Help! Killed my RS-HFIQ!

[WA2EUJ]

Hi Rob,

There are two sources of power for the RS-HFIQ. All of the digital and baseband circuitry is powered by the USB which sounds like it's working OK. All of the analog RF circuitry is powered by 13.8V. There are two green LEDs (one for each power source), are both of them lit? Have you checked to see if the RS-HFIQ is getting 13.8V?

73,

Jim WA2EUJ

[Björn]

As far as I understand the digital baseband and analog are powered by a 5V derived from the +14V via  U17, T4 and U7, i.e. an isolator followed by a linear regulator.

So even if your USB-stuff works you might miss the +5A as there are no LED on that power. 

Another 5V source is 5V_RF which is linearly regulated directly from the 14V in U5.

You should, as Jim pointed out, check for both green LEDs as one of them are the +5A that is the derived 5V.

Better is if you find all regulated powers and measure them using a multimeter.

If ALL power are good my most intelligent guess is U9, the low noise amplifier that is probably the most sensitive component for power surges and are involved in the RF-path for both RX and TX.

Good luck!

I fried my power transistor and physically destroyed the bias postentiometer just before leaving home for the summer house and completely messed up my construction of a good transceiver. The summer house is on an Island where QRM is more or less unknown :)

Now I have got another rig to get on the air while getting ready to take my RS-HFIQ on the air or I had missed my opportunity to start my HF-experience out of the noisy city and all QRM. A wonderful rig, but I WANT to build my own...

[Björn]

EDIT: the LEDs show the 5V regulated directly from the +14V and not the isolated 5V from the 14V.

[rob...@thefrenchfamily.org]

Jim,

If I plug in just USB, I get one green LED.

If I plug in a (working!) 13.8V power supply, and turn the power switch on, then I get the second green LED.

No receive audio.

No apparent transmit power.

As an aside, I did discover that, although my antenna tuner appeared to be working at first blush, it is not correctly calculating SWR... it is much lower than my antenna analyzer shows (e.g. when I tried to tune up my 40m dipole on 80m... analyzer shows high SWR, tuner thought it was < 2:1!!!).  I checked the forward voltage drop cross the two rectifiers in the tuner's SWR bridge, and I get ~270mv and ~365mv respectively (they're BAT41's).  Don't know which one is bad, but I imagine this indicates that at least one got damaged by the surge.  Hopefully the ADC in the tuner's processor didn't get zapped... Anyhow, seems obvious that the path into the radio was via the antenna, through the tuner, into the radio,, and then from there it got to the +14V and zapped the power supply.

Back to the RS-HFIQ, given that I have two green LEDs, I understand that at least V+ and 5V_RF are good. 

- I double-checked 5V_RF via U5.  Good 13.8V in, good 5V out.

- I measured the input and output of U7 relative to AGND (at the base of R44).  7.4V input, and 5.8V output... I think that is a good +5A?

Any thoughts as to what is next on my list?  As I look at what is between J7 and the nearest path to +14V, I'm definitely concerned with the number of components.  E.g. do I need to worry about things like C110 being bad?  All the various diodes?  (I would have thought it would be, path to GND, but given that it killed my 13.8V power supply...) 

The radio would have been in receive when it happened, so I'm guessing all of those switching diodes would have opened a path through to U9 and beyond.  Is there a way (without any fancy test equipment) to generate an internal signal to verify whether or not the I/Q (audio) section is intact?

Thanks in advance for any help you can provide.  I'll probably be able to replace an SMD part or two, but much beyond that, I'll have to be looking for someone with the right tools/skills who I can pay to fix this... I fear it will rapidly exceed my abilities!

It's a bummer... this has been a great rig.  Lesson learned about lightning, even with an attic antenna.

-Rob KC4UPR

[rob...@thefrenchfamily.org]

Hallelujah!

I ordered a 4 pack of RD16HHF1's from HobbyPCB (I wanted 2, 1 for a backup, but anywhere else I could buy them from, by the time I paid for shipping and handling, I was going to pay just as much).  I replaced the final amp transistor in my non-transmitting RS-HFIQ, and voila!  5 watts output!

To recap what brought me here... several months ago, as I was doing some miscellaneous thing in the shack, I noticed that my RS-HFIQ appeared to be stuck in transmit.  In addition, there was an overload indication on my antenna tuner.  And furthermore, the power supply for my RS-HFIQ was shot (I'm assuming the LEDs were all on due to the Nano).  Anyway, the sum total of the damage was:

- Switch power supply was dead.

- Antenna tuner was bad, or rather, the SWR and power indications were wrong.  (I ended up fixing this by replacing one of the two diodes on the Stockton bridge.)

- The audio input on my sound card was dead.

- My RS-HFIQ would not transmit.  (I did determine it could receive, after I replaced the sound card.)

Warren, WA8TOD, was super helpful and corresponded quite a bit by e-mail.  He showed me the trick about measuring the RS-HFIQ current when the USB was unplugged, as a check of the bias, and based on that, it appeared that either the bias was bad, or the final was bad.  As best I could, I measured the bias voltage, and it seemed to be okay.  I also check what was happening at the EXT_PA connection, and found that if I touched a probe to that while transmitting, (a) I got voltage, and (b) my SDR receiver picked up the transmission.  So clearly transmit was happening, but wasn't getting passed the final amp. 

So, replaced the amp, and it appears to be good!  I am still going to do a few more checks.  Also, if anyone has any advice on any additional adjustments I should do, that would be appreciated... do I need to do anything with the bias pot?  Or should it be close enough from the previous final amp transistor?  (If I do need to adjust the bias, are there instructions for that?) 

I will say that replacing that transistor was a bear. Specifically, unsoldering the center pin of that transistor, with all the copper that it is connected to!  I ended up using some tip files to get the holes big enough to get the new transistor into.  I'm guessing that is also lead-free solder... ack.  Anyway, I got it done.

Excited to start using it again!

Rob KC4UPR

[rob...@thefrenchfamily.org]

One aside... so I am still curious what other components, if any, might also have been killed on my RS-HFIQ.  There's no visible damage, but who knows.  At some point, it would be good to get someone with a spectrum analyzer to make sure something in the filtering didn't get killed, so I can be sure I'm not splattering trash all of the bands. 

One thing I did was check continuity across C110, the coupling capacitor at the antenna jack J7.  I was surprised, because I actually measured 4.56k ohms across C110, with no antenna plugged in.  I don't see any components that would cause this... is the capacitor bad?

Thanks

Rob KC4UPR

[rob...@thefrenchfamily.org]

Another observation.  I notice that when I initially transmit, I get a brief blink of the clip LED, and then it goes away (when I have the TX audio adjusted for about 4.5 W output).  This is something that I typically have observed on my uBITX whenever I transmit, but it only occurs when the T/R relay switches.  With my repaired RS-HFIQ, I also see this whenever I key... but since it is able to do full break-in when I use Quisk, I see it every time I key.  This is NOT something I observed in the past with my RS-HFIQ, so I am guessing something else has changed too... either some other bad component, or some adjustment that I need to make.  Any thoughts?

Thanks.

-Rob KC4UPR

[F5NPV Amateur Radio]

Hello Rob

Happy you manage to sort out your issue.

- i am powered my RS-HFIQ with 13.5v

-It is not recommended to power only the USB , because it involes some current peak on the digital cirtcuit, when you switch off the main switch , the USB be disconnected or not powered

-Both Green LEDs are ON when the main power supply + USB are ON. 

-According the Clip LED , most of a time this led is blinking for me when transmitting . At the early stage i was afraid it will affect my signal quality , but in fact after a carefull check i did not notice any issue. As far i can remember i manage to address this issue with some adjustment on the sound card input level. But nowadays i do not pay much attention. In order to ensure everything is ok , try to monitor your transmit signal quality with your UBITX , just in case.

So far so good , i am still using your CW mod with Quisk 1.57 . 

After one year using the RS-HFIQ and associated software likely Quisk is a great experience. Great and reliable device. All reported a good voice quality (i am using a very cheap USB microphone and i am very surprise regarding the audio quality)

Most of the time i am using the device with a 40w Mosfet and EL519 300w tube amplifiers (Homebrew/homebuild) .

Since the propagation is back bit by bit , hopefully we will be able to chat on the Air.

Merry Xmas From France  guys

F5NPV / Didier

[all...@gmail.com]

Hi Rob

I am doing this from memory but I think I have it right. Jim may want to confirm or correct my deductions.

The clip light is driven by the software. A circuit detects the output voltage from the PA and compares it with pre-determined values that are stored in an array in the software. There is one value for each band of operation. The light indicates that the output has reached a level that was previously determined to be one where the distortion produced by the PA exceeds a level that the author has deemed to be acceptable. 

In 2016 when I first received my RS-HFIQ I did extensive testing and found that the level at which the clip light came of varied widely by band from as low as 3.8 watts and as high as 6 watts depending on the band selected.

It is possible to change these values to make the light come on at whatever power level you wish. Normally doing so would require use of a spectrum analyzer and a two tone test signal. The drive levels are increased until the distortion becomes too high and then the constant for that band is altered to make the light come on at that power level. This is not something I would personally attempt without a spectrum analyzer and the knowledge to use it. 

The light should be considered a guide. Many factors can affect the power output such as gain of the PA etc. Since the power level detection is purely voltage driven a bad SWR match can make the light come on at a lower power level than when the match is perfect.

WA8TOD

[rob...@thefrenchfamily.org]

Thanks for the inputs, guys!

Regarding the clipping LED, the reason that got my attention was that this is the first time I have observed it with my RS-HFIQ.  Note that this is NOT about the clipping LED coming on during steady state transmission:  I usually adjust the output of my sound card until the clipping LED goes out, and then adjust it down just a tad more.  I usually have to do this separately for CW and digimodes, but big picture, during a steady state transmission, the clipping LED is OFF.

What I have noticed after my final amp transistor repair, however, is that the clipping LED flickers briefly every time I transmit, at the beginning of the transmission.  For FT8, I don't even notice yet... it might flicker on, and then goes out for the duration of the transmission.  However, for CW, the clipping LED comes on at the beginning of every dot/dash.  This is definitely NEW behavior post-repair.  That is why I am wondering if some other components, besides the final amp transistor, were damaged in "the event".  (Especially given the other things that took damage... sound card, power supply, antenna tuner.)

Monitored via my power meter the behavior is very similar to what I normally see with my uBITX.  With the uBITX, whenever I transmit, the T/R relay clicks over, and for just a second the power on my power meter spikes a couple watts higher than expected, and then settles out at the expected power output.  I have always just attributed this to some artifact of the T/R relay... my RS-HFIQ never had this happen, the power would just go smoothly to the expected level.  However, now I have noticed that the clipping LED and the initial spike in output appear to be correlated.

This couldn't be at all related to the fact that I can measure continuity across C110, could it???

Regards,

-Rob KC4UPR

[rob...@thefrenchfamily.org]

Extra weird... so the clipping LED is not just associated with an initial switch from receive to transmit... it appears to be associated with the leading edge of a pulse. Again, steady state signal, I get normal indications.  For example, I have my audio output from the sound card set to give me about 4 watts output power.  If I transmit a continuous CW tone into a 1:1 load, I get an initial spike in power up to about 5-6 watts on the meter, and then it settles out at 4 W.  The clip LED lights just briefly with that initial spike.  Same if I transmit FT8. 

But, I was doing Feld Hell via Fldigi, which is obviously like a very fast CW signal (but transmitted using SSB with Fldigi), and the power output on my meter is popping all over the place, and the clipping LED is flickering constantly.

So of course the leading edge of a pulse is full of a bunch of harmonic content.  Are there any particular circuits in the TX chain that would be responsible for shaping the leading edge of any kind of TX pulse?  I realize the RS-HFIQ is just supposed to take in audio I/Q and convert that to RF, but clearly (in my mind at least) there is some harmonic content getting through from the leading edge of my transmit pulse.

Rob KC4UPR

[jcve...@gmail.com]

Hi Rob,

For the clipping LED, the RF voltage at the drain of the MOSFET is capacitively coupled (C8) into the the voltage detector formed by D27 and D28. The detector circuit is coupled to the digital ground not the RF ground so the Nano can read it but that is a hokey setup and is prone to digital switching causing the LED to blink. And the conditions under which it might erroneously blink can change with your setup.

I would also comment that the RX/TX switching in the RS-HFIQ was NOT designed for any pulse shaping. It's designed to be activated when no I and Q are present then the rise-time of the signal is controlled by the DSP creating the I and Q waveforms. Applying a constant I and Q 1 KHz signal to the IQ in and switching the signal on and off with the key line would be a BAD way to make CW.

The IQ Keyer makes a quardature 600 Hz signal that modulates for the I and Q input and ramp those signals ON and OFF in accordance with the ARRL's definition of a 'good' CW signal. You can look at the Arduino cod eto see how that happens.

73,

Jim WA2EUJ

[rob...@thefrenchfamily.org]

Jim,

Thanks for the info.  In my case, the CW is being generated (and presumably shaped) by Quisk.  So, not to worry, I am not using a constant tone switched via the key line.  I am using I/Q audio generated by Quisk.  I will say, I do use the key line... I have a GPIO key setup where I key the key line (rather than using the serial commands) to begin transmitting, and then have a slightly delay after the last CW key down time before I unkey the key line.  So I am using the key line for T/R switching, with about 0.25" delay for turn off.

As I noted, I see this same behavior (clipping light at the beginning of "pulses") when I'm sending Feld Hell using Fldigi.  So the pulse leading edges that seem to cause the clipping LED to light, are the leading edges of the audio pulses. What I was wondering is, if possible as this gets mixed to RF, if the harmonics on those pulse leading edges are not getting filtered, resulting in extra RF being read by my power meter.  Again, this would be if there was some other damage in the RS-HFIQ besides the final amp transistor I replaced.

There are some differences with my setup from prior to the "event" (static damage?) however.  Since the previous sound card was killed, I'm using a similar but different card... I may try changing sound cards too, to see if maybe this card is being noisy when it has a large change in audio output level.  Definitely will do that before I think about replacing anything else in the RS-HFIQ!

Rob KC4UPR

[Jeff S]

Can anyone advise as to what the part number is of U9?  Think this might be a problem on mine -- want to replace it, but can't reference it via Mouser.

Thanks -

Jeff

[Björn]

That should be  Mouser number 841-MMG3H21NT1  i guess?

I bought mine from Digikey with order number MMG3H21NT1CT-ND

Good luck!

[rob...@thefrenchfamily.org]

Out of curiosity, how would failure of that component be manifested?  I know someone had mentioned the LNA as something that could likely have been damaged on mine when it took its power surge/static/etc hit.  But I am able to both receive and transmit now, so I would think the LNA is likely fine.  (My current outstanding issue is still this weird little power spike on the leading edge of any "pulses", whether CW or digital.)

Rob KC4UPR

[WA2EUJ]

Here's a link to it on Mouser:

https://www.mouser.com/ProductDetail/NXP-Semiconductors/MMG3H21NT1/?qs=%2Fha2pyFaduhJ3KB97xW83K5d7fSIJ3lz7yWYSJJQkCBiUTRnGF7Sgg%3D%3D

Here's the line from the BOM is you want to source it elsewhere:

U9

TRANS HBT 20.5DBM 19.3DB SOT-89

FREESCALE

MMG3H21NT1

MOUSER

841-MMG3H21NT1

You can find the complete BOM here:

https://sites.google.com/site/rshfiqtransceiver/home/technical-data/rev-d-bom

[IZ3HRM]

Hi,

the component is a Low Noise Amplifier used both in the RX and TX path through some diode switches. In RX amplifies the signal, in case of failure you can hear only noise. In TX is the driver for the final amplifier (RD16HHF1), in case of failure no power is delivered to the antenna.

As per my experience (I did a copy of the design from scratch making some ingenuities in the pcb path design) the cause of fault is a self oscillation. Typically the LNA heat up al lot and it's power regulator U5 (MCP1703-5002E/DB) goes in a cyclical protection due to the overcurrent. looking to the LNA power pin with an osclloscope you se a square wave due to the described behaviour.

As soon as my design will be stable I will publish all.

73

mauro

IZ3HRM