AliExpress / eBay JL Hood 1969 Class A Amp

Discussion in 'Power Amps' started by purr1n, Apr 25, 2017.

  1. ruinevil

    ruinevil Acquaintance

    Contributor
    Joined:
    Nov 17, 2019
    Likes Received:
    51
    Trophy Points:
    18
    Location:
    Capital District, NY
    Home Page:
    Wait for @Priidik, but from my Googling, its a voltage divider that reduces gain. Basically it makes the signal entering the amplifier less loud and reduce preamp/DAC hiss.
     
  2. Priidik

    Priidik MOT: Estelon

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    2,153
    Trophy Points:
    113
    Location:
    Estonia
    The original does not have anything before the input cap.
    There are certainly infinite ways to mod the circuit to make it 'specul'.
    I doubt that adding input resistors (shunt or in series) it makes it sound any better.
     
  3. Mihnea

    Mihnea New

    Joined:
    Aug 25, 2020
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Romania
    Hello guys, I want to build this amplifier myself , starting with this amplifier board:
    https://www.audiophonics.fr/en/ampl...lifier-bipolar-2x15w-8ohm-a-pair-p-11129.html
    Looks quite similar to what you guys have. I am planning to perform the tweaks you have recommended, but I still have some questions... What I am most concern is the PSU. It bothers me that it is on the same board as the amplifier. Also, for a class A, I would expect a bigger capacitor rail (see Hiraga). Do you think there is room for an impruvement on the PSU?
    Also, do you think it would worth using two toroidals (one for each channel)?
    Thank you in advance and please don't mind if I'll throw a few more questions your way.
    All the best,
    Mihnea
     
  4. dBel84

    dBel84 Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    3,230
    Trophy Points:
    113
    You don't have to use the on board rectifier if you don't want to, just add the DC from your external supply where the big capacitor is not fitted. Alternatively, add the biggest capacitor you can fit in there or use these point to pull out a decent CRCRC filter. Many options. If you do use the on board rectifiers, make sure you bolt them to the chassis or a heatsink as they will get hot.

    Re dual transformer, it may offer some benefit but not sure it will be worth the extra work.

    .. dB
     
  5. Katalyst

    Katalyst New

    Joined:
    Jul 30, 2020
    Likes Received:
    0
    Trophy Points:
    1
    Location:
    Indiana
    Using two transformers would allow a Mono block setup.
     
  6. Mihnea

    Mihnea New

    Joined:
    Aug 25, 2020
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Romania
    Thank you guys for your replies. The hardest decision is if it worth going for mono block setup (I know that in theory it does, but it's hard to tell if the results will pay off for all the trouble).
    Fitting a bigger capacitor seems to be the easiest upgrade on PSU. Adding a CRCRC filter implies having a different transformer, I need to think this one properly (can you recommend some resources on this topic?). Did anyone of you performed such upgrades on this board?
    Also, have anyone of you reversed engineered the schematics of this board? It the PSU regulated?

    All the best,
    Mihnea
     
  7. Priidik

    Priidik MOT: Estelon

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    2,153
    Trophy Points:
    113
    Location:
    Estonia
    Best free resource is LTSPICE. Real handy with such tasks before you waste solder and time on the bench.

    The psu on the pcb you linked probably has a capacitor multiplier (cap mpr), hence the third power transistor.
    It could instead be series regulation. One eBay pcb that I got had cap mpr on it. I doubt there are too many different ones.
    Personally I wouldn't put a cap mpr into such circuit. This would be better served for steady current draw and micro power. (read: it sucks for large transient surges) A practical cap mpr for several amps of current draw the voltage drop would be severe, and yes, it should be compensated by a higher voltage transformer. Even more than for LC or CRC filtering.

    I have built two of the JLHoods, one with LC filtering in psu, other with a cap multiplier. The LC one sounded better. It wasn't honest apples-to-apples as the one with LC has much beefier psu and is meant for speakers, whereas the other was meant for hp-s and operates on much lesser rail voltage and has less gain.

    Monoblock has one inherent disadvantage, not that it applies too much to JLHood, but for monoblock you'd use say 2 times smaller transformer. Where while using one big common to both channels the likelihood of having powerful transients on both channels simultaneously is not 100% --> it has more reserves for the one channel that has transients. This gives you better headroom, so to speak, and also the impedance of the secondaries would be lower --> in theory better bass. Again, bass is not a strong suite of JLHood amp to begin with, so do not fret.
     
  8. Mihnea

    Mihnea New

    Joined:
    Aug 25, 2020
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Romania
    Thank you again. I have to fill a lot of gaps here, both in experience and on tools&parts. I'll take it slowly. I'm going to buy the kit, with a good pair of caps and see how it goes. Reverse engineer the schematics to fully understand what they have done on the PSU, then try to design a better one (using LTSPICE as you've recommended). Maybe I can still use that 2N3055 that is in there, or maybe I can source a pair of LM338K like in the original design.
    As for the transformer, I've decided not to go for the full mono block setup. Then I am not sure if using two transformers in the same chassis will make any difference. My options are one transformer 2x24V 450VA or two transformers 1x24V 250VA.

    Thank you,
    Mihnea
     
  9. Priidik

    Priidik MOT: Estelon

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    2,153
    Trophy Points:
    113
    Location:
    Estonia
    Practically any modern high power NPN works in the output there.
    Mine have Toshiba 2SC5200 in older and newer the TCC5200 in them and work great. Theoretically there is not a huge difference in using modern high speed devices there, but the modern ones are more available and probably cheaper. These Toshibas, for example cost peanuts from reputable shops. I expect the newer ones to end up having better measured performance due to better hfe and thus more feedback, better linearity, less noise etc.
    You can roll the output devices and report back. I guess nobody has done this so far properly.

    Probably not a good plan to use LM338 as output device. It's a voltage regulator.

    In general I do not encourage the idea of using the L-bracket to connect the transistors to heatsink. It's high thermal resistance path. (this is somewhat offset by the TO-3 transistors in general having better SOA).
    I prefer something like this and is equally simple to attach to a heatsink: an eBay JLHood link
     
    Last edited: Aug 27, 2020
  10. Mihnea

    Mihnea New

    Joined:
    Aug 25, 2020
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Romania
  11. Mihnea

    Mihnea New

    Joined:
    Aug 25, 2020
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Romania
    Hello guys, do you happen to know what value the volume pot is? Also, I presume it is wired at the input signal.
    Thank you,
    Mihnea
     
  12. je2a3

    je2a3 Almost "Made"

    Joined:
    Feb 12, 2019
    Likes Received:
    380
    Trophy Points:
    53
    Location:
    USA
    Home Page:

    50K input level control

    [​IMG]

    + 10K at the RCA jack
    [​IMG]

    input Z = approx. 8.3K ohms
     
    Last edited: Sep 23, 2020
  13. Mihnea

    Mihnea New

    Joined:
    Aug 25, 2020
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Romania
    Thank you je2a3!
     
  14. Katalyst

    Katalyst New

    Joined:
    Jul 30, 2020
    Likes Received:
    0
    Trophy Points:
    1
    Location:
    Indiana
    These are not installed in mine.
     
  15. je2a3

    je2a3 Almost "Made"

    Joined:
    Feb 12, 2019
    Likes Received:
    380
    Trophy Points:
    53
    Location:
    USA
    Home Page:
    Assuming your input level potentiometer is also a 50K and nothing else is in the path, then that'll be your input Z which makes it an easier load for higher output Z preamps.
     
  16. Mihnea

    Mihnea New

    Joined:
    Aug 25, 2020
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Romania
    Hello guys, I'm asking here the same question I've asked on https://www.diyaudio.com/, hoping to get a more clear answer...

    I'm building this dual rail linear regulated power supply for a class A Hood
    amplifier (I have ordered a kit - the version with 4 output transistors/channel). I have purchased almost all the parts for the PSU and now I try to figure out what kind of cooling I need for the diodes and for the regulator.
    I have run the LTSpice simulation and I am quite puzzled by the results. Given an output current of ~4.5 Amps, the Power consumption on each diode is as much as 40W, while on the regulator is ~20W.
    The diodes are MUR860 (not available in Spice so I have used a close match) and the regulator LT1083CP.
    Given these figures, the radiator on each diode needs to be 3 deg/Watt and this is way bigger than I have expected.
    The question is: am I missing something here? Do I really need this kind of cooling in normal working conditions? I was planning to use one cooler for each two diodes, placed back to back, but now I don't think this is a valid option...
    Schematics and simulation data bellow:

    [​IMG]
    [​IMG]

    Thank you,
    Mihnea
     
  17. Priidik

    Priidik MOT: Estelon

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    2,153
    Trophy Points:
    113
    Location:
    Estonia
    How did you calculate or read these power values?

    Sanity check is to visualize what goes on in a diode in such scenario: current through the diode is combination of the load current and current shunted by filter capacitors, usually it is about twice the RMS of the load current for the whole diode bridge half. This is true, when you have no shunt elements. Capacitors mostly generate reactive power, but still add to the current through the rectifier. However, this current will not be wholly aligned with the voltage drop over the diode, so there is significant reactive power being produced.
    A ball park diode bandgap drop is half a volt, that is for 5A of current in phase you have about 2.5 W of power dissipation.
    - edited a few mistakes out

    For comparison, here is one of my prototype amp psu:
    upload_2020-10-14_15-36-50.png

    The regulator can easily waste tens of watts, this is based on how much voltage you are willing to sacrifice for the power and better regulation, cleaner dc.
     
    Last edited: Oct 15, 2020
  18. Mihnea

    Mihnea New

    Joined:
    Aug 25, 2020
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Romania
    Hello Priidik, thank you for your answer. The values that I've mentioned are computed by LTSpice... I am also puzzled by those figures. I've looked again today: Spice shows spikes of 20 Amps through each diode... Something is awful wrong here.
    Not sure how can I attach the asc file, but I would be very grateful if you could take a look to see what is wrong there.

    Thank you so much,
    Mihnea
     
  19. murray

    murray Friend

    Pyrate
    Joined:
    Sep 29, 2015
    Likes Received:
    478
    Trophy Points:
    63
    Location:
    N.Z.
    In general, bigger filter capacitors will give you bigger current spikes, but shorter duration. The overal RMS will still work out about the same.
     
  20. Mihnea

    Mihnea New

    Joined:
    Aug 25, 2020
    Likes Received:
    1
    Trophy Points:
    3
    Location:
    Romania
    Hello! I think I have found out what is wrong with this simulation.. There is no Serial Resistance and no Parallel Capacitance for the Voltage Source. That is why the diodes figures were way too high...

    [​IMG]

    And I am not sure what values to use. I am using this transformer, hoping to get clean 18 V DC.. I'm starting to think that I might have picked the wrong transformer for the job.

    [​IMG]

    What do you think ?
     

Share This Page