1. Things you need to know about the new ‘Conversations’ PM system:

    a) DO NOT REPLY TO THE NOTIFICATION EMAIL! I get them, not the intended recipient. I get a lot of them and I do not want them! It is just a notification, log into the site and reply from there.

    b) To delete old conversations use the ‘Leave conversation’ option. This is just delete by another name.
    Dismiss Notice

TA2020 Tripath Amp

Discussion in 'd.i.y.' started by a.palfreyman, Nov 5, 2019.

  1. a.palfreyman

    a.palfreyman pfm Member

    Bought a mega-cheap TA2020 amp from ebay to add to some old Wharfedale Diamonds for TV duties. Had to correct some errors due to (amongst other things) the input caps (10uF electrolytics, replaced with 2µ2 films) and the charge-pump cap (1uF electrolytic) being the wrong orientation as well as the speaker terminals being marked wrong so that the outputs would be out of phase. Having 'scoped the outputs, with about 0.6Vp-p on input and the volume pot adjusted to give about 4Vp-p on the outputs, the output waveform is a sine-wave but with about 1V of HF "splatter" on it. Have checked the output filtering which is now to the tripath application spec after changing some 0.1uF for 0.22uF (3 per channel). Just wanted confirmation that despite the output filtering, this HF "noise" is normal.
  2. Arkless Electronics

    Arkless Electronics Trade: Amp design and repairs.

    Yes it's normal.
  3. Tony L

    Tony L Administrator

    Also worth noting that the published specs are highly misleading, e.g. it is often sold as a “20 Watt amp” or similar, when in reality it produces 5-10% distortion at that output. Well implemented and with a good PSU they can be a surprisingly nice little 5-6W amp, where you get 0.1% or better (I have an Amptastic knocking around), but if you need more than that 5-6W forget it. It sounds surprisingly good into the Tannoys and I used it for ages in the TV rig driving the LaScalas, it is a useful cheap spare amp for me as my speakers are such high efficiency.
  4. a.palfreyman

    a.palfreyman pfm Member

    Excellent, thanks guys.
    Yes, I noted the specs are "abysmal" near rated maximum but reasonable at about 3/4 max output and decent at about 1/2 max output. Won't be driving high volumes, just to try on TV duties so no massive expectations. The drivers in the LG TV are very "buzzy" on certain human voice frequencies.
  5. a.palfreyman

    a.palfreyman pfm Member

    I have run the little Wharfedale Diamond speakers via a CD player and my Obelisk 3 for a few hours before trying the T-amp. After listening to the T-amp, the sound is quite a bit darker i.e. recessed mid / upper mid and more rolled-off top. Is this typical of Tripath amps? Have read that the inductors can make or break these little amps. Anyone experimented with 10uH output inductor types?
  6. Mike Hanson

    Mike Hanson Really?

    In my experience, yes. I've generally found them to be very detailed at low volume, but "reserved". They've never sounded punchy to me. Great for a little system in a bedroom, intended for pleasant music in the evening, not played very loud.
  7. Arkless Electronics

    Arkless Electronics Trade: Amp design and repairs.

    As virtually always "type" or "brand" counts for diddly squat! Class D amps generically have/had an issue in that the output filter network is optimised for a certain load impedance and it will be either bright or dull in top end if not used with correct impedance speakers.
    I say "had" above as many are now using much higher switching frequencies and so the filter is operating way beyond human hearing... if it then rolls off at 50KHz rather than 100KHz due to a mismatch then it ain't a big deal!
  8. a.palfreyman

    a.palfreyman pfm Member

    Thanks chaps, good to have some real-world experience to back my own initial findings.
    Will give them some use over the w/end and see how I go. Strangely all the filter caps after the inductors were 0.1uF and I changed the three per channel to 0.22uF as per the tripath documentation to match 8 ohm. My own hearing begins to roll-off about 5k and is gone by 9-10k anyway!
    What did you say? ;)
  9. Julf

    Julf Facts are our friends

    Some newer designs (particularly the Hypex and Purifi ones by Bruno Putzeys) have the output filter circuitry be part of the feedback loop, dealing with the impedance issues.
    Arkless Electronics likes this.
  10. stevec67

    stevec67 pfm Member

    Dunno about the 2020 but the earlier T amps of 15 years ago were really sensitive to power supply. I had one on a wall wart and it was weedy, but with a lead acid battery to prop it up it was very convincing. Up there with most budget integrated amps at least. I never worked out whether this was a psrr issue or simply the power rails drooping under load and causing soft clipping.
  11. Arkless Electronics

    Arkless Electronics Trade: Amp design and repairs.

    Almost certainly the latter. Most class D amps are bridged and therefore have to deliver double the current. 8Ohms appears as 4 etc.
  12. a.palfreyman

    a.palfreyman pfm Member

    Found a problem with my Wharfedales at the weekend. These were active ones back from the early Eighties but the cases were battered after Uni so with my Dad'a help we made some cabinets out of 3/4 block-board of the same nominal dimensions. Cross-over consists of mid-bass run open i.e. directly across the input terminals whilst tweeter has 3R9 series resistor feeding 2u2 in series with 0.21mH with tweeter connected across inductor. 8kHz seems quite high for a cross-over (many early ones state cross-over of 5kHz on back of speaker) but main problem was tweeter was connected same phase as mid-bass where it should be reverse polarity. Also noticed that these are paper woofers (1109A) rather than the "pickled" poly ones more commonly seen (1103A) so wondered if the high cross-over was because of these mid-bass units?

Share This Page


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice