Johnson Ranger audio driver mod

Discussion in 'Amplitude Modulation' started by N8FVJ, Apr 14, 2020.

ad: L-HROutlet
ad: l-rl
ad: Left-3
ad: Left-2
ad: abrind-2
ad: Subscribe
ad: L-MFJ
  1. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

  2. W1BR

    W1BR Premium Subscriber QRZ Page

    Did you look at the link? Electret element conversion with W3AM phase rotator and AD audio processor chip. The board is in the D104 base.
  3. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    I did not see the link.

    What kind of audio response does this combination have? Is it fairly linear or does it have dips and valleys in the low frequency area?

    I only ask because you said you saw a lack of LF response in the first stage and your mic is feeding that first stage.

    Have you made any component changes in the speech amplifier say starting with C52? Where in the first stage do you see this lack of LF; at the right of C52 or at the plate, Pin1 ?

    Last edited: Apr 24, 2020
  4. W1BR

    W1BR Premium Subscriber QRZ Page

    Pheel, I am using a HP generator for test audio.
  5. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    I read your analyses, but you have talked all around but STILL haven't answered my simple question:
    If the Ranger worked in the original circuit with +300 volts and stock grid bias, why won't it still work with +300 volts on the plate after the modification, using stock bias voltage?

    300 volts is 300 volts, regardless of the source. Furthermore, the tube parameters are based on initial static conditions, with no signal present. How does changing from the transformer to a plate choke, or returning the 12.5 k resistor to +550v, increase the plate dissipation or change anything else at the tube if the plate voltage is the same is it was before?

    Give me a simple direct answer to that question without detouring though the entire engineering procedure leading up to re-designing the class-A speech amplifier in the Ranger, and I'll let it rest.
    W8KHK likes this.
  6. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    Did you actually print out the GE 12AU7A data sheets and plot the data points I pointed to, and then do the few simple calculations?

    Apparently you read it but have not grasped the details of the analysis as the answer to your question has been answered multiple times.

    Reply #99 has the latest technical response with all of the information containing the detailed procedure, analysis and technical explanations.

    We can't use special pleading here but must use technical data to explain a technical situation.

    Last edited: Apr 24, 2020
  7. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    "EFJ engineers obviously designed the circuit to drive an inductive load with a certain current flowing through the tube and transformer so they biased it accordingly and they simulated the circuit design using a slide rule and pencil and paper."

    We are only concerned with DC and tube plate dissipation; whether or not the plate current flowing from the source is through an inductance or a pure resistance wouldn't have anything to do with the plate current or plate dissipation.

    'Since the circuit requested changed a number of basic parameters in the circuit one has to modify those "parameters" and make appropriate component changes.'

    None of the basic parameters were changed; Vg, Ip and Vp are the same as the original.

    'I gave a reference to the GE tube specs and curves as to why the design component values were as stated. If you don't change the component values and properly bias the circuit, you could have a situation where the input power to the tube and plate dissipation would have been exceeded.'

    Looking at your drawing of the circuit diagram showing Vp = 300 and 11v bias, these are EXACTLY the parameters of the original circuit, although calculating the bias voltage through a cathode resistor can be tricky, since bias voltage affects plate current and plate current affects bias voltage. If the tube is indeed running at 3.45 watts total plate dissipation (which wouldn't surprise me), hundreds of Rangers have been running that way for well over 50 years, apparently without noticeably short tube life. There are other under-rated components in the Ranger as well, such as the infamous "Chernobyl" resistor in the VFO section. I'm not sure where the 15 mA total plate current figure came from in this discussion; I seem to remember we starting out at 20 mA, which would most likely be in excess of the manufacturer's rating for the tube. WE ARE NOT TRYING TO RE-DESIGN THE ENTIRE DRIVER STAGE TO CORRECT EFJ'S ORIGINAL SHORTCOMINGS; THIS IS ONLY INTENDED TO TAKE THE DC OFF THE DRIVER TRANSFORMER, LEAVING THE REST OF THE CIRCUIT INTACT FOR NOW, HOT TUBE AND ALL. The owner is interested in improving the quality of the audio from the Ranger, not correcting every aspect of EFJ's cost engineering. EFJ's design is what it is; other inherent problems are beyond the scope if this discussion and can be addressed later if need be.

    Now, let's look at the problem, step by step.

    Let's say you set up a test circuit on your workbench, using a 12AU7, a 300 volt power supply and a 550 volt supply. Take the appropriate components and duplicate exactly the stock Ranger circuit, including the driver transformer. You don't need the modulator stage, since the 1614s are operating class-A and draw no grid current, so their presence or absence from the circuit will have no effect on the parameters of the driver stage. Use a 820-ohm cathode resistor, and 150k grid resistor just like EFJ did.

    1. Connect the grid (pins 2,7) to ground through a 150k resistor, ignoring the feedback winding from the mod xfmr.

    2. Insert a milliammeter in series, between the tied-together cathodes of the tube sections and the 820 ohm resistor.

    3. Connect the bottom end of the transformer to the +300v supply. Now turn on the supply.

    4. Now, read the total cathode current of the 12AU7. It reads ___? (Ip)

    5. Measure the plate voltage at pins 1 and 6 of the tube. It reads +300 volts? (Vp)

    6a. Measure the cathode voltage at pins 3 and 8. This is the grid bias on the tube, relative to the cathode. It reads ___? (Vg)

    6b. The plate-cathode voltage (Vpc) on the tube is 300 - Vg volts. The total plate dissipation of the two sections is Ip × (Vpc) . According to RCA, the maximum plate dissipation of a 12AU7A is 5.5 watts.
    Remember, this is the stock Ranger. Does the measured plate dissipation exceed this wattage?

    7. Now, insert a 28H choke in place of the driver transformer primary. Ground the bottom end of the driver transformer and connect the top end to pins 1,6 through a 1 to 2 mfd capacitor.
    Turn on the +300v supply, and re-check all the readings. Cathode voltage? Voltage reading taken between cathode and grid? DC plate voltage? DC plate current? Plate dissipation?

    8. Were any of those voltages or currents different from the initial measurements? If so, which readings changed and why?

    9. Now remove the 28H choke and replace it with a wire-wound slide-tap resistor, 20 or 25k @ 10w. Adjust the tap to 12.5k ohms. Connect the other end of the resistor to the +550 volt supply.

    10. Now re-take all the above readings again. Which readings changed?

    11. If the plate voltage is significantly higher, re-adjust the tap to increase the resistance, to bring Vp back to +300. Likewise, if the plate voltage is lower, re-adjust the tap to decrease the resistance, until the plates are brought back up to +300v. What measured resistance brings Vp to exactly 300 volts?

    12. At Vp = 300v, does the measured cathode current still read the original Ip? If not, what is it and explain why it changed?

    13. Is the voltage drop across the 820-ohm cathode resistor still the same as the original Vg volts?

    14. If any of the voltages or currents measured during the three sets of readings described above are different from original, what changes occurred, and why?

    15. So, in conclusion, why reduce the voltage on the plate of the tube, if it has been operating that way for decades without any obvious problem?
    ranger audiodrivermap.jpg

    Attached Files:

    Last edited: Apr 24, 2020
    W8KHK likes this.
  8. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    Try moving the resistor to the 550 volt source and see what happens. That scope pattern looks like the distortion you see when a single-ended stage is driven to saturation and flat-tops at one-half the audio cycle. I'd almost guarantee that the distortion is occurring in the 12AU7 stage. Probably insufficient plate voltage to produce the needed voltage swing. Flat-topping in the push-pull modulator would look symmetrical on the positive and negative peaks.

    Those flat-topped peaks are severely tilted because of phase shift in the transformer. If the 12AU7 gets noticeably hotter than it did before you first attempted to modify the stage, try increasing the value of the resistor while closely monitoring the modulation with the scope.

    The stock Ranger has that roll-off towards the high end, and the lows begin to roll off below 1 kHz, so the response curve looks sort of midway between an upside-down V and an upside-down U.
  9. W1BR

    W1BR Premium Subscriber QRZ Page

    For the heck of it I am going to hook up a decent Bogen amp to drive to the interstage transformer--I'll use a decent SE audio output transformer backwards between the audio amp and
    driver transformer primary. That will eliminate DC saturation issues and any effect the feedback winding has.
    KD2ACO likes this.
  10. W1BR

    W1BR Premium Subscriber QRZ Page

    Well, she can do quite well at 100 Hz modulation level using a solid state amp to drive the existing interstage. But I ran into major instability and self oscillation problems
    during the tests. Too much RF I suspect. Anyway, after using tons of 2.5mH chokes and bypass caps I was able to achieve the modulation depths shown below. There
    is some distortion that I think is more related to RF getting back into the Bogen amp than problems in the audio chain. I used a high quality SE audio transformer backwards
    of the 8 ohm Bogen amp with the secondary going to the primary on the existing Ranger interstage. View attachment 653538 View attachment 653539 View attachment 653538

Share This Page