Phase Rotating Asymmetry Eliminator

Discussion in 'Amateur Radio' started by W5HRO, Jul 10, 2016.

  1. W5HRO

    W5HRO Administrator

    Has anyone ever used one of these? They are available on eBay for only $20

    Also a reference link:


    Power Supply:
    +/- 5VDC to +/- 18VDC ( < 200mA )
    Recommend best +/- 15VDC

    Input Impedance:
    5.5k OHM ( at 1kHz Sine wave )

    Input Level:
    2000mV RMS ( at Power supply +/- 9VDC).
    2500mV RMS ( at Power supply +/- 12VDC).
    3000mV RMS ( at Power supply +/- 15VDC).

    Frequency Response:
    10Hz - 100kHz ( -3dB )

    Attached Files:

  2. W1VTP

    W1VTP Member

    This type of technology has been strongly suggested by a respected AMer, Clark N1BCG for some time now. It should make reception of AM signals using older diode detection more tolerable. There is a paper out there but I couldn't find it on a quick search. Here's something put out by Orban:
  3. K4TQF

    K4TQF Member

    I just turn my head sideways so many degrees... in phase audio every time ! :lol:
  4. W1VTP

    W1VTP Member

    I think the FM stations use something like this technology to keep the sidebands equal on both sides. This mostly affects speach.
  5. W5HRO

    W5HRO Administrator

    You would need to stand upside down on your head to eliminate "Asymmetry" :mrgreen:

    The only problem I see with those $20 boards is they can only handle tiny voltage swings. Wouldn’t work too well with a hot high-Z D104 and vacuum tube circuits.

    Although, I do remember something I posted on the old board about vacuum tube op-amps :icon_shh:
  6. WD5JKO

    WD5JKO Member

  7. W5HRO

    W5HRO Administrator

    You might be able to build the phase rotator using like 4 or 8 of these in a similar configuration in a tube audio chain. Each stage would just need to be well regulated.


  8. W1VTP

    W1VTP Member

    Is he still incorporating that damn gating circuit? Apparently, the LOUD fans in the transmitter made the gating circuit necessary. That would be a show stopper for me - I've heard the gating action and it's annoying. Otherwise, it's a good solution and the DYY rig which is oversensitive to asymmetrical speach and may I suggest some of the older vintage transmitters that were deliberately designed to limit positive peaks to barely 100%. The designers of commercial amateur radio gear sure has some dumb ideas about the effect if positive peaks of an amateur radio transmitter.

    So, if you are using a D-104 and if you going to use stock EFJ for for that matter, Collins transmitters the DYY board might be just the ticket - as long as you can get rid of that gating circuit. That should be a feature that should be operator selectable .


    PS: Looking at the circuit, I see there is a jumper to select or deselect the gating function so - if you need a phase rotator - this might be the ticket - for $80.
  9. W5HRO

    W5HRO Administrator

    I think it would be more fun using eight 12AX7A's and build a speech amp phase rotator that hooks directly to a D104.

    Just get a small rectangle aluminum chassis box along with 8 bayonet tube shields and build it. It would take a little experimentation to get each set of the four stages configured correctly, but I bet it would work. You would basically just configure each stage as op-amps, but in the same configuration as those solid state boards. Then it could be used with any tube transmitter into its audio driver stages, etc. You would defiantly need to use a regulated +200V to 300V DC supply with it and maybe small zeners or the little NE lamps on each stage like in the drawing I posted to keep it stable. The other trick would be to use new modern production 12AX7A’s of the same brand to make sure they are all the same and fairly well matched.
  10. WD5JKO

    WD5JKO Member

    Maybe I could add this circuitry to the Astatic UG-8 base.

  11. KE7TRP

    KE7TRP Member

    I purchased that board a case, the Power supply board (needs - and + 15 volts. I even purchased the connectors. I never got around to putting it together. There was an device used on AM BC stations that was rack mounted. I understood it was filled with epoxy so nobody could see the circuit. Maybe someone here will remember the name of it.

    I really like the K7DYY board. If you need low impedance input, Its a great way to go. I used one on my T368 to the 600 ohm input and had great success. I think I have the early version. Did he add the phase rotator on a later version? I dont remember any jumpers on the board. Its been at least 3 or 4 years.
  12. W5HRO

    W5HRO Administrator

    Those K7DYY boards really don't work very well with high Z mics that produce high voltage swings like a D104. Jim already discovered that after having to attenuate the mic's swing way down by decreasing that stock 6.8M resistor. He had nothing but square-waves. I think he had to reduce it down to like a 150K resistor or some similar value. It will work like that, but you lose a ton of the mic's frequency response doing so.

    Steve, WA1QIX created a similar circuit years back using one of the ADI chips and he even said those chips really don't work very well even with lower Z mics. It's a shame he won't come here anymore. I think he's afraid of what his buddies will think up in that neck of the woods ::)
  13. KE7TRP

    KE7TRP Member

    Which board was this an issue on? I see he has revised this thing over the years. My board does not have that problem at all. Its clean and clear with a non powered D104. Its not square waving at all. I have spent hours and hours using it while watching the scope and analyzer in my shack. Maybe the operators voice is stronger then mine and he is making more input to the board or maybe its a different board all together. I think I have an early board with less features that the new ones. I can look at it and see which one I have. I have even run the thing into my audio rack and have had no issues at all.
  14. W5HRO

    W5HRO Administrator

    Ask Jim, I remember he scoped out the one he had and I think at it's input and the D104 was hammering it. You have to look at it with a scope at the input and the various stages. Think about it, those chips and that first transistor only have a +5V supply rail. The voltage swing coming from a D104 it way too high for it.

    Here's the link to what I was also saying: Re: Using high Z mics with 5V processor chips

    "I have used those chips (the SSM2166P) with high level inputs. I have a circuit which has an input level control (ahead of the chip). You could add a resistor in series with the input (ahead of the gain pot) for additional attenuation.

    The SSM2166 is not a real good compressor chip, but it sort of works.

    The schematic is at"
  15. W5HRO

    W5HRO Administrator

    Here, to show how hot a non-amplified D104 is, I hooked mine up with clip leads to a X10 scope probe.

    My D104

    With the scope set to .1V/Div (100mV) I can exceed all 8 divisions if I give it a loud HEY... 8 divisions (800mV) with a 10X probe is 8Vp-p. That's pretty dam hot. You can see why Astatic used a 9V battery (supply rail) when they added that 2 transistor amp board later on and even that wasn't really high enough.

    And if you notice with my voice the negative peaks are more pronounced and go down much farther than the positive peaks on average. These mics and with my voice could really use a phase rotating asymmetry eliminator, but those little solid state boards would never cut it. It would have to be a tube type rotator so as not to lose the great wide frequency response the D104 will provide.

    A normal "HEY..." (.1V/Div w X10 Probe)

    A loud "HEY..." (.1V/Div w X10 Probe)

    There's actually a slight voltage swing loss with the clip leads and the scope probe as well. I've seen some D104's easily produce 10Vp-p to 12Vp-p in the past. Some are just hotter than others.
  16. W5HRO

    W5HRO Administrator

    You know, you might be able to buy one of those $20 boards on eBay and change out the 4 chips with ones that will handle a bit higher supply voltage, maybe like +/-20V to +/-30V and then just add the FET buffer transistor with the 6.8M or 4.7M resistor. If you could get the supply rails up to at least +20V to -20V it might be fine.

    With that current board and chips it says using +15V and -15V supplies the input max is 3000mV RMS (3V) so that's 8.48Vp-p. That's just a little shy of what it needs to be at least with my mics. My D104's are really old like from the 1940's and 50's so they might be much hotter and closer to 20M. That one in my pic and voltage examples is a 1940's model. I'm trying to remember, but I think sometime around the mid to late 1970's Astatic reduced the elements down to maybe half of their original level. I seem to remember = or > 4Vp-p for some reason. I worked on a ton of those mics for CB'ers back then and because of the CB craze and the increase in sales someone at Astatic may have changed them so they wouldn't clip into those two transistor amps because of only having the 9V supply rail and because CB's then had lower Z solid state inputs anyway. You would just need to check your mic to see if it has the older = or > 8Vp-p swing or the newer = or > 4Vp-p swing. I think up until at least 1970 most D104's were still closer to 20M and then they went down to only 10M after that. Anyway, that's what I remember, its just been a very long time ago.
  17. W1VTP

    W1VTP Member

    I think this subject is worthy of good implementation. I don't worry about asymmetry with my 32v1 or my class E transmitter or for that matter my Flex 5000 'cause all these units can easily handle my voice asymmetry characteristics. However, there are lot of EFJ, Collins transmitters - oh yes, the GKs that may have difficulty handling asymmetry greater than 120% positive peaks - that's where this phase rotating technology would save the day and give greater talk power and at the same time accommodate older diode detectors.

    I'm looking forward to future contributions where this technology is successfully implemented.

  18. W5HRO

    W5HRO Administrator

    I'd actually like to try building a rotator using the 12AX7A's like I mentioned, but I just have too many other projects going on. Maybe someday after everything else is done I can try it.

    I think what would work best is to eliminate the asymmetry then just use the negative peak limiter to increase the positive peaks instead of trying to do all of that in the audio chain instead.
  19. KE7TRP

    KE7TRP Member

    My D104 has no clipping at all on the input or output. The schematic shows that there is a buffer transistor. My board does not use the above chip. I guess my voice just wont drive it as hard as jims. I also have the early board. I see he has three versions now. Here is the one I have. I think its also possible that some D104 heads are having issue. I have about 20 D104 heads here and when I swap them out, Every damn one of them sounds slightly different and have different output levels. So you have to deal with that.
  20. K4TQF

    K4TQF Member

    The NJM072B show about -20 degrees @ 20KHZ. Just a few degrees at 3 or 4KHZ... but that's from the data sheet, not in circuit.
  21. WD5JKO

    WD5JKO Member

    Clark, I think your board was the best of the series. I had the next version that changed processor chip type, and eliminated the pot on the input side. I had to put the pot back, which lowered the input sensitivity. This reduced the overload, but also made the noise gate threshold too high. The noise gate was cutting off my voice, so I then had to adjust the noise gate components to lower the threshold. I still had two problems:

    1.) During the processor attack time, the output was clipped severely. This was apparent after a pause in speech and then suddenly speaking. I added a clipper on the output to keep this occasional spike in check.
    2.) Close talking the microphone would still cause clipping. I have learned to just speak about 6" away from the element.

    After all these issues, I am quite satisfied with the results, and signal reports are excellent.

    When I listen to a station running this setup hooked to the K7DYY transmitter, I often hear the transmitters crap out on a modulation peak, and then need to be reset to get back on the air. I have little doubt that this is caused by the occasional voice peak coming out of the processor that spikes to the limit. Perhaps the latest version with the phase rotator helps minimize these issues?

  22. W5HRO

    W5HRO Administrator

    Yeah, but have you clipped a scope probe across the resistor to ground right at the transistor's source connection and looked at it? That buffer transistor only has a +5V supply rail and most D104's are too hot for only a +5V rail.

    Those boards would probably work better if the +5V regulator was removed and just used the +9V directly from the battery instead. Don’t know whether the chips could handle it, but if so it would probably extend battery life too.

    If the board has the pot to reduce the gain down to the next stages then maybe just run the buffer at +9V and the chips at +5V instead. That might work too.
  23. K4TQF

    K4TQF Member

    The NJM072B/082B & NJM072/082 are dual JFET input operational amplifiers. They feature low input bias and offset currents, high input impedance and fast slew rate. The low harmonic distortion and low noise make them ideally suit for amplifiers with high fidelity and audio amplifier applications. The NJM072/082 may cause oscillation in some application like voltage follower. ■ FEATURES:
    ● Operating Voltage ( ±4V~±18V ) ● J-FET Input ● High Input Resistance ( 1012Ω typ. ) ● Low Input Resistance ( 30pA typ. ) ● High Slew Rate ( 13V/µs,20V/µs typ. ) ● Wide Unity Gain Bandwidth ( 3MHz,5MHztyp.) ● Package Outline DIP8, DMP8, SIP8
  24. KE7TRP

    KE7TRP Member

    Ok. Glad we cleared that up. I am now pleased I ordered a spare board at the time. Mine seems to work great and does have input Pot. It really works great. Glad you fine tuned your newer board. Did you document changes?

    I have heard many of the transmitters trip on the air. I am sure you are correct in that it was a voice peak that took it out. Those transmitters really sound great. I have never heard a bad one. At one point I was very tempted to order one but at the time I spent most of my AM on 40 and they where 80 and 160 only.

  25. KE7TRP

    KE7TRP Member

    Yes. I have a Tech scope in the rack and hooked the probe up after reading your test. I have no clipping on input or output. I dont yell into the microphone though. I would think if you did that, then you might be able to over power the front end. My board has an input pot to reduce input as needed. I guess my early board did not have the issues Jims board had.