QDX rev3 from QRPLabs

Probably yes. I saw 440 but thought that 60 already sold.

He keeps 60 kits in stock for spare.

I’ve bought several kits from QRP Labs and always been impressed by them, not least by the comprehensive instructions. But I think Hans needs to put his prices up - except for the things I want to buy of course! :slight_smile:

3 Likes

Hi all,

For those who wouldn’t have had the chance to order the new QDX, this is where it happens :wink:

“”
QDX batch #2 availability, Friday 17-Dec-2021 1800Z
From: Hans Summers
Date: Wed, 15 Dec 2021 08:36:36 UTC

Hi all

QDX batch #2 sales will commence at 1800Z on Friday 17-Dec-2021, at QDX 4-band 5W Digi transceiver

QDX batch #2 is 375 kits. Some of these have to be allocated to batch 1 leftovers due to the QRP Labs shop over-selling in last few seconds of the frantic 14 minutes on 11-Oct-2021. There are also several promises to be fulfilled and various other special cases. So the actual number that will be available for general sale will be a bit less than 375. The QRP Labs server has been upgraded to a higher performance machine and so maybe we won’t see the same overload issues as last time.

Regrettably, I have to announce a price increase from $60 to $65. The original batch 1 QDX kits were manufactured in January 2020, a whole different world ago. Unfortunately many aspects of the manufacturing and logistics have increased in price. Even the main microcontroller chip STM32F401RBT6 procurement cost me in excess of $5 extra now, compared to the price in January 2020 before the global semiconductor shortage crisis hit. Just check Digikey, Mouser, Arrow etc., and you’ll begin to understand the difficulty of finding these at all, at any price. And this is just ONE part of the QDX. So I have no choice, but to pass some part of the cost increase on, by increasing the price of the kit; $65 is the best I can do at this time. The price of Aluminium has also seen large increases since batch 1 and this has affected the manufacturing cost of the enclosure too; however at this time I am not increasing the price of the QDX enclosure. I ask for your understanding on this small increase and let’s all hope for better days in the not too distant future.

QDX batch #2 has a new Revision 2 PCB. In particular, Rev 2 PCBs no longer need the four 47uH SMD inductors to be replaced with through-hole components, nor do they need the 220uF capacitor at the DC input, nor do they need the 25MHz TCXO daughtercard installation. So the assembly is easier to do. The full list of changes for PCB Rev 2 is as follows:

  1. The SMD pullup resistors R5, 10, 11, 12 and 14 that pulled up MOSFET gates to +5V were deleted, as they were not necessary.
  2. QDX batch 2 boards will be supplied with firmware 1.03 ready installed (which will be made available for download shortly), which drives 5 I/O pins in Push-Pull mode rather than Open-Drain mode, this is required to be compatible with the pull-up resistor removals mentioned above.
  3. L1, 5, 7 and 9 (47uH SMD) have been replaced with through-hole axial molded inductors that are installed in the PCB properly, like all other through-hole components, rather than tacked on the bottom as a modification.
  4. IC2 (5V regulator) has been replaced by 78M05 which is a very robust regulator and this removes the need for the added 220uF capacitor. Additionally IC1 (AMS1117-3.3), the 3.3V regulator, is now powered at 5V from the 78M05. It is therefore working much less hard with only a 1.7V drop.
  5. The 25MHz TCXO, decoupling capacitor, and coupling capacitor are no longer on a separate tiny TCXO daughtercard; they are now SMD components installed directly on the underside of the board by the SMD assembly factory.
  6. A P-channel MOSFET is added at the DC input socket as reverse polarity protection. The device used is AO3407. It is rated 4.1A and has a specified typical ON-resistance of 34 milli-ohms. At 1A (say) on a 9V QDX the voltage drop will therefore be 0.034V which is of little consequence.
  7. A potential divider circuit added to bias the BPF switch IC3 at mid-rail; this was neglected in the original and does not seem to cause any issues but perhaps in the case of very strong signals, could have a few dB impact on dynamic range.
  8. IC7 has been moved a little further from the board edge following some reports that it could conflict with the PCB rail in the aluminium extrusion; this seems to be a manufacturing tolerance thing since I did not see it on my builds here, but it did affect some people.

Please let me know any comments or questions.

73 Hans G0UPL

“”

73 and good luck
Éric

PS :
Hans @G0UPL is working on a high band version …
Stay tuned

4 Likes

Eric thanks for the update!

Additionally the TRX can be used on 60m! See info on webpage.
So 60m-40m-30m-20m will fit well with SOTA.

4 Likes

Andy,
There are CW programs that use the keyboard and USB audio tones just as WSJT and most RTTY programs do. So I can’t see why CW is ruled out.

I’m sure a few CW contacts I’ve had were with the other op using a keyboard.

Personally I find their decodes are full of errors but in low noise environments it might be viable.

I think you could figure out how to hear the CW and send it with a key with minimal extra hardware.

73
Ron
VK3AFW

1 Like

le père noël n’est pas loin … pourvu que ma ligne ADSL tienne !!!
73 Nicolas

2 Likes

I heard he was working in a shopping centre where there was a Covid spreader and now he has to isolate for 14 days.

73
Ron

1 Like

Maybe such personal information / gossip should not be spread on the Internet.
My 2 cents
73

2 Likes

Sigh. I was referring to Santa as referenced in Nicolas’ post. I think you misunderstood? Santa’s personal details are public property are they not.

Okay sorry I missed that. Santa is a tough lad and will manage.

2 Likes

That’s OK Joe. One of my life’s burdens is making jokes that aren’t understood. In this case I was implying that Santa with his sack of new rigs would not be calling until New Year’s Eve.

A joke that has to be explained is a poor one. Most of mine are in that category.

73
Ron
VK3AFW.

5 Likes

Hi all,
Last new on the 2nd batch of QDX.

“”

QDX batch #2 availability, Friday 17-Dec-2021 1800Z
From: Hans Summers
Date: Thu, 16 Dec 2021 05:24:49 UTC

Hi all

I don’t have a technical way to enforce one per customer. Checking each order to make sure they hadn’t ordered in batch 1, and contacting them if they had to ask why, would be an excessively time-consuming task that I nowhere near have resources for.

We will have to rely generally on good nature,

However, don’t overestimate the size of the problem. This isn’t a QCX which is a single band radio available in 7 different band versions. QDX is one radio, one version. Why would anyone in general, want to buy more than one? They might want to build two together with a friend. Which really is the same anyway as one per customer except we don’t know who the other customer is. Similarly if someone already has a working QDX, in general they don’t need another one.

I should focus on the REAL problem. Which is: in our difficult world today, how to get the parts to be able to produce enough QDX kits so that, like QCX, everyone who wants one can have one.

The current difficulty is the AK5386 (ADC chip) which isn’t just marked as out of stock, but following the fire at the AK factory in Japan, is now also marked obsolete. Like most other electronic parts dating back to the 807 and further… There’s always going to be a few somewhere, but that isn’t the same as being able to buy thousands at a reasonable price to make a batch of kits. I’ve identified a possible alternative ADC chip and procured samples for testing. If I can get that to work then I can make more QDX.

73 Hans G0UPL

“”

73, Éric

1 Like

Good luck with your efforts. You are providing a great rig at a give away price. The only thing we can be sure of is that everything could change tomorrow.

73
Ron
VK3AFW
PS I like the idea of avoiding the audio modulation.

Fast!!!
I got my QDX 18:02 UTC :blush:.
Available:
18:02 180
18:03 124
18:04 66
18:04:30 24
18:04:45 Out Of Stock

about 360 QDX in less than 5 minutes.

73, Ludwig

2 Likes

Hans should put his prices up. :slight_smile:

1 Like

The QDX’s were sold out in under 5 minutes! :open_mouth: Fortunately, I was able to get one!

Dave, AE9Q

1 Like

Today in the very first hours I finished my QDX …

(I had to use 2 old capacitors instead a missing part - Hans will send this c in the next time.)

… and in the afternoon I checked the parameters. (There are also some fine built in tests!).

The output is around 4 … 6 W @ 11 … 13 V.

Then I made a real test on 30 m with a fast success - hurrah!

Now I’m waiting for a chance to test from a summit.

Hope to meet you then!

73, Ludwig

6 Likes

Hi all,

For those who did not have the opportunity to have this little digital QRP during the last QRPLabs sales, you will be able to order one because Hans @G0UPL has just made this announcement for a new batch of nearly 1900 pieces.
For the SOTA this QDX rev3 with a Raspberry Pi you will have the possibility with little weight to activate many summits :stuck_out_tongue_winking_eye:

I have no interest but I love Hans’ work :+1:

“”
#QDX: 01-May-2022 Sunday 1800Z feels like a good time for another pileup!
From: Hans Summers
Date: Sat, 30 Apr 2022 09:05:41 UTC

Hi all

The QDX boards arrived here yesterday lunchtime. I’ve been building and checking and I’m happy.

So let’s open sales, 1800Z, Sunday 1st May 2022!

QDX 4-band 5W Digi transceiver

The trifilar transformer is a bit close to the board edge. I don’t think the solder pads would touch the enclosure rail but I put a piece of tape over it just in case. Other than that, everything is perfect. See attached picture.

First QSO was PD0WAG on FT8 40m last night who answered my first CQ immediately.

Testing, I used my calibrated -60dBm 7MHz signal source, connected to the QDX Rev 3 input, and a QDX Rev 2 PCB for comparison. The audio gain setting in the configuration of each one was set so that -60dBm appeared at -60dB FS audio on the PC. Recording with Audacity at 48ksps and spectrum analysis (Hanning window, 8192 points); comparative results attached. Rev 3 has a 6dB lower noise floor than Rev 2. In other tests, it appears to give that up at the top end i.e. the dynamic range is similar. So broadly speaking, Rev 2 and Rev 3 are similar, but the lower noise floor in Rev 3 (meaning improved sensitivity), will come in useful when considering higher frequency bands in future.

QDX Rev 3 replaces the now-obsolete AK5386 ADC chip with PCM1804 which has a differential input; I had to re-design the baseband amplifier / ADC driver to suit this. The new circuit therefore uses two op-amps in instrumentation amplifier configuration. In some scenarios the improved and excellent balance may also help performance (noisy environments, supplies etc).

Rev 3 also has a PTT output connector (3.5mm jack); the tip and ring provide two ways of using PTT, either as a grounded output (suits most amplifiers) or as a positive-going +5V output which suits the QRP Labs 50W PA. Use of the PTT will require firmware 1_04 (soon).

Rev 3 has pads for the Band0, Band1 signals that select the BPF band, and also the LPF 0, 1, 2 signals that switch the three LPFs. These can be used if you are planning to drive external equipment.

Rev 3 as the PCM1804 ADC and the two op-amps on the top side of the board; the SMD components are therefore on both sides of the PCB because I couldn’t fit them all on the lower side. Well I could have done but it would have been at the expense of groundplane and I do like a good groundplane. A few of the other components are moved around so that the signal path is shortened (shorter connections between logical stages of the transceiver).

There are 2,000 Rev 3 PCBs.

I will initially release 1,900 for sale. That will therefore keep 100 in reserve some for failures during flashing/testing here (manufacturing yield 100% would be nice but rare), any over-orders, and such.

Also at the point of ordering, everyone will have to click a mandatory check-box acknowledging that they understand: it is likely to take some weeks or more than a month to get them all shipped! We are a small team, and it is a huge amount of work flashing, testing, packing, preparing import and export documentation, shipment bookings, etc. A big proportion of the kit parts are already packed and ready here, and the enclosures are all ready and prepared. Enough toroids and NP0 capacitors to complete the 2,000 aren’t here yet, they are with FedEx… due here in a few days. Next week we have public holidays Monday/Tuesday/Wednesday here so won’t be able to start in earnest until Thursday. In other words, patience will be needed. First shipments will go to the people who over-ordered in the last batch in December and have been waiting patiently since then.

Please note that the enclosure has a change to the rear panel, the BNC RF connector and the USB connector are in the same place as Rev1/2 but the power connector moved slightly and there is a new hole for the 3.5mm PTT jack.

Price change is a $1 increase, from $65 to $66… there is an increased parts count (extra PTT connector, extra op-amp for the differential drive); increased parts cost (PCM1804 ADC compared to my pre-pandemic AK5386 stock), increased manufacturing cost (double sided PCB), but slightly lower STM32F401RBT6 cost. All in all, considering the design changes and the state of the world, limiting the price increase to $1 is a surprisingly good result! I also still have 1,000 chipsets in stock for a future batch #4!

73 Hans G0UPL

“”

adjust your alarm clocks and smartphones if you are interested !
It will go very quickly - GL

73, Éric
F5JKK

5 Likes