Distorted audio v4 Main Board?

An initial report from Mike (callsign not known) of distorted audio on a v4 main circuit board was tracked down to the audio amplifier.  The v4 main board is the new board shipping from HF Signals from around the beginning of June 2018.

Mike solved his problem by bypassing the audio amp and using an output amp as indicated bottom left in the photo above.   You can see that he has also incorporated a mic amp board in his prototype.

We will be watching out (or rather listening out) with interest to see if others are experiencing the same distortion issues from the audio amplifier section.  Hopefully this is just a one-off!

Reference

Follow up

Mike has now fixed the problem with crossover distortion.  It was a biasing issue in the audio final.

Raj VU2ZAP suggested lifting up D15 or 16 and disconnecting one end from the board. Then, solder the center and one end of a 47 or 100 Ohm preset between the two diodes.   Make sure the preset is at minimum. Power up and slowly increase the preset till distortion goes away, while also checking that  Q72  and Q73 don’t get hot when the volume is low.

 It is possible that this problem may reappear on other new v4 rigs.

Redoing the biassing as suggested by Raj VU2ZAP solved the issue.  Mike  took a 500 ohm pot and paralleled that with a 125 ohm resistor.   He  says, “The mod doesn’t look so good (LOL) but it works. One of the transistors became warm if the voltage difference due to the pot became too large. I adjusted this while feeling the temp and listening to a CW carrier”.   The setting point was where he observed low/no distortion AND low temperature.

The photo below shows how Mike fixed the issue.

Reference

17 Replies to “Distorted audio v4 Main Board?”

  1. I’m surprised by the position of the authors and sellers of the transceiver.
    despite the obvious shortcomings main board of version 4, they did not recognize the problem, do not comment on the situation, do not offer compensation for inconvenience, do not do not remove the main board from the sale.
    those who bought main board v4 did not even receive an apology for the inconvenience.
    why is that?

    1. I am 100% with you Andy.

      I am still trying to hear the “sparklingly” clear audio, the “big station sound, and all the other “embellishments” . The only sparking I got was the loud bang I got when the audio ic blew up. It scared the hell out of my granddaughter. I am researching case law in regards to any action. I gave my unit to a colleague in the US Department of Commerce.
      I strongly advise constructors to adhere to the FCC regulations, especially in regards to the spurious and or harmonic outputs,
      I assure you that a litigation exercise with the Federal Government is a very expensive undertaking, besides being stressfull.
      Finally “caveat emptor” what can you expect from a low cost kit in which the parts were sourced from the cheapest supplier some are counterfeit and the quality control needs improvement.
      As for meeting FCC SPECIFICATIONS please show me the evidence, and not from some hobbyist, rather from the NTIA Labs.

      Good Luck friend!
      Will

      1. Well, Wilhelm R Holzl JD Esq (retired) I think you’ve missed the point. The ubitx is sold as a relatively low cost component that forms the basis of an hf tranceiver but it is stressed that it is intended that users experiment , modify and ‘mess about with it. It’s not an expensive finished product but a vehicle to encourage people to get stuck in often at component level. Any shortcoming are admitted on the suppliers website with suggestions as to how they might be tackled. As for rushing off to a lawyer as soon as a component fails (a problem which is highlighted on the suppliers website ) I just find that really depressing and sad.

  2. Mike initially fixed the problem by substituting an off-board audio amp. Then, ‘Mike has now fixed the problem with crossover distortion. It was a biasing issue in the audio final.’

    What was his final fix?

  3. My oscilloscope clearly shows crossover distortion as well. Also I was looking into ways to get 5-6 dB more audio gain when I measured both sides of R51. The high side was 11.7 and collector side was 1.94 V. I was thinking of changing the collector resistor from the 4.7 K to 10K but with 0.69 Volts on the base it implies about a beta of 400 on the 2N3904! Anyway that may require two resistor changes. Not sure yet if I need it however but that transistor is NOT biased half way.

  4. Crossover distortion occurs whenever the amplifier load is lower than about 200 ohms. At 8 ohms it sounds like a torn speaker cone, at 60 ohms (headphones in series) it is quite tolerable but noticeable. A 8 ohm-to-200 ohm transformer (5.6:1 turns ratio) solves the problem if you have an efficient speaker. A 100 ohm resistor in series with series-connected ‘phones is perfect. Since I will employ a switchable CW filter I will bypass the existing amp and use one with a lower current drain.

  5. Please ignore my comment above. Audio is fine. A wiring fault on my part. Audio is low as some comment above but I feed it through small amp and all is good enough.

  6. My V4 board sounds terrible, and the behavior described up to now does seem to match the problem I hear. I decided to check this out with spice until I can get back to the hardware.
    A spice model shows that the audio will show crossover distortion if the bias voltage is not high enough, such as if the two diodes are BAT54 schottky’s. If replaced with 1N4148, the amplifier models very well. I will test this on my radio, but this may just be a bill-of-material problem.
    Besides the distortion, it tends to stop the lowest level signals, being buried in the crossover, making the gain appear to be too low most of the time.
    Do we know what the diodes are soldered to the board?

  7. Following up on my previous observations with spice modeling, I did not find the MMBD4148 diodes in the right package to fit the PCB, but I did find other silicon diodes with similar forward voltage that would fit into the footprint. The results in the audio were very good to my ears. I might still have a tinny speaker, but the sound was much improved and lower level signals were now copyable.
    The forward voltage on the original diodes in-circuit was 5.9V each (schottky-like), the new drop is 6.8V (silicon-like). I think this confirms that the problem is the diodes placed on the board are the wrong type but proper diodes will fix it. I hope someone else can try this out to confirm it works generally for V4 boards.

    1. Bob – I originally liked this soluton. But on closer examination, you are off by an order of magnitude. typical Si Vf = .7 V not 7 V. Schottky is like .2 – .4V.

  8. My new uBitx with V.4 board also has the low and distorted audio problem. Since this is a clear design (or bill of materials) flaw I feel it is up to the supplier of the uBitx to come up with a definitive solution and not us as clients.
    I wonder if and when we will see an elegant solution to this problem?

  9. I also have very low and distorted audio on my brand new ubitx v4. Tried 4 ohm 8 ohm speakets and a couple of headphone same on all. The audio is unreadable. Pretty depressing for something that has been getting such good press. Though ubitx is sold as an experimenters kit, mods and hacks would be expected to enhance basic OEM functionality, not allow for it. One should not expect to be required to self engineer and execute a fix for a prob that prevents basic operation especially when it is caused by an obvious design or component defect. HF should be engineering a fix and offering support for those to whom they have sold these defective v4 boards.

Leave a Reply

Your email address will not be published. Required fields are marked *