QDX Deaf with high current. #qdx


Dave, M0OOR
 

Hi,
I have just finished building my QDX Version 3 for 9v operation.
The first thing I did on powering up was to update the firmware to v1.04.

Firstly, the current consumption on receive is 180mA and not the 100mA as suggested in the manual, I then notice that the 5 volt regulator gets quite hot and the 3.3 volt regulator is warm. connecting to WSJT-X the receive indicator barely reaches 2-3dB and very few stations are decoded (the Mic input of the PC is set to maximum).
I have tried increasing the receive gain in the Band Configuration command menu from the default 54dB to 70dB but with no improvement.
When I run the Input Analysis command, I get no waterfall and No Signal is shown at the top of the display.

I have visually checked the PCB for faults and all looks fine. I get the correct voltages on the LPF pin diodes.
If I transmit on 20m I get abut 3.5 watts output and I get reception all over Europe on PskReporter so I guess the LPF's are all ok.

Any help getting it working would be greatly appreciated.

73's

Dave M0OOR


Alan G4ZFQ
 

On 30/07/2022 18:33, Dave, M0OOR wrote:
Firstly, the current consumption on receive is 180mA and not the 100mA as suggested in the manual,
Dave, about 150mA has been reported here. Is your meter truly accurate? An extra 30mA would mean something like wrong components rather than a definite fault.

I then notice that the 5 volt regulator gets
quite hot and the 3.3 volt regulator is warm.
Not sure, never taken much notice. What is "quite hot"?

connecting to WSJT-X the
receive indicator barely reaches 2-3dB and very few stations are decoded (the Mic input of the PC is set to maximum).
I have tried increasing the receive gain in the Band Configuration command menu from the default 54dB to 70dB but with no improvement.
Again, not sure. The Audio filter sweep was OK? Confirms T2 is probably correct. The fact that the sweeps work tests the signal path. If you have a capacitor, say 1000pF +/- not critical you could try connecting an antenna via that to T2 primary, see what happens.

When I run the Input Analysis command, I get no waterfall and No Signal is shown at the top of the display.
Input analysis requires WSJT to supply a signal. This is not reallly a diagnostic test.

73 Alan G4ZFQ


Evan Hand
 

Dave,

One comment to Alan's post: Audio and RF Sweeps need to have a 50ohm dummy load connected to the antenna to work correctly.

As Alan does suggest, using the terminal Audio and RF Sweeps help in isolating where the problem lies.  I would run them for each band to verify the opposite side band supression.

73
Evan
AC9TU


howard winwood G4GPF
 

If the 5V and 3.3V regulators are hot, it would follow some other components will also be hotter than they should, especially I/C’s.
Finding same should point you in the right direction.

On 30 Jul 2022, at 19:33, Dave, M0OOR <cedar8@...> wrote:

Hi,
I have just finished building my QDX Version 3 for 9v operation.
The first thing I did on powering up was to update the firmware to v1.04.

Firstly, the current consumption on receive is 180mA and not the 100mA as suggested in the manual, I then notice that the 5 volt regulator gets quite hot and the 3.3 volt regulator is warm. connecting to WSJT-X the receive indicator barely reaches 2-3dB and very few stations are decoded (the Mic input of the PC is set to maximum).
I have tried increasing the receive gain in the Band Configuration command menu from the default 54dB to 70dB but with no improvement.
When I run the Input Analysis command, I get no waterfall and No Signal is shown at the top of the display.

I have visually checked the PCB for faults and all looks fine. I get the correct voltages on the LPF pin diodes.
If I transmit on 20m I get abut 3.5 watts output and I get reception all over Europe on PskReporter so I guess the LPF's are all ok.

Any help getting it working would be greatly appreciated.

73's

Dave M0OOR


Dave, M0OOR
 

I did do the sweeps with a dummy load on all bands and the USB graphs look fine albeit that the LSB graphs don't look as low as possibly they should do, so maybe the unwanted sideband rejection isn't what it should be? (I don't know what affects this).
The 5v regulator gets hot enough that you can't keep your finger on it, but I can't work out what is causing it, IC3 and IC4 get slightly warm and I'm worried that one of these may be faulty.
I think I am going to rewind the trifilar transformer and L12 and see if that makes any difference. Meanwhile, if anyone else has any ideas...

73's

Dave.
--
David Thomas M0OOR


Alan G4ZFQ
 

On 01/08/2022 12:42, Dave, M0OOR wrote:
I did do the sweeps with a dummy load on all bands and the USB graphs look fine albeit that the LSB graphs don't look as low as possibly they should do, so maybe the unwanted sideband rejection isn't what it should be? (I don't know what affects this).
Dave,

This test shows if T2 is wound and installed correctly.
How much down is LSB? If it is around possibly 50dBthen the secondaries must be wound and connected correctly. Fundamentally there is no adjustment. Seems to me if the secondaries are correct then the primary must be.

Again, if you get good RF sweeps it shows L12 and the capacitors are resonant.

The 5v regulator gets hot enough that you can't keep your finger on it, but I can't work out what is causing it, IC3 and IC4 get slightly warm and I'm worried that one of these may be faulty.
My QDX uses about 140mA RX when I feed it 7.5 volts. Others have said 150mA, I wonder how Hans got 100mA?

But I have just checked mine, the regulators run warm out of the case.
I cannot see how L12 or T2 could cause these to run as hot as you say.
It does seem something that requires investigating. Apart from measuring the 5 and 3.3 volts outputs I'm not sure how best to check. The normal procedure is to lift the relevant pin of an IC and note the difference in current.
Make absolutely sure there is no solder splash or wire shorting something.

If IC3, 4 are at a similar temperature they are either both faulty or like many ICs just at normal working temperature.

73 Alan G4ZFQ


Mike Easterbrook
 

I think the 100ma RX current value was from early QDX history and with the various upgrades most recent builds have been averaging out in the 140-150 ma range. I believe there was a note to this effect but can't locate it right now - away from home with poor search facilities (v.poor internet connection but great scenery - in Langkawi).
73 Mike 9M2LXM

On Mon, 1 Aug 2022, 22:24 Alan G4ZFQ, <alan4alan@...> wrote:
On 01/08/2022 12:42, Dave, M0OOR wrote:
> I did do the sweeps with a dummy load on all bands and the USB graphs
> look fine albeit that the LSB graphs don't look as low as possibly they
> should do, so maybe the unwanted sideband rejection isn't what it should
> be? (I don't know what affects this).

Dave,

This test shows if T2 is wound and installed correctly.
How much down is LSB? If it is around possibly 50dBthen the secondaries
must be wound and connected correctly. Fundamentally there is no
adjustment. Seems to me if the secondaries are correct then the primary
must be.

Again, if you get good RF sweeps it shows L12 and the capacitors are
resonant.

> The 5v regulator gets hot enough that you can't keep your finger on it,
> but I can't work out what is causing it, IC3 and IC4 get slightly warm
> and I'm worried that one of these may be faulty.

My QDX uses about 140mA RX when I feed it 7.5 volts. Others have said
150mA, I wonder how Hans got 100mA?

But I have just checked mine, the regulators run warm out of the case.
I cannot see how L12 or T2 could cause these to run as hot as you say.
It does seem something that requires investigating. Apart from measuring
the 5 and 3.3 volts outputs I'm not sure how best to check. The normal
procedure is to lift the relevant pin of an IC and note the difference
in current.
Make absolutely sure there is no solder splash or wire shorting something.

If IC3, 4 are at a similar temperature they are either both faulty or
like many ICs just at normal working temperature.

73 Alan G4ZFQ






Alan G4ZFQ
 

On 01/08/2022 15:12, Mike Easterbrook wrote:
I think the 100ma RX current value was from early QDX history and with the various upgrades most recent builds have been averaging out in the 140-150 ma range. I believe there was a note to this effect
Mike

Ah yes!
I too remember now. Hans did say somewhere that the different audio ICs in the v3 took more current.
If Dave's measurement is accurate his is taking a little more than normal. But not a lot.

73 Alan G4ZFQ


Dave, M0OOR
 

** Update **
In the interests of sharing my findings, I have now got it working.

What I have done:
- Reworked L12 - this improved things greatly ( I must have had a fault there). 
- Reflowed joints on trifilar transformer.
- I am now in the process of fine tuning the LPF toroids.
- So far I have added one turn to L3 for 40m and it now scans perfectly.

I have found that the radio is very sensitive to the toroid windings for the LPF's as they are in circuit on transmit and receive, I even get different readings with the board in and out of the case.

Anyway, I am getting there, I have over 4 watts out on 40m which I am happy with. I will continue tuning 20m toroids. The 80m LPF seems way out but as I don't use 80m I will leave that until last.

Thanks to everyone for their suggestions, this is a great forum for a great radio. Thanks Hans!
73
Dave.
--
David Thomas M0OOR


Dave, M0OOR
 

By the way, my current draw readings are from my bench power supply so may not be that accurate. After running the radio for a couple of hours and making several QSO's, nothing got too hot (apart from me that is)!
73
Dave.
--
David Thomas M0OOR


KEN G4APB
 

Dave,
you seem to be getting Exactly the same results as me. rx current 180mA, low audio out (btw, mic input is wrong input for wsjt-x - select qdx ip and op), poor LSB separation. When you have selected the correct audio input for wspr, can you let me know if you see multiple interference bands on the wspr window, another ‘feature’ of this set of problems.
I have rewound toroids, removed finals, measured voltages-all fine, to no effect believing I have a fault causing the high rx current and other problems. I have never seen a note saying the current should be any higher than 100mA. V3 QDX 1.04 firmware. The latest manual graph even suggests lower than that!
73 Ken g4apb 


Pim Niessen
 

Mmm, it appears I got something quite similar after finishing the build today.
QDX build for 12V, firmware 1.04
.
- Current consumption: 173 mA
- Low audio level in WSJTX, only ~ 10 dB above noisefloor, orange indicator
- Spikes in Audio spectrum every 100 Hz and in between
- It does decode only strong FT8 stations on 20 m
- RF sweeps, way out of wack
- Sideband suppression  < 30 dB, wanted sideband at ~ -30 dB
- ADC gets a bit warm, about body temperature.

So far I've found that there is something not correct around IC9. Pin 5 is at ground level causing it's output , pin 7, to go close to the supply rail. All the other 3 opamps are at around 0.7 -0.8 V at their outputs.

So next is investigating why IC9B-5 is at ground. First glance did not reveal a short.

Is there a list of normal voltage available for that part of the circuit?

73,
Pim, PA2PIM


Ronald Taylor
 

Hi Pim. IC5 is a quad AND gate. If one input is low the resulting output should be low (not Hi). On 5B, inputs are pins 4 and 5 and the output is pin 6 (not 7). Pin 7 is the ground lead for the chip. In receive mode, input pins 2, 5, 10 & 13 are all held low by the lack of a transmit signal from the microcontroller. In receive mode you should be seeing low outputs on pins 3, 6, 8 & 11. 

Good luck ... Ron

On Mon, Aug 1, 2022 at 9:31 AM <pa2pim@...> wrote:
Mmm, it appears I got something quite similar after finishing the build today.
QDX build for 12V, firmware 1.04
.
- Current consumption: 173 mA
- Low audio level in WSJTX, only ~ 10 dB above noisefloor, orange indicator
- Spikes in Audio spectrum every 100 Hz and in between
- It does decode only strong FT8 stations on 20 m
- RF sweeps, way out of wack
- Sideband suppression  < 30 dB, wanted sideband at ~ -30 dB
- ADC gets a bit warm, about body temperature.

So far I've found that there is something not correct around IC9. Pin 5 is at ground level causing it's output , pin 7, to go close to the supply rail. All the other 3 opamps are at around 0.7 -0.8 V at their outputs.

So next is investigating why IC9B-5 is at ground. First glance did not reveal a short.

Is there a list of normal voltage available for that part of the circuit?

73,
Pim, PA2PIM


Pim Niessen
 

Hi Ron,
I did not intend to mention IC5.
So far I've found that IC9 pin 5 is at ground level and it should not. It should be at about mid-rail since it's getting ~ 2,5 V from the VCOMR pin of the ADC. It's a high impedance circuit there so levels might vary a bit but overall it should be symmetrical with IC9A and  IC7A/B.

Currently looking for the largest magnifier I have.....

B.T.W. Tx appears to work. Only tested Terminal mode Tx test and that produced  > 3W on all bands. Fine for now.

Pim


Dave, M0OOR
 

Ken,
When I say mic input I actually do mean the QDX Transceiver input, sorry for the misleading terminology.
Yes, I do see multiple interference bands on the waterfall compared to a similar setup using my FT817. I assumed it was due to my power supply (it's definitely worse with a switch mode supply). So I don't know if more smoothing/RF suppression is needed on the supply rails.
I have found that the unit is extremely sensitive to the way the LPF toroids are wound, small compression or expansion of the windings having a big effect on the frequency received. I am at the moment using the RF Scan utility to fine tune my filters, but it's a bit tricky as you think you have it right only for it to change when the board is back in its case. I assume this is because one set of coils is very close to the case.
Incidentally, one thing I have found during my tinkering is that if I leave longer leads on the toroids so that they are spaced a little higher off the PCB, they do seem a little less touchy.
I hope this helps, if you have any better mileage, please let me know.

73

Dave.
--
David Thomas M0OOR


KEN G4APB
 

Thanks for that Dave and Pim,
it confirms to me there is more than build issues here. I see the same interference on well proven analogue psu or battery, regardless of PC used or USB lead etc etc. My FT897 and QCX-mini all have clean waterfalls so this is really bugging me. This new USB sound card interface is supposed to be far superior noise-wise to standard USB audio dongle types, but I have yet to see it in action.
I don’t have the coil sensitivity you mention. I did have to add 4 turns to the main rx inductor to get the rf sweep to reach 80m band, but that does not change in or out of the case. Did you add the recommended insulation tape next to the tri-wound toroid?

73 Ken g4apb 


Dave, M0OOR
 

On Mon, Aug 1, 2022 at 07:00 PM, KEN G4APB wrote:

Thanks for that Dave and Pim,
it confirms to me there is more than build issues here. I see the same interference on well proven analogue psu or battery, regardless of PC used or USB lead etc etc. My FT897 and QCX-mini all have clean waterfalls so this is really bugging me. This new USB sound card interface is supposed to be far superior noise-wise to standard USB audio dongle types, but I have yet to see it in action.
I don’t have the coil sensitivity you mention. I did have to add 4 turns to the main rx inductor to get the rf sweep to reach 80m band, but that does not change in or out of the case. Did you add the recommended insulation tape next to the tri-wound toroid?

73 Ken g4apb 


 Ken,
Yes, I added the insulation tape as per the instructions. I haven't tried on battery yet but if you are getting the same interference using a battery supply then it does rather point to the soundcard, perhaps a future firmware update will address this problem??
I haven't tuned my 80m coils yet but I did have to add one turn to L3 to bring 40m into band.
I don't get the interference with my FT817 using a homebrew interface, nor with my ADX that I built whilst waiting for my QDX to arrive and that uses a cheapo USB soundcard dongle.
Let's hope we get there in the end as this looks like a useful rig for using portable.

73's Dave.
--
David Thomas M0OOR


Alan G4ZFQ
 

Ken, Dave and Pim,

Hans has not said anything yet but it does look as if there is a fault in your QDXs.
Pim in particular has found a voltage wrong that should not need to be tested.
The spurious signals are not normal.
Dave's coil sensitivity is not normal. Something may be oscillating.
A firmware update, possible but only if he knows, it does not seem to affect many QDXs.
Try a direct approach if he does not see this thread.


73 Alan G4ZFQ


Pim Niessen
 

Alan, All,
I do not think firmware will fix my particular issue, but one never knows.:-) 

Earlier this evening I've lifted one side of IC9, i.e. pin's 5-8.
The junction of C74 - R37 is still at ground level and I also measure a voltage drop over R39.
I need to investigate and document measurements further but these SMD parts are tiny and I'm lacking the proper tools at the moment to see enough and to remove/replace parts.

What does not add up yet is the higher current consumption while receiving. A short in this part of the circuit with  >100K parts should only cost fractions of a mA.
So there might be more going on.
Ah well, I remember the Elephant consumption instructions.

73 PA2PIM


Milt
 

What voltage do you measure at the junction of R17 and R18?  It should be at half of the VCC.  If it is not, it could account for the lack of voltage at the junction of C74-R37.  Could be a problem with T2

Milt W8NUE

On Aug 1, 2022, at 3:17 PM, pa2pim@... wrote:

Alan, All,
I do not think firmware will fix my particular issue, but one never knows.:-) 

Earlier this evening I've lifted one side of IC9, i.e. pin's 5-8.
The junction of C74 - R37 is still at ground level and I also measure a voltage drop over R39.
I need to investigate and document measurements further but these SMD parts are tiny and I'm lacking the proper tools at the moment to see enough and to remove/replace parts.

What does not add up yet is the higher current consumption while receiving. A short in this part of the circuit with  >100K parts should only cost fractions of a mA.
So there might be more going on.
Ah well, I remember the Elephant consumption instructions.

73 PA2PIM