开云体育

The USB, LSB, CWL & CWH modes were only detecting AM. I found that there was no LO signal on the 455 kHz SSB Mixers. The CAR-DDS should put 453.5 kHz to 456.5 kHz depending on mode.

All other functions on the FT-817ND (bought early 2006) are working. CPU was reset. I replaced the AD9835 DDS IC. Nothing at o/p which is available direct at capacitor that feed the 455 kHz transformer feeding the two mixers. On my set the 3x Texas parts on schematic are not used, but there are PCB adaptors for 2x JRC made parts for SSB TX and SSB TX and the 68 MHz to 455 kHz mixer is an SA602, All SMD parts on DIL adaptors.

The CAR-DDS uses the same 5V as the PLL-DDS (7.2 to 8 MHz) and the same 22.x MHz master reference. I statically checked with ohm / Diode test DVM every DDS lead to places on PCB, the VS-analogue feed resistor and VS-digital feed inductor. The clock seemed to be shared with the Octo DAC/Volume control IC and serial data with the PLL. You can tune signal gen at 14 MHz across band and see S level go up and silence for about 3 kHz, or keep SG constant and tune main dial in 10 Hz steps for about 3 KHz. If a second SG feeds 200mV Pk to Pk to SSB RX Mixer LO in pin (453.5 or 456.5 depending on USB or LSB), then you can get a changing tone as you tune 14 MHz (radio or SG) in filter pass band.

Definitely no LO on SSB at 455 KHz from DDS (AD9835) and it's been replaced with new part direct from ADI. Everything works on all bands and modes except the LO. but I've not tried SSB TX.

Double checked ribbon cable.

What I have I not thought of?

Plan B, if I can't get CAR-DDS working is Clapp Oscillator using a 455 kHz resonator. I have that working on a breadboard with two trimmers (with parallel caps all in series with resonator and a switch at roughly 453.5 KHz and 456.5. I'd sacrifice CW mode so LSB & USB are USB and CWL & CWH are LSB, replacing the mechanical test switch with a VFET driven by the shift register that generates SSB & CW mode signals from CPU serial data (one of two 4093). Actually, 453.5 and 456.5 kHz crystals used to be a stock part, but the resonator based Clapp osc will at least get SSB working.

I've 30 years experience replacing/soldering SMD ICs and check connections after, so I'm confident about the AD9835 replacement. Actually I was sceptic that it was faulty, but I can't find anything else faulty.

Well after two weeks or more, there is no CAR-DDS. Even after changing the AD9835 ic.

I've implemented "plan b". It's a slim LO board that sits against the front of the alloy chassis between the ribbon and the earthing spring at the corner. A Clapp oscillator using 2N3904 as emitter follower (collector to supply). There is a cheap 455 kHz resonator, maybe from a remote control handset. For 456.3 kHz there is a 33pF with a parallel trimmer in series with the resonator connected to base. The capacitor divider Base-Emitter-Ground is 2x 680pF. There is a 68 pF and parallel trimmer also in series with the resonator and it is switched to ground by a second 2N3904. It has 220K bias from 6V supply to collector. A VFET was found to have too much off capacitance and is like a varicap! The transistor has 180k base to emitter/ground and 22k in series to logic control. The underneath of the main PCB has CW and SSB logic levels decoded by a 4093 feeding two diodes. The SSB logic signal switches the transistor and thus oscillator from 456.5 (LSB) to 453.5 kHz. All the various offsets on the FT817 seem to change main PLL-DDS, not CAR-DDS (unless maybe if the narrower option filter is fitted?). So USB - LSB - CW - CWR: USB: USB mode and display frequency is correct. LSB: Still USB mode and display is off by 3 kHz. CW: LSB mode and display is out by 3kHz and CW pitch. CWR: LSB mode and display is reading low, by CW pitch setting (minimum 300 Hz, max 1000 Hz).

I couldn't find any decoded USB/LSB logic signal.

Radio now working brilliantly on SSB receive (AM, FM, FM N and WFM were all working as the CAR-DDS is only used for modes using SSB mixers). I'll check TX next.

If anyone knows how to extract a USB vs LSB logic level on the FT817ND, then I can make it work more sensibly on LSB.

The IF spectrum is inverted at the 68 MHz IF, but is obviously the "normal" way round on the 455 kHz, as 453.5 kHZ is USB and 456.5 kHz is LSB. Those are both in reality just outside the pass band of the filter, but that is normal.

Silly me. Of course CWR or CW doesn't allow voice modulation, so using SSB mode vs CW mode for switching USB/LSB only works for receive. TX working on SSB in USB in USB or LSB mode. RX is USB in USB or LSB and LSB in CW or CWR. I may have to put a manual switch for USB/LSB. Obviously on LSB currently (either CW or CWR) the display is out (reading low) by value of CW tone setting (300 to 1000 Hz).

On Tue, May 27, 2025 at 02:37 PM, paul wrote:

maybe use the filter switch lines as voltages for USB/LSB?

I considered that. I can't imagine buying the filters at current prices. I had a CW filter on my FT101ZD MkIII I passed on to a friend recently.

I hadn't thought of the shortcut. I usually have A=IPO and B=Att, or else A=Power set.

I'm thinking I can add 1 more transistor (or likely a small VFET) and make the transistor switch used to add capacitance for USB into 1/2 of a SR Latch. Then make the mic Up & Down set USB and LSB. The VFET would drive an LED at the speaker grill holes (as it's offset) to indicate LSB mode. The power on default can be USB (likely but can be ensured by small capacitor on gate.

I've also tried all forms of the "Reset", but nothing from the Car-DDS. I have a decent Rigol SA, 20 MHz Hameg and 100 MHz (2 Gs/s) digital scope. Using a scope probe on the Rigol SA at 10:1 as it's 50 / 75 Ohms.

A more exotic solution would be monitor user selection of Mode via CAT and then setting the oscillator shift. It's certainly stable enough for SSB. I don't do FT8 anyway :D

On Tue, May 27, 2025 at 05:19 PM, paul wrote:

Did several mods to my 817 but the size of components makes it....ehm...less relaxing :)

Well, I first encountered SMD on Thick Film modules in 1970s! Then in mid 1980s "SMD introduction research". I just use a regular iron and bit, but a mag-lamp. However my old stock of SM caps and resistors is about twice the size of the ones common in the FT817, though bought 2 years later.

The Mic buttons was over thinking it. I have a miniature toggle switch near power jack for USB/LSB. It only puts +6V on the switch transistor that adds extra capacitance (so no RF on it). Most LSB & USB stations are perfectly tuned on the 1kHz steps. Thinking about it, I could have an "auto mode" that uses the FT817 band voltage as LSB and USB is mostly per band. However I had to turn a knob on the FT101ZD for USB or LSB!