The Fisher Phase Inverter Noose Revisited (for 500C/800C receivers)

Discussion in 'Fisher' started by dcgillespie, Feb 9, 2019.

  1. dcgillespie

    dcgillespie Fisher SA-100 Clone Subscriber

    Messages:
    9,905
    Location:
    Ball Ground, GA
    Nearly 6 years ago now, I posted a thread entitled "Improving the Fisher 400", which straight away jumped into a discussion about a specific resistor used in the design of the 400 receiver (and indeed many of Fisher's other receivers and integrated amplifiers. These resistors -- 150K in value as both R96 & R97 in early versions, and R115 & R116 in later versions -- I nicknamed a "noose" as their presence certainly acts to limit the output of the phase inverter stage by strangling it. For those interested in that early discussion, it can be found here in post #1:

    http://www.audiokarma.org/forums/index.php?threads/improving-the-fisher-400.511867/

    Because of the greater scope of that thread, there was not a lot of effort put into the whys and wherefores of Fisher's intentions for these resistors, other than to suggest that they were for protection of the output tubes to prevent them from being seriously over-driven. Part of that thread's effort presented an near completely new design of the power amplifier section, wherein the noose element was removed and eliminated. Besides the remake of the power amplifier section itself (which included the addition of EFB™ control of the output stages), the noose removal broke out to become its own mini modification for those wishing to improve the power amplifier section, but not wishing to go the full mile in modification. This was easily accomplished by:

    1. Removing the noose resistors,

    2. Powering the AF Amplifier section (only) of V12 and V13 (later version, V11 & V12 early versions) from the next lower (295V) B+ tap and,

    3. Following my steps that Larry posted in the stickies section to reset the AC Balance control.

    Over the years, two things happened. First, the phase inverter modification as a stand alone piece was well proven out in the 400, as it became known that using some of the modern manufactured Russian tubes in the stock circuit simply would not allow it to perform correctly -- not by a long shot. The stock circuit operates the phase inverter section at a low quiescent current level, causing tubes that did not have close conformance with the original plate curves to be over biased. By following the steps above, it eliminates that problem, allowing both vintage and modern tubes to operate properly in the otherwise stock design. But another trend developed along the way.

    The basic steps above simply grew to become the basic go-to procedure to improve the phase inverter section of any Fisher piece that employed a phase inverter noose. While this was an idea I never "officially" endorsed in other pieces, neither did I warn against it, nor were there any particular voices of concern expressed of poor performance resulting from implementing the procedure in other pieces, either. Heck, I even installed the procedure in my own 800C, where at the time of installation, routine testing indicated no problem.

    But then, another AKer (Vinelectra) installed the noose modification into a 500C/800C receiver he was working on (I forget which, but they are identical for the purposes of this discussion), and encountered a number of problems, not the least of which were reduced power output and increased distortion over that which the stock design produced. At the time, I did a quick recheck of my 800C and again found no problems, but none the less, because of the problems Vin noted, and because I had never made an official study into the performance of the phase inverter section and removal of the noose in the big receivers, I made a note to revisit the Fisher noose when time permitted in those applications. I'm glad I did, and that Vin sounded an alarm.

    In a nutshell, whereas the modification allowed the driver circuit in 400 receivers to move from a position of being rather sensitive to tube characteristics to one of being largely insensitive, the same modification in the big receivers causes the circuit in those units to move from being rather sensitive, to notably more sensitive. In my 800C, the particular tubes I have installed in the driver positions continue to operate fine with the modification in place, but apparently in Vin's case, the tubes he had installed did not work well with it. Accordingly, in my test bed 500C receiver, I too was able to find perfectly good tubes that performed over a very wide range from properly, to marginally, to in fact very poorly, in BOTH the stock design, and when modified with the procedures above. There are reasons for all of this, and a good resolve for the issues as well. But I don't want to get ahead of myself, so we'll get started on all of that, next time.

    Dave
     
    tcdriver, baconbadge and 1rebmem like this.

     

    Please register to disable this ad.

  2. dcgillespie

    dcgillespie Fisher SA-100 Clone Subscriber

    Messages:
    9,905
    Location:
    Ball Ground, GA
    The noose resistor does in fact limit the output capability of the phase inverter stage, but as opposed to my original thought as to why Fisher found it necessary to use it (to limit output stage overdrive), it turns out that the noose is nothing more than a bias adjustment for the phase inverter stage. However, there are consequences in adjusting the bias in this way, as will be shown later on. But first, why is it necessary to adjust the bias for this stage in the first place?

    The answer starts with Fisher's decision to use a 12AX7 for the driver stage to begin with. This was done no doubt to limit SKUs and help standardize much of the design throughout their equipment. Now the 12AX7 is in fact capable of performing the duties of the Fisher driver stage -- but only if circuit conditions are carefully controlled.

    Back in the day, with lower line voltages and output tubes that biased properly on less negative grid voltage than most do today, the phase inverter stage had to produce a total peak-to-peak signal swing of 34 volts total between both sides of the two push-pull signals. That's a peak signal swing of 17 volts per side, which will then excite the output tubes to full power against the -17 volts of bias the stock design operates the output tubes at. But that's literally only half of it.

    Because the two push-pull drive signals are out of phase with each other, a 17 volt peak signal swing per side means that actually, a 34 volt peak-to-peak (Class A) signal must be developed on each side of the push-pull signal, with each output tube then only responding to (about) half of that voltage (17 volts peak) because they are operating in Class AB1. However, because these two separate, out of phase 34 volt peak-to-peak signals are being developed by a single tube, it means that the phase inverter stage is really developing a 68 volt peak-to-peak signal to drive the output stage to full power output. In more general terms, it ultimately means that the voltage swing between the inverter tube plate and cathode elements is 68 volts at full power output.

    An additional requirement is that the plate and cathode resistors for the inverter stage together should not total more than about 100K (50K each) -- this so the stage represents a sufficiently low drive impedance for the output tubes to "see", which helps to maintain good HF stability characteristics in the overall design. As well, the following (output) stage grid return resistors (330K) are effectively in parallel with the inverter's plate and cathode load resistors too, meaning that now, a single 12AX7 stage must create a 68 volt peak-to-peak signal across what amounts to an 82K resistor with low distortion. What's more, it can be shown that it must do this while effectively operating from no more than a 260 volt B+ supply. So what's the point of all of this?
    A single healthy 12AX7 section is capable of this task, as long as circuit conditions are ideal for it to do so. And therein lies the beginnings of the basic problem: as long as circuit conditions are ideal for it to do so.

    Just like output tubes (or any other amplifying tube for that matter), a 12AX7 has characteristics that vary from one tube to the next due to manufacturing tolerances. Because of the greater current flow involved and potential for damage, output tube bias is adjusted to optimize performance and ultimately minimize/eliminate damage to the tubes and surrounding components. Why? Because just like the 12AX7 driver stage, the output tubes of Fisher's big receivers are capable of their task, as long as circuit conditions are ideal for them to do so. By adjusting output tube bias, circuit conditions are in fact being optimized for the particular tubes installed so they can perform their task optimally, and this ultimately because there is no leeway -- the power output and distortion specifications of the unit requires that their "task" takes them right to the limits of their capabilities. Similarly, because Fisher is asking so much of a 12AX7 tube in the phase inverter stage, nearly the same exact scenario exist: conditions must be optimum for the installed tube to perform its task with optimum performance.

    Fisher of course could control the tube characteristics issue by specifying the characteristics of the tubes they ordered since they purchased them by the tens of thousands. Today however, conditions are very different with higher line voltages, and use of tubes with wider manufacturing tolerances and resulting characteristics that are all over the map -- this to populate a unit originally designed to operate on bogey tube characteristics.

    Now no doubt all of this may seem somewhat strange, because whoever thinks about adjusting the bias of a 12AX7? And why is the phase inverter stage effectively only operating from a 260 volt source when it is clearly connected to a 350 volt source?

    Answering those questions starts bringing us back to the topic of this thread (the noose), and the roll it really plays. But first, a quick peak is needed at the stage preceding the phase inverter, the AF Amplifier stage. Because it is direct coupled to the inverter stage, how it operates has a very great influence on how the inverter stage operates.

    The duties of the AF stage are simple: provide maximum practical gain (about X80) with an output capability of 34 volts peak-to-peak -- easy stuff for any 12AX7 section to produce. A 350 volt B+ source, large plate load resistor, direct coupling, and bypassed cathode resistor fills this bill. Since the inverter section has a gain of about 2, this then produces the 68 volt swing needed across the inverter tube to drive the output stage to full power output.

    The operating conditions of the AF Amplifier stage are such that the characteristics of the actual tube section used have little effect on its duties: most any 12AX7 section (even a weak one) can produce the desired results. However, one other point is of significant concern. Note that Fisher specifies the voltage at the plate of this tube as 130 vdc. For the inverter section to operate properly, it is important for this voltage to be on target, which in fact, a 12AX7 of bogey characteristics will produce in the circuit as designed. However, tests with a large number of real world 12AX7s of various condition have shown that the actual plate voltage of the AF Amplifier stage ranges from a low of about 108 vdc, to a high of nearly 150 vdc -- and that's if the B+ for this stage is only 350 volts. So now, the characteristics of both sections of the driver tube have become critical for the proper operation of the driver stage as a whole.

    Today, this is all complicated even more by higher line voltages, and output tubes that require even greater drive from the inverter stage to produce full power output. It's no wonder then that any modification of the stage must be well thought out if performance is to actually be enhanced.

    So the stage is set, all of which starts to close the circle, by bringing us back to the bias for the inverter stage, the noose, and what can be done about all this schtuff.

    Next time.

    Dave
     
  3. nj pheonix

    nj pheonix AK Subscriber Subscriber

    Messages:
    10,272
    Location:
    New Jersey
  4. dcgillespie

    dcgillespie Fisher SA-100 Clone Subscriber

    Messages:
    9,905
    Location:
    Ball Ground, GA
    With everything laid out now, here's the problem with the driver stage then in a nutshell: In a classic configuration, a 350 V+ supply for both tube sections and 130 AF Amplifier plate volts (necessary for the desired amount of gain), combined with 47K plate and cathode load resistors for the inverter section (necessary to produce the low output impedance), is simply not workable with a 12AX7 tube. In a classic configuration (no noose), that would result in nearly 2.8 mA of current through the phase inverter tube section, which would leave hardly if any extra current flow capability in the tube to produce any negative going peaks at the plate (or positive going peaks at the cathode). In short, the tube would be way, way under biased. But operated correctly, a 12AX7 tube is capable of the needed drive, so what to do? Enter the noose.

    By attaching a resistor between the cathode of the inverter stage and the 350 volt B+ supply, the collective current draw of this resistor and that of the tube through the 68K cathode resistor, is more than the tube can easily draw on its own. The effect of adding the noose then is elevate the DC return of the inverter tube to about 90 volts above ground (which is what would otherwise appear at the cathode were the tube removed from the circuit), allowing the inverter tube to then re-bias itself more normally in the middle of its load line, which will then allow for maximum un-distorted output from its push-pull outputs............assuming that the plate of the AF amplifier stage is sitting at 130 volts, that is. It is because of the noose then that the inverter effectively only operates from a voltage source of only 260 volts, which does certainly restrict what the stage could produce were it biased properly and allowed to operate from the full 350 volt supply. But in this configuration, that is simply not possible with the circuit values chosen, so the decision was made to effectively restrict the B+ voltage with the noose, which in return, then allows the inverter stage to bias itself properly. And, by choosing the right values for the noose and the cathode resistors, the proper voltage restriction could be produced (90 volts), at an impedance that is in line with the target value for each side of the push-pull connection, as 68K (cathode resistor) // 180K (noose resistor) = 49.355K. Then, by placing the Phase Inverter AC Balance control in the plate leg, the 33k fixed resistance and available 25K resistance of the pot places a matching 49.355K plate resistor nearly in the middle of the pot's rotation. So, the proper DC bias, AC Balance, and output levels have all been achieved now in the stock design, as long as you have a tube installed that represents bogey tube characteristics in both sections. How many of those have you got lying around??

    The huge down side of all this is that also effectively (because of the noose), the phase inverter stage no longer operates with cathode bias, but now largely operates with fixed bias -- which is fine, until you start plugging in driver tubes that don't closely mimic bogey characteristics -- (just like in an output stage as well) -- or apply a noose modification that ultimately throws the inverter tube into virtual saturation mode with some tubes depending on their individual characteristics. So we know that the modification developed for the 400 receiver's phase inverter stage returns notably inconsistent results when applied to the bigger receivers, making it completely unacceptable for these units. As a starting point then, how does the stock circuit perform with different tube examples installed? Take a look. These shots were were taken under the following conditions:

    1. Bone stock driver circuit (including 330K following stage grid return resistors), with the Phase Inverter control adjusted as I outlined in the sticky that Larry posted for that process.

    2. The scope is un-calibrated in these shots, used for visual purposes only.

    3. A 1 kHz sine wave is used, adjusted to the onset of clipping as discussed with each picture.

    4. The plate and cathode scope shots are superimposed one on top of the other.

    5. Meter indicates cathode AC voltage in RMS values. Peak levels would be 1.414 X the indicated level.

    6. Driver stage B+ = 350 volts. Eh = 6.30 vac.

    7. Test signal applied remained constant between all shots except as noted.

    BELOW: A near new American made 12AX7A with true bogey characteristic values in each section (a rare breed indeed!). AF Amplifier stage plate voltage = 130 volts. Output level is 23.0 vac at either push-pull output. This translates to 32.5 peak volts, which is more than enough to drive even high bias voltage Tung Sol 7591A tubes to full power output with low distortion. Fisher done good.
    SAM_2685.JPG

    BELOW: Same conditions, but now a new Tung Sol 12AX7A Re-issue tube is installed. AF Amplifier stage plate voltage is 120 volts, coupled with however the inverter section's characteristics deviate from bogey as well. The effect can clearly be seen on the scope. When the drive is reduced to the onset of clipping, the output level is now just 16.8 volts, translating to 23.8 volts peak per push-pull side, being just barely able to drive some Tung Sol 7591A Re-issue tubes to full power output without distortion. Woooops.
    SAM_2687.JPG

    BELOW: Same conditions, but now, a brand new Amperex 12AX7A is installed (made in Germany). AF Amplifier stage plate voltage is now just 110 volts, combined with however the inverter section's characteristics deviate from bogey as well. When drive is reduced to the onset of clipping, the output level is now just 11.0 vac, translating to 15.6 volts peak volts per push-pull side, which is not even enough signal to drive American made bogey 7591 tubes to full power. Uh oh......
    SAM_2686.JPG

    So the stock circuit works, and works well -- as long as you bought your tubes from Fisher. The pics above were just random choice tubes. There were worse pics, and some not so bad. These do a good job of representing what many typical tubes perform like in the stock circuit today. As you might guess too, as AF Amplifier stage plate voltage goes over 130 volts, then the bottom side of the waveform becomes compromised. Therefore, tubes with bogey characteristics in both sections produce best possible performance in the stock driver stage design. But besides the basic driver stage issues, there's one other little problem we haven't even touched on yet.

    We'll get to that, and some answers for all of this, next time.

    Dave
     
    UncleBingo likes this.
  5. AlTinkster92

    AlTinkster92 AK Subscriber Subscriber

    Messages:
    3,343
    Location:
    NC
  6. thornev

    thornev Well-Known Member

    Messages:
    777
    Location:
    Mid Hudson Valley, NY
    I've got this one (noose removed and PI readjusted per Larry) in my 500-C. Interest peeked. I've done a lot of tube rolling and I could clearly hear how some tubes produced significantly better audio than others. And I did log all my tube rolling. So now I can go back and see what combination of tubes produced what sounds.
     
    Last edited: Feb 9, 2019

     

    Please register to disable this ad.

  7. larryderouin

    larryderouin Turn it UP, POP? PLLUUEEEZZZZZEE Subscriber

    Messages:
    22,226
    Location:
    Glen Burnie Md.
    Looks like a BIG EDIT / Modification to the Easy Phase Inverter Adjustment Thread MAY be in order. Maybe just delete it and re-WRITE the whole thing limited to the 400, with this one dedicated to the 500c/800c and maybe the other "BIG" units.
     
    Last edited: Feb 10, 2019
    baconbadge likes this.
  8. thornev

    thornev Well-Known Member

    Messages:
    777
    Location:
    Mid Hudson Valley, NY
    Uh oh. Does all this mean we 500-C folks that implemented the noose-removal and PI adjustment tasks are going to have to undo it all? I guess we'll wait for Dave's final post on the subject. I didn't notice a degradation in sound after I implemented so maybe I just happened to have the appropriate tubes as Dave suggests.
     
  9. ncwalz

    ncwalz AK Subscriber Subscriber

    Messages:
    111
    Location:
    Alberta, Canada
    Oh oh
     
  10. dcgillespie

    dcgillespie Fisher SA-100 Clone Subscriber

    Messages:
    9,905
    Location:
    Ball Ground, GA
    The other issue - not nearly so in depth to resolve, is that it is not uncommon for the cathode of the phase inverter stage to operate at (with some tubes) nearly 150 volts above ground. That's a pretty serious breech of limitations imposed on a 12AX7, and needs to be dealt with as well. There have been occasions where driver tubes have internally flashed and that was the end of their life, when the heater/cathode insulation within the tube could take no more. Fortunately, that can be resolved pretty easily as part of the fix offered for the biasing issue, as well.

    So what is the fix for the driver stage biasing issue? In concept, the answer is simple enough: Convert the phase inverter stage so that it operates with pure cathode bias. The usual way this is done however requires the use of a coupling cap between the AF Amplifier stage and the phase inverter stage. This is easy enough to do, but then it destroys the beauty of Fisher's original direct coupled design between the two driver stages. As well, an additional coupling cap potentially adds LF phase shift than can reduce or cause LF stability problems. That too can be addressed, but the approach then begins to take on becoming a thread you pull on that has no apparent end.

    An alternate and somewhat unique approach is to leave the direct coupled connection in place, and then reconfigure the cathode circuit of the inverter to "stand" on an appropriate R/C tail network, that will drop the necessary DC voltage so that the stage will bias itself correctly, and decouple the network as well, so as to maintain an effective AC ground for the inverter to operate from. Such an approach has a number of advantages:

    1. Unlike the noose approach, it draws no additional current from the power supply.

    2. The tail resistor is large enough such that in combination with the normal cathode load resistor, the phase inverter stage starts to take on constant current characteristics, which is exactly what is needed in view of the potentially wide range of AF Amplifier stage plate voltage levels possible. This then allows the phase inverter stage to bias itself correctly, according to its individual characteristics, as well as the circuit conditions present, as determined by the characteristics of the AF Amplifier tube section.

    3. Each channel requires its own R/C tail network (so they can each adjust independently of each other for the tube they support), one of which however is made up with two resistors to form an appropriate tap where they meet for biasing the audio heater circuit to about +75 vdc. This then keeps the heater/cathode voltage relationship well within specification for 12AX7 tubes. The total resistance of both R/C networks is identical.

    The net parts count grows by 1 part per channel (plus two more for the channel with the heater bias tap), and takes a couple of new T-strips to make a neat installation of it. Other than that however, the installation is rather straight forward.

    The results of the modified design speak for themselves, as shown in the following scope shots. The three tubes used and test conditions are exactly the same as before, including the voltage on the AF Amplifier stage plate in each case, except for the circuit now operating with the modification installed:

    BELOW: Same American made 12AX7A with true bogey characteristics in both sections. Output level is now 25.0 vac at each push-pull output. Slightly more than produced with the stock bias design.
    SAM_2688.JPG

    BELOW: Same Tung Sol 12AX7A Reissue tube. Output level at the onset of clipping is now a healthy 25.0 vac, versus 16.8 vac before.
    SAM_2690.JPG


    BELOW: And finally, the very same Amperex 12AX7A. Output level at the onset of clipping is now 23.75 vac, or over double the 11.0 vac produced under the same conditions before in the stock circuit -- a major improvement indeed! Note the meter is showing 24.5 vac as recall that the drive level remained constant in all of the scope shots presented. Close examination will show that the top crest is very slightly compressed, so when backed down, onset clipping voltage is still 23.75 vac -- which represents a nearly 50% reserve in drive capability for even the highest bias voltage Tung Sol Reissue output tubes, so even this tube is now way more than capable of driving the output stage to full, undistorted power output.
    SAM_2689.JPG


    A few final shots show the ultimate outcome of the modified design:

    BELOW: The Right Channel at 1 kHz -- fully stock, as shown at the 4Ω output, operating into a 4Ω load, at the onset of top wave crest clipping, using the problematic Amperex 12AX7A. When backed down to the onset of clipping on the bottom wave crest, power output is then a mere 16 watts RMS.
    SAM_2693.JPG

    BELOW: Same conditions, same tube, in the left (modified) channel. Power output at the onset of clipping is now a healthy 30.25 watts RMS, with output clipping defined by the output stage as it should be, rather than by the driver stage wheezing and coughing in the process of trying to drive it in the stock design. Power output and distortion are the same as when using the American true bogey tube, or the 12AX7A Tung Sol reissue tube. Fisher tubes are no longer required to deliver Fisher performance!
    SAM_2692.JPG

    The final post will provide a few pics of the Left Channel installation, which is also the channel I chose to build the tap into to properly bias the heater circuit to eliminate the stress on the heater/cathode insulation within the driver tubes. A schematic will also be posted showing the relevant portions of: (1) The original phase inverter circuit, (2), The modified phase inverter circuit, and (3) The modified phase inverter circuit that includes the tap for biasing the heater circuit.

    Audibly, comparing basically any tube in the modified channel to the stock channel using the American true bogey tube produces absolutely no discernible difference at all. Both are powerful, dynamic, and responsive. As well, frequency and transient response tests are unchanged between the modified and stock designs as well. I must say however that with any kind of demanding material, the Amperex tube in the stock channel makes its presences known rather quickly. Breakup on LF information is hardly hard to spot, as is a general restrained element of the sound.

    Finally, one last benefit of the modification is that for some tubes, those that end up under-biased (i.e., drawing too much current) in the phase inverter section that is, they will last much longer with the modification in place, since their proper operation will be assured.

    For those that installed the 400 receiver phase inverter modification into a 500C or 800C, I highly recommend that you remove it, and either (A) use only good Fisher branded tubes in the phase inverter position (or tubes that you know perform properly in that location), or (B) Install the properly developed modification I'm presenting here, and then forget about what tube you use. For any of those groaning at the thought of having to rework their receiver, I'm with you -- I need to rework mine as well. In the long run however, the results will be more than well worth the effort.

    Dave
     
    nj pheonix likes this.
  11. nj pheonix

    nj pheonix AK Subscriber Subscriber

    Messages:
    10,272
    Location:
    New Jersey
    Dave,
    does this follow the integrateds (X, KX200?)
     

     

    Please register to disable this ad.

  12. 1rebmem

    1rebmem AK Subscriber Subscriber

    Messages:
    175
    Location:
    Wilmington, NC
    Dave....first, thank you for taking the time to publish the above information!

    I have a 500B and am wondering how this latest info affects it.
    You told me back in a November thread:
    I previously removed my noose, resistivity balanced the PI anode to cathode circuits, put the AF stage on the next lower B+ node and elevated the heater to about 64vdc.
    So: If my 500B is good to go with what I've done, what would be the ideal AF and PI plate voltages?
    If I need to redo my AF/PI stage please let me know the direction.

    500B Schematic:
    http://akdatabase.org/AKview/albums/userpics/10004/Fisher 500-B Service.pdf

    Thanks so much Dave.
    Jef
     
  13. dcgillespie

    dcgillespie Fisher SA-100 Clone Subscriber

    Messages:
    9,905
    Location:
    Ball Ground, GA
    OK. Wrapping up, some final pics include the installation of the new R/C tail bias network and heater bias tap in the left channel of my test bed 500C receiver, and a schematic of the modifications. In an Epilogue, I will then discuss why the original modification worked well in 400 receivers modified with EFB, and even those that had not been fully modified to that extent. And, what's changed in the last 6 years since that modification was published. In light of all that the testing for this project turned up, I also hope to have the results of some additional re-testing done on the 400/B receiver's phase inverter circuits, which are for all intents and purposes, are basically identical with regards to the concerns discussed in this thread.

    Dave

    SAM_2694.JPG

    SAM_2695.JPG

    BELOW: In this final shot, you can see how the original 220Ω heater balancing resistors have been moved from their original location over in the power supply area, to now at the Left Channel phase inverter tube area, where the new heater bias voltage source is located. They still connect to the heater winding connections via this tube's heater pins (4/5 & 9), but their center connection now goes to the heater bias source, rather than ground.
    SAM_2696.JPG

    BELOW: The schematic of the modifications should be relatively self explanatory. I installed the network with the heater bias tap in the Left Channel driver circuit (Fig. C). The Right Channel will get the simpler modification of simply removing the noose, changing the cathode resistor to 47K, and then adding in the 91K resistor and 22 uF cap (Fig. B). The tail that includes the heater bias tap only needs to be installed in one (either) channel, since it biases the heaters of all tubes connected to the winding. As can be seen, the total resistance of both tails is identical. Not shown on the modified inverter drawing is the two coupling caps to the output tubes that continue to be connected to the plate and cathode phase inverter elements as in the stock design. Finally, with the modified design, adjustment of the Phase Inverter control is now as follows with the unit unplugged and turned OFF:

    1. Use a jumper clip to short the 350 volt B+ source to ground as before.
    2. Use a second jumper clip to short the positive lead of the new 22 uF cap to ground.
    3. With the black lead of your Ohmmeter connected to the chassis, measure the resistance at pin 3 of the phase inverter tube you are adjusting, and record the reading.
    4. Now measure the resistance at pin 1 of the phase inverter tube you are adjusting (black Ohmmeter lead still grounded), and adjust the Phase inverter control so that the reading at pin 1 matches that at pin 3.
    5. REMOVE BOTH SHORTING JUMPERS, and you're finished.
    Fisher 500C:800C P.I. R:C Tail Network.jpg
     
    gadget73 likes this.
  14. tubeactive

    tubeactive AK Subscriber Subscriber

    Messages:
    322
    Location:
    NJ
    Dave, YOU, sir, are THE MAN !
     
  15. thornev

    thornev Well-Known Member

    Messages:
    777
    Location:
    Mid Hudson Valley, NY
    Dave G said: "For those that installed the 400 receiver phase inverter modification into a 500C or 800C, I highly recommend that you remove it"

    I removed the noose in my 500-C and adjusted the PIs. Does this thread mean I have to (well, not HAVE TO) put the noose back and instead install this new circuit and use the NEW PI adjustment process?
     
    Last edited: Feb 12, 2019
  16. larryderouin

    larryderouin Turn it UP, POP? PLLUUEEEZZZZZEE Subscriber

    Messages:
    22,226
    Location:
    Glen Burnie Md.
    Reverse what you did for the 400 mod (including putting back the noose resistor). Back to Stock. Then optionally you can leave as is, or go with the updated 500c/800c mod.
     
    thornev likes this.

     

    Please register to disable this ad.

  17. Rockyhill

    Rockyhill In the meantime, which is a groovy time Subscriber

    Messages:
    3,242
    Location:
    Maine
    Looking forward to the “B” test results.

    Thank you, Dave, for sharing your expertise once again.
     
  18. ncwalz

    ncwalz AK Subscriber Subscriber

    Messages:
    111
    Location:
    Alberta, Canada
    Probably add a resistor and change a resistor to put it back to stock. I think I had to change a resistor to bring the pot back into effective range once I removed the noose on the 500c.
     
  19. thornev

    thornev Well-Known Member

    Messages:
    777
    Location:
    Mid Hudson Valley, NY
    That's right, nc. Same here. Had to add resistors to bring the PI pots into an adjustable range. In the case of my 500-C, I had to add 45K resistors at V12 and V13, pin 3. I remember some resistor that was attached to ground, some kind of divider next to the PIs.
     
  20. dcgillespie

    dcgillespie Fisher SA-100 Clone Subscriber

    Messages:
    9,905
    Location:
    Ball Ground, GA
    In my comment to remove the 400 receiver modification, the thought is to either put the unit back to its stock configuration, or, install the modification that was in fact developed for the 500C/800C receivers.

    Dave
     

Share This Page