Kenwood TS440S Too high voltage in VCO 5 (7.8v versus 5v)

Ghetto said:

Thanks Brandon. Just starting doing voltage probes vs schematic voltages on the 'pink' oscillator chain you described as well as IC13. They all look pretty spot on other than Q21 emitter is 4v where it should be 3.4v. Q21 tested OK

Click to expand...

do you have 36mhz at Q21?

Do you have a radio that can receive it (or its 3rd harmonic at 108MHz) just to see if it is there? Or a different HF rig that can receive 9.000MHz?

edit
A Baofeng UV-5R can pick up 108MHz. Since it is just a CW signal, all it does quiet the FM receiver, but at least its a way to check for the signal.

Although unnecessary, a DIY H-field probe (its just a loop of wire with optional shielding commonly made of coax cable) attached to the Baofeng makes it a non-contact probe that only detects nearby signals. This makes it much more easy to pinpoint signals and it also prevents other signals from opening the squelch on primary channels causing it to exit the broadcast band mode. I just tried it on mine and it worked a lot better than using the antenna.

If you have a HT and a junk coax for it, you have everything you need.

You may need to turn your RF gain down. I recommend a half inch loop made of thinner more flexible coax if you have it, but if all you have is 1/4" stuff, it will work, just keep the loop small.

Honestly, for just seeing if the signal is there, you don't need a H-field loop at all, just a sniffer coax (E-field probe) with very little center sticking out. A cut off coax should do.

Ghetto said:

Hi Brandon, Sorry to disappear a while ago, I've been really busy on everything but radios - I run a beef cattle farm by myself and sometimes get caught up with big workloads.

Fast forward to today, I got a chance to try and sniff the 9mhz harmonic on Q21 (for 36mhz oscillator) as we were talking about. I dont hear anything but I thought I would post a video of the test in case I'm doing something wrong. I think I deleted a previous post that you read about VCO 2 being 7.8v on its test point (TP5) as well as extra info.

- I have since tested Q21 out of circuit - OK
- All electrolytics replaced on PLL Board.
- Checked every solder joint/traces from IC13 back to Q21 etc. All look good/passing low resistance tests etc.
- Checked D8 (OK) out of circuit for fun in VCO 2

That frequency shift on power on is across all frequencies so I understand its not a single VCO problem and its a mixer for everything as your saying.

Anyway heres the video:

Click to expand...

I think I was wrong saying you could hear a subharmonic like that. I recently had the opportunity to try that again. I must have had a radio that had the right mixer trash in it to lead me to that conclusion. I apologize for the bad procedure.

There is still looking for the higher harmonic, which I know works good. An analog FM radio that might go a tad past 108MHz would do it. I need to go back and refresh my brain on where we were.

It might see it, but it wont be of normal amplitude if it does. My 200MHz scope can see 460MHz, barely, but the amplitude is in the weeds. At the rated bandwidth, signal amplitude shown on the scope trace is 3dB down from true, Gets worse the higher you go.

We have 36MHz providing the 9MHz reference to the phase detectors, and since VCO 5 seems stable, the reference signal should be fine. In my mind, this limits the drifting problem seen in VCO 3 and 4 to those VCO's.

Only because VCO 5 is stable do I want to assume the 9MHz is coming out of the divider IC13. If that wasn't stable, we would he using that HF radio to sniff for 9MHz 3rd harmonic of 27MHz around those chips. But, I am assuming it is there since VCO 5 is stable. Moving on.

VCO 4's 91MHz is divided down and used as a last LO in the SSB detector stage and VCO 3's 99-104MHz is used for the RIT shift. Two totally independent things (apart from that common 9MHz reference). Check both of these to verify that they are indeed drifting. Start with the 91MHz signal. Should be 91.3 in USB and CW, and 90.7 in LSB.

We are approaching oscilloscope territory.

Because it sounded like the sweep brought the IF through the passband of a filter, I want to say that these VCO's are walking right past where they need to stop and that suggests a PLL issue. Maybe the chip itself, maybe the data lines going to it, maybe the data itself from the micro. These PLL's use digital inputs that use 4 lines to clock in 7 groups in sequence.

Lets start by verifying that VCO 4 is indeed sweeping. Set your FM receiver to 91.3, the ts-440 on USB, and sniff VCO 4 while you do a power cycle and listen as it passes through to see if the static quiets and stays quiet, if it starts with static, then quiets, then goes back to static, or no quieting at all.

You're welcome! Wish I knew what was wrong.

Now that we know the 91MHz PLL is walking around and that the reference is good, it must be between the PLL IC and the VCO buffer.

First thing I would do is ensure the VCO's buffer is working and that its output is making it to the PLL IC. I would do this by using the handheld radio to sniff for the 91MHz at pin 15 of the PLL (be fast, catch it when it quiets or track the shift on the handheld). The cap at the output of the buffer (Q4) is large for 91MHz, so the signal should be just as strong at pin 15 as it is at the buffer.

PLL chips are usually diagnosed with an 'oscope, however...

If you had a sensitive audio probe (like a Carlson Superprobe (see youtube) with an H-field probe attachment), you can listen for the divided down PLL signals since they are in the audible range. I just came up with that idea so I tried it on my President Washington before posting this. Theory confirmed, I hear 10kHz!!!). It don't work with the E-field probe that Mr. Carlson designed the probe for, but toss a loop on it and it works for this***.

Both the 9MHz reference and the 91MHz are both divided down to 5kHz. Sever the connection from the VCO buffer (pull C53 C15 for example) and listen for the 9MHz/1800=5kHz. And since your VCO is slowly walking past, you should be able to reinstall that cap and hear the other 5-kHz signal heterodyne with the divided reference as it passes. Remember, the loop needs to be right on the PLL chip and loop orientation matters, so turn it until you hear the 5kHz.

Next question, does the phase detector output track frequency changes? Its moving, we know that, but in response to anything? Lets give it something to respond to. Probe the test point with the volt meter and, while ignoring the slow drift you see, take something metallic and touch that varactor. Remember how probing the wrong side of that inductor before caused it to rail? Did the test point rail again or get squirrely?

That leaves one last aspect of the PLL to check, the digital inputs. I am going to write them off as being good since VCO5 is getting its data and they share the same data lines. What is different is the clock signal. Each PLL gets its own clock signal from the same chip that produces the data (IC2 on the control board). This means all you gotta do is verify there is an AC signal on pin 9. Easy enough (just catch it when it loads data, probably at power up).

If you have a solder sucker and experience removing IC chips, swapping the PLLs for VCO 4 and 5 around could give us the same answer much faster (since we know 5 is good).

***I will admit, I believe my washingtons PLL is TTL, yours is CMOS, so I am not sure there is enough current flow in yours for an H-field probe to detect it. But with those gates, maybe the E-field is worth a try???