TPA3250 somebody is listening?

FX-Audio has introduced an amp based on the TPA 3250 chip. Under $70 got to be kidding right? Following the diy efforts on the TPA 32xx chips I think our Chinese friends presented this particular amp with the note,[ preamp uses OP amp NE 5532 : Replaceable design suitable for DIY,] for our benefit .- Maybe-. Has anyone checked this out? thanks
 
Fanatic
Joined 2009
Paid Member
Very good price, and it`s including a 24v 4a psu. Caps are 35v it seems.
They claim 35w 8ohm and 70w 4ohm at 0.01% non-specified distortion?
Is that a voltage regulator/smoothing circuit near the power socket?

Might be a tpa3116 killer. Tempted to replace the one in the children's room, it will have to wait until January/February sometime.
 
Last edited:
Fanatic
Joined 2009
Paid Member
In short: The TPA3250 is better than the TPA3116.

Is it much better? That depends on your application.
It is slightly more powerful, you can get probably 2x25w nearly clean power (guesstimate somewhere between 0.01-0.001% distortion) in 8 ohm with the included PSU.
Whereas the TPA3116 also gives 2x25w in 8 ohms but at 21V max, and 10% distortion.

It seems to be a slightly better designed PCB, there may be some power conditioning going on (not tested), there is a pre-amp stage containing the well known NE5532, seemingly well optimized for modern use at 0.775mv specified input voltage.
Many TPA3116 seem to be set to very aggressive gain settings making it more noisy than it has to be, and there is often no pre-stage on the stuff sold cheap on ebay.
 
Last edited:
TPA3250 is from the top-of-the-range series. As I would expect from what I see in the advertisement, cooling may be a problem. A rather small heatsink enclosed in a small cabinet without vents - technically not optimal. There is this general perception that class D amplifiers do not have cooling issues. This is not true, they are just less than for class AB amplifiers.

As someone wrote about the TPA3116D2: it may make noise ("hiss"). I have such a TPA3116D2 amplifier board and initially a 50K potentiometer was used. I shunted the TPA3116D2 inputs (at the output from the potentiometer) with 4K7 resistors and the "hiss" disappeared.
 

ICG

Disabled Account
Joined 2007
If the delivered amp is identical with the ebay photos, it got a few very nice features, the preamp OPs (the NE5532) are socketed, that's remarkable. You can easily replace them without any soldering, experiment with different/better OPs etc. The output inductances are air cores, they don't get saturated and start harsh distortions like ferrite ones do if they are too small. Also a nice-to-have is an error-/protection LED.

I haven't heard it but it looks quite promising. On high volumes it will likely profit from a bigger power supply though.
 
Or rather, if you require more undistorted power get the TPA3251 or the TPA3255.

I do not know about the performance of the supplied psu, but 24v 4a may be a very balanced choice for this chip. You get "good enough" performance up to 25w/channel in 8 ohm, and 40-45w/channel in 4 ohm.
Otherwise the distortion starts setting in just over 30-35w/channel in 8 ohms and 60-70w/channel 4 ohm anyway, regardless of power supply and voltage. Need more power? Get something else.

Edit:
The numbers are a bit better at 30V supply,still very good for 4ohm loads, but the difference is not humongous, and 24V supplies are probably cheaper.

Edit2:
Just ordered two of these.... Splurged 4$ on shipping to get it just a little bit faster.
 
Last edited:

ICG

Disabled Account
Joined 2007
Otherwise the distortion starts setting in just over 30-35w/channel in 8 ohms and 60-70w/channel 4 ohm anyway, regardless of power supply and voltage. Need more power? Get something else.

While that (distortion) is true, I honestly doubt you'll get 2x70=140W (without any losses!) out of a 96W power-brick. And why add the distortion from the PS and the amp if you can avoid it? Sure, you'll rarely use that much power but that's saving at the wrong end IMO.
 

ICG

Disabled Account
Joined 2007
That's still rather tight for such a power brick, remember, a lot of speakers got quite the dips in the impedance at some point which means it will draw more power with some 'luck'. That will not burn up the power brick but that's saving on the wrong end, why waste potential? These PS aren't very expensive. If you say 'that's enough for me', that's fine. I, personally, would want to use the full potential and I don't like driving PS over the edge, a beefier PS usually lives longer. I like relyability.
 
I agree.
But for my intended application I doubt I will ever see 20W on peaks in 8 ohms. And was considering doing this "dual mono" with two amps, each amp uses one channel for the 8ohm load, and one channel for the 4 ohm load. The 24V PSU is free, and it gives me more power than I need from this amp.

If I'm not happy then there are other places they are needed. Need a replacement for my 70's Nikko amp while it gets refurbished, need some new caps, switches and potmeters (both volume and channel panning is worn). Basic cleaning does not work any more.

Edit:
If I'm not happy with the supplied power unit, I got 4x12V SLA sitting around doing nothing that will work great.
 
Last edited:

TNT

Member
Joined 2003
Paid Member
I agree.
But for my intended application I doubt I will ever see 20W on peaks in 8 ohms. And was considering doing this "dual mono" with two amps, each amp uses one channel for the 8ohm load, and one channel for the 4 ohm load. The 24V PSU is free, and it gives me more power than I need from this amp.

If I'm not happy then there are other places they are needed. Need a replacement for my 70's Nikko amp while it gets refurbished, need some new caps, switches and potmeters (both volume and channel panning is worn). Basic cleaning does not work any more.

Edit:
If I'm not happy with the supplied power unit, I got 4x12V SLA sitting around doing nothing that will work great.

Please report back on your findings once delivered and tested.

//
 
The 325X distortion number seems pretty good but does anyone know its harmonic profile? What about its IMD numbers? To me, these class D amps are not really meant for home hifi. Their application should be limited to portable sound reproduction. I will admit that when you start getting under 500 ppm distortion, you're getting into the 'acceptable' range but like in all things, the devil is in the details.

FauxFrench
I've heard lowering the input impedance on these amps helps with the noise. I've been looking at some cheap 3116D2 boards that are all SMD. Their ~25x37mm size allows you to mount a legacy CPU heatsink of any size. The chip is the tallest part on the board. Have a look here.
 
Thats a nice little amp, even if they use "low cost" dcdc converter parts and do this strange opamp swap-mount thingie. I first thought, there's a differential linedriver on the middle right (pcb) but this seems to be some kind of microcontroller?

@dellama, the TPA324X/5X series does <500ppm with ease at a wide range of listening levels.
 
doctormord
Yes, .01% is the specification claim for that chip but depending on the mix of the harmonic profiles, this can sound less than great. It's certainly inferior to the single digit ppm distortion, low noise floor and organized harmonic profile efforts achieved with class AB and composite amps. The other problem I foresee with ALL class D chips is the high frequency non-linearities (>6K) which are bound to diffuse the ambience and soundstage of playback. I haven't heard enough of these to make any final judgment on them but so far, I'm inclined to reject class D for anything other than portable sound. I just ordered a pair of those mini boards to play with. I suspect you'll be interested in their development if you haven't bought a pair for yourself already. Did you note my comment on the main TPA3116 thread about the meager capacitance on this mini board?
 
Last edited: