TPA3116D2 Amp

[rhing]

Biased at 3V means that there will be 3 volts DC on the inputs to the IC, which is why input capacitors are required. I don't think you want your transformer volume control to be biased as well, keep the caps. Or replace them, if you really think it will make a difference. I don't think it will, unless they're very expensive, in which case the placebo effect will kick in and it will sound much, much better.

I had contacted Kevin Carter of K & K Audio, the U.S. supplier of Lundahl transformers regarding this very same issue. He said the following:

The key to using a small signal transformer at the input to do what you want to do without the coupling caps is the presence of absolutely equal DC bias voltages at the inputs connected to the input transformer secondaries. Any significant DC current flow (uA levels) due to different input terminal potentials will saturate the core of the transformer and result in very poor bass response.

So the question is whether the pair of input terminals ("+" and "-") for a given channel (right or left) are absolutely equally biased at 3V.

He later recommended that the caps could be left in, but some of the sonic benefits of having the transformers might be compromised, especially if they are not required. He also recommended either the Lundahl LL1592 or Lundahl LL690 input transformers--one per channel.

These are expensive placebos indeed.

 

[Lo_Tse]

I've ordered two of the blue YJ boards (to drive some two-way speakers with a Nanodigi as an active crossover). Would this PSU be good for powering the YJ amps (one for each board)?

24V 5A 120W Watt AC/DC Power Adapter for adapter connector 2.1 & 2.5 Charger PSU | eBay

Based on the requirement of the chip, 24 V is fine. I am using a 24 V brick for my red YJ board now and I enjoy the sound. I did a brief comparison between a 19 V brick and this 24 V SMPS, I could not detect much difference in my system. Please do not flame me just because I report what I experienced.

You can spend a few more bucks and get one of those SMPS which has adjustable output voltage. Typically, the voltage can be changed by + 10%. So if you do not like the sound at 24 V, turn it down by 10% to 21.6V and you should feel better.

Regards,

 

[KJA 2013]

... I suppose one could stick the Danzz blue board inside a humongous aluminum rackmount case with huge heatsink fins. Add some serious ballast weights inside to simulate the toroidal transformer. Connect some serious resistive 500 watt class electric cartridge heaters to the fins and put some cool lights and big galvanometer VU meters on the front panel. Make everything machined aluminum billet stock. Then ask an audiophile reviewer to test the thing (telling him it costs $12,000 list price). Then wait for the review to come back....

Lol! I've been tempted to do something similar....

 

[jean-paul]

I did too but it is still not finished. The case is more expensive than the amp. Even the feet are more expensive than the amp itself.

Meanwhile I replaced the blue input caps for Wima MKPs as YJ used different ones for L and R. For the power supply I use an Rcore 12 V AC with Kendeil 22.000 µF 40 V so no SMPS.

 

[Saturnus]

Based on the requirement of the chip, 24 V is fine. I am using a 24 V brick for my red YJ board now and I enjoy the sound. I did a brief comparison between a 19 V brick and this 24 V SMPS, I could not detect much difference in my system. Please do not flame me just because I report what I experienced.

You can spend a few more bucks and get one of those SMPS which has adjustable output voltage. Typically, the voltage can be changed by + 10%. So if you do not like the sound at 24 V, turn it down by 10% to 21.6V and you should feel better.

Regards,

Why would anyone flame you? If it's your impression then it's your impression. My impression is that it sounds a bit more stressed and starts to lose subtle details but I must point out that it's a very minute and only have a discernible effect when playing complex music. It doesn't seem to matter what the actual output level is though.

 

[xrk971]

I did too but it is still not finished. The case is more expensive than the amp. Even the feet are more expensive than the amp itself.

Meanwhile I replaced the blue input caps for Wima MKPs as YJ used different ones for L and R. For the power supply I use an Rcore 12 V AC with Kendeil 22.000 µF 40 V so no SMPS.

Please show pics when done. Feet costing more than amp is funny though.

 

[dbarker]

I believe it was Saturnus that proposed that the chip starts to sound stressed above 21V. Later, McDark supported this and said that he preferred a 12V power supply, regardless of type.

I think the preferred voltage for best sound quality is between 19v to 21v, and should not exceed 21v. People have said that a big 6 amp linear power supply sounds better. Many of us use SMPS laptop bricks at 19v - I think that is the defacto ps. Ultimate sound quality I would imagine will come from a battery. But then you are limited to either 12 volts or 24 volts.

My linear regulated 12v 1a plug type psu does the job just fine. At my usual listening levels it rarely consumes more than 2.2W according to wattmeter it constantly plugged in. You could probably be surprised how loud it sounds at say 5w. My speakers are alpair7.3 (88 dB afair). So these 21v 3a bricks are overkill for me as I found them less pleasant compared to battery or linear supply.

Based on the requirement of the chip, 24 V is fine. I am using a 24 V brick for my red YJ board now and I enjoy the sound. I did a brief comparison between a 19 V brick and this 24 V SMPS, I could not detect much difference in my system. Please do not flame me just because I report what I experienced.

You can spend a few more bucks and get one of those SMPS which has adjustable output voltage. Typically, the voltage can be changed by + 10%. So if you do not like the sound at 24 V, turn it down by 10% to 21.6V and you should feel better.

Thank you denizens of the internet. I'll get a couple of 19v laptop PSUs, and then I can experiment with higher voltages (with different PSUs) when I have everything working.

 

[BuildMeSomething]

My two YJ Blue boards arrived today... just visited evil-bay and bought two laptop bricks for power duties Just need to get an appropriate enclosure for these now, hmmm...

Might order a couple more of these boards for any future 4-way builds with the miniSharc at the fore...

P.S. miniSharc new firmware is out with Slave/Master I2S mode

 

[mightym]

My Mini DSP arrived today.

I have to say I'm impressed with the rapidity that all the things I've ordered from China have arrived. My YJ order arrived in podunk eastern Ok in 10 days, I need to check but I think the Mini DSP was a little bit quicker..

I bought a couple of power supplies from Amazon that are advertised as "Univerasal", they both came with 10 extra adaptors to fit the 2.1mm connector, and are adjustable from 12 to 24V, and have a USB charging port to power the Mini DSP....

Now I need to finalize my order at PE so I can butcher some wood and make some amps!

Here's the link for the power supply:Amazon.com: New Universal Charger AC Power Adapter Laptop Notebook USB/10 Tips 12V-24V 90 Watt Max: Computers & Accessories

I have no commercial interest here, just trying to share.....

John

 

[luxury54]

a question for the experts,
can this class D kit be used also as a headphone amplifier output? in affirmative case is there any need to place some power resistors at the output for impedance matching? I'm asking this because I remember that when trying this scenario with a class T tripath, I needed to make a custom wired headphone because there it couldn't be used any ordinary headphone with a common GND so I had to wire the 2 channels separately with it's own GND wire each and with coresponding jacks so that the GND was separate for both channels.
Is this also the case with class D amps so that you cannot have common ground ? thanks

 

[HaroldHill]

a question for the experts,
can this class D kit be used also as a headphone amplifier output? in affirmative case is there any need to place some power resistors at the output for impedance matching? I'm asking this because I remember that when trying this scenario with a class T tripath, I needed to make a custom wired headphone because there it couldn't be used any ordinary headphone with a common GND so I had to wire the 2 channels separately with it's own GND wire each and with coresponding jacks so that the GND was separate for both channels.
Is this also the case with class D amps so that you cannot have common ground ? thanks

Yes each channel will need its own ground. And you may want to lower the gain with a resistor or resistor network so you don't blow up your ears (and possibly headphones) when a small tweak in the volume pot dramatically raises the volume.

 

[jean-paul]

I think you're better off with a dedicated headphone amplifier like TPA6120A2. These TPA3116 simply are designed for speakers.

 

[Saturnus]

I think you're better off with a dedicated headphone amplifier like TPA6120A2. These TPA3116 simply are designed for speakers.

Especially given that the normally (like most class D amps) have an LC output filter which is designed for a specific load impedance which is totally dissimilar to a typical headphone.

 

[HaroldHill]

I think you're better off with a dedicated headphone amplifier like TPA6120A2. These TPA3116 simply are designed for speakers.

Ah, quite true. I'm using a Sure TPA3110D2 to drive my headphones, but they are notoriously inefficient orthodynamic cans and driving them off the speaker taps of an amp is quite common.

 

[Saturnus]

Ah, quite true. I'm using a Sure TPA3110D2 to drive my headphones, but they are notoriously inefficient orthodynamic cans and driving them off the speaker taps of an amp is quite common.

The Sure TPA3110 doesn't have an output filter so it's possible on that. On the TPA3116 amps however, I wouldn't recommend it unless it's a filterfree version (which I haven't seen yet).

As noted above however, please note that all speaker outputs on a BTL amp are hot. So the headphones must have individual grounds for each side.

 

[HaroldHill]

Thanks for that info, Saturnus. I just ordered a YJ TPA3116 blue board. What might I expect sound wise if I were to use it to drive headphones with an impedance of about 40 ohms? Might there be other negative consequences?

 

[Saturnus]

I suspect it will sound quite ghastly in the whole upper range but especially at around 2-4Khz (where the ear is most sensitive) the sound will be quite horrific.

 

[HaroldHill]

Well, the parts are all in the mail, so I'll give it a try. Worst case scenario, I'll have a new amp for my neglected bookshelf speakers! Thanks for the insight, Saturnus.

 

[xrk971]

Saturnus,
Why do you think it will sound ghastly as a headphone amp? If one were to match the impedance with say a resistor network, would not the response, HD, and SNR values of the amp still hold true?

Haroldhill,
I have done some research into headphone amps and it looks like the O2 by NwAvGuy - sold prebuilt by JDS for $100 looks tough to beat based on specs. I think it can be DIY built for $40.
http://www.jdslabs.com/products/35/objective2-headphone-amplifier/

 

[Doctor J]

I know this was mentioned earlier in this thread, but can someone explain why at roughly 11 and 3 o'clock both my stock red and blue YJ boards have noticeable hiss, but in other positions (including full volume) it sounds dead silent? Are most people using the TPA 3116 with a dedicated preamp?