Heathkit DX-40 Restoration and Mods
  



DX 40 MODIFICATION AND DESIGN NOTES AND OPERATIONAL HINTS

The photo above is the station I often use in Novice Rig Roundup. The receiver is a Lafayette HE-10, often mistaken for a Hallicrafters S-38. The DX-40 and DX-35 were very popular Novice rigs back in the day. There were lots of them out there, but many have been scrapped out, due to rotary switch and transformer failure. For sure, lack of a fuse a primary cause of transformer failure. If the filter capacitors fail or the 5U4 arcs, there is nothing to protect the transformer. The inclusion of a fuse would have saved many of them from an early demise.

If you have the bad luck to acquire one of these with a bad transformer, it is not the end of the world. You CAN use the Heathkit HP-23 power supply to operate this rig from an external power supply.

The original target market was new Novice operators who needed a 75 watt input rig for a year, and then might keep it as a transitional rig with some phone capability when they got their General license. Who would have thought that this rig would still be popular more than 50 years later? If you are lucky enough to find a DX-40 intact and can repair the symptoms of age, it makes a nice piece of nostalgia in your shack, and has hours of fun to offer.

I do not recommend that you operate a VFO from the transmitter power supply; Heathkit did not provide enough reserve capacity for that extra load.

In the event of filter capacitor failure, the current can go up and burn out the transformer. The bleeder resistors can go open, causing the voltage across the capacitor to soar. This capacitor then shorts and blows the transformer. If you replace filter capacitors, be sure to check the bleeder resistors. A shorted capacitor can also blow a bleeder resistor.

The DX-40 and DX-35 exciter stages also are quirky, in that they are operated in series for DC purposes. If just one of those tubes goes bad in a way to increase current, it also causes the other tube to fail in a cascade fashion. Because they are connected to the HV supply, they can cause transformer failure. The 5U4 rectifier tube is operated outside its recommended ratings. Heath even states this in the manual. If the 5U4 fails and arcs or shorts, that could damage the transformer. If a 6146 gets gassy and draws too much current, it can take out the transformer.

The meter slide switch is problematic. These are cheaply made, and get dirty. An application of Caig De Ox It will be needed periodically to make the meter work reliably. A mini toggle switch can be installed to upgrade it. If you do this, I recommend a center off toggle switch. In the center position, this saves wear on the meter from bouncing during CW.

One set function switch contacts switch the center tap of the high voltage winding to ground, to enable the B+ supply. This has an RC network across it to try to minimize the glitches when this is switched. The repeated sparking at these contacts also eventually causes switch failure. There is no easy way to work around this contact failure without a new rotary switch, which is unavailable.

There is some speculation that the transformer insulation also can fail at the center tap. A short article in Electric Radio showed the failure analysis to be true. Now you can attribute the failure to poor transformer manufacturing procedures. There is an RC snubber network to catch spikes caused when the switch opens and closes. Dozens of contemporary tranamitters and receivers switched the center tap to ground, and they did not fail like the DX-35 and DX-40. I speculate that the transformer failure is due to the rotary switch making bad contact, and arcing, causing transients that the snubber network cannot control. You can believe what you want about that. I am an optimist. I like to think if you can treat the root cause, the switch, you can successfully extend the life of these rigs.

Probably the most common source of failure is the rotary function switch. This part is no longer available anywhere. If only they had used toggle switches, this would not end the life of these popular rigs. Many of the rigs of that time used these rotary switches: Eico 720, Johnson Ranger, DX-60, Knight T-60 and T-150. The contacts cannot handle the AC surge current when the rig is powered up. I sometimes fix these rigs by wiring around the contacts so that the rig is always on. Then it can be plugged into a switched outlet strip.

If the final cannot be dipped on 80 meters only (but all other bands are OK), the fixed padding capacitor in the 80 meter position may be bad. This is likely due to operation into a short or open or a bad antenna.

If the final cannot be dipped on 80 meters or 40 meters, it is often due to inadequate loading capacitance. A 1000 pF or even a 2000 pF capacitor may be required across the output coaxial connector on this band only. A simple modification would be to add a center off toggle switch next to the antenna tuning knob on the front panel, to switch these additional fixed loading capacitors.

The DX 40 seems to like about 2 mA drive for maximum output. Nominal grid current according to the tube manuals is 2.5 to 3 mA. NEVER EXCEED 4 MA! This is above the rated current of the final and will cause 6146 failure. Also observe Heathkit instructions for Plate Tuning and Current for various conditions. Operating the tube off resonance will cause rapid failure of the final tube. A dirty or defective meter switch can cause misleading readings, which can cause inadvertent misadjustment of the tuning, and cause tube failure. There is no protective negative supply or clamp tube in the DX-40; it depends on rated grid current for grid leak bias.

The DX-40 uses a pi network interstage coupling circuit that was popular in that time. There is no provision for neutralizing the final amplifier. As a result, on 10 meters certain 6146s will self oscillate. Selection of the final tube may help this problem, but it may not work. 15 meters usually works OK. I tried to install neutralizion as the DX-60 has, but was unable to get it to work right, without removing turns from the 10 meter coil to get the grid tuning range correct. I did not do this mod in the end.

 

For more discussion on this topic, see my web page topics:

Be sure to inspect/replace the resistor in the parasitic choke on the plate of the 6146. This MUST be a NON INDUCTIVE resistor like the original carbon resistor; if it is +/- 20%, leave it alone. If you must replace it, use an Ohmite OX or OY series non inductive resistor, available from Mouser electronics. Failure of this resistor can result in instability on bands above 20 meters.

If you use a VFO, you will not likely have the following problems. Additional instability can occur with certain crystals (dropping out of oscillation) on 40 meters and above. The Heathkit manual suggests removing turns from the coil in the plate lead of the first 6CL6, to tune it to 20 meters. It is possible that selection of the 6CL6s may correct this problem. Be sure that both the 6CL6s are matched on a tube tester; they are in series for DC. If they are not matched, it is possible to get a cascade failure as described above.

If you do nothing else in addition to your restoration, please add a fuse to protect the transformer!

GALLERY OF REPAIRS

             

CHECK 27K RESISTORS
AND ALL 1 & 2 WATT RESISTORS
         

DX-40 CRYSTAL DOOR MODIFICATION

The crystal door is often missing on the DX-40 and DX-35, because people did not want to have to unscrew it and remove it to change crystals. INSIDER TIP: THE DOOR ON THE BACK OF MOST HEATHKIT OSCILLOSCOPES IS A PERFECT FIT! The old scopes can be found at hamfests cheap. But I have seen them on Evil Bay, if you want to spend that kind of money.
I found a solution for both problems. Make a new door with sockets for both types of crystals and plug it into one of the internal sockets.

IMPROVING DX-40 AUDIO AND MODULATION

The DX-40 and DX-35 do not come anywhere near making 100% modulation on AM phone. This has given screen modulation a bad name. Also, the Peak Carrier Control is similarly considered bad. I disagree, and here's why. Peak Carrier Control reduces carrier level during periods of no modulation. If a small screen modulated rig is run through a linear amplifier to boost the power, Peak Carrier Control reduces stress on the large tubes used in the linear amplifier.

There are a lot of complicated modifications, some requiring a negative voltage power supply, to get full modulation. Such measures do deliver good sound, often better than plate modulation (which is subject to limitations imposed by the transformer frequency response). If you are looking for something simpler and easier to install in your DX-40, I suggest you consider reconfiguring your DX-35 or DX-40 screen modulator to the modulator circuitry in the DX-60B. The earlier transmitters used a high value resistor between the modulator and the 6146 screen. This did not take advantage of the low impedance output of the 6DE7 cathode follower stage. The DX-60B circuit adds a resistor to ground and drives the screen directly from a low impedance. The modifications on the web change the resistor to ground into an adjustable pot, to allow adjustment of the carrier level to the correct value for the linear amplifier, so the resting carrier does not exceed the plate dissipation of the tubes in the linear. While this is a good idea, the 4 watt pot is hard to obtain. One of the web modifications specifies the value of the pot as 50 ohms; the correct value is 50K. I cannot imagine how many people have burned something up playing with this otherwise good modification. Beware of stuff you find on the web, it may have errors in publication, or worse yet, design errors. You can break the fixed resistor to ground into two inexpensive and readily available separate fixed resistors to simulate the adjustable carrier pot, if you need to trim the carrier level on AM down to a safe value for your linear. It is much cheaper and easier to obtain. Presumably, you won't be changing it that often, once you have found the "sweet spot". It just takes a little cut and try experimenting.

Once you get this part working, try some things on your own. This is about learning the technology, something that is a bedrock principle of amateur radio. Go ahead and fatten up some capacitors, along the lines suggested in the web modfication, if you like. Study the DX-40U, the European version, which includes inverse feedback, which will lower distortion.

To prove my point, I show below scope photos of a stock DX-40 and a stock DX-60B. Note the flat topping on the positive peaks on the left display of the DX-40, indicating that the modulator design is insufficient to provide adequate modulation in either the positive or negative peak direction. To the right, observe that the STOCK UNMODIFIED DX-60B has nearly perfect waveform, and nearly 100% modulation. The DX-60B can be overmodulated, in fact. Why go to all the expense and hassle, when the improved design of the DX-60B will give you good modulation, without figuring out how to add a negative voltage power supply to the DX-40?


Stock DX-40 — Undermodulation

Stock DX-60B — 100% Modulation

Heathkit finally really got it right in the DX-60B. The modulation was vastly improved. It included grid block keying for nicely shaped click free CW wave form. It included a separate low voltage power supply so they could get rid of the kludgy RF driver series connection for DC. The DX-60 had sufficient loading capacitance, so you did not have to cob on external fixed loading capacitors. And the transformer was designed with enough capacity to power an external VFO. The switching of a center tap of the High Voltage transformer was gone. This was an almost perfect Novice rig, which could carry into your early General class years. It was comparable to the excellent Eico 720, except the Eico did not have grid block keying. Sadly, the one thing that did not get fixed was the horrible rotary function or mode switch. There is a way to correct that, if you still have a good function switch.

USING THE D LABS RELAY KIT TO PRESERVE THE DX-40 FUNCTION SWITCH

If you still have a good function switch, the best way to keep the DX-40 in good working order is to install the D LABS relay kit from N6TLU. It uses the function switch to switch the low current coil of a relaym, which then controls the operation of the B+ plate supply. The installation instructions are straightforward and there are Youtube videos showing the process. Get this done before the function switch fails.

Contact Terry, N6TLU to order the kit at n6tlu@comcast.net

Take a look at the details for more information. This kit was originally developed for the DX-60, but it has been adapted with installation instructions for a wide variety of vintage transmitters.

Be sure to investigate the new K2 module before ordering. The K1 module switches the B+ plate supply and provides 125 VAC to the antenna relay. The K2 has extra contacts you can use for switching an external linear amplifier and other functions you may want to implement. The K2 module only costs a little bit more than the K1 module.

NOTE: The D LABS kit does not switch the AC power on and off. You may want to take a look below at my extensive modifications, and just use the AC switch portion of that modification.

IF THE DX-40 FUNCTION SWITCH IS BAD, REPLACE IT WITH TOGGLE SWITCHES

If the rotary function switch is bad, and there is no junk rig to salvage it from, there IS an solution. Remove the key jack and function switch from the front panel. Using their holes as endpoints, drill two more equidistant holes between them. Mount 4 bat handle toggle switches in these holes. From left to right, the functions of the switches is: AC POWER, TUNE/OPERATE, B+ ON/OFF, and PHONE/CW. Connect the existing wires to the new switches. In the beginning, it would have been better to have used toggle switches the way the DX-20 is arranged.

If generic rotary switch wafers were still available, you could use the method below to make a new rotary switch that could drive the relays. I remember when centralab and stackpole had engineer's kits with these wafers and shafts and detents to build up anything you could imagine. Those days are sadly gone.

DX-40 FUNCTION SWITCH MODIFICATION NOTES

BEWARE OF THE LETHAL VOLTAGES INSIDE YOUR EQUIPMENT!

EVEN EXPERIENCED AND KNOWLEDGABLE TECHNICIANS HAVE GOTTEN CARELESS AND BEEN KILLED WORKING ON VACUUM TUBE EQUIPMENT. EVEN SOLID STATE EQUIPMENT CAN KILL YOU IF YOU DO NOT KNOW THE SAFETY PROCEDURES.

DO NOT WORK ON YOUR RADIO GEAR UNLESS YOU ARE REALLY SURE OF ALL THE SAFETY TECHNIQUES REQUIRED. THESE INCLUDE, BUT ARE NOT LIMITED TO:

  • DISCONNECT THE POWER CORD WHEN WORKING INSIDE.
  • DISCHARGE ALL FILTER CAPACITORS.
  • BE CAREFUL HANDLING TUBES, ESPECIALLY THE PLATE CAPS.
  • REMOVE FRAGILE COMPONENTS WHEN SERVICING, AND STORE THEM SAFELY TO AVOID DAMAGE.
YOU ARE RESPONSIBLE FOR THE DECISION TO MODIFY YOUR EQUIPMENT, AND NO LIABILITY IS ASSUMED MY ME OR ANY OF THE LINKS I MENTIONED. THE SOLUTIONS PROPOSED MAY NOT BE APPROPRIATE FOR YOUR RADIO SYSTEM.

At the time I did this modification, the D LABS kit was not available. I recommend you use Terry's kit, not the method I describe in this section. This modification reduces wear on the failure prone function or mode switch by using relays to do the actual switching. This modification also reduces the failures associated with the AC power ON/OFF switching. DO NOT DO THIS MORE COMPLICATED MODIFICATION UNLESS YOU UNDERSTAND FULLY THE TECHNIQUE, AND HAVE A GOOD BIT OF EXPERIENCE WORKING ON THIS EQUIPMENT Please do not write me for what part numbers to order, etc. This is an expert level mod, and quite complicated, and you will need to work independently.

I have devised a method of relay switching of the necessary functions. However, it requires that you have a good switch to start with, so that the control logic of the function switch works out. It also can be used for Push to Talk on AM, and a foot switch T/R for CW. The mike jack is changed to a two circuit stereo connector. The tip is PTT and the ring is mike audio input. A single circuit standard mono ¼ inch phone plug is used for a foot switch. If neither of these functions is desired, a standard mono phone plug MUST be inserted in the mike jack, with a short from tip to ground. Then the front panel switch controls the rig as explained in the Heathkit manual.

In addition to the AC line switching and HV switching, certain other features are installed in this DX-40.

The ¼inch phone plug attaches to the external VFO like a VF-1 or such. It keys the VF-1 with the DX-40. There is a red button on the box mounted to the back of the DX-40. It keys ONLY the VFO. This enable sensitive spotting without keying the exciter stages; it also work in any mode without switching to TUNE mode.

It is not necessary to close the PTT talk switch for TUNE mode. This circuit must be closed to use PHONE or CW transmit modes. It saves wear on the switch to use a foot switch for T/R, and makes things almost like full break in CW. On phone, it enables fast turnover and PTT. If you don't want this feature, install a ΒΌ inch phone plug with the contacts shorted in the mike jack.

There is a center off toggle switch that places either a 1000 pF or 2000 pF or nothing in parallel with the antenna, to facilitate better loading on 80 meters with some antennas, as described above.

There is a T/R switch included. Attach the T port to the DX-40 output coax. Attach the R port to the receiver. Attach the A port to the antenna. DO NOT INTERMIX THESE CONNECTIONS OR RECEIVER DAMAGE MAY RESULT. If there is any doubt, verify the relay path with an ohmmeter before transmitting.

There is an RCA jack marked MUTE. It is normally closed in receive mode. It opens in transmit. Obviously, one side is grounded. This is the correct sense for a Drake 2B or 2C. It also works for a SB-303. The receiver must be set per its manual for this function to work. You could leave this MUTE port disconnected, and use the receiver for monitoring your transmitter. You can change the sense of the contact closure inside the box for other receivers. There are spare contacts on one relay, if these do not meet your need.

The external relay box MUST be attached to operate the DX-40, since its mode functions are controlled by relays in the box. Plug the T/R relay into the octal plug on the back of the DX-40. DO NOT PLUG ANYTHING ELSE IN THE OCTAL PLUG. IT NO LONGER AGREES WITH THE HEATHKIT MANUAL.

                 

This modification does not address the problem of the 6CL6s being in series for DC. They had to add a 6 volt filament winding to the transformer; this winding was to prevent the heater to cathode voltage on the second 6CL6 from exceeding ratings. It would have been bettter to add a second low voltage winding for the exciter stages. But that would have required a second rectifer tube like a 5Y3. These days, you would do that with a silicon diode, but they didn't have them cheaply back then. It was a cost saving measure that they did it this way. Remember, the market for Novice rigs was very competitive on price. This was a beginner's rig as conceived. I thought about removing the modulator stages and mounting a separate transformer with LV B+, 6.3V filaments, and a 5V rectifer winding for the 5U4. That would have reduced functionality at the expense of a dubious gain. After all, we added a fuse to protect against 6CL6 related failures. I considered it good enough, and stopped work, to enjoy it for what it is, a fun Novice rig.

  

73,
Janis
AB2RA
Text size:  +   –      Back to Vintage Equipment
Back to the Home Page