Arjen Helders TA2020 with horrible base line

[t04fr]

Hello,

I am new to this forum but I hope that somebody might be able to help me.

The Problem is that my TA2020 gives a horrible bass line in combination with my speakers. This does seem to be characteristic of this chip (or board - Arjen helder's MK II) as I can reproduce it with all my TA2020 amps. The problem is particularly present when I play Peter Fox's "Alles Neu". Here it sounds as if the speaker is over-driven even at low levels.

I have been building tree amps with Arjen Helder's TA2020 MK II boards and these have been working fine with my old DIY 2-way 4Ohm BR speakers (with relatively high sensitivity but no impedance correction).

Now I have been building two small BR speakers called MIVOC Mini Sats using the XAW 110 HC and the XGH 258 Alu. These are relatively high impedance. The x-over incorporates an impedance correction so that the impedance lies between 7 and 12 Ohm over the whole frequency range.

I have also modified the output filter where I have increased the inductance to 22µH and reduced the Cs slightly to suite the higher load impedance. This did not make any difference to the problem.

Has someone seen a similar behaviour. Does someone have an idea how to solve this problem (utilizing the TA2020) or have I just reached a limit of what is doable with the Tripath chip?

Cheers,
Frank

 

[whaleman]

Check your speaker output wiring, you may have one speaker out of phase (reversed polarity) in which case one speaker would cancel the other's bass response and vise versa.

 

[t04fr]

Thanks for the note.

Well it is not just a cancelation of the base signal. The frequency response is fine and I have cross checked it with some another AB amps where it also sounds just fine.
It is as if the supply voltage drops when the base starts on the TA2020.
I have buffered the board with an extra 30mF. The SMPS does not cope with much more than this.
I have also checked the supply voltage on a scope and it looks fine. There is very little Ripple and noise.

I believe that due to the relatively high impedance of the speakers the cut off frequency of the board's output filter is increased. Could it be that this causes some extra frequency bands (multiples of 44.1kHz ) to be present which again is demoduleted to a lower frequency band?

Cheers,
Frank

 

[kristleifur]

Have you tried NOT buffering the SMPS? I understand that some SMPSes really don't like having capacitance on their output.

 

[Pano]

Well this is strange. I've built and heard a lot of TA2020 amps and none had a bass problem. Quite the opposite, really, that's one of the strengths of this chip.
I would guess a power supply problem. Can you run it off a big battery to check?

The output filter only effects the top end response - unless you've done something really strange, so that is not the problem.

BTW, base and bass sound the same, but don't mean the same thing. "Bass" is the word you want.

 

[t04fr]

Thank you for your ideas and comments.

I will try to power the amp with a battery next (and also try to remember the difference between "bass" and "base").
I let you know about my progress.

Frank

 

[t04fr]

Hi there,

I did not even have to cross check with a battery supply.
When I connected a different set of speakers with the same old speaker cable the problem became apparent.
My Amp sits in a ICY box NAS1000 housing and has DIN sockets (to save space) to connect to the speakers. These seem to be corroded and generally make a bad contact with the speaker cable used on my new set of speakers.

So after I had spend about 100EUR on x-over parts to linearise the speaker impedance it turns out it's just the sockets for 0,50EUR each.

Thanks again for your help

Frank