R.L. Drake

Drake Mods - R4 Mods and Tech

Authored by VE3EFJ


    I've had a few folks ask me which is the most desirable Drake receiver? Well, it depends. First, its a personal item and therefore open for debate. I've had folks from various backgrounds provide personal reasons for the R4A, B or C.

    For reasons that are contained within this document, I prefer the B or the early model C - the one with the MOSFET mixer.

  • Voltage Regulation
    Early R4C receivers employed a 12 VAC secondary power transformer. Later model R4C used a 14 volt power supply by changing the power transformer. The early model supplies could fall out of low limit spec with marginal 110 VAC. While there is a Drake fix for this, the best option is to use an IC audio amp and fix the 12 volt regulator.

  • R4B Manual Trivia

    The front cover depicts the operation of the noise blanker.

  • Low TX output in Transceive
    This may also show up as low(er) sensitivity of the receiver when transceived with the transmitter (B and C series). The cause is usually alignment or the injection cables. All RF cables (the C series had 2) must be RG/62 low capacitance cable. It is best to align the receiver and the transmitter when slaved together.

  • T4any/R4any Intermittent Transceive
    Sometimes the crystal/VFO switch on the side of the R4B can become faulty. In most cases it is left in the VFO position 'forever' and is easily overlooked. As a matter of course, it should be cleaned and cycled a few times.

    Pay some attention to the quality of the female RCA jacks on the back of the receiver and transmitter. Quite often these connectors no longer make firm contact with the cables due to wear or abuse. Make sure that the center pin of the interconnect pin is pinched by the socket. In some other cases, you may find that the outside ring of the cable is similarly loose.

    Cable quality may also be a factor, especially if you are using the original cables. Remember - in most cases you are using cables that are about 20 years old.

  • Notch Filter
    The notch filter should take the calibrator completely out of indication on the S Meter and almost completely out of the audio. When you get the notch at this point, the trim pot gets quite touchy.

    Adjustment is not at all difficult, but initially you can walk right by the T Notch null point if the trimmer is way off. Rather than try and see it at this point, listen for the null by a change in 'character' of the background noise or the calibrator. Start with the trim pot at about center.

    If you get no null at all, more than likely someones been mucking with the slug in the notch can and the coil is not resonating at 50 kHz.

  • R4C Audio (all series)
    Change C100 from .22 uF to .68 uF. Do not use a higher value. This will remove a lot of the raspy audio and clean up a fair bit of distortion. Use a tantalum and observe polarity.

  • R4C Audio Resonance
    The antiVOX line output resonates with the output transformer secondary at a frequency outside of the receivers passband range creating a high frequency hiss sound. Bypass this line with a .005 uF capacitor. You may need to touch up the anti VOX adjustment after this change.

    This change has been done by a number of people and is said to improve the audio considerably, especially when combined with the C100 change mentioned else where.

  • R4C Pass Band Bleed Through
    Replace first IF crystal filter with 8 kHz GUF-1. HIGHLY recommended. The stock Drake filter is 4 pole with 65 db stopband and a very poor shape factor. The replacement International filter has a stop band greater than 80 db and a good shape factor. This one simple change will improve the receiver considerably.

  • Crystal Filters
    A stock R4C came with 2 crystal filters - an 8 kHz wide first IF and a 2.4 kHz second IF SSB filter. The first IF filter does not do the R4C justice. Replace this filter with the GUF-1 if at all possible. You can add a sharper SSB filter. I use 2.1 kHz. Why not a 1.8? Well, the 2.1 has nice 6/60 db figures. It puts up a nice flat band- width plateau without killing fidelity. For CW, a 250 Hz width is about right. The 125 Hz is just a bit too narrow and the 500 Hz is too wide for current band conditions. The 125 Hz makes the tuning and PBT control somewhat touchy. It rings surprisingly little and is a good CW filter. The 250 Hz is not all that much different except that the tuning requirements are more relaxed.

    deleted -------------
    In mid May, word was received that International Fox Tango had been bought out. The above detail has been deleted.

    International Radio
    13620 Tyee Rd.
    Umpqua, OR. 97486
    (541-459-5623) 9AM to 1PM Pacific Time
    E/Mail inrad@rosenet.net

    Crystal Filters - Tech

    • TR3
      The TR3 does not use filters similar to the TR4. Crystal filters used in the TR3 are held in a metal box under the chassis. I've only had one TR3 cross my path and that was for service for a friend. I therefore took no liberties and did not 'crack' the box.

      Based upon the behavior of the filter and the age of the unit, I suspect that the filter used in the TR3 is composed of discrete crystals.

    • - TR4 et al
      These are 500 ohm 9 MHz filters. The TR4any are single conversion transceivers. A dedicated SSB filter is used for USB and LSB, although either filter may be used depending upon the band. The reason for two filters is to not have the transmit frequency shift between the sideband selection.
    • - R4C
      The first IF crystal filter is 5645 kHz at 1000 ohms. The second IF frequency is 5695 at 50 ohms.
    • - TR7/R7
      The crystal filters are 5645 at 50 ohms. For this reason, you can't put in the R4C first IF crystal filter into a TR7 for an AM filter. It would of been so nice if the TR7 etc took R4C crystal filters, but noooooooo. It looks like this was purposely done in the TR7. The PTO/frequency on this radio is not inverted ala R4any and the lower band edge corresponds to 5.05 MHz on the PTO. The TR7 PTO is essentially the same PTO used in the 4 line.

  • R4C S Meter Balance (early model)
    Some early R4C receivers could not balance the S meter after properly setting the AGC threshold (sensitivity control). Early model R4C receivers have only one trim pot for setting the S meter zero.

    Replace R33, a 470 ohm 1/4 watt with 680 ohm 1/4 watt.

  • Sensitivity - Late Model R4C (26K and higher)
    Poor sensitivity on this series receiver can be attributed to lack of PTO signal. On some series late model R4C a pi network on the output of the PTO line had a 620 pf cap. Replace this cap with a 390 pf. (W8CS)

  • Sensitivity - R4B
    The sensitivity adjustment affects S meter balance and sensitivity. Ensure it is no higher than -1.35 volts and no lower than -1.2.

  • BFO Bleed through R4C
    Early models could deflect the S meter while the passband tuning was moved across the IF frequency. Ensure all tube shields are in place.

    In extreme cases, check wiring harness layout and add a 47K 1/8 or 1/4 watt resistor from the base of Q5 to ground.

  • R4C Intermittent Crystal Calibrator
    The mounting screws for the blanker brackets are held by two #4 sheet metal screws from underneath the chassis. One of these is a short screw. Ensure that this screw is positioned for the plate near the chassis edge towards the back of the receiver. A normal length screw will short the calibrator when the calibrator is seated down.

  • R4C T7C IF Transformer
    The purpose of tapping the IF transformer from the output of the third mixer is to reduce noise bandwidth on the narrower filters. You'll notice a difference in the S meter reading of the calibrator should you have another SSB filter installed in one of the CW positions. The S meter will increase in reading when this filter is selected. The difference could be as much as 20 db indicated. While it may be annoying, it is not an indication of reduced sensitivity so long as you can get a noise peak from the preselector as outlined further in the text.

  • 50 kHz filter (late model)
    Seems to be applicable to serial nos 21000 and higher. Make a a 50 kHz network of a 10 mH choke and 1000 pf capacitor in parallel. Add a .01 uF 250 volt cap in series with one end. Install this network from pin 7 of V6 to ground. This should be the plate pin. The receiver in question should have a T7C (not a T7 IF can).

    Lead dress for this mod can be critical and is noted by an increase in audio hash and hum.

    This change is applicable ONLY to R4C receivers that feed B+ to the plate of the 3rd mixer through the notch filter. My own experience is that this change seems to do little, but it does no harm either. Its a very popular 'secret mod' that you may want to try, just for the hell of it.

    Its *supposed* to act like a tuned IF transformer for the plate of the 3rd mixer to minimize mixing products.

  • 50 kHz IF Coax (late model, early series)
    Late model R4C (tapped IF transformer T7C) could have some additional CW crystal filter loss due to use of high capacitance shielded cable running from T7C. Replace this audio cable with RG/174 and change the value of C49 to 430 pf.

    You don't need this change if there is no CW filter installed in the radio. Your receiver is eligible for this change IF you have a T7C (NOT a T7) IF transformer and if C49 is currently 390 pf in your radio. In this case, the coax in question will have a slightly larger diameter than the replacement RG/174 and will have a white colored center conductor.

  • R4C Audio (all)
    The audio stage in all R4C receivers is out of context with the rest of the reciever. It causes a lot of heat to be generated, distorts, and has terrible frequency response. There really isn't much you can do about this except substitute an audio IC for this. Some minor updates can be made by changing C100 (detailed elsewhere) that will provide improvement.

    An LM380 could be mounted on one of the support brackets for the noise blanker or, if you build a circuit board for it and use ground lugs ala Drake, you could use the 2 audio output transistor mounting screw holes.

    Some folks have used the LM383 and this chip will provide a bit more audio output. I have no personal experience with this chip, but I've heard that it can be tricky to deploy without having it oscillate. Commercial users of this chip should not have this problem, but home constructors using the LM383 should be aware that the LM383 is a high gain, high current and high output linear audio power amplifier.

  • Sartori Passive L/C Filter
    Sartori provided an L/C filter that was inserted between the volume control wiper arm and the input of the audio amp ostensibly to cure 3rd mixer noise and audio amp frequency response. I had one in an R4C that I had purchased and I removed it. I didn't like what it did to the recovered audio at all. If your receiver has this after market change, you may wish to make some of the changes - particularly the C100 value change if your receiver is stock. Bypass the Sartori audio filter and see which you prefer.

  • Power Supply - R4C
    This is another R4C weakness. It is not a good design. The 2 resistors at the right hand edge of the board get very hot and will eventually cook the circuit board. One of these is the dropping resistor from B+ 150 for the PTO(!) Drake does this all the time in their equipment and it is a terrible design philosophy.

    You cannot properly fix the power supply unless you make the audio changes because the class A audio output stage draws 1/2 amp (!!) and hauls the power supply down. Once you replace the audio stage, the low voltage supply will climb and you can use the EP487 as a pass transistor or install an electronic regulator.

    Once you have made the audio amp AND regulator change, eliminate some of the heat generated from the PTO dropping resistor by power- ing the PTO from the low voltage 12/14 line instead of the 150 volt line. The PTO already has a series 100 ohm 1/2 watt dropping resis- tor so .... no problem to run with the 7812 regulator.

    There is another mod circulating that uses the filament supply as a voltage boost for the low voltage line. Do not do this mod and if your receiver has had this change I strongly recommend you remove it and revert the supply to original. This mod cures nothing and actu- ally generates as much, if not more, heat. What it was supposed to have done was raise the input voltage above the 7812 input threshold so the regulator can work with the 1/2 amp load of the stock audio amp. While this does work and does reduce hum and noise considerably, it also creates a lot of additional heat from the regulator. This mod is on the right track, but the 'cure' is as bad as the disease.

  • Accessory Crystals
    Band crystals for the Drake and just about any other radio ever made may be purchased from:
    LesMith Crystals Ltd.
    Oakville, Ontario,

    These folks do small - read single - quantities and have historically dealt with amateurs since Day 1. They offer a high quality product at a reasonable price (abt $17 Cdn).

    Crystal specs are series, 20 pf and HC6/U for band tuning. You can also use the crystal positions for fixed frequency operation, but this would most likely be used for MARS etc and I won't bother with detail. The crystal specs are different between these two applications.

    The TR4any uses overtone crystals.

    Band range crystals are interchangeable between the entire R4any and T4any and even between the receiver and transmitter.

    Another source is Jan Crystals in Fort Meyers, FL.

  • CW Operation
    If you intend on operating the 4 line on CW only, service life of the 6JB6 finals may be extended by turning the idle bias down to the point that the PA cathode current meter just moves. Please do not run SSB at this setting. Be a pal.

  • R4C LM380
    I enclose this for those that wish to experiment a bit. For those that want to solve this problem using commercial avenues, Sherwood Engineering offers an audio kit, or they will upgrade and service the receiver as per your requirements.

    The National LM380 is near perfect for this application and will, together with the voltage regulator change, clean the audio and reduce current consumption.

    LM380 Pin outs

                  Pin   Use              Pin  Use
                  ===   ===              ===  ===
                    1   bypass             8  aud out
                    2   +ve input          9  N/C
                    3   grd               10  grd
                    4   grd               11  grd
                    5   grd               12  grd
                    6   -ve input         13  N/C
                    7   grd               14  Vcc

    I have included a schematic drawing program - SKEM. This program is evaluation software only - you can load it up under DOS and view and print enclosed schematics. This version of SKEM demonstrates a lot of potential. Its most frustrating area of operation is in the 'undo', for it backs out the schematic in reverse order to input. A lot of input gets wiped out to fix an earlier mistake.

    All electrolytics have -ve terminal on the ground unless stated otherwise. Observe good wiring practices, as this is a high gain, high current and high output IC. None of the parts are especially critical except perhaps the 2.2 and 220 ohm reststors. If all you have is a 4.7 K resistor instead of a 5.6K, then by all means use it. Instead of a 2.2 ohm, if all you have is a 2.7, then use it. Value changes such as this will not stop the circuit from working. Much the same applies to the voltage regulator changes.

    Remove the audio output transistor and socket from the R4C. You could mount it on brackets right where the original transistor was. If you do this, simply connect the volume control center pin to the 4.7 uF input coupling capacitor. Connect B+ to the LM380 from pin 14 to the 12 volt line on the original BFO/audio board. Output from the LM380 (through that 220 uF coupling cap) goes to the headphone jack where the black wire is.

    I suggest you build the circuit on Radio Shack perf board and run it stand alone from a seperate 12 volt supply. If more gain is required, you can increase the value of the 2.2 ohm resistor slightly. If the ratio of the 220 and 2.2 ohm resistors exceed the gain of the LM380, it will oscillate. And loudly! Parts layout is not critical so long as you keep in mind that this is a high gain, broad band power amplifier.

    Next, on to that 'voltage regulator'. You may as well use the EP487 as it was intended - as a regulator. Remove R116 from the circuit board. Remove CR18 and CR19. Remove C201. Connect a 12.6 volt 1 W zener from one of resistor pin holes that went to the base of Q12 and ground. Cathode going to base, anode to ground. Please don't use the wrong R116 hole. The 'wrong' side has 160 volts on it! Zener diodes make very smelly firecrackers. Last, install a 180 ohm resistor from base to collector of the EP487 'regulator'.

    The above changes make a useful difference in the R4C audio and will clean it up nicely. You need to make both of these changes at once. The LM380 needs clean DC to operate from. The regulator change for the pass transistor will not work with the stock audio amp because it draws too much current.

    Some have noted that an LM380 produces 'cross over distortion'. This is simply not true. The distortion of an LM380 at 13 volts nominal is around 1%. The TR7 uses this IC for its audio output and I can testify that the audio is *clean*. What an LM380 is prone to do, though, is oscillate at super audio frequencies. You may not hear the oscillation, but you will hear its effect on the normal audio frequencies. That is why there is considerable bypassing in the above design.

    Included SKEM files:

    All of the above are in SKM.ZIP, plus the original sample files.

    The schematic drawing program is in SKEM.ZIP.

  • 160 Meter Operation
    160 meters on the R4any/T4any was an option and enabled through installation of a 12.6 or 12.9 MHz crystal. The crystal to use depends on operator preference. It all depends where you want the band edge and what you want the frequency readout to display. If you want '8' to indicate 1.8 MHz, use the 12.6; if you want the band edge to be '0', use the 12.9.

    Some R4any that allow for 160 operation may have either the 12.6 or the 12.9 installed - I've seen both. If your R4 and T4 both included 160 but use different 160 meter crystals, it can cause some con- fusion until you get used to it.

  • AGC Transients R4C (early model)
    Verify that there is a network of a .01 uF and 1 Meg ohm resistor connected in series installed on the AGC board between the wire connect points on the board. This network is installed between the green/white wire and the yellow/white wire on the foil side.

  • AGC Pumping With Sharp Filters
    When good shape factor crystal filters are employed the AGC will pump when the calibrator is tuned right on the filter edge and the AGC is set to 'fast'. Excerpt July, 1976 Ham Radio pg 12:

    ".... designs with shape factors between 1.4:1 and 1.2:1 have two unpleasant side effects:

    • 1. The extremely sharp skirt selectivity presents a problem for the AGC circuit because of high group delay and phase shift, which cannot be compensated for. In almost all cases strong inter- fering signals at the edges of the filter response band will make the AGC pump. This instability introduces distortion and overshoot.
    • 2. Because of their high Q and ... the filters ring."

    Continuing, Rohde says ".... SSB reception should be between 1.9 and 2.4 kHz to limit operator fatigue .... (The) bandwidth on the famous KWM-2 was restricted to 2.1 kHz for this reason."

    Ignore it. You can't fix this without hurting the otherwise wonderful AGC.

    Many theory books show 'ideal' filter passband as an oblong box on its edge. This is not inaccurate when confined to desirable IF bandpass characteristics. 'Practical' filters have skirts. Some filters with sharp skirts will not cause severe AGC pumping but they may have quite severe ripple, depending on the response type. In general, the 90's approach is for large stopband attenuation and filter shape factors of around 2:1. IF DSP can clean up the skirt problem. This is overkill for Amateur applications, but does illustrate the move away from 1980-think of severe skirt roll off being desirable.

    There will always be trade offs.

  • 3rd Mixer Noise R4C
    Amateur 're-engineers' have claimed that an R4C weakness is 3rd mixer noise. Some of these amateur engineers have had a considerable 'go' at the third mixer inventing theoretical problems that generally do not exist.

    It is the 1st mixer that sets the sensitivity of the receiver. It is the third IF amp that provides a significant amount of the receiver gain. By the time the signal gets to the 3rd mixer it should be processed enough to easily overcome 3rd mixer noise. And it does. If your receiver works well, leave the 3rd mixer alone.

    Drake employed considerable changes over the years to this area throughout the R4C series. Improvement in an early R4C can be rendered by installing a pair of back to back diodes from the junction of C53 and C52 to ground. Use 1N4148.

    When Drake employed the 6EJ7s as mixers, the injection was moved from the cathode of the 3rd mixer to the control grid. It is this connection that some Drake enthusiasts assume to be 'noisier'. This connect point isn't 'noisier' (white noise). It *is* very capable of creating hash and is much less tolerant of sloppy lead dress. I have a mod for this further on with much greater detail.

    There was also a Sartori mod that injected the LO signal into the third mixer from the bottom end of the secondary of the 3rd mixer grid input transformer. This mod follows good engineering practices and one of my R4C receivers has had this change. My other R4C, an early one, has the injection to the cathode of the 6HS6. I cannot tell much difference.

  • R4C Noise Blanker
    Do not use the blanker gain trim pot to make the receiver 'more sensitive'. It won't. Ensure the S meter deflection on the calibrator is exactly the same on 10 meters both with the blanker and with the 9 pin jumper plug. Excessive blanker gain will degrade the AGC by compromising the gain balance in the receiver and possibly allow the BFO to bleed into the IF strip.

  • Blue Dial Filters
    The heat from the dial lamps will eventually turn the blue dial and meter filters clear. You can restore the color by dipping the bulbs in nail polish, specifically Artmatic USA #163 Peacock Blue Nail Enamel (With Hardener) (Dec 1993 QST pg 86, A. Ross W2NXC).

    The file 'NEWR4C.SKM' contains a schematic of the changes to use the EP487 as a real voltage regulator.

  • R4C IF and RF Amp Resistor Changes
    There have been previously published mods to change screen grid resistor values to improve sensitivity and allow for S Meter zero on early R4C receivers. This is a bad mod. It does not improve sensitivity and is overkill as a method for S meter balance. It increases the receiver gain and consequently alters the good AGC charac teristics of the receiver. Do not do this mod and if your receiver has been modified, revert to original factory values.

  • R4C Late Model 3rd Mixer
    As explained above, this 'flavour' of R4C with the 6EJ7 3rd mixer is claimed to be noisier. Well, it isn't noisier. My late model R4C was not original in this area; it had been modified to inject the LO through the bottom of T6 into G1 of V6. It worked well, but there was a lot of hash - power supply 120 Hz spikes in the audio. When I placed a screw driver blade near G1 (or C199) of the 6EJ7, the garbage increased. This is no good. No good at all.

    The following will not cure white noise in the 3rd mixer, but if you have the above problems, it will kill this hash, buzz and assorted garbage once and for all. All mixers make white noise - consider it incurable. Basically, this mod changes the 3rd mixer to cathode injection and allows G1 to be DC grounded. The verbal text describes the end result of the circuit changes and is not a step by step procedure.

    Change CR20 and CR21 to 1N4148. Change C52 to .005 uF. Remove C200 and C199. Replace C199 with a straight piece of wire. Where C52 connected to pin 1 of V6, connect it to pin 3. This essentially reverts the 3rd mixer of late model R4C with the 6EJ7 to the circuit used in the early models.

    I cannot give you a before and after comparison, for I never have had a chance to play with a stock late model R4C. However, after this mod my '6EJ7' R4C is dead quiet with the stock audio amp and power supply. With no antenna and normal volume I'd swear the speaker was disconnected. Additionally, the problem with the S meter moving as the PBT control was rotated was all but eliminated. Signals literally jump out of the speaker from nowhere. This change will not make the receiver more sensitive, but it did clean up significantly the garbage in the audio (for me). The reason for this change revolves around the need for a DC path for G1; cathode injecting the LO is the easiest way of provide injection. Additionally, the concept is proven from the earlier R4C designs. Indications are that the 6EJ7 is a pretty 'hot' pentode mixer.

    I cannot testify to what an original late model R4C was like. Before you try this change, I'd expect that you have some audio hash that gets almost unbearable in the AM position. When you place a screwdriver near C199, the hash and 'junk' increases in amplitude. If you do not suffer these symptoms than rule 1 of modifications takes precedence ....

    "If it ain't broke, don't fix it."

  • R4C 3rd Mixer 6EJ7
    In the later series of R4C a 6EJ7 was used in the 3rd mixer. It is this tube and surrounding circuit that can be responsible for a considerable amount of hash. Try substituting another 6EJ7 in this circuit location.

  • R4C Audio Oscillation
    Some R4C audio pre amplifiers will oscillate in the 100 kHz range contributing to audio fuzzyness and other problems that you think are sourced in the RF stages.

    Connect a series network of 4.7K 1/4 W and .0012 uF across R83.

  • R4C Audio Hash
    R4C receivers are notorious for making not only harsh audio, but also having some hum and power supply hash thrown in for good measure. Most of this is curable, but not without some effort. The later model using the 6EJ7 3rd mixer can be the worst of the lot.

    Noticeable improvement can be made by returning the power supply secondary grounds to the canned filter caps. Some folks have put a copper strap under the circuit board ground lugs on the circuit boards, running a strip of copper under the whole length, grounding the lugs. I'm skeptical about the long term. Copper corrodes.

    The low voltage supply/regulator/audio is marginal, at best. Measure the voltage on the audio output transistor emitter. If its above about .5 volt (assuming the proper emitter resistor), the transistor could be drawing too much current or be going into thermal runaway, hauling that marginal supply down.

    The previously listed mod changing the 3rd mixer 6EJ7 to cathode injection helps considerably, for it grounds G1 to DC. This (original) floating grid can be responsible for an incredible amount of crud.

    For a simple solution to the inherently lousy audio response in the stock audio amp, the C100 change makes it much more pleasant.

    For the price this receiver sold for, it shouldn't have these problems in the first place. What makes it worthwhile is how good the receiver becomes once these marginal and frankly unacceptable characteristics are attended to.

  • R4B Sensitivity
    Tune in the calibrator and then pull V10, the 12AX7 noise blanker clamp. If the S meter rises, replace the tube. A gassy 12AX7 will drive the NB clamp diode partially on, killing IF gain.

    If a new 12AX7 still does not cure the problem, it could be caused by the clamp diode. An acceptable substitute is a 1N270.

  • R4B Crystal Filter
    An R4C first crystal filter can be used if T5 and T6 in the R4B are replaced with R4C part number 251-9285. The filter would be installed on the preselector bracket and coax run from the low impedance windings of the replacement transformer to the filter. T6 in the R4B is part of the crystal filter. There would be some sheet metal work involved for brackets and shielding to insure that the filter stop band attenuation was not compromised. You'd do this if you were to purchase a GUF-1 for your R4B.

    Table of Contents Previous Page Next Page Drake List Web Site

    Layout and Design Copyright © 1997 by Thom LaCosta-K3HRN
    All Rights Reserved