New DIY Altoids Tin Amp - The Class A "Wintergreen Handwarmer" by xrk971

  1. xrk971
    You should read the thread where that circuit was developed:

    It's too much to discuss what each one does, but this topology is nothing new. Lookup single ended source follower amplifier.

    Rl is to provide a nominal load to the output so that the capacitors are driving something otherwise when the phones are disconnected it is a pure capacitive load. RL also helps to bleed off the caps so that after initial turn on transient, there is no "pop" when you insert the headphone jack.
  2. xrk971
    The multi JFET output circuit I posted above was a design by another member and the row of JFETs on the bottom are the CCS's. Same thing can be done with a MOSFET. Basically, a MOSFET can be biased to flow current and it sets the class A bias current by setting its gate voltage. You can see this in many single ended class A designs. You need to read up on constant current sources using a MOSFET.

    Here is a design for a tube headamp that is single ended class A with a MOSFET CCS that is controlled by a TLM431 programmable zener. You can replace the tube front end with the BF862 front end that I am using and it would work quite well. It looks very similar doesn't it.

    This is a Sodacose WTF amp design:

  3. funch
    I fired one of my amps up today, and can confirm that those little heat sinks do work. I can only hold my finger on one for about two seconds.
  4. xrk971
    That means your bias is running real hot.
  5. funch
    I should mention that it was driving my DIY 'phones with T50RP planar drivers, which take a good amount of current. At idle, the 'sinks were much cooler.
  6. xrk971
    That's interesting because SE Class A amps should burn about same amount of heat whether playing music or not. What are T50RP planars?
  7. funch
    T50RP's are Fostex headphones. A few years ago several Head-Fi'ers started either modifying them or taking them apart and using the planar drivers to make their own headphones.

    I find that my version is a bit harder to drive than even the Audeze LCD planars.
    Last edited: Jun 17, 2017
  8. macky112
  9. macky112
    Hi X,

    Does it make sense for this analogy: a Honda Civic (as an ideal but less powerful amp) and a Ferrari (as an ideal more powerful amp), and both are cruising on the freeway at 60 mph (as my comfortable listening level at X db) then when sub bass are called for (as the car acceleration), the Ferrari amp though playing at same listening level would produce better sub bass?
  10. xrk971
    I think the analogy is partially right. You are describing dynamic headroom as being important to hearing bass dynamics. The "Ferrari" can also maintain a higher speed which requires more power. So lower impedance phones can be played louder with less strain.
    macky112 likes this.
  11. macky112
    hi @xrk971 is there any scenario that the AMP may damage a DAP? I tried changing the rail caps and I think my DAP is damaged?!
  12. xrk971
    If you connect the DAP to the input of the amp it goes to a capacitively coupled input to a JFET gate. There is zero possibility of any excessive DC voltage going back from the JFET to DAP to cause damage. Even if you connect the amp output to the DAP output, the capacitors AC couple to prevent any DC voltage from damaging the output stage of your DAP.

    Describe symptoms of your DAP and why you think amp may have damaged it - it's highly unlikely except in rare event that you have a bad capacitor. So measure the D.C. Volts at the input connnected to your DAP - should be 0v.
  13. macky112
    I thought about it, and it may have happened during my 1st build attempt, it was the scenario with bad soldering on the resistors that caused really weird DMM readings, and without knowing better, during troubleshooting, I tried 2 DAPs (Benjie S5 and Xduoo X02) with my first build, and immediately the S5 went into a reboot cycle (and at the time, I didnt know why it did that). After the S5 stuck in a reboot cycle, I checked the X02, it booted up fine and was able to navigate the menus, so I tucked it aside without listening to it, and gave up on troubleshooting my first build.

    yesterday, I replaced rail caps on my 2nd build to 2200uF and wanted to see if I can hear any difference in bass performance (and now I know better, after reading through the Objective2 thread that a non-functioning Amp can damage both DAP and headphones, so I pulled out my cheapest DAP, Xduoo X02), and that's when I found out one channel of the X02 was dead, and the working right channel has random static sound.

    So during troubleshooting my first build and trying to get some sound through it, that was when it may have happened, probably some AC/DC current gotten into the DAPs (due to short or open circuit somewhere on the amp?) and fried something.
  14. macky112
    Hi @xrk971 to reduce size, what would be the lowest voltage rating caps I can use for input caps and output caps for a battery build?
  15. xrk971
    If you look at the D.C. Levels at those caps then add the max voltage swing.

    For input probably not more than +/-2v plus 1.5vdc so a 5v could theoretically work. For output D.C. is about 7v. Max output swing is about +/-5v so 12v would be minimum for output. But for lower distortion you don't want to have a cap too close to limit. I use 35v Elna Silmic and film cap is 50v rated.

Share This Page