UREI BL-40 Modulimiter

Discussion in 'Amplitude Modulation' started by K4KYV, Jul 13, 2021.

ad: L-HROutlet
ad: l-rl
ad: MessiPaoloni-1
ad: Left-3
ad: HRDLLC-2
ad: abrind-2
ad: Left-2
ad: L-MFJ
ad: L-Geochron
  1. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    Just to clarify, the quiescent DC voltage at the base of Q7 should be between 0.65 volts to 0.85 volts "greater" than (with respect to) the voltage at its emitter if it is expected to conduct current.

    Pheel
     
    Last edited: Jul 16, 2021
  2. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    This discussion is very helpful and much appreciated.

    That was my first thought, to test the circuit with Q10 and Q11 removed from the circuit, if conditions could be replicated to simulate the transistors being there.

    I checked the two bias diodes with DVM diode test mode and they read OK. I have some unused 1N4003s and tested those to make sure, and get the same readings, so no reason to suspect failure there. I don't think semiconductor devices would likely be intermittent; once gone they're gone.

    What about substituting a pair of resistors in the emitter-collector circuits of Q10 and Q11 to see af a.c. and DC voltages at the base of the transistors would appear to be something near normal? Of course, the quiescent current through Q10 and Q11 would have to be known to estimate what value of resistances would simulate the transistors, plus DC base current to Q10 and Q11 would have to be considered.

    Other than an internal fault in the transistors, what else besides excessive emitter-base current in this circuit would cause them to pull enough emitter-collector current to self-destruct?

    Here's a link to the BL-40 manual. It includes a stage-by-stage description of circuit operation and troubleshooting, plus a complete schematic. Comprehensive, but the one thing missing is a voltage chart at critical points along the circuit.

    https://www.steampoweredradio.com/pdf/urei/manuals/UREI Model BL-40 Modulimiter Manual.pdf
     
  3. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    I don't blame you for taking a cautionary approach.

    So here on Page 1 is a suggested dummy load implementation.

    Assumptions;
    A. Q10 and Q11 each source and sink 20 mA peak.
    B. The peak output voltage is +, - 15V.
    C. From B, this means each output transistor drops ~ 3Vpeak; 3V/40 mA = an equivalent resistance of 390 ohms with each conducting 20 mA, except for RQ11.
    D. I agree with Rick that a small negative bias should exist at the output, so RQ11 = 383 ohms.
    E. Q10 and Q11 each have a dc beta of ~ 50; if this is the case, then the base drive current = 20mA/50 ~ = 0.5 mA.
    F. Base peak drive voltage = 17V; 17V/0.5 ma = 34k or 33k RBDL.

    Others may have an even better approach.

    Pheel
     

    Attached Files:

    Last edited: Jul 17, 2021
  4. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    Editor timeout before finishing.

    Assumptions;
    A. Q10 and Q11 each source and sink 20 mA peak.
    B. The peak output voltage is +, - 15V.
    C. From B, this means each output transistor drops ~ 3Vpeak; 3V/20 mA = an equivalent resistance of 150 ohms with each conducting 20 mA on each half-cycle. To keep the supply current down, we will use RQ10 = 390 ohms and RQ11 = 384 ohms, since each RQ will be conducting at the same time.
    D. I agree with Rick that a small negative bias should exist at the output, so RQ11 = 384 ohms.
    E. Q10 and Q11 each have a presumed dc beta of ~ 50; if this is the case, then the quiescent base drive current = 20mA/50 ~ = 0.5 mA.
    F. Base peak drive voltage = 17V; 17V/0.5 ma = 34k or 33k RBDL.

    Others may have an even better approach.

    Pheel



     

    Attached Files:

    Last edited: Jul 17, 2021
  5. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    Again, thanks to everyone for your help and comments, both by e-mail and on this forum.

    Not precisely related to the demise of the final transistors, I see what appears to me a misprint or error on the schematic, or am I overlooking something hiding in plain sight? In the processor sections of the unit, I don't see how IC 1, 3, 4, 7, 10 and 12 are getting their power (there is no IC 8 or 9 in mine, since my regulated power supply uses simple +18 and -18v 3-prong regulators). IC 2, 5,6 and 11 show pin 7 connected to a positive voltage source, and pin 4 to negative. The rest show no pin 4 or 7 connection at all.

    I have not yet attempted to trace the physical layout on the circuit board to see if there actually are connections to pins 4 and 7 on those chips.
     
  6. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    Your op amp IC's were packaged in the early T0-5, 75 or T0-99 "can" type package with leads arranged in a circular (radial) fashion rather than in the DIP package.

    https://www.ti.com/lit/an/snoa033/snoa033.pdf


    The power supply pins were pin 7 plus V and pin 4 minus V.
    https://www.ti.com/product/LF356
    https://pdf1.alldatasheet.com/datasheet-pdf/view/8610/NSC/LM310.html


    Many times just a few IC's are marked on the schematics with the power pins and it is assumed all other op amp IC's of the same type will follow the same PS pinouts.

    Pheel
     
    Last edited: Jul 20, 2021
  7. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    That's what I was thinking, although for example, IC11 is fed with +/- 17v. I would ASSume (sometimes that gets you in trouble) the default to be +/- 18v and the ones marked differently to be exceptions. I didn't see any reference to that in the circuit explanation, but I skimmed through it pretty rapidly so it I should study it more carefully. In any case, closely following the physical circuit board traces should answer the question.

    It would be possible to make a radial-to-DIP adaptor, but that would be painstaking, unless someone has produced such a thing commercially. And I suspect not all DIP to circular adaptations would necessarily be identical. A year or two ago I checked and some, but not all, of the circular op amps used in the BL-40 were still available from sources like Digi-Key and Mouser. I have at least one spare LM-301A available; I pulled IC-11 out to disable the automatic phase-flipper which I found to be a nuisance for my use.
     
  8. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    In the "Phase Optimizer" block they do have a 1 Volt voltage dropping circuit for +, - 17V.

    Pheel
     
  9. WZ5Q

    WZ5Q Ham Member QRZ Page

    Good luck finding it Don; ohhh wait, it's sitting over there next to the guy wire clips.... ;)
    Now if you could only find out where that Exit Portal is for the wandering Mini Black Hole you got in your shack.
     
  10. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    The saddles to the guy wire clips never did show up, but I managed to scrounge up enough loose ones here and there to finish the job. I did find the footlocker-size cardboard box full of tropical hardwood scraps out in the g̶a̶r̶a̶g̶e̶ junk shed (unlabelled), next to a pile of empty boxes. The op amp chip is in a marked container next to the spare T4-C. If only they will still be there in case I need them. Despite many attempts and much effort to "organise" my parts collection, it is very difficult to keep one tiny piece in a place where I can put my hands on it 5 years later when I need it, and one of the surest ways to have something disappear is to re-locate it to a "safe" place.

    Regarding the BL-40, I noticed right away that R60 and R61 had shown signs of overheating. The parts list indicates 1/2-watt, but I replaced them with 1-watt. Then after testing it with new capacitors replacing the obviously defective swollen ones and it still didn't work, I noticed R96 and R97 had overheated as well, and one of them checked open. Still thinking maybe they had used undersized resistors (like in the PC boards in the Gates BC1-T and G), I stuck a pair of 2-watt carbon resistors (the only 10-ohm resistors still in tolerance that I could find in the junkbox). Right away they also began to overheat and that's when I checked further and found the shorted output transistors. I'll probably leave the 4.3 Ω ones in place, but definitely will put 1/2 watters back to replace the 10Ω ones, and maybe even go to 1/4 watt ones if they will still just barely be within nominal power rating.

    Something I noticed about the newer resistors is how incredibly small they are for a given wattage. It had been a while since I had purchased any new resistors of any kind since I have a well-stocked junkbox, and I was surprised that the new 1-watt film resistors (that's all I could find in that ohms range) are about physically the same size as a traditional quarter-watt carbon composition resistor. Maybe with plenty of ventilation they will dissipate one watt without burning up, but the surface is bound to be much hotter than with the older larger size resistors. You have the same amount of calories being radiated from a much smaller surface area; the only way that can happen is for the temperature to rise. Kind of like what happens when you focus the image of the sun onto a piece of paper with a magnifying glass. Not so good when the resistor is mounted directly in close contact to a composition PC board. If the resistor in the circuit dissipates any substantial amount of power and lead length isn't critical, it might be a good idea to clip the leads a little longer so that the resistor is raised a couple of cm off the board, rather than lying flat right on it. Smaller isn't necessarily always best.

    In the BL-40, I replaced C41, the 2000 μfd blocking capacitor feeding the output transformer with a 16v one. The parts list recommends 15 volt rating, but after removing the old one, I noticed it was made in Germany and was rated at 40 volts. I'm pretty sure it came that way from the factory, since all solder connections appeared to be pristine, with no evidence of re-soldering anywhere. Once I get it going, I may replace that 15v with another 40-50 volt one just to be safe, although in operation the DC voltage across it shouldn't exceed a couple of volts if it's working properly. One advantage with the sub-miniaturisation of modern day components is that one can substantially upgrade power and voltage ratings with plenty of overkill, and still leave room to spare.
     

Share This Page