The TS-530/830 mod and upgrade thread

[KG4CGC]

WOW!

[N8YX]

Tip #6: Product Detector upgrades

Also present on AG6K's "Circuit Improvements" page. Several modifications are made in this area. One involves swapping the OEM 1N60 diodes used in the product detector for low noise-figure Schottkys. The LO side of the bridge is unterminated; a resistor to ground fixes this. Next, a few components in the IF bypass RC network are changed - opening up the received bandwidth a bit, and as Rich noted "...you'll get the full effect of sky noise". The modification makes for a more robust sound from the speaker - and if less bandwidth is desired, the Narrow filters can be activated...or an accompanying SP-230 with its built-in audio filters can be used...or both.

Component changes on the '530's IF Board are as follows. These are applicable to the TS-830 as well, though depending on production run, not all parts may need to be swapped. The '830's component designators are different and will be covered in another post.

- R41 is changed from 6.8K to 2.7K
- C35 is changed from .033uF to .01uF. This is a polystyrene unit
- R42 is changed from 47K to 3.9-4.7K (I used 4.3K)
- D16-D19 are changed from 1N60 to 1N6263 or similar Schottky parts. I got a bunch of these from K4EAA and matched them to .001v forward drop before installation

Prod-Det-1.jpg

A 100 ohm, 1/4w resistor is added from the junction of R39 and R40 to ground:

Prod-Det-2.jpg

(Insert obligatory Linda Hamilton "...terminated..." quote here.)

Follow the service manual procedure and adjust the product detector via TC2. I used a scope and got the null down to the point that the LO energy was unobservable at the test point, the drain of Q4. I've tried using 1N5711s in my FT-90x projects and found the weak-signal performance to be a slight - but not a great - improvement. The 1N6263s seem to work better, at least in this application.

FWIW, I checked sensitivity with a signal generator at 14 and 28MHz. I cannot turn it down to the point that the signal isn't detectable (.01uV or thereabouts), and the recovered spectrum is very low in noise.

[KA2PTE]

Tip #6: Product Detector upgrades

- R41 is changed from 6.8K to 2.7K
- C35 is changed from .033uF to .01uF. This is a polystyrene unit
- R42 is changed from 47K to 3.9-4.7K (I used 4.3K)
- D16-D19 are changed from 1N60 to 1N6263 or similar Schottky parts. I got a bunch of these from K4EAA and matched them to .001v forward drop before installation


A 100 ohm, 1/4w resistor is added from the junction of R39 and R40 to ground:

Ok decided to try this mod on my 830S. It seems to me the area on its schematic is near D20-D23, which are all 1N60 type diodes. I have taken a photo showing them on the pcb.

HPIM0810.JPG

R42 in the 530 seems to be R78 and R41 is R77, but its already been changed to 2.7K in the 830S.
C63 in the 830 appears to be the same component as C35 but its been lowered to 0.022.

Not sure if I ought to change them to match, and from what I can tell R75 and R76 are the equivalent to R39 and R40
and when you follow the connect trail, there's no termination resistor I can see.

UPDATE: I terminated the product detector via R75 and R76 to ground with the 100 ohm resistor, put everything back together for a listen.
It does sound improved on rx somewhat and I am told audios good so I guess it was a success.

This is a photo of the 100 ohm resistor terminating on the foil side:

HPIM0813.JPG

[N8YX]

Tip #7: Balanced Modulator update

Another one borrowed from AG6K. The modification cleans up the transmitted signal a bit at the balanced modulator. In the 530, diodes D27-D30 (1N60) are changed to the same type as used in the previous modification. Follow the service manual procedure and null the carrier after the IF Board is reinstalled in the radio. The mod is applicable to the '830 too, albeit with different component identifiers.

I found it easier to do all IF Board mods at once, and in addition I replaced the large filter capacitors at the rear of the board with 450v, 105c-spec units while I had things apart.

Bal-Mod-Diodes.jpg

[N8YX]

Tip #8: Fix TS-530 "S meter slam"

According to users, assorted 530 series rigs will slam their S meter's needle to the right on power-up. The meter in the rig on my bench was hitting its pin hard enough to be audible. Engaging the RF Attenuator, turning the AGC to Fast (or Off) and disconnecting the antenna can either reduce or eliminate the condition - if one remembers to do so.

User group comments reflect that the condition is 'normal' - although neither of my 830s do it. This prompted me to study the '830's AGC and S meter circuitry, and to attempt to identify a root cause.

A couple things become evident: The TS-530 series really IS a cost-reduced version of its big brother, as many key circuits are implemented in the same way but are lacking in a few areas. In the case of the '530's IF Board, a meter snubber circuit present in the '830 is missing. Then there's the matter of the AGC timing capacitor being installed backwards relative to that of the '830.

The fix involves pulling the IF Board, removing and re-installing C84, then adding a few components to the foil side. Easiest to do while you have the board out for other modifications:

AGC-Fix-1.jpg

This is the "after" pic, with the capacitor installed as per the '830 ('+' terminal to the RF Gain control).

Next, add a 33uF 16-25WVDC capacitor and a series 1.5k 1/4w resistor from the 'M(eter)' line (pin 2 of connector 10) to ground. Circled is C84 - foil side - for reference.

AGC-Fix-2.jpg

The orange tantalum capacitor next to the yellow circle is an addition. It's across R106 (10k) and connects to the (new) negative side of C84. Again, present in the '830. I used a 0.39uF 35WVDC part - though I may swap it for a 0.1-0.22uF some day (the latter value being used in the '830).

The S meter still pins on power-up, but with much less force - and somewhat less duration. AGC and ALC functions were checked and are not affected by the mod...only the action of the meter itself, and only from a cold start.

Further work may be done, at which time I'll edit the post as needed.

[N8YX]

Adding a band to your 530/830 AUX position

As shipped, both series of rigs had provisions for adding an auxiliary band...according to Kenwood, for receive only. This band spot can also be made to transmit, and does not involve butchering the radio in any way.

What it does involve is a decent amount of resourcefulness, patience and a careful touch with the soldering iron. A well stocked junk box and test bench helps too - as does a well stocked beer cooler, for use with your friends whose junk boxes have all the required parts.

This isn't a so-called 'screwdriver' mod. Some component fabrication is required, as is removal of the RF Board and tinkering around the final amplifier assembly.

WARNING: VOLTAGES INSIDE WILL KILL YOUR ASS DEAD IF NOT CAREFUL. Bear in mind this ain't your grandfather's transistorized PA set.

I've chosen to add 27.0-27.5, as later on we'll look into how to make the '+0.5' switch work with it too. This will net us complete coverage from 27-30MHz, and will be useful in driving transverters whose IFs are at 27 or 28MHz. I'd refrain from adding 60M to the rigs - their VFOs run at 5.5-6MHz and there will possibly be an issue with reciprocal mixing products. Counter and PLL programming theory will be covered and is useful in adding other band ranges.

WARNING: USING THE COMPLETED RIG IN A SERVICE WHICH REQUIRES CERTIFICATED EQUIPMENT CAN PUT YOUR LICENSE IN JEOPARDY. EMERGENCY, RX OR TRANSVERTER USE ONLY, PLEASE.

There are several steps which are required to add a new band, and each will get its own section:

-Program the Counter Unit
-Remove the RF Unit and gain access to the PA compartment
-Wind coils
-Install coils (and on the '530's RF Board, add Transverter connection components)
-Configure the PA band switch
-Update PA compartment components as required
-Replace removed components and assemblies
-Align RF Unit and test

[N8YX]

Program the Counter Unit

The x30 series rigs (including the TS-130) departed from previous Kenwoods in that they use a programmable PLL/VCO arrangement to derive the HFO - this instead of a design which uses a crystal controlled oscillator for each band segment; either directly or as a PLL reference. Both versions have their advantages, but the biggest one here is that adding a band range to the PLL merely requires a few switching diodes. The analog portion of the radio requires more work - which will be covered later in the thread.

It's worth mentioning one of the "hacks" which a person may run into when servicing a '530 or an '830. Kenwoods of this genre used a 4-bit BCD value to program the PLL for a given band range. Along the way, someone figured out that if you cut one of the logic lines going into the PLL Board, you can change the programming.

Yes, this buys you all of the 11M range (including the freeband areas above and below the Class D allocation). What it costs you is the use of 10M. Not the correct way to gain spectrum, especially if transverter use is the goal.

Shown here in an '830 is the requisite lead of J2 (PLL Board) after being cut to enable the mod; it was subsequently repaired by a former owner. Note that the frequency display DOES NOT TRACK with the actual frequency range if the mod is performed, as the counter readout is programmed independently of the PLL. Worth noting if you're restoring a rig which won't transceive where it's supposed to:

PLL-Wire.jpg

Counter Unit theory and programming

There are two main functional areas which the Counter Unit is responsible for: Provide a readout of the actual operating frequency, by means of both preloaded band logic (for the three Most Significant Digits) and a circuit which counts the VFO output frequency and supplies it as a value for the three Least Significant Digits. Multiplexing of this information is handled by a single IC, and the band preload values for the readout are accomplished by diodes D105-D110 (absent) on the Counter Unit.

Similarly, the PLL Range Select programming is handled by diodes D101-D104 (also absent).

Note that 'absent' = logic HIGH (1). The circuit uses complementary programming, so we must install a diode to make a given line LOW (0).

One may study existing programming information in the Counter Unit and PLL sections of the '530/'830 Service Manuals to get an idea of how Kenwood implemented the various band ranges. For purposes of this thread, I've detailed a band with a start range of 27.0MHz.

You must first remove the Counter Unit from the radio (a number of wiring harnesses require unplugging) then take the shield off the assembly to gain access to the foil side of the board. Use a solder sucker to clean out the holes for D102, 106, 108 and 109 then insert and solder 1N4148 or similar silicon switching diodes, following the silkscreen mask for anode/cathode orientation:

CounterBoard.jpg

Use a pair of end cutting nippers to remove excess lead length then reassemble the Counter Unit and replace in the radio. Turn the set on and rotate the band switch to its Aux position, turn the Cal(ibrator) On and rotate the VFO to one of the marker frequencies. If everything was done properly, you'll hear the marker tone in the receiver (albeit weakly) and the displayed frequency will indicate the correct (new) band.

[N8YX]

Remove the RF Unit

Arguably the most tedious part of the process. Given the fact I've removed the assembly a gazillion times while researching material for this thread, I can complete the task in about 10 minutes - assuming both top and bottom covers are already off the rig. Follow the process below and it's pretty easy:

- Be sure the rig is disconnected from mains power
- Unplug all harnesses from the Counter Unit and remove it from the rig
- Remove the Counter Unit bracket from the rig (2 screws)
- Remove the PA Compartment cover
- Use a shorting stick to short the top of the PA choke to a nearby ground
- Label the front-most PA tube then remove both tubes and set in a safe place
- Remove the PA Load knob and shaft (2 2mm set screws at front of flexible coupler)
- Remove the PA Tune shaft (loosen 4 screws on front coupler, front one on rear coupler, slide front coupler so as to permit shaft extraction)
- Loosen the Drive shaft coupler screws (4) then push the coupler as far forward as it will go
- Disconnect all header plugs and TMP plug (1) from the RF Unit. If unsure about where they go, take close-up photographs before removing them
- Remove band switch shaft tension springs (3) from the shaft. One is located between band switch wafers and the other two are attached to bulkheads
- Rotate the band switch to 1.5MHz. Locate the PA Tank bandswitch coupler (adjacent to front tube socket). Loosen both 2mm hex set screws
- Rotate the band switch to 10MHz. Loosen both 2mm set screws, then pull the bandswitch shaft all the way out. Avoid misaligning the wafers as you remove the shaft
- Remove 7 round-head sheet metal screws from the RF Unit
- Use an ~800 degrees F iron to remove the white wire that's soldered to the rear-most variable capacitor frame (530/830)
- De-solder the NFB wire from the PA Compartment feed-through insulator and remove (830 only)
- Stand the rig on its left side, locate the DRV wire at the rear right of the RF Unit (foil side) and de-solder (530/830)
- Lay the rig back down and lift the RF Unit out of the chassis. Remove the Driver tube and shield (optional). Set both rig and RF Unit aside.

[N8YX]

Wind Ant/Mix and Drive coils - Part 1

There are several types of tuning coils used on the 530/830 RF Unit. Two are Toko shielded models, while the Driver coils are open-form, slug-tuned RF types for which I cannot find a modern-day equivalent - or source of surplus parts. Thus, we must dive into the junk box. Fortunately, the forms are the same type used on the TS-820's RF Unit...and there's no shortage of those boards in the Kenwood hybrid parts market.

I obtained the coil forms needed for the work in this thread - along with a lot of other useful parts - from a scrapped TS-930S Signal Board. Lodestone also has the forms (L30 series):

http://www.lodestonepacific.com/shieldedcoilform.php

Taking a look at the Ant/Mix coil area of the x30-series RF Unit, we see the following:

RF-Unit-Coils-1.jpg

The inductors for the new band will go in the areas circled yellow. The orange circled area is L12, one of the 10M inductors. I de-soldered the component for purposes of measuring its value.

A note about de-soldering: Don't attempt high volume component removal via heat and solder wick with any Kenwood hybrid's circuit board, else you'll end up lifting the foils from the board. They're made of phenolic material - not glass substrate, like Drake's and Cubic's gear of the day - and as such don't tolerate a lot of heat. Nor does the adhesive used thereon. At a minimum, buy a Hakko rework station which incorporates a temperature-controlled vacuum de-soldering handpiece. (Yes, I'm a Pace fanboy ... but Metcal, JBC, Etneo and the higher-end Weller rework stations are all good choices.)

Here's the 10M 'Ant(enna)' coil compared to one which fits into the Aux-band Ant/Mix spots. The difference is apparent:

Coil-Types.jpg

We cannot reuse the OEM 10M coil for purposes of this exercise, but we can re-use its value. Said value can be measured with an LCR meter, or a small capacitor can be tack-soldered across the coil and the tank then grid-dipped. We then approximate the required value and use the coil calculation info at

http://www.carnut.info/star-parts/pcbs/toko-winding.htm

although for this experiment we can use a value of 6 turns...derived empirically. Whatever gets us there. Bear in mind that for other band ranges, you'll have to calculate the required inductance - or measure the OEM coils used for a nearby ham band then determine the required inductance based on existing, known preselector-circuit tank values as a starting point.

[N8YX]

Wind Ant/Mix and Drive coils - Part 2


Next, the Ant/Mix coils are wound. If you're using a new Toko or Lodestone form, the shield and ferri-cap will be separate from the coil tower and base pending assembly; you'll need to source some #40 magnet wire. If you're recycling a coil, it must be disassembled if the inductance is too high. The easiest way to do this is to use a small jeweler's screwdriver to gently lever out the indentations which lock the can to the base, by means of wedging the blade between the two. The tuning slug is then screwed flush with the top of the form and a small nut driver is used to push the form out of the shield. It takes patience and a bit of care.

Once everything is apart, unwind the "finish" side of the coil by de-soldering the winding from the base pin, then unwrapping it from the form. Leave the "start" side attached to the other pin:

Ant-Mix-Coils-1.jpg

Wind a total of 6 turns starting in the second groove (3t) then skipping to the third (3t) then back down the center of the form to the pin which originally held the "finish" winding. Loop one turn around the pin then solder the wire to the pin. It'll take a few seconds for the insulation to burn away. A decent amount of flux helps, as does an iron temperature of ~700 degrees F:

Ant-Mix-Coils-2.jpg

Take note of the 'center-tapped' comment. If the start winding point is on the side of the coil with no center pin, no big deal. Just remove the center pin with pliers and push it into the center hole in the other side of the form. It's only used as a locator in this application.

Use an ohmmeter across the coil to verify continuity between pins. Next, install the ferri-cap and shield can then check for shorts between coil pins and can. If everything checks out, set the completed coil aside and wind another just as you did the first.