开云体育

QMX+ Power Amplifier

Roelof Bakker All Messages By This Member #142767 开云体育 Hello all, Last week I finished the build of a rev1 QMX+. No problems were encountered and it worked fine on powering up. During the build, I have measured the frequency response of the low pass filters.These appeared to agree very well with a simulation carried out with help of RFsim99. One anomaly was encountered: the series resonance of L507, a RFC feeding DC to the centre of the 60M-80M low pass filter caused an additional loss of about 0.5 dB just in the 60M band. To get rid of it the Q of the RFC was lowered by mounting a 2K2 SMD resistor in parallel of L507. I have been reading that many builders experienced a drop in output power on the higher bands and my build was no exception. As I like tinkering a lot, this appeared to be a nice challenge to see if this 'issue' could be improved upon. My build is for a supply voltage of 9V. Reading up on this issue on this list, it appears that the output transformer L501 is critical. Muhsin, TA1MHS, wrote that he had replaced the BN43-202 binocular core with a BN-61-202 core, which improved performance on the higher bands. Apparently a lower inductance for L501 helps. I did not have stock of a BN-61-202 core, but I? had a BN-43-302 at hand and wound a WTST on it for 9V operation. The change in output power was negligible, though. The next thing I tried was using 2 x 3 BS170 instead of 2 x 2. Powered with 10V, gave these results: Band?PWR-W?I-mA 160?6.55?1150 80?5.35?928 60?4.99?925 40?5.73?975 30?5.68?1363 20?4.86?931 17?3.57?1258 15?5.73?1613 12?3.89?1553 10?5.05?1282 6?3.87?1080 This more or less solves this issue for me, though I don't think this is recommended in actual operating. I wonder what the reason is for the much higher current consumption on e.g. 15M compared to 80M, whilst delivering about the same power? A matter of lower impedance? Anyway, this is great fun. To be continued. Test gear: Sensor Monitor Power Meter 100W dummy-load Rigol DSA-815-TG spectrum analyzer Lab power supply 0-32V 2A 73, Roelof, pa0rdt -- Roelof Bakker, pa0rdt
mux_folder2001 All Messages By This Member #142768 开云体育 Hi Roelof, Interesting result with 2x3 BS170s. What is your 2x2 output? Something I have been thinking to try is stacking a second ACT08 on top of the existing one. My simulations seem to show that gate drive is very important. Is your build a 12 volt or 9 volt build? Tony AD0VC toggle quoted message Show quoted text From:[email protected] <[email protected]> on behalf of Roelof Bakker via groups.io <roelofndb@...> Sent:?Tuesday, April 8, 2025 8:13 AM To:[email protected] <[email protected]> Subject:?[QRPLabs] QMX+ Power Amplifier ? Hello all, Last week I finished the build of a rev1 QMX+. No problems were encountered and it worked fine on powering up. During the build, I have measured the frequency response of the low pass filters.These appeared to agree very well with a simulation carried out with help of RFsim99. One anomaly was encountered: the series resonance of L507, a RFC feeding DC to the centre of the 60M-80M low pass filter caused an additional loss of about 0.5 dB just in the 60M band. To get rid of it the Q of the RFC was lowered by mounting a 2K2 SMD resistor in parallel of L507. I have been reading that many builders experienced a drop in output power on the higher bands and my build was no exception. As I like tinkering a lot, this appeared to be a nice challenge to see if this 'issue' could be improved upon. My build is for a supply voltage of 9V. Reading up on this issue on this list, it appears that the output transformer L501 is critical. Muhsin, TA1MHS, wrote that he had replaced the BN43-202 binocular core with a BN-61-202 core, which improved performance on the higher bands. Apparently a lower inductance for L501 helps. I did not have stock of a BN-61-202 core, but I? had a BN-43-302 at hand and wound a WTST on it for 9V operation. The change in output power was negligible, though. The next thing I tried was using 2 x 3 BS170 instead of 2 x 2. Powered with 10V, gave these results: ... This more or less solves this issue for me, though I don't think this is recommended in actual operating. I wonder what the reason is for the much higher current consumption on e.g. 15M compared to 80M, whilst delivering about the same power? A matter of lower impedance? Anyway, this is great fun. To be continued. Test gear: Sensor Monitor Power Meter 100W dummy-load Rigol DSA-815-TG spectrum analyzer Lab power supply 0-32V 2A 73, Roelof, pa0rdt -- Roelof Bakker, pa0rdt
Roelof Bakker All Messages By This Member #142776 开云体育 Tony, I have added the results for 9V: 10V???9V? Band?PWR-W?I-mA?PWR-W?I-mA 160??6.55??1150? 5.19?1047 80???5.35?? 928??4.24? 851 60???4.99?? 925?3.92? 859 40???5.73?? 975?4.49? 898 30???5.68??1363?4.45?1256 20???4.86?? 931?3.89? 872 17???3.57??1258?2.95?1189 15???5.73??1613?4.86?1532 12???3.89??1553?3.45?1486 10???5.05??1282?3.97?1182 6????3.87??1080?3.61?1068 Regarding lack of sufficient drive, I don't think that is the case here. DL4JAL has designed a CW only 5W transceiver, which uses the same set up as the classic 'Sierra' using band modules to change bands: https://www.dl4jal.de/cw_qrp_trx_bm10/cw_trx_bm10.html The power amplifier is inspired by the QMX+, but he uses Mitsubishi medium power fets with bias power control instead of BS170's and claims 5W out on all bands. 73, Roelof, pa0rdt -- Roelof Bakker, pa0rdt
dl8lrz Reiner All Messages By This Member #142834 I also use 2x3 BS170s on my 9V QMX+. The power increases by almost 1W, and the heat on the BS170s decreases slightly. However, the larger capacity of the 3x BS170s affects the LPF tuning, especially on my 20-17-15m LPF and higher too. I had a lot to tuning. ? 73 Reiner
Roelof Bakker All Messages By This Member #142845 Hello Reiner, ? Thanks for your comments. This is an interesting issue to play with! ? 73, Roelof, pa0rdt ? ?

Previous Topic Next Topic