes-Lite2/wiki/Repair

73,

Steve
kf7o

[Richard Wyatt]

Hi Steve, I only have a DVM, not a scope. But I can verify that with only 12v to the HL2, no ethernet, no n2adr, Q5 removed. There is 3.041-3.042v present with reference to ground at TP4.

As far as common ground, both my HL2 and Neptune amp share the same 13.8v 30A PSU and surely the rf cable between the HL2 and Neptune amp would ground both devices via the braid of the coax anyway. 

Everything was working and stable. Then in the middle of a QSO on 40m my power dropped from 100 to 3w and I noticed there was no red tx led on the amp. I investigated and found that I could manually make the amp tx by removing the phono cable from the HL2 and shorting it.

Ive just connected everything and the HL2 is reading 17.7C which is the ambient room temperature and I can hear the relays and see the band filters switch on the N2ADR board within the Thetis setup menu. As far as I'm concerned other than the EXTTR failure, HL2 is working as it should.

I really appreciate your offer to write custom gateware with a hardware mod to overcome this issue Steve, but for me I'd rather make sure everything is 100%, after all the radio is less than 2 months old.

I have already received a reply from Makerfabs to return the board for repair or replacement at the cost of shipping back and forth. 

I think I'll take this option, I want things to be 100% in case I sell HL2....It just wouldn't sit right with me selling it with a known issue.

Rich M0RDW

[Steve Haynal]

Hi Rich,

It is great that Makerfabs is doing a warranty exchange. I still have concerns that some are operating the EXTTR out of spec. There have been maybe a dozen problems I recall associated with EXTTR. The same exact circuit is used for the INTTR (Q2) and I don't recall any failures with that. Please be sure that your EXTTR is in spec, including transients, RF and high SWR mismatch between the PA and antenna. The datasheet for the nud3124 is here:

https://www.onsemi.com/pdf/datasheet/nud3124-d.pdf

Also, take a look at this post about HL2 and PA transients:

https://groups.google.com/g/hermes-lite/c/LDwyAgLK2uk

73,

Steve

kf7o