Wiltron 6409 RF Scaler Network Analyser

[dazz1]

I read all your development!
Use lm337 with 2 resistor in place of 7918.
Check for heatpipe, is perfect device to spread heat on large area.
Flat tube are easy to glue on aluminum heatsink and boost the heat conduction.

Switching regulator today are also very good, drop-in for 78xx and some time for 79xx regulator.
You same a lot of heat and energy.

Hi
It is nice to know that at least one person reads my posts.     and thanks for the advice. 

I can get a LM337 and I could graft one into replace the 7918 but I have a queue of other projects and no particular hurry to fix the 6409.  I will just move on to the next in line until I build up a large enough parts order to get free shipping.  Then I will return to the 6409.

Heat pipes will definitely work but I will try my tack-on heat sinks first.  This equipment has been operating since 1992 with not so good thermal management so I could have just left it factory original.

I am reluctant to install switching regulators because of the risk of inserting noise. I will stick with the analog inefficient regulators.

 I have switched to a project to make a low High Voltage probe for my multi-meters.    The multi meters only have a 10Mohm input impedance which loads down any voltage on valve equipment I measure, giving a false reading.
I am making a probe, rated up to 1000VDC pk, with a 1:10 gain and input impedance 1Gohm.    I have been buying parts from Aliexpress and making parts in my workshop.

[ve2ulu]

For digital part, 5v, most power go there, switching psu with good low esr capacitor, is good way to save heat without trouble.
I know the switching regulator are noisy but most time are not really a problem.
I update my dvm Solartron 7061 like that and now, the 7061 run cooler and use less watt.

[dazz1]

For digital part, 5v, most power go there, switching psu with good low esr capacitor, is good way to save heat without trouble.
I know the switching regulator are noisy but most time are not really a problem.
I update my dvm Solartron 7061 like that and now, the 7061 run cooler and use less watt.

Unfortunately the 5V regulator runs cold.  No benefit in adding a custom heat sink or switching regulator.

The BBQ hot regulators are +18V, -18V and +12V regulators, all supplying key analogue circuits. 

The 6409 is not an instrument I will use often, or for very long, so switching regulators offer no significant benefits.    I am sure there are lots of cases where switching regs make perfect sense, just not in this case.

The root cause of the hot regulators is the input voltages supplied from a toroidal transformer.  Not practical to modify that.

[dazz1]

Hi
The 6409 includes battery backed nvRAM.  I rechargeable ni-cad is the battery.  When power is applied, the ni-cad is charged through a resistor.  When power is off, the battery discharged through the resistor to keep power on the RAM.  There is more to it than that, but this is enough to identify the problem that the 6409 spends most of it's life off.

I removed the ni-cad and found some minor damage to traces.  The battery is about 35 years old so leakage is not a surprise.

Rather than fit a new ni-cad battery, I decided to fit a coin cell. I replaced the resistor with a diode to prevent the coin cell being "charged" when the power is on.

The nvRAM doesn't hold important information.  I know this because there is a jumper on the PCB to enable/disable battery backed RAM.  I think it only stores the current front panel settings.  When the 6409 is powered up next, it remembers the settings.

Photos show:
The encased ni-cad battery
some corrosion of the track and pad
The coin cell fitted.

[dazz1]

Hi
The N-type connectors on the Wiltron SWR Autotester were badly corroded and needed replacing.  In addition, I couldn't find exact copies of the connectors.  I had to machine away some PTFE at the back of the connectors.  As a result there was a significant risk of the pins being axially located out of spec.  I needed a N-type calibration kit to measure the pin (M) and socket (F) locations.  Not something I could find at an supplier in my country, so I have made one.

Is is probably not obvious how it works, so I plan to make a YT video to show its use.

Images show:
The components that make up the N-type connector gauge kit.
A cross section of one gauge.
This gauge kit measures "G" and determines if "G" is within spec.  This is what the gauge kit does.  It measures "G".
The setup in the lathe, including the use of the digital test indicator to accurately machine the depth of bored holes.
The 2nd gauge works only because the thickness is accurately measured and known.

[dazz1]

Hi
I have made a video on the N-type pin gauge in use.
You can see the video at this link:  https://youtu.be/A-BSFsCUibk

[dazz1]

Hi
So after waiting months to build up a long enough parts order to avoid paying shipping, I finally put and order in and got some parts.

I received a 7918 -18V voltage regulator. I cannot buy a 7918 device off the shelf in New Zealand.  I had to put an order in to one of the big suppliers.
I installed the 7918 without issue. 

I also received a DIN connector to replace one damaged on the front panel. 

Replacing the DIN connector required something approaching open heart surgery.  It became evident that the 6409 was assembled from the front panel, working backwards.  Removing the front panel to get to the DIN connector was not easy. 

[dazz1]

The end result is a 6409 SNA back in operation again.
The smooth curve on the screen indicates the YIG oscillator is working.  Very important because this SNA is built around the YIG oscillator.

I have also replaced the leaking  NiCad rechargeable battery with a Lithium coin cell. To do that I replaced the charging current limiting resistor with a diode.

When I start up equipment for the first time, I use my power meter.  The power switch on top allows me to switch the instrument on/off without going anywhere near the instrument.   If things go wrong, I can just sweep my hand over the switch to isolate power.

Next step is to calibrate the 6409 to gain assurance it is telling me the right answers.

[dazz1]

One of the things I found when I operated on the 6409 was under the shiny knob. 
This shadow of dark gray would appear to be sun related.  My hypothesis is that the 6409 sat on the same bench for years.  The shadow of the knob stopped that area fading. 

Sun exposure has also caused the protective film to deteriorate and flake off. 

I need to find a replacement 6409 front panel (not likely at all) or have a new front panel artwork created and printed (expensive).

[edpalmer42]

The front panel isn't graphically complex.  You should be able to draw up a new version with any drawing program.  Talk to a local print shop about what oddball sizes and graphical formats they can print.  Then have them print your file on the back of a transparent sheet.  This protects the ink or toner from contact.

If necessary, you could split the print in two pieces.  The vertical white line in the center of the panel is an obvious place to do the split.

[dazz1]

The front panel isn't graphically complex.  You should be able to draw up a new version with any drawing program.  Talk to a local print shop about what oddball sizes and graphical formats they can print.  Then have them print your file on the back of a transparent sheet.  This protects the ink or toner from contact.

If necessary, you could split the print in two pieces.  The vertical white line in the center of the panel is an obvious place to do the split.

Yes, you are right.  The graphics isn't complex, but it has to be accurately sized and positioned.   That is the tedious part (I have made a start).  I have talked to a local print shop and they can accept vector graphics which I can produce with GIMP. 

The front panel is grey tinted acrylic.  It does a great job of improving contrast of the CRT at the expense of brightness. 

If I just get a printed film, then positioning it over the panel to exactly match the existing cut outs will be a challenge.
If I start with a new sheet of forest green tinted  acrylic, I will get improved contrast and brightness.   I can apply the decal, then get the panel laser cut or water jet cut.  That will require some experimentation to make sure the process works.  No option will be cheap.

Before I go down that road, I want to ensure I can achieve something close to factory new performance to justify any further investment in time and $$$

[dazz1]

Hi
The Wiltron 6409 Service Manual has a whole section on performance verification.  Another on calibration (tuning) of the circuits.
I have read through these sections and determined I don't have enough test equipment to complete the process.

I have fixed all faults detected and the 6409 appears to be working.  I am going to close it up and call it repaired and operational, but performance not verified.
I will come back to verification/calibration at a later date when I have the capability.