HX-50

 


Home Up HX-50 HQ-110A HQ-129-X

 

 

 

This is a Hammarlund HX-50 transmitter just into the shop for recapping and alignment.  Capacitors, some 2 watt and power resistors are on order from Mouser.  Initial cold checking found quite a few high wattage (over 1/4 w) out of tolerance.  

 

IM000105.JPG (233400 bytes) IM000106.JPG (237068 bytes) IM000107.JPG (275345 bytes) IM000108.JPG (283557 bytes)
Manual at BAMA

160m mod

You will need an RF probe on the VTVM for the Alignment.  Build one. IM000085.JPG (59993 bytes)  IM000084.JPG (18671 bytes)  

Description, Schematic and construction details.

I used a 1n191 diode, 0 .001 uf 630v cap and a 4.7meg resistor.  The diode did just fine in place of the 1n34a when checked over the full range of a HP-606a generator.

 

Big Caps replaced for 750v supply

capold new.JPG (221448 bytes)New caps are smaller. cap cutting.JPG (158978 bytes)Protect all sockets and adjacent items with tape. cap cutoff.JPG (178293 bytes)Cut off tops of caps just above the wider part of the cap base leaving the insulator. cap guts.JPG (218049 bytes) Pull out old cap guts so cans can be glued over new caps for appearances.
capleads.JPG (220306 bytes)Extend cap leads.  Drill holes in can cap insulators nest to solder tabs,  to pass new wires through.   cap soldering.JPG (231805 bytes)Solder extended cap wires on to existing terminal posts. cap fitted.JPG (198958 bytes)A dab of hot melt to hold the caps from falling out when chassis is flipped over.  The original can covers may be glued back on for appearances sake.  I may order some clear heat shrink tubing to help strengthen the can covers.

 

Issues in the HV RF cage

L115 R181.JPG (236428 bytes)  This is the parallel combination of L115 || R181. IM000026.JPG (196372 bytes) This is L115 placement in the RF cage IM000029.JPG (262584 bytes) A second angle.

The HV RF cage was full of cigarette tar.  When unsoldering components the smell of cigarette smoke wafted up from the soldering gun.  All the components have been cleaned with contact cleaner and denatured alcohol.  The chassis and RF containment box was also cleaned.  

These are pictures of the parallel combination of L115 and R181.  It appears that someone changed the turns arround the 47 ohm resistor and shortened it.  Additionally it appears that two windings were shorted with solder.

This must be resolved before over 700 volts of EMF is restored to this unit. Schematic here: http://bama.edebris.com/manuals/hammarlu/hx50/

Power Supply Choke

Well am back on the case this day in Feb 07.  It has been a year letting this rig "breath".  While going back and checking all voltages I discovered the negative bias voltages to be quite lacking on several tubes.  So checking back to the negative bias supply I found there to be insufficient negative voltage on the half wave rectified bias supply. 

Checking further back I found 190 vac across the choke (L112).  Well that would detract from the negative voltage power supply.  The choke measures 97 ohms cold.  No luck on Google for the L112 choke specs.  

 

L112 had 190 volts of AC across it.  The DC resistance is 97 ohms.  Seems a little high to me. 

You can find L112 on the high voltage secondary center tap, to  ground. 

For a full PDF schematic click this hyperlink. 

 

Update on Choke L112  "swinging chokes"

You can't beat the new groups!  Great guys willing to help.  Just ignore the Flame wars and Spam.  I received several private emails and a bunch of replies in the news groups.  Below is a helpful sample.

rec.radio.amateur.boatanchors,       rec.radio.amateur.homebrew

Hi,

I checked the L112 choke in my HX-50, and it is 75 Ohms d.c. resistance. That is in the ball park of your 97 Ohms measurement. That same choke also serves as the high voltage choke, since it is in the common center tap return for the medium voltage and H.V. taps as well as the bias tap.

In standby, I measure 195 VAC across the choke. So, it sounds like your choke is innocent of the low bias
readings.

Have you put a scope across the bias supply capacitors C176 and C177? They may need changing out if there is excessive ripple. The 680 Ohm resistor in series with the bias supply may have increased value.
Did you check the value of R142 and bias adjust pot R143? They may be dragging down the bias voltage.

You might measure the bias voltage with the 6DQ5 final tube removed. If the tube is gassy, it would tend to put a + voltage on its control grid, which may be working backward
in your bias supply, making the supply less negative.

73,
Ed K.

..........Not exactly.

In a choke-input filter, the inductance must be above a certain critical value or the filter isn't really a choke-input filter.
This critical value (called critical inductance) is directly related to the total load resistance. The lower the load resistance, the less inductance is needed. The load resistance is just the output voltage divided by the total current delivered by the rectifiers.  

In a transmitter, the load resistance and current vary all over the place with keying, modulation, loading, etc.


The inductance of real-world iron core filter chokes depends to a certain extent on the direct current through the choke, which magnetically saturates the iron. One way to reduce this effect is to include an air gap in the iron core. The wider the gap, the less the inductance variation. But such a gap reduces the overall inductance.

If a choke with constant inductance were used, it would
need to have enough inductance for the lowest-current/highest load resistance condition, yet enough current capability for the highest current condition. That means a wide air gap. Such a choke would be large and expensive.

Instead, a choke with a narrow air gap is used, Its inductance varies with the current - more inductance with less current, less inductance with more current. This is exactly what is needed with a choke input filter. Such chokes are called
"swinging chokes".

The power supply for my 150 watt homebrew rig uses a swinging choke in
the high voltage supply. This choke has 25 henries inductance at 30 mA
DC but only 5 henries inductance at 300 mA DC.......

73 es GL de Jim, N2EY

I turns out this bias circuit worked after replacing R 168 and R172.  i had originally replaced these resistors.  So perhaps I had a cold solder joint.  Well come on now ....that happens to the best of us once in a blue moon.

 

PSC

IT'S ALIVE!   Feb 23

After replacing R172 and pulling out an added (mod) series resistor with R143 rheostat (Idle tube bias) the Screen bias on the final came up to -76 volts.  Not all negative bias voltages are as documented.  BUT putting the rig on 20 meters and tuning up the output produced 20 watts of clean CW.  I have been reversing numerous modifications to restore this unit to original.  This was just one more mod to remove. 

The 80m band was dead but the scope showed adequate input to the final 6DQ5 sweep tube.  All bias voltages looked good.  The problem was a shorted C207 100pf 1000v cap switched in parallel to the Plate Tuning variable air capacitor.   I temporarily replaced that cap with one of 500v rating.  I will order a new 1000 volt cap soon.  Once replace the 80 meter band showed 20watts output CW. 

The 10m "B" and the 40 meter bands have dead crystals in the oscillator section.  All other bands have output from 10 to 15 watts.  The unit needs a good re-alignment at this point and at least two new crystals.

I think there may be a few other bad caps in the output network.  I have to rotate the plate control full clockwise to peak the output into a 50 ohm 100 watt dummy load.  But for now I am quite happy to see watts output. 

Hint  .

IM000115.JPG (182287 bytes)

Hang labels to identify the transformers during alignment.  It really helps. I wasted a lot of time flipping through the manual to the Coil ID diagram.  

 

VFO Dial Scale adjustment 

Adjusting the VFO so it lines up every 100khz can be tricky.  Knifing of the variable cap was needed.  Knifing is where the outer two rotor veins are bent in or out to smooth out the response of the oscillator.  All the adjustments interact with the maximum capacitance position of the cap.  But with patients and care it is doable.

IM000114.JPG (93330 bytes) Build a pick up for the MFJ or oscilloscope with freq counter combination. IM000110.JPG (174334 bytes) Set the high side of the scale by adjusting the oscillator coil  IM000113.JPG (111268 bytes) Se the low end with the trimmer cap.  IM000109.JPG (178898 bytes) 100khZ steps checked after knifing the variable cap. 
IM000107.JPG (195856 bytes) Looks good. IM000111.JPG (157271 bytes) Tight baby! IM000112.JPG (114813 bytes) Specs allow for 1/2 needle width.   cat small.JPG (297557 bytes)  Neuterizing...No, Neutralizing is next.  Hang in there baby. 

 

Neutralizing the Sweep tube

Wow.  What a task.  I am not done yet.  There seems to be a stability problem.  I have the scope monitoring the output for a clean sine wave.  That helps a lot.  If not for the scope I would have walked away thinking Job Well Done to only hear the wrath of other hams about interference.  I am chasing the parasitic suppressor found in the plate circuit.  

 

IM000029.JPG (262584 bytes)  Remember this original parasitic suppressor from above? The resistance has risen above 75 ohms during burn-in & alignment.  That is the problem with most resistors in vintage gear.  It is just best sometimes to replace the power resistors that get hot.
IM000116.JPG (51174 bytes)  Winding a new choke for the parasitic suppressor.  
With the help of 1967 ARRL Handbook and measuring the original inductor I wound this new coil using a gauge bigger wire and a Bic pen as a form.  I did not have the matching gauge wire on hand.  This matches the original inductance.  

Finding a non inductive 47 ohm 1w resistor for the suppressor is next.  The original 10% resistor is a bit toasty and measures above 75 ohms.  I will order a carbon comp for the final installation. A metal oxide resistor are alleged to have little inductance at this frequency.  Carbon Film resistors with the spiral trace of carbon wrapped around a ceramic core will show inductance at higher frequencies.    That in parallel with the inductor drops the overall inductance.  That is where the harmonic instability is occurring.  

Finding new 1 watt carbon composition resistors was a challenge.  I found plenty at ham fests.  However, they are typically way out of tolerance.   I just refuse to use NOS resistors and capacitors in restorations.  You are just installing  "Time Bombs" waiting for a quick failure.   If you demand originality there are ways to create restoration looking capacitors and resistors.  See the bottom of this hyper linked page on reproduction resistors.

I came across an OHMITE offering of carbon ceramic  at Mouser.  Advertised to be a non inductive replacement for carbon comps.  The carbon ceramic resistors came in and were installed in the suppressor.  I reused the original coil that was wrapped around the 47 ohm resistor.  It works fine.  

I tried the unit out and made a 75m contact to VA during their March VA QSO contest.  What a thrill to log that one in!  Well 10m c band is not working.  So back out to the shop.  

 

Second mixer and RF driver stage alignment page 30. 

An important discovery.  When performing alignment 7-8-1 on page 33 (you can find the manual at BAMA) Second mixer and Driver stage, one must complete the adjustments of both plate slugs (The 2nd mixer transformer and the drive transformer) at one particular frequency then change to the higher frequency and adjust the output slugs. The plate slugs are the top slugs nearest the chassis.  The output slugs are closest to the bottom or the under side of the chassis. 

You must peak (bounce back and fourth) from the 2nd mixer plate slug and the driver plate slug with out touching the frequency as though you are adjusting the top and bottom slug in an AA5 IF transformer to find resonance.  I made this mistake at first. 

Once the low frequency of that band is peaked change the generator frequency and peak both output slugs.  alternating from 2nd mixer to driver stages until output is peaked.  Then return the the lower frequency, take a sip of coffee and spend about 5 to 10 minutes repeating the low and high frequency adjustments until you coffee needs refilling and no more improvement can be seen.

I had to back out all the slugs and start from scratch.  Move the plate or top slugs up and move the output sluge to the bottom.  The first peak of the slugs is far from the final position.  

Example:

10-A band with the 10-C band

 

page 30

Adjust both T111 and T118 plate slugs at 28.2Mhz for peak output.  You must bounce back and fourth from T111 and T118 to obtain best max output at 28.2.  There seems to be interaction between the stages especially at the high frequency bands.

Change to 29.5mhz (change the band switch) and peak the output slugs. Again bouncing from T111 and T118 until no more improvement can be had.  

Now go back to 28.2 (change the band switch) and peak the plate slugs again. You have affected the plate slug peak when you moved the output slugs near the plate winding.  Refill you coffee cup.

AGAIN go to 29.5 and repeat the process at both frequencies until no more improvement can be had.  Refill coffee cup.

You may now proceed to the other bands repeating the above procedure.  Make more coffee......

The amp seems a lot more stable.  I can easily tune up into a dummy load and see decent wattages.  A tweak of the neutralizing cap C188 was needed to keep the final from running away in oscillations.  Use the Neutralization procedure in the book.  

Here is a link where some great Hammarlund cabinet restorations have been accomplished. 

Hit Counter

 

 Contact me including your thoughts and comments. 

Please see the LINKS page.  I have collected many resources over the years.

135,912 unique web site visitors (14,499,000 hits) from October 2004 through August 2011.

Copyright 2004 - 2012. All rights reserved.