Spectal purity paranoia

Post Reply
KD7MW
Posts: 41
Joined: Sun Feb 06, 2022 4:18 am

Spectal purity paranoia

Post by KD7MW » Sun Dec 18, 2022 8:26 am

I recently got a TinySA spectrum analyzer. So of course, I've been playing with it and measuring my various radios harmonic output. I have several VHF and UHF radios, and all but one are compliant with the FCC spectral purity requirement that transmitted harmonics and spurs must be, at most, -40 dB below the level of the fundamental, for a radio of 25 watts or less.

That one exception is my TYT MD-9600 (Version 1 hardware). It's fine on 70 cm. But not on 2 meters. The 2nd, 3rd and 4th harmonics are -30, -29 and -37 db below the fundamental. When I patch in an old DCI 2-meter mechanical bandpass filter, the harmonics drop to better than -50 db below fundamental. But that isn't a good solution for a dual-band radio. The situation is disappointing. We've all heard about "dirty" Baofengs, but I'd never heard a peep about the MD-9600.

Those of you who have been in the innards of the MD-9600, do you have any suggestions? Is there anything I can tweak inside the radio to improve the harmonic suppression, or is it just the way things are? Were there any changes made in the later hardware versions 2-5 that might improve the situation?

G4EML
Posts: 919
Joined: Sat Nov 16, 2019 10:01 am

Re: Spectal purity paranoia

Post by G4EML » Sun Dec 18, 2022 10:20 pm

Unfortunately there is nothing that is adjustable in the MD9600 PA stage so there is nothing much you can do to improve things without hardware modifications.

I have today checked the output of my two V1 radios and my V4 radio.
I measured better results than you but they are still not great.
Are you sure you were not overloading your analyser input? That could partially explain your results.

At full power 144MHz (40W)

V1 Radio 1 288Mhz -50dB 432 MHZ -30dB 576MHz -45dB
V1 Radio 2 288Mhz -45dB 432 MHZ -32dB 576MHz -46dB
V4 Radio 288Mhz -61dB 432 MHZ -33dB 576MHz -42dB

The radio uses the same PA transistor for both bands with diode switched low pass filtering for VHF and UHF. The two low pass filters are in parallel with the unused one being isolated by diode switches.

A bit of experimenting showed that the 288MHz response is due to the VHF filter perhaps being a little too high in frequency. The V4 radio performs well here but it does have a different PA arrangement. I don't have any schematics for this version so I don't know exactly what was changed.

The poor performance at 432 and 576 appears to be because the UHF low pass filter remains in circuit when on VHF and is only isolated by a couple of diode switches in the off state. These probably don't have a very high isolation and are allowing the UHF frequencies to leak around the VHF filter.
I managed to reduce the level of these two harmonics to undetectable levels by temporarily shorting out the UHF filter path. It might be possible to add some extra switching diodes to do this but it would need some time to develop.

Colin G4EML

KD7MW
Posts: 41
Joined: Sun Feb 06, 2022 4:18 am

Re: Spectal purity paranoia

Post by KD7MW » Mon Dec 19, 2022 2:49 am

Thanks, Colin. I re-ran my measurements, this time at 1 watt and 5 watts output, directly into the TinySA through a 90 DB pushbutton variable attenuator. My attenuator's limit is 5w. The RF sampler I have for HF is not flat at VHF/UHF, so I can't use it to measure accurately at higher power. It seems pretty flat between 144 MHz and the 1st harmonic of 2m, but it attenuates the higher harmonics drastically.

Previously, I'd fed the TinySA with about -27 to -30 dBm. The TinySA Wiki says to keep input below -25 dBm "for best results." But this time I attenuated the input down to about -40 dBm. Here's what I got, using the U.S. standard simplex frequency of 146.52 MHz. Results rounded to integers.

1 watt, input -40 dBm to analyzer
Deltas: 293 MHz: -33 dB, 439: -41, 586: -45

5 watts, input -41 dBm to analyzer
Deltas: 293 MHz: -34dB, 439: -39, 586: -54

My previous results were at the 250 mw setting. So perhaps the lower power settings make something non-linear with respect to frequency. Anyway, my results are now better but still not good. In contrast, my old Kenwood TH-G71 and our beloved Radioddity G77 both show harmonics at -50 dB below fundamental or better. As does a Yaesu FT-7800. My 2m SSB/CW/FM radio, a Yaesu FT-290 RII is the best of all--the harmonics are all down in the noise, below 92 dBm. Not surprising as that rig has helical filters.

Anyway, now that I know what I know, I'm not comfortable using the "9600" on 2 meters without an external filter. But it's a pain to have to switch between filtered and straight through to change bands. Too bad. It was nice to have a radio that I could leave connected to an outside ground plane and use for local 2m FM, local UHF DMR repeaters, and my hotspot (at very low power).

Post Reply