Date
1 - 20 of 28
#QDX V3 filters and low output power #qdx
|
Esa Nieminen
Hello
I am using 9V power supply. All output powers are bit low compared to graph showing Output vs Supply voltage. I have been able to make some test QSOs on 40m band (my antenna is half wave dipole). Rx is OK on all bands, I am wondering how the filter outputs from internal RF test should look out. In my QDX on 20m band, peak -10 is from 14MHz to 15.5MHz. Output is very very low. On 30m band, peak -10 is from 10MHz to 11.5 MHz. Output is low and 2nd harmonic is -29dBm only. on 40m band, peak -5 is from 8Mhz to 8.3 MHz, on 7MHz it is -10. Output is about 4W. On 60m band cannot be measured but there is the highest output on spectrum analyzer and dummy load. On 80m band, peak is -15 about 5MHz. Output is about 4W. By the way, I used MAC to get them. The QDX terminal program was very strange to use. When trying to get out from the selection I had to use cmd+q to get out then restart QDX and terminal. Mac told every time: Disk not correctly removed?? Otherwise it works fine with wsjt-x. Any advice? 73 de oh2awg, Esa N |
|
|
|
I've been using 8.5 to 9 volts and seeing 4.5 to 5 watts on my QDX3 all bands. Yesterday, I saw the power falling off to 2 or 3 watts. Cause was current limiting on the power supply. Power back to normal now.
Beforehand, I noticed the power reading peaked at 5W then very quickly dropped off about a 1/4 watt. 73, Rich nj1a |
|
|
|
Esa Nieminen
thanks But not in my case, My current limit value is 1,3A. Currents on different bands: 80m 580mA, 60m 630mA, 40m 550mA, 30m 450mA, 20m, 330mA. 73 de oh2awg, Esa NOn 29.6.2022 20.43, Richard Lee wrote:
I've been using 8.5 to 9 volts and seeing 4.5 to 5 watts on my QDX3 all bands. Yesterday, I saw the power falling off to 2 or 3 watts. Cause was current limiting on the power supply. Power back to normal now. |
|
|
|
Esa Nieminen
Thak you very much Mike I have very different from mine. I can send them later. But I have good reference now. 73 de oh2awg, Esa N On 29.6.2022 21.20, Mike GM1RHV wrote:
Hello, |
|
|
|
Jim G7FRI
Hi Mike, In terms of power output, I would expect some variation and drop off at the higher frequencies. Using the QRP Labs dummy load I measured 23v on 80m down to 20v on 20m which equates approximately to a range of 6W-4W. I’m happy with that given all the variables involved… |
|
|
|
Alan G4ZFQ
Your RF sweeps look very different to the ones I posted for my 12v v3 QDX… https://groups.io/g/QRPLabs/message/87810 <https://groups.io/g/QRPLabs/message/87810>Jim, I read most posts, I see large variations on different sweeps posted from various QDX rigs. I'm beginning to think the actual figures are meaningless although the relative values might make sense. None of the traces is necessarily relevant to transmit power. They essentially check the RX BPFs although the test signal does go through the LPFs. Esa says RX is good on all bands so we can assume the LPFs are at least passing signals on RX. 4 watts is acceptable, low power on 20/30m would indicate a problem with the LPF or it's TX switching. I suggest careful examination of the 30/20m LPF soldering and components. Also check the LPF switching voltages conform with the description in the manual. 73 Alan G4ZFQ |
|
|
|
Roelof Bakker
Hello Mike,
The cut-off frequency of the 20M low pass filter might be too low, which attenuates the 20M output somewhat. This does not effect 30M. 73, Roelof, pa0rdt |
|
|
|
Mike GM1RHV
Hello Roelof,
Many thanks for the feedback. In fact I think that the 40m LPF cut off is starting to fall off before the 40m band starts. Slightly less induction or a tiny amount of additional capacitance might sort it. For 20m, that first dip, like the indentation in a tooth, could be responsible for reduced output. I seem to think that the high frequency fall off should start around 17MHz? Mike |
|
|
|
Alan G4ZFQ
On 29/06/2022 20:41, Mike GM1RHV wrote:
In fact I think that the 40m LPF cut off is starting to fall off before the 40m band starts. Slightly less induction or a tiny amount of additional capacitance might sort it.Mike The traces are an indication of the RX filter. Ideally it will show an inverted U trace of the RX bandpass filter. This would hide most of any effect of the LPF unless it was really a long way from design. That "tooth" can not be significant. If it really was the LPF then the response would drop further, not rise again. If the LPF is responsible for "only" 3.5 Watts on 20m that would mean it was reducing the output by ~2dB which would hardly show on the trace. Esa's thread has been well and truly hijacked:-) Have you checked the 30/20m LPF? There were some posts about using a terminal with a MAC. As I recall a different key combination is required. Maybe here? https://groups.io/g/QRPLabs/topic/91778412#87011 73 Alan G4ZFQ |
|
|
|
Esa Nieminen
Hello again
I finally got my test pictures. If I understand correctly: I have to increase either capacitance or inductance on 80m and 40m bands, but how much? My problem bands 30m and 20m the filters look fine so what is going on? Here is also pictures taken on HP spectrum analyzer. As you can see the low output values compared with lower bands. The shown dBm values are not accurate because I have only non calibrated attenuator between QDX and spectrum analyzer. 73 de oh2awg, Esa N |
|
|
|
Evan Hand
Hi Esa,
The L12 adjustments and the 4 sweeps are for receive only. Based on the spectrum pictures, you have an issue with the 20/30 meter low pass filter (LPF). I would start by reflowing the solder connections on the coils for those bands (L4 and L10). If that does not improve the output, then verify that there are no shorts or broken traces around that LPF. Next is to confirm that only the 20/30 LPF is selected. D3 and D6 Banded sides would be at zero volts when the 20meter or 30meter band is selected. The others should be at whatever voltage you are supplying to the radio. The last step would be to verify that the correct capacitors are in the correct locations. I would not mess with the L12 windings unless you have reception issues. 73 Evan AC9TU |
|
|
|
Esa Nieminen
Thank you Evan I will try again to solder again all connections and do a better check. 74 de oh2awg. Esa NOn 30.6.2022 20.49, Evan Hand wrote:
Hi Esa, |
|
|
|
Jim G7FRI
Is there really a problem here on 20m/30m LPF? The manual states that for 30m the 2nd harmonic is only about 12dB down (see attachment). regards, |
|
|
|
Evan Hand
Jim,
The filter has a -12db effect at 20.2MHz. The Push-Pull design would lower the content at that frequency than the fundamental. You should still see a value that meets your country's purity requirements or not operate on that band. Esa posted pictures of the spectral output of the QDX, not a sweep of the filter. also, look at the main signal strength. 20/30 are much lower in magnitude than 40 or 80. I will still stand by my suggestion that there is a filter issue on 20/30meters LPF. |
|
|
|
Esa Nieminen
Thank you Evan
I did a better check. And there was a solder blob which was not easy to see. So I got 30/20m filter in order. Then I started testing with my test load and analyzer. I found that power levels on 80m and 40m had gone down about 2dBm and the power level on 30m and 20m are equal. But then I noticed that RX only current had also risen from 140mA to 240mA which was very strange. I started to suspect a burned PA FET. I took them off and one of them was just a resistor. I checked all with my component tester and replaced the faulty one. I soldered them back but the RX only current is still 240mA and the FETS seem to get warm without TX request. What is going on? May be some of the FETs is leaking? By the way, the component numbering around IC3 has changed on REV3 board, C39,41,43,40. I have supply of SMD capacitors so I would like to know which one is which so I could try to improve 80m and 40m RX sensitivity by adding capacitors on top. 74 de oh2awg. Esa N |
|
|
|
Evan Hand
Esa,
Others found that the AO3407 (Q7) reverse voltage protection had failed when the BS170 failed. The MOSFET failure creates a short across the input connector. That is one possibility. For the Rev 1 board, when one or more of the BS170s failed, it took out IC5. That shows up as a load on the 5volt regulator IC2, and the regulator limits the current and drops the voltage. So I would measure the voltage out of IC2 and IC1 while you are at it. The one complaint I have with the QDX is how easy it seems to have cascading failures. The failures start around the BS170s failing in the shorting mode. The BS170s in this implementation are sensitive to SWR mismatch. That seems to be the starting of the cascade. Do you still get the Red LED blinking and then turned on when you power it up? 73 Evan AC9TU |
|
|
|
Alan G4ZFQ
On 02/07/2022 18:46, Evan Hand wrote:
For the Rev 1 board, when one or more of the BS170s failed, it took out IC5.And it has happened with the QCX, just needs a BS170 to fail "the wrong way":-) 73 Alan G4ZFQ |
|
|
|
Esa Nieminen
Thanks Evan I will check those items tomorrow. Yes, the red led is blinking and receiver is OK. 73 de oh2awg, ESa N On 2.7.2022 21.46, Evan Hand wrote:
Esa, |
|
|
|
Esa Nieminen
Hello Evan
I removed all FETs and current went down to 140mA. As I thought one of the FETs had turned to be a resistor. I tested all of them before installing them back. What killed them ( total 3 pc) is a mystery because I have used only dummy load?? May be it was "at the edge of burning" though tester did not find it bad. I am using 9V as supply voltage from my laboratory power supply. The other thing: To make 80m and 40m RX filter give better frequency response I added 130pF to 80m and 10pF to 40m band. Here are the results. 73 de oh2awg, Esa N |
|
|
|
Esa Nieminen
Hello
My bad luck seems to continue. I soldered the FETs back and started transmitter testing and the result was: The input current when TX is ON only 450mA compared to previous about 600mA which is very strange. May be some of the FETs is still dead? I am running out of my stock of BS170 for my uSDX project. Which is a partially succeeded project even I burned the small fet on 8 band filter board. I replaced it to 3 BS170 and made one band filters for 80m and 40m. I get about 4W out of them. RX has been OK all the time. 73 de oh2awg, Esa N |
|
|