Kenwood TS-530S Hybrid arcing under driver tube

[KA2PTE]

Pulled the RF board for some component replacement, and decided to clean the bandswitch wafers while it was apart. I have a TS-830S
in working order nearby and copied by eye the wafer positions with respect to the dot on the plastic wheels in the center. However when powered on
theres arcing under the RF driver tubes socket. When I move the band switch to AUX (then re apply power) , its not arcing anymore.

I heard one of the wafers has the B+ for the tube on it, so I am thinking maybe one of the 2 in the back were not aligned? Its hard to see inside between the boards
to confirm. I had marked on the front panel where the arrow for the knob was pointing when I removed the shaft so that the final cage wafers would re-align right.

I also looked under the board to make sure no hardware was floating under the socket, looks ok.....

[N8YX]

I have an RF Board for the x30-series hybrids in Ye Olde Junque Box. Let me have a look later today and see what's what with regards to B+.

[KA2PTE]

Roger that.

I turned off the lights and the arc is from the tube socket foil right to chassis , which is almost a qtr inch gap. Looks like the plate only has 300Vdc on it, so am bewildered what can jump that
distance...unless its RF ? The send switch is off, heaters are off.

I had re-soldered those foil pads for the tube socket and while I was in there replaced the (2) 100k half watt resistors nearby the tube, plus R38. R38 (470k) was ok but it was carbon and I hear they are noisier than metal film types.

[KA2PTE]

So took the board out again for an inspection. No signs or marcs of arcing to the metal chassis frame so maybe its just arcing on the socket pins. There was some flux
left on the pins from a flux pen I used so I cleaned that off good, and cleaned and defluxed the board some more but after reinstalling, still have the arcing. Someone else suggested
to check the switch foil pads for cracks, as they crack easily. I guess will have to take the board out once again.

[KA2PTE]

Just put the board under the magnifier again, and I saw some black stuff
between Pin7 (the plate) to the common foil.
I scraped at it and now theres a tiny groove you can see in the pix if you
zoom in.

As I rubbed on the nearby ground foil it started to expose the copper
foil...there is a layer of some stuff coating lots of the boards in this
radio , probably niccotine as it was likely owned by an ex smoker. So maybe
its become conductive near this area of the tube and the voltage potential
ignited it, and ate up some of the board.

They did put that pin pretty close to the ground plane in the design.

[N8YX]

Nicotine is definitely conductive in the presence of HV.

It's a runaway process...if an arc starts it carbonizes the board material, which lowers the electrical resistance of the area, which allows more current to flow...which heats up additional material...

What I would do is to thoroughly scrub that area with 91% isopropyl alcohol after de-soldering and removing the tube socket. Then carefully inspect the board. ANY carbonized areas need to be removed and repaired. Use a Dremel tool with a very small ball end mill. My board repair equipment includes a Pace PPS-400 and a Microchine handpiece with various bits - I use it to clean out burned or damaged sections of a board. I'll then grind up a bit of the original board material from junk boards kept on hand (phenolic, in the case of Kenwood's hybrids), mix with non-conductive epoxy until a thick slurry is obtained then spread this into the milled-out area. Smooth it down as flat as possible, let cure 24h then repeat for the other side of the board if necessary. Then gently sand the repair down flush with the rest of the board and re-install any removed components.

If any circuit traces need repaired or replaced, I have ways to do that too. Let us know what you find.

ETA:


They did put that pin pretty close to the ground plane in the design.
Look at those two pins on the schematic. What are they? Is there any reason why the ground plane should be there - such as to shield it from the other nearby pin?

[KA2PTE]

Yea thats what I thought too, the niccotine as a "fuse" more or less.

Anyhow all back together and working.

I put a dab of wood glue over the pad to further insulate that plate pin.

After re-installing, no more arc.

Aligning the bandswitch was not too bad. I have it running now and RX for
sure is greatly improved. I replaced L41 and the nearby FET and D2 as per KD7DNY's site...he supplied the parts.
Many signals over S9 where as before, it was rare to see a S9 on the meter.

The CAL signal level is for sure lots higher as well and the drive (rf front
end) tune is no longer flat on 40m.

[KA2PTE]

Was running this rigg again today and noticed that 10m was not working. Turns out the drive level when set to the [29] band setting
just moves the alc meter slightly when you try to peak it with the drive control. 28 and 24 are dead, but I can peak 21 and other bands with
good alc. I had noticed after the bandswitch rework, the alc was more 'snappier' on the 80 and 160m bands, and 40m showed some improvement, but its not as robust as the others. Wondering if I put the switches back together right....cant remember if these bands worked before I began the task.

[N8YX]

Look at the timing of all the decks. Trace out the infamous L41, which is part of the 160M circuit and then verify the PCB connections to the various 160M front-end coils. Look to see where the wiper of each wafer is in relationship to one another - it would be almost impossible to get the deck off by 180 degrees and have the radio work at all.

One of those decks controls the logic which changes the radio's bands via counter and PLL units. Do the others line up with it when the band switch is set to a given band?

If yes to all the above, you need to align the RX front end and TX drive preselector tuning per the service manual. Start with 10M and work your way through each lower band successively. It's important to do them in order.

[KA2PTE]

Did some more checks, and after tapping on one section of the bandswitch, I saw the ALC level changing during tune. So I shot it some more with deox, and the level is improved but still low. Also changed the driver tube but no real difference. One site called hybridrestore had an article that says on the 530, 10m needs more carrier level, and that helped but there's still not a really high level of reading on the ALC scale, perhaps half way between the markers.

I also put the cal function on, and checked the s meter levels, and I get about a 7 on all the bands, except 80 and 160 which are about a 5 even when tuning the drive control. I think I was previously told this is not right, and to perform the alignment in the SM as you said. The band switch seems aligned ok, so I guess I will try the SM alignment. What page is it that I ought to start with in the manual?


Look at the timing of all the decks. Trace out the infamous L41, which is part of the 160M circuit and then verify the PCB connections to the various 160M front-end coils. Look to see where the wiper of each wafer is in relationship to one another - it would be almost impossible to get the deck off by 180 degrees and have the radio work at all.

One of those decks controls the logic which changes the radio's bands via counter and PLL units. Do the others line up with it when the band switch is set to a given band?

If yes to all the above, you need to align the RX front end and TX drive preselector tuning per the service manual. Start with 10M and work your way through each lower band successively. It's important to do them in order.