Best AM detector?

Discussion in 'Technical' started by BBurns, Jan 11, 2015.

  1. BBurns

    BBurns Member

    Hello All,

    Thanks to the generosity of Larry Baker WB5OFD, an HBR-8 arrives by UPS tomorrow. This is the entry level model of the HBR series of homebrew receivers designed by Ted Crosby back in the 1960's.

    The stock HBR-8 comes with a product detector, though a later QST article showed an infinite impedance detector that could be switched in place of it for AM reception. Very likely I will be needing to add some sort of detector for AM reception, and I'm looking for advice on which circuit to choose for all those qualities that we love about AM, like low distortion, to come through.

    Cheers,

    Brian
     

    Attached Files:

  2. W5HRO

    W5HRO Administrator

    The easiest thing might be to just throw a solid state diode in there for a detector and be done with it.
     
  3. BBurns

    BBurns Member

    Hi Brian,

    Yeah, I thought that I might do that, but I figured that if I was going to go to the trouble to build a receiver, I might as well put in the best AM detector.

    By the way, the deal on the old RCA broadcast transmitter seems to have fallen through---oh well...

    Cheers,

    Brian
     
  4. W5HRO

    W5HRO Administrator

    Well, is there room inside of that receiver to add a diode tube? It would be easier to just add a SS diode right before the 6BY6 PD tube that you could switch to for AM reception.

    AM tube detector designs are somewhat limited and there's not a whole lot you can do. What I found works best is using a soft limit resistor and feeding a small voltage into it and adjust the offset. It greatly reduces the distortion during high IF output levels. I did that to my HRO-60 and it really helped.
     
  5. BBurns

    BBurns Member

    Hello Brian,

    "What I found works best is using a soft limit resistor and feeding a small voltage into it and adjust the offset."

    Thanks for the suggestion (:->)...Since I'm at the "solder practice" stage, I would really appreciate a schematic with component values.

    The HBR-8 turns out to be a parts rig, so I'll be building from scratch. I've got plenty of desk room for a receiver, so whatever I build will allow acres of room for trying out ideas.

    Cheers,

    Brian
     
  6. W5HRO

    W5HRO Administrator

    When I get around to working on the HRO-60 again I will look at it and draw the circuit up. The receiver needs a few caps and resistors replaced. That's one of my next projects after I get my transmitter done.

    Right now I cannot remember 100% for certain what I did to make it work. There's at least one resistor, a voltage, and I may have added a SS diode in conjunction with the diode tube. With my HRO-50 I just temporally added a small value resistor to ground and nothing else. The HRO-60 is the one where I made everything work correctly and still keeping the detection sensitivity at low IF output levels.

    What happens is at high IF output levels the audio will usually distort in most receivers using the old tube diode scheme. Part of the problem is that they completely cut off the one half of the cycle during detection. The trick is to try and eliminate that.
     
  7. BBurns

    BBurns Member

    At your convenience Brian.

    I can see by the amount of work that went into this "simple" receiver that I'm going to be at it for a long time.

    I'm thinking of the "modular" approach, with a group of small chassis that are built and tested separately. Does that seem like a good idea?

    Cheers,

    Brian
     
  8. W5HRO

    W5HRO Administrator

    The only issue with doing that is sometimes when the grounds on the different chassis's touch each other or move around it can shift and de-tune things a bit. One example would be the xtal calibrators that plug into many of the old receivers. I don't know how many times I've seen where people have drilled hole's in the chassis beside the octal socket so they could bolt the chassis and the calibrator's housing together to get a better ground and to keep it from moving around.

    It would probably be ok as long as you kept certain things on the same chassis. I'd keep the Mixer, HFO, CWO/PD, and any other frequency converter tubes, xtals and their tuning all on the same chassis.
     
  9. W5HRO

    W5HRO Administrator

    I did find a partial hand drawing last night of what I think is in my HRO-60. The only difference is that I know I used a SS diode and an extra resistor in series with it to equal the same voltage drop as the 6H6. If using a 6H6 you could just use each half and it would not need a SS zener and the extra resistor. You could also use the smaller 6AL5 instead too.

    The below will forward bias the diodes and I think it will still extract the intellenge in this configuration even though the diodes are turned on. If not just remove the supply and maybe place a very small resistor value to ground at the one diode instead. It still should provide full wave detection. I think what I remember now is that keeping the load stable does more good than anything else.

    detector.png
     
  10. W5HRO

    W5HRO Administrator

    Brian,

    I quickly glanced at the HRO-50 and 60 schematics this morning and I think I actually have the diodes turned around in the other direction. I think that's why I was remembering a negative voltage in the beginning. It just depends on whether you want the voltage swing to go from 0V down to a negative voltage or swing from 0V up to a positive voltage. I think feeding it into the grid of the 1st AF amp you would want it to swing negative instead, but it would just depend on what class the AF amp tube was biased in.

    Anyway, the HRO's do the below for detection.

    HRO_Detector.png

    Then what I temporally did to my HRO-50 is below. I added a resistor to soft limit things during high IF output levels and the resistor is a very small resistance value.

    HRO_Detector_w_Soft_Limit.png

    I think the below is what I actually did to my HRO-60, but again I'm not 100% positive. It's possible I'm only biasing one diode, but for some reason I'm remembering doing both.

    HRO-60_Modification.png

    Anyway, I think what I remember now is that creating a stable load on the detector, thus creating a stable load for the IF's output is what works the best so I hope this info is useful. You would probably want to use a 6AL5 or similar tube and do some experimenting. Both my HRO-50 and 60 badly need caps replaced now, including all of the 47K screen dropping resistors, etc. The last time I turned on the 50 I could tell it was starting to have problems and the 60 hasn't been turned on in the past 5 years.
     
  11. K5JXL

    K5JXL Member

  12. BBurns

    BBurns Member

    Hello Ken and Brian,

    I'm sorry for the long delay in getting back, but between a seriously long stretch of teaching (guitar making), and a major emergency generator screw-up by yours truly...you really don't want to know the details, except that nobody got hurt.

    So, many thanks Brian for all the detector information...I really appreciate your taking the time (:->).

    And yes Ken, I've done a major research job on the HBR series, and have gathered about everything that is available, including a couple of CD's that contain a lot of hints and kinks. My only reservation about the design is the regenerative front end.

    Regenerative anything is subject to overload by strong adjacent channel signals, and I'm looking hard for a way to avoid that annoyance. One of the main experts on HBR's says that Ted Crosby never did solve the strong adjacent channel signal problem. I ran across an interesting solution to the problem in a 1970 QST article on an experimental receiver for 75 meter DX work. He has a tuneable bandpass filter ahead of the RF amplifier stage. I can email a copy of the article to anyone interested.

    Cheers,

    Brian
     
  13. K5JXL

    K5JXL Member

    Thanks for the response Brian B.
    I would like to see that information .
    73
    Ken k5jxl
     
  14. BBurns

    BBurns Member

    Hello Ken,

    Could you send me the appropriate email address to send the article to?

    brian at lessonsinlutherie dot com

    Cheers,

    Brian B
     
  15. K5JXL

    K5JXL Member

  16. W5HRO

    W5HRO Administrator

    Did you get it?

    By the way, I've updated the Classified Ads forum with a new classifieds system just like the download system. The two or three ads that were in there had to be removed so you might want to create your HBR ad again. It's done now thru the classified menu selection next to the downloads selection and each listing will appear in the old Classified Ads forum just like in the New Downloads forum. The "Items For Sale" category can also have prices, etc. where people can buy the item and if you have a business PayPal account you just put in that accounts email address and get paid that way.

    Anyway, this is all going to be in the new v2.1 upgrade when I make the change and I think the new Classifieds addition is the answer instead of the Store I had installed a while back. I think the Download and Classifieds systems are exactly what we need and nothing else. The Gallery was never really used and neither was the Store.
     
  17. BBurns

    BBurns Member

    Hello Brian,

    Yes I did get the HBR-8. I'm afraid that it's just a parts rig, as the builder didn't pay much attention to the "keep the leads short" dictum. There are some nice parts in it though---ceramic sockets, tuning cap, transformers etc. And once again it was free for postage thanks to the generosity of Larry Baker WB5OFD.

    Cheers,

    Brian B.
     
  18. BBurns

    BBurns Member

    Hello All,

    Back at the question of AM detectors, I ran across an extensive article on AM detectors in the January 1961 issue of 73 magazine:

    https://archive.org/details/73-magazine-1961-01

    ALL of the back issues, and a searchable database are on this site. 73 Magazine went for 43 years, what a resource!

    Cheers,

    Brian
     
  19. W5HRO

    W5HRO Administrator

    All of those were originally downloaded from here (at the old url location) then they were archived there shortly after Wayne Green died. They are still here too. As a matter of fact that is one of the last two publications I'm reformatting and re-uploading in single pdf's instead of the huge zip files. The others are already done.

    Another thing, I downloaded a $1.99 app from the apple store which is a DjVu to pdf converter so I can finally convert all of the DjVu files from BAMA. Years ago a few hams started using that stupid DjVu format for whatever reason when pdf had always been the norm. Unfortunately tons of stuff was uploaded to the web by hams using that non-standard format. The good thing is I can select 1000 files or more at the same time and walk away from the computer, then come back after it's all finished. Not bad for $1.99

    Anyway, once all of the big zip files are eliminated I will probably set the time between file downloads to like only 1 or 2 minutes max.
     
  20. W5HRO

    W5HRO Administrator

    I read through the attached article you referenced and it describes the main issue and that is when the load on the last IF stage goes from no load to full load. I'm assuming that's why you originally mentioned a variable impedance detector. Anyway, that is the main problem and despite what many think trying to come up with a variable impedance detector is the wrong approach. The trick is to create a stable load for the last IF stage and not try to create one that varies with it which is usually an impossible task. Incorporating the cathode follower circuit can also help with that.

    Another way to also improve the problem is to use enough RF & IF stages in the receiver so the output coming from the last IF stage is closer to the same level regardless of the receiver's RF input level. Meaning, make the IF's output level closer together when the receiver's input level is 1uV or 100uV. That can also be better accomplished using a good AVC control design, but it still requires enough stages to keep the lowest input level like 0.7uV up high enough. The only issue is don't try to control any S-meter from the output of that detector. Detect S-meter control from like the first IF stage or etc. instead.

    1. Create a detector which provides a stable load for the IF’s output.

    2. Make the IF output's level closer together regardless of RF input level.

    Doing those two things with a HB receiver design is not all that difficult because you can build what you want. The issue deals mainly with the majority of the old manufactured vintage receivers. About the only thing you can do with those is improve the AVC control a little and then place a stable load on the detector itself. Trying to add extra stages along with moving the S-meter control on those is usually not practical.

    Basically I disagree with some of what the article says except for the last cathode follower part because it's geared more towards the problems associated with the old manufactured vintage receiver designs and not how you should do it regarding a HB project where you can do what you want and do it right. And, using the cathode follower circuit alone wont resolve the other issues like detection at very low IF levels or the clipping that occurs overall.
     

    Attached Files:

  21. W5HRO

    W5HRO Administrator

    What you might try is something like the attached 12AX7A biased cathode follower detector below. The trick is to just use enough negative voltage where the open/close condiction ratio offers the best rectification of RF input signal and to help eliminate the clipping. Without the negative bias voltage almost half of the detected envelope will be clipped off.

    Biased_Cathode_Follower.png
     
  22. BBurns

    BBurns Member

    Hello Brian,

    Thanks very much for following up on this! I've got your suggestions and schematic tucked away on disk, and printed out and in a three ring binder (:->)...

    Cheers,

    Brian B.