PH2LB shack on the internet.

QRP-Labs QDX

This page was last updated : 22-Jul-24 23:05:36

This page is work in progress and may be updated frequently (last update : 2024-06-20 12:23 UTC). 

Page content : 

• Introduction 
• The manual
• Let start building  
• Powering up
• Does it work ? 
• Output power and current draws
• 9V? Yes 9V
• QDX Mosfet mod
• What's next

Introduction

Loving the kits from QRP-Labs, and when Hans Summer announced his new QDX digital signal tranceiver I knew I wanted to buy and build one. And when he added it to his webshop I didn't hesitated for a moment  and was to late. The next batch, I forgot, but when the 3rd batch was sold I didn't and I ordered a QDX incl the housing.  

This little kit with a stock of 1000 was sold out in under 4 days. Showing that hamradio kitbuilding is here to stay.  
So after a litte more then a week the parcel service dropped of a package from QRP-Labs. 

Content :

• QDX - Digital Transceiver
• http://shop.qrp-labs.com/enclosures?product_id=164  

FYI : I sorted the components and put them in the plastic storage box (it's not how QRP-Labs ships there parts).

I got a revision 3 PCB. All SMD components are installed 

Front side 

Back side. 

 

The manual

As with every kit which Hans sells, the also was a very nice manual for the QDX - Digital Transceiver (link).  Chapter 1 contain a Introduction  Chapter 2 contains a list building general guidelines, images of the PCB layouts (track and component placement),  technical specification and the full Parts list of all components in the kit and photo's of a finished QDX allowing to kit builder get a idea about what to expect. And after sub chapter 2.3 Inventory parts helping the starting kit builder to identify parts, the fun starts.

Let start building 

Because my eyes are getting worse, I used  a magnifier to sort the capacitors and write the value on the paper tape.  That way I don't have to continuously use the magnifier.  

Section 2.4 describes the placement of the capacitors and after a 15 minutes of soldering all the capacitors where in place.  

 Section 2.5 describes the placement of the 6 diodes and section 2.6 describes the placement of the 4  47uH inductors

 Section 2.7 describes the placement and installation of the 4 BS170 fet's. 

 As of section 2.8 the fun really starts. I don't know why, but a lot of OM's hate winding toroids, coils etc. For me it's a very relaxing thing to work on.

Hans writes in his manual : 

It is a good idea to de-burr the toroid because sharp edges can damage the insulating enamel on
the wire – however be careful since the ferrite is brittle and can easily be broken. So GENTLE
scraping of the edges with something sharp like a knife is possible.

I took a closer look at the BN43-202 binocular ferrite and indeed there were a few sharp edges. To remove them I use my old scraping tool which I own for decades. 

 I used have to place the tip of the tool into the BN43-202 binocular ferrite and turn it a few times and all the sharp edges where gone. 

And following the instructions in Section 2.8 I wound the BN43-202 binocular ferrit.

 I remove the enamel off the wire I use a X-tal knife and a stainless steel plate (see also this how-to).

By just turning the BN43-202 binocular ferrite 4 times, all enemal is scraped off and 

 Hmmm looks like a kind of animal. 

 The BN43-202 binocular ferrite is installed.

 For some reason I skipped winding the other toroids and started with installing the connectors as described in section 2.14. To be sure the connectors are in the correct place, I used the back plate from the optional enclosure.

to keep the DC chassis part in place and aligned I plugged in a DC plug to keep it in place.  

A day later I started again in section 2.9 and took time to wind all the toroids (but forgot one).
Quiz question : which one did I forget ? 

 

After removing the enemal from the wire with X-tal knife and a stainless steel plate (see also this how-to) I soldered the toroids in place. 

Answer, I forgot the trifilar toroid T2 as mentioned in section 2.12. But no worries, I found out before applying power ;-) 

T2 installed so we are almost done. 

 The status LED is soldered in place and the tape as mentioned in section 2.15 is also installed. 

Time to box it all up acording section 2.15 Optional enclosure.

Powering it up : smoke test

Then the big moment was there. USB cable connected, Antenna connected, 9V power applied, current draw 140  and  . . . no magic smoke.

That's a good starting point. 

 Because I build the QDX for a 9V supply voltage I added a label on top of it, that I should be powered with 9V only. 

 I followed to operation manual  (link) and used a terminal program to look around in the configuration menu see Section 4.

Does it work ? 

After everything looked good, I configured a instance of WSJTX as described in section 2 of the   operation manual  (link) and after a while RX on variouse bands, I decided it was time to try to TX and I called MQ3FON. MQ3FON replied and gave me a report. So TX also seems to work.  

Monday 20 june 2022, I was a day off and left the QDX on to monitor FT8 on 20m. Below are images from a 6 hour RX / TX report from PSKReporter. 
(click on image to get more detailed image)

Note : Once in a whille I replied to a CQ (interrupts from working in the garden).

Conclusion : yes my QDX works. And with a RX current of about 140 mA, I can leave it on cheaper then my FT897D for monitoring WSPR / FT8.

Output power and current draws

To measure the output power, I the use of my Rigol DS1054 scope with the following settings : 

• Coupling : AC
• BWLimit : OFF

For every band I used the FT8 frequency and clicked Tune in WSJTX, wrote down the UMax (V) value and used the following formula to calculate P output (W)

Also I wrote down the RX and TX current (at 9V).

IMHO it doesn't need much tweaking (for now).

Latex code for formula (I used https://www.codecogs.com/eqnedit.php as a editor ) : 

9V? Yes 9V

I have made my QDX to be operated at 9V (6 x 1.5V AA battery compatible) and because my main DC in the shack is 13.8V with powerpoles I needed a little step down box.

For that I used a  L78S09CV from STM which is a fixed linear voltage regulator capable of handling 2A. The schematic is very straight forward as shown below.

The PCB was cut to the size of the  small die cast aluminium case and used my Proxxon Micromot to cut the tracks and for the PowerPole connector I used the PA4Q powerpole chassis part.

 

2 stickers made with a labelwriter finishes it. 

Does it get warm ? Well a little (3.84W) but the metal box is dissipating enough

The QDX can be used for long duty cycles although not recommended (unless you reduce power). The peak currents are described in the QDX manual page 72 (max 1.1A).

To reduce output power I have plans to build another box like this which output's 6V. That one will need some kind of heat sink  because it will generate aprx 6.25W in the peaks. 

 

QDX Mosfet PA mod which didn't work for me (yet)

Note: I described this PA mod, it didn't work out for me. Please read my experience below. Note that it doesn't mean that this MOD couldn't been working for you.
For me at least it didn't. Why ? I couldn't find out. :-(

While browsing to The Tindie Store, I came across the JasonKit QRP shop and saw that he was offering a QDX Mosfet PA mod Kit. It was designed by Barb WB2CBA and uses two FDT86256 mosfets. 

In short, the PA mod accommodates two FDT86256 Mosfets and provides some heatsink PCB space. The layout matches the motherboard BS170 footprints so the mod can be easy installed.

The FDT86256 offers several advantages over the BS170.

• higher breakdown voltage (Vds), which stands at an impressive 150V, in contrast to the BS170's 60V.
• ability to handle higher power levels, with a rating of 10 watts, compared to the BS170's 800 milliwatts.

The drawback of the FDT86256 is that when compared to the BS170, and that's its higher input capacitance (Ciss), which measures 55 picofarads (pf). This higher capacitance can potentially impact its performance at higher frequencies. However, in a Class D configuration, it performs comparably to a pair of BS170s. Typically, the Ciss capacitance of a BS170 is around 24pf. When used in parallel pairs, the total capacitance reaches 48pf, which is not significantly different from that of a single FDT86256. So the FDT86256 is a good replacement for a BS170 pair.

An additional benefit of this PA Mod is that it allows for the operation of the QDX under 13.8 Volts. Under this voltage. QDX can comfortably reach the 5-watt limit without worrying about damaging the finals. However, it's important to note a caveat: the 5 Volt regulator on the QDX board tends to get hot at this voltage, so exercise caution when operating at 13.8V.

And looking at the price and reasonable shipping cost, I didn't hesitate and order a set. Then the waiting started but after a 2 weeks, a envelop was dropped of containing the PCB with a Mosfets. 

Next thing to do, find time to install it in my QDX.

22 July 2024

On my day off between the house hold task, I found some time to apply this modification.

First I removed the 4 BS170.

By adding a bit more solder and sweeping across the pins. Then I removed the left over solder with 3S-Wick.

Which always leaves some residue which needs some cleaning. But with a cotton tip with a few drops of isopropyl alcohol on it, that residue is easy to get off the PCB.

And after a visual inspection if I didn't damage something, I applied a small piece of the double sided tape I also used for the QMX+ tilfeet.

Using 0.6 silver plated copper wire, i made 2 bows to align the PA board to the main board. First solder the bottom part of the bows and after that bending the bow pins a little, the boards are held together. Then the top pins are solder and trimmed.  

4 out of 6 done, almost there.

PA board installed and ready to taste the pudding.

I connected the USB, antenna and power cord (connected to my current limited powersupply set to 9V with 1.5A max).
After doing a final check if nothing was out of the ordinary, I powered up the QDX and the powersupply noted a 146mA current draw on RX. 

For testing I monitored for a few minutes, replaced the antenna with my dummy load and let WSTJX call CQ for a few time while keeping a sharp eye on the current readout on the powersupply.

Peaks of 850mA on 20m which was less then the previous measurements (see above). But ok, less could be better (???).

The MOSFETS became (not to) hot to the touch and using a thermocoupler I measured 54 dgr C. aprx 130 dgr. F. Not to extreme but I'm gong to add some small heatsinks to it like Barb did.

Then hooked up my antenna and tried to make a contacts by replying to a CQ. No takers, but atleast we getting out of there.

 

How much does it throw?

Using my OZ2CPU digitale RF mW-dBm-mV meter in combination with the -40dB RF Sampler connected to a dummyload I measured the output power of this new configuration.

I will measure the power output for the various bands.

• 80m - 4.17W
• 60m - 3.89W
• 40m - 3.47W
• 30m - 3.24W
• 20m - 2.69W

Ok, these values are much different from what I had before (see above). That a thing I have to investigate. 

How much does it throw? continue . . .

I kept me thinking what could be the reason why the output power was much lower then expected. It seems that it frequency depended. And quite different from the original configuration. 

Could it be that the tape I used would have some negative impact ? The LPF reacted different to these PA FETS ? Stray capacity ? Faulty components? I don't know.

First I try to re-tune the 20/30m LPF, I couldn't squeeze anything out of it. So that wasn't the cause.

Second, I de-soldered the PA board and wanted to solder it on female pin headers only to find out that . . . . the pinning wasn't 2.54mm / 1 mil :-(  So I placed a small piece of material underneath it, 3 x the size of the tape and soldered them back.

Again no difference. 

For 2 hours I thought about it, tried peeking and poking the LPF's, the BPF etc. No cigar. Only thing I found out was that the PA board was getting to hot to handle, so I might have done something wrong or damaged something.
Out of idea's, I decided it was time to remove the PA board and re-install the BS170's (luckily I have them in stock).

I measured the power output for the various bands using my OZ2CPU digitale RF mW-dBm-mV meter in combination with the -40dB RF Sampler .

• 80m - 4.78W
• 60m - 4.79W
• 40m - 4.57W
• 30m - 5.24W
• 20m - 4.90W

These values are close enough to the original value and the BS170's don't get hot. So for now I leave it as it is. 

The PA mod PCB is in my "need to investigate" tray for future inspection. At-least I'm going to measure the FET's with my Peak Electronics DCA55 and compare the outcome with the datasheet.
Maybe one of the FET's it's defective or just a bad quality one. Time will tell.

To be continued. . . . . 

 

What's next ? (2 years old plan)

Setup a Raspberry PI to make a permanent WSPR monitor (using my PA0RDT mini-whip antenna as the RX antenna).

 

 

---