Today I've connected mine "V3.1" to a 12V supply. It behaves well enough for the purpose I bought it for. I'll use it in quite a small room so don't need all that much volume.
So far the chip is not getting too hot.
I will start a dispute on Aliexpress anyway. They stated a specific chip in description and I got something entirely else.
BTW: Mine didn't have any warranty or antitamper stickers to begin with (which is what encouraged me to open it in the first place)
So far the chip is not getting too hot.
I will start a dispute on Aliexpress anyway. They stated a specific chip in description and I got something entirely else.
BTW: Mine didn't have any warranty or antitamper stickers to begin with (which is what encouraged me to open it in the first place)
The open c45c46 has been noticed by several buyers and mentioned in multiple threads multiple times. SMSL is selling you SA36aPRO, changes to specifications and parts can occur without prior notice, you know, it is the beauty of branding we like. You read a review of Macbookpro and screen is fabulous, then you buy same typenumber and screen is flawed, you go to Apple service point and they say, yes now it is eg Samsung screen, it used to be eg LG screen and the new screen is up to specJust useless for your purpose. etc etc
Thanks - I missed all those posts somehow in the billion posts on the 3116 thread
Do you think it is worth it to populate those spots with good caps?
I think it is easier not to have to remove the little caps when you want to put oscons or fpcaps there anyway People that did mostly say they hear improvement, even the ones that didn't like the pro compared to tripath, after filling the missing caps they like 3118 better than tripath, so they seem to be listening to it, doctormord will say they do not listen blindly
If all SMSL come with warranty or hologram sticker and the 7492 doesn't there is a second odd thing about it, but could still be same reason, the printer could be under few meter of mud after the landslide for example, or another cost cut, they buy in dollars I think and sell in yuan mainly, could be they had worse year than expected...
People that did mostly say they hear improvement, even the ones that didn't like the pro compared to tripath, after filling the missing caps they like 3118 better than tripath, so they seem to be listening to it, doctormord will say they do not listen blindly If all SMSL come with warranty or hologram sticker and the 7492 doesn't there is a second odd thing about it, but could still be same reason, the printer could be under few meter of mud after the landslide for example, or another cost cut, they buy in dollars I think and sell in yuan mainly, could be they had worse year than expected...
Taking Irribeo's point that a product line may have substitutions or small changes that don't represent a decline in specification, substitution of the core component is an entirely different matter. That's like ordering a Intel Mac that is advertised as such, and delivering an AMD Mac. Both crunch numbers, both have legitimacy in the market, but you buy an Intel product you expect an Intel product.
This forum seems built around the very discriminating and dedicated hobbyist, so when people make recommendations on products here, when products are advertised with specific specs in the marketplace, everyone here I think expects that's what is delivered.
Hopefully its not something tragic like the landslide. But obviously they can still manufacture product and fulfill the supply chain, so they should be able to communicate a spec change as fundamental as the core amp chip, unless they don't want to because they know what they are doing....the initial post was to get some determination on whether this is a counterfeit product - it's that different.
Anyway that's my rant on a yank's expectations on getting what has been advertised.
This forum seems built around the very discriminating and dedicated hobbyist, so when people make recommendations on products here, when products are advertised with specific specs in the marketplace, everyone here I think expects that's what is delivered.
Hopefully its not something tragic like the landslide. But obviously they can still manufacture product and fulfill the supply chain, so they should be able to communicate a spec change as fundamental as the core amp chip, unless they don't want to because they know what they are doing....the initial post was to get some determination on whether this is a counterfeit product - it's that different.
Anyway that's my rant on a yank's expectations on getting what has been advertised.
Some people buy a computer for the display, it is the only reason they buy, the display. Then Apple changes supplier for that display without communicating the change, without writing to all magazines that tested the old spec computer, without changing the typenumber. When you compare the displays color temperatures differ, towards edges new screen also has deformations. These things happen every day. Apple is still selling the Macbook because they decide what the macbook is. Now when they sell something as quad core computer and deliver a dualcore, like HP does/did, then consumers get upset when 100 out of 1000000 sold find out and they refund the 100, case closed.
Wow, I did not hear about the quad vs dual core debacle. Glad I have the 3118 - that was my first TPA, the 3118 which I did as a flip chip with legs up so I could solder dead bug style to the pins. It's a great sounding amp and it's too bad SMSL is changing the specification without telling us. It was a nice board with a lot of the fixin's already applied. One question I have is does the 7492 substitution sound as good in any event? Could you notice a difference in SQ?
HP Pavilion 17 Cheap but typo?/mistake in all official HP material too, not just resellers advertising. Don't know, was december 2015 when discovered I believe.
Here power and specs could even be better for 7492, at least for most speakers used, still it is sloppy work by resellers, they didn't pay attention and smsl itself only sells to chinese agents I think, those might have missed it or covered it up because demand for sa36apro was good Still it looks as if sa36a was intended name, no pro added on pcb.
Cheap but typo?/mistake in all official HP material too, not just resellers advertising. Don't know, was december 2015 when discovered I believe.Here power and specs could even be better for 7492, at least for most speakers used, still it is sloppy work by resellers, they didn't pay attention and smsl itself only sells to chinese agents I think, those might have missed it or covered it up because demand for sa36apro was good
Still it looks as if sa36a was intended name, no pro added on pcb.This is an amp, not a computer. It has only one functional purpose, to produce sound - unlike a computer which has many functional purposes besides supplying an image (although I agree it's a common crappy practice that frustrates a lot of people) the comparison doesn't fairly represent the nature of this substitution.
Whatever the legal implications would or would not be, it's incompetence at least, deceptive at worst - and it should be called out, which is what we are doing here.
Whatever the legal implications would or would not be, it's incompetence at least, deceptive at worst - and it should be called out, which is what we are doing here.
Mod suggestions?
This is, after doing research, exactly why I bought it.
Looking at the 2 layouts (v2.2-v2.2) vs (v3.1-v3.2), I'm wondering what mods one would make on the TDA configuration to more closely match the quality of the TPA configuration.
This is, after doing research, exactly why I bought it.
Looking at the 2 layouts (v2.2-v2.2) vs (v3.1-v3.2), I'm wondering what mods one would make on the TDA configuration to more closely match the quality of the TPA configuration.
Last edited:
you can't but do you need to ? compare datasheet to see if there is a difference you find troublesome or get 3118 and compare, I think xrk likes ST in general, for me the ST's have trouble driving 1 or 2 or 3 ohm impedances, the TI has less or no trouble doing that, again what most people likely find is they dislike the power on pop and off pop off 3118 and haleluja, gone with 7492 oh P version is aimed at lower voltages, might have less heat issues the other had, ST have gigantic heat issues with low impedance speakers I have, I mean classA runs cooler, they just can't drive low impedances imo.
http://www.st.com/st-web-ui/static/active/en/resource/technical/document/datasheet/CD00211320.pdf
oh P version is aimed at lower voltages, might have less heat issues the other had, ST have gigantic heat issues with low impedance speakers I have, I mean classA runs cooler, they just can't drive low impedances imo.http://www.st.com/st-web-ui/static/active/en/resource/technical/document/datasheet/CD00211320.pdf
In my experience so far (bought several types of amp boards, some times the same amp board on several occasions), you never know what *exactly* you will get - particularly if you're buying on ebay (or other such sources, e.g. alibaba etc.). Different caps, even different chips, and usually slightly different boards - all quite "normal" for buying from ebay or other "usual sources".
But, if SA amps are advertised as having a particular chip, it is logical to expect to receive the amp with the specified chip, so we're not talking about the usual ebay roulette...
For anything serious you simply cannot rely on the Far East suppliers/manufacturers. All of the things we've seen here (different chip!!! no kidding?) are the usual "Far East" ways of "doing business".
I'm not denigrating the Chinese makers (I have lots of their amp modules with which I'm quite happy!), I'm just saying that you can't expect reliability and repeatability unless you're willing to spend more money.
Most of the manufacturers make their amp modules using whatever parts they can get the cheapest on that particular day...
Exceptions (better made, better quality control) usually cost more - for example, Sure boards... (although they also have issues sometimes, like mislabelled speaker polarity, etc...).
Rgds,
Denis
No such reliability with western brands, only in your mind, you don't look for different pcb's in your Nad or B&O or Harman or anything, they change during production as well, just the chinese get the blame, it is normal for any brand, any unfortunately or fortunately, things sometimes improve, SMSL fixed the pop issue their first 36pro had and they fixed the issue that when buyers power up without speakers connected or use speaker switch, because of pop lol, the ampboards get defective, major improvements for 99% of buyers
*** I think they fixed the no load oscillation 3118 can have in 7492 version with 4 diodes on each side of 7492, I don't know, but think that is what those diodes are for. Decoupling is very very close to chippins, big Panasonic is very close in comparison to old pro and low esr in comparison again. parts look about the same, inputcaps a little lower value but 7492 has more stable higher inputimpedance than 3118, easier on potchoice for volume, most likely better in this version than in old too. it is different, parts are different and surely with low impractical impedances this new amp won't work well, but 99% of buyers will be glad I think, but I could be wrong
AMEN.
irribeo, your attention to this post is appreciated, but the sum total of your comments so far seems to be "who cares". This is an enthusiast forum isn't it? There is lots of information and opinions on this forum about the different qualities of these chips, and how their implementations on the board affect the output. So it seems to me that those sharing info here, or coming here for information would care and would want to know that this product is being specifically advertised as having the tpa3118 amp and is being shipped with the tda7492p amp. Why bother with this forum if it doesn't matter?
Was that a consistent problem with the v2.1-2.2 config? I'm not seeing that as more than incidental. Most people seem very happy and impressed with that version.
Are you saying that the TDA config on this particular product is better than the its former TPA config? If you are please say so, we can work from that.
Just saw this post, that is exactly the kind of comparison I was looking for and helps answer my initial question - thank you. This is helpful.
Are you saying that the TDA config on this particular product is better than the its former TPA config? If you are please say so, we can work from that.
*** I think they fixed the no load oscillation 3118 can have in 7492 version with 4 diodes on each side of 7492, I don't know, but think that is what those diodes are for. Decoupling is very very close to chippins, big Panasonic is very close in comparison to old pro and low esr in comparison again. parts look about the same, inputcaps a little lower value but 7492 has more stable higher inputimpedance than 3118, easier on potchoice for volume, most likely better in this version than in old too. it is different, parts are different and surely with low impractical impedances this new amp won't work well, but 99% of buyers will be glad I think, but I could be wrong
Just saw this post, that is exactly the kind of comparison I was looking for and helps answer my initial question - thank you. This is helpful.
Last edited:
(3.2 seems Panasonic FC, 3.1 seems chinese to me)
All sa36apro with 3118 have a pop when you power on and power off. All sa36apro with 3118 can burn components in output when powered without speakers connected, like you do when you use a speakerswitch to not hear the pop sound. All 3118 in higher gainseting especially, like smsl offers, have inputimpedance related volumecontrolproblems. Fixing this last requires buffer opamp imo, everything is fixable but requires additional parts. So could have been an updated 3118, but needed redesign for that, beyond tiny version difference you see for previous pro's. You did notice the price hike for 3118pro during 3118pro lifetime/productiontime, dollar yuan is difficult I guess, they were spending too much money and needed to spend even more to raise component count further. They also stopped production of 3116 sa60 so it seems, inputside pro2 3118 would have been much like sa60, if they felt like improving the 3118pro volume controll, diodes next to 7492 are like what was needed for 3118 too I guess, some additional ,4 components? I believe for doctormord pop fix, or other popfix, but they do not mount wires components above pcbs like we do so real new pcb. They could have and maybe they will, maybe not for all "problems", but diodes next to 7492 to me indicate that was really problem they experienced and hurt by, burning output without speakers, resellers claiming money back or free new amps, which possibly burn again when no speakers connected
All sa36apro with 3118 have a pop when you power on and power off. All sa36apro with 3118 can burn components in output when powered without speakers connected, like you do when you use a speakerswitch to not hear the pop sound. All 3118 in higher gainseting especially, like smsl offers, have inputimpedance related volumecontrolproblems. Fixing this last requires buffer opamp imo, everything is fixable but requires additional parts. So could have been an updated 3118, but needed redesign for that, beyond tiny version difference you see for previous pro's. You did notice the price hike for 3118pro during 3118pro lifetime/productiontime, dollar yuan is difficult I guess, they were spending too much money and needed to spend even more to raise component count further. They also stopped production of 3116 sa60 so it seems, inputside pro2 3118 would have been much like sa60, if they felt like improving the 3118pro volume controll, diodes next to 7492 are like what was needed for 3118 too I guess, some additional ,4 components? I believe for doctormord pop fix, or other popfix, but they do not mount wires components above pcbs like we do so real new pcb. They could have and maybe they will, maybe not for all "problems", but diodes next to 7492 to me indicate that was really problem they experienced and hurt by, burning output without speakers, resellers claiming money back or free new amps, which possibly burn again when no speakers connected
Thanks for the background irribeo, that does seem to suggest better justification for the change than grabbing what's laying around. Resellers should still be told so product listings can be updated.
I had planned on using higher voltage, I have a 19v 4amp brick. What recommendations for replacing that cap? I have the 3.1, single big cap - could go with another single or 2, the pcb appears to accommodate both. abraxalito recommended two 25v chemicon kzh 2200uf, those seem a little hard to find.
I had planned on using higher voltage, I have a 19v 4amp brick. What recommendations for replacing that cap? I have the 3.1, single big cap - could go with another single or 2, the pcb appears to accommodate both. abraxalito recommended two 25v chemicon kzh 2200uf, those seem a little hard to find.
Last edited:
kze kzh kzm kzn all are similar I think, kyb is a little different, ky also a little different, but I like all the chemicons and prefer the sound, treble improves imo, some will laugh I strongly believe treble superiority chemicons compared to all wet electrolytics other brands I have had next to tpa3116 and 3110 too and some ST Micro chips too, don't have 7492. Some Rubycons real close. SMSL placed two decoupling ceramics per side, which seems better than STM typical datasheet implementation of just one 100nF, I don't know what values they chose, I would measure and use 100nF and 2 to 10 uF, I look towards 10uF, but I think internally the 7492 is different, I think AD modulation looking at outputfilter and then the switchingfrequencies in datasheet are more real than 3118 datasheet figures, I mean the AD datasheet figures are the actual speed, the TI 3118 datasheet is what you find in output, the BD modulation is like 2 times oversampling I think I read, meaning real speed internally is half or externally first measured, ripple peak is one harmonic higher, the first frequency ripple cancels in BD. So speed in 7492 is higher than normal 3118 application I think, don't know if that changes decoupling needs much. AD output filter has critical capacitor, the one bridging the two outputs, quality of that one is very important, also trevor marshall wrote in his class D 7492 description of making a Sure7492 decent, that one is unstable coming from china with 3% distortion instead of 0.3% distortion, SMSL seems to be more carefull than Sure electronics. Trevor marshall article is good read when you own this amp I think, outputfilter design for the speaker you chose, real important imo with all these amps, outputcurrents can walk away, explode, dragging performance down. Google Trevor Marshall class D or tda 7492, read couple of times, download simulation files, look at current peaks at inaudible frequencies, educational I think.
I strongly believe treble superiority chemicons compared to all wet electrolytics other brands I have had next to tpa3116 and 3110 too and some ST Micro chips too, don't have 7492. Some Rubycons real close. SMSL placed two decoupling ceramics per side, which seems better than STM typical datasheet implementation of just one 100nF, I don't know what values they chose, I would measure and use 100nF and 2 to 10 uF, I look towards 10uF, but I think internally the 7492 is different, I think AD modulation looking at outputfilter and then the switchingfrequencies in datasheet are more real than 3118 datasheet figures, I mean the AD datasheet figures are the actual speed, the TI 3118 datasheet is what you find in output, the BD modulation is like 2 times oversampling I think I read, meaning real speed internally is half or externally first measured, ripple peak is one harmonic higher, the first frequency ripple cancels in BD. So speed in 7492 is higher than normal 3118 application I think, don't know if that changes decoupling needs much. AD output filter has critical capacitor, the one bridging the two outputs, quality of that one is very important, also trevor marshall wrote in his class D 7492 description of making a Sure7492 decent, that one is unstable coming from china with 3% distortion instead of 0.3% distortion, SMSL seems to be more carefull than Sure electronics. Trevor marshall article is good read when you own this amp I think, outputfilter design for the speaker you chose, real important imo with all these amps, outputcurrents can walk away, explode, dragging performance down. Google Trevor Marshall class D or tda 7492, read couple of times, download simulation files, look at current peaks at inaudible frequencies, educational I think.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.