New TK2050 board

[sendler]

Originally Posted by alkasar

very interesting board. I am tempted

Coming from gainclones, let me share a concern, maybe a bit off-topic, about smps.
Many people here seem to prefer smps.
Linear PSU (transformer + rectifier + caps) have large caps, to filter and also provide an energy tank to supply power instantaneously when needed. Frequently 50.000 µF or more for powerfull amps. This provides the amp "authority" to drive the speakers.
SMPS do not have this large tank. Neither do the amp boards here.

I just don't get how SMPS can compare to linear supplies with large caps ?

You will get it once you try an SMPS. I made the same mistake myself for years, stubbornly adhering to the incorrect theoretical hype against switch mode power supplies to the point of never even trying one. You can never know what something will sound like based on speculation. The only way to know is to actually compare. The cost is less than a linear supply at the give away price they are being offered, and the sound is much better. More open and faster. You can always add more tank caps. That will be my next test but the stock SMPS already sounds much better. When I first got the Sure 2X100, I hooked it up to one leg of my big dual mono active filtered linear supply that I had been using for all of my AB amps. Dual 24-0-24 150va transformers, Schotky diodes though I later switched to MUR860 in my 8 channels of TDA7250 amps as they sound a little smoother. 220uf of smoothing precedes the Linear Tech voltage regulators which drop 4v down to 29v rails. I added these as a last step and the active regulation does take away a slight edge from the sound. This feeds 40,000 uf of Panasonic filter caps, bypassed with stacked film. I had Pi filter series coils in my supplies between the two big caps but comparative listening revealed that there was no advantage and the slight resistance was in fact vaguely rounding the leading edge of kick drum so I jumped them. Don't bother with Pi filters. They are a good way of wasting most of the upstream cap. Heavy wire, star ground, to the amps. I have used the same power supply to evaluate all of my amps over the last 12 years and a $40 SMPS came along off the shelf and killed it.

[tsharp]

i found what looks like an exact replica of the meanwell 27v 350w power supply on the bay, its unbranded but carries the same specs (according to the listing anyway) but at half the price of the meanwell. has anyone purchased or had any experience with these smps?

350W 27V DC 13A Regulated Switching Power Supply [K007] - eBay (item 170431332887 end time Jul-11-10 00:15:10 PDT)

[kc7wdg]

Ordered a v1.0B with the CTR pot as I will be using with polk monitor 5jr+'s.
Cannot wait to get it.

[AP2]

Originally Posted by alkasar

very interesting board. I am tempted

Coming from gainclones, let me share a concern, maybe a bit off-topic, about smps.
Many people here seem to prefer smps.
Linear PSU (transformer + rectifier + caps) have large caps, to filter and also provide an energy tank to supply power instantaneously when needed. Frequently 50.000 µF or more for powerfull amps. This provides the amp "authority" to drive the speakers.
SMPS do not have this large tank. Neither do the amp boards here.

I just don't get how SMPS can compare to linear supplies with large caps ?


ps: great to see the board designer respond and take into account the users comments

Hi,
You're absolutely right about the function of the large capacitor on linear PSU, this is the soul of the amplifier (if amplifier has low output impedance).
This is the difference between generic SMPS that disregard of this important feature (besides other problems, noise, and the introduction of thd especially at low frequencies).
if the SMPS is regulated, then there is another problem (not able to respond quickly to requests for current) is useless to try it with constant resistive load. example. 68V 8Amp continuous load, the voltage drops to 2V But with 8amp burst (60Hz) 10V voltage decrease. this decrease is sufficient to deform the seal of the amplifier.
on regulated SMPS, audio really is not easy to solve this behavior.

regards

[alkasar]

Thanks Sendler. I actually tried an smps in a small ClassD system with TDA8920BTH amp from connexelectronic. I must say I was not impressed by the tda8920 amp. Cristi boards are excellent, the sound is ok, but in y opinion, not at the same level as Peter Daniel's LM4780 gainclone. Clearly bass are missing. I though this is linked to smps.
I have not done direct comparison smps vs linear on the same amp.

I'll give classD a second chance with this board, given the wonderfull comments maybe with the new mono amp board with ST517B... what will be the recomended voltage of this new amp ?

[hifimediy]

Originally Posted by alkasar

Thanks Sendler. I actually tried an smps in a small ClassD system with TDA8920BTH amp from connexelectronic. I must say I was not impressed by the tda8920 amp. Cristi boards are excellent, the sound is ok, but in y opinion, not at the same level as Peter Daniel's LM4780 gainclone. Clearly bass are missing. I though this is linked to smps.
I have not done direct comparison smps vs linear on the same amp.

I'll give classD a second chance with this board, given the wonderfull comments maybe with the new mono amp board with ST517B... what will be the recomended voltage of this new amp ?

Large range of voltage can be used for ST517B(10~56V from datasheet), but we suggest use a voltage no lower than 30V DC, to achieve larger power output. So, may be 30~50V DC is the range we suggest.

[abraxalito]

Originally Posted by alkasar

I just don't get how SMPS can compare to linear supplies with large caps ?

The size of the cap depends on how long it has to hold its charge for. With a linear supply that's around 8mS before it gets recharged because its running at mains frequency. With a SMPSU running at say 40kHz, then the output cap only has to hold up the supply for 25uS before it gets recharged. So the values of output cap can be more than 200X smaller. Does that help in understanding the advantage?

[teamacc]

Do you have any estimations on when the sta517 board will be ready?
thanks

[alkasar]

Originally Posted by abraxalito

The size of the cap depends on how long it has to hold its charge for. With a linear supply that's around 8mS before it gets recharged because its running at mains frequency. With a SMPSU running at say 40kHz, then the output cap only has to hold up the supply for 25uS before it gets recharged. So the values of output cap can be more than 200X smaller. Does that help in understanding the advantage?

hey... that's the theory answer I could not find anywhere ! thanks, very clear.

[Don Jerry]

Originally Posted by alkasar

hey... that's the theory answer I could not find anywhere ! thanks, very clear.

Finally tried my board, works just fine!
I used a couple of opportunely modified 350W ATX smps in series to obtain 27Vdc.
I checked the dc offset and was 99% perfect, I had to make a very little adjustment. Now the dc offset is = 0 (very good!).
The heatsink screws were a bit loose and I tighten them. I planned to add 2 extra screws and a little washer for each screw (I think that's better to don't risk to bend the board).
Now I've to buy a cabinet and to do some other modifications (bridge removal + Mundorf M-Cap input capacitors).
The only thing that I don't like on the board is that there's no room for replacing the potentiometer with a better (and usually bigger) one, like for example an Alps blue. No problem for removing the input signal connector to obtain more room, but how to do with the dc offset trimmer?
Anyway my opinion is Very good (8/10), thanks to hifimediy and Nick122147!!!