So I just got my SA-36 but it's the TDA7492PE version.
I opened it and it has 2 4400uf 16v capacitors. Would there be any benefit in swapping them for 2 4400uf 25v ones? And can I have those 2 and add a 2200uf 25v extra?
I'm currently using a laptop power brick 19v 3.41A but I'm guessing it's not delivering all of the power. The chip is supposed to handle up to 26v.
Will post pictures of the replacements I bought.
I opened it and it has 2 4400uf 16v capacitors. Would there be any benefit in swapping them for 2 4400uf 25v ones? And can I have those 2 and add a 2200uf 25v extra?
I'm currently using a laptop power brick 19v 3.41A but I'm guessing it's not delivering all of the power. The chip is supposed to handle up to 26v.
Will post pictures of the replacements I bought.
Run the chip at 25V and get some 25V caps in there. Best would be Nichicon or NCC or Samyoung - get as many as you can physically fit in there and go for lowest ESR.
These are the ones I currently have (2 of each). Would these be good enough or should I order from the ones you listed. I would prefer to use the ones I have here because shipping takes a lot to Mexico.
Couldn't figure out how to edit but I got a couple of TEAPOs 4700uf 25v, it says SH105°c, 12/15, A3 TD.These are the ones I currently have (2 of each). Would these be good enough or should I order from the ones you listed. I would prefer to use the ones I have here because shipping takes a lot to Mexico.
I googled TEAPO and they seem to be "good" with low resistence, I guess these would be better? They were more expensive than the ones in my previous post.
I don't recognise the brand markings on those so quite likely they're no-name Chinese brands. Which might possibly be OK (at best).
Are you saying the shipping cost exceeds the cost of capacitors? If so then it might pay you to order even more expensive capacitors - like OSCONs for example. They're the gold standard but they don't come in very high values, you'd need to parallel smaller values.
The thing to do is experiment with what sounds best, learn as you go. There's no 'right answer'.
Are you saying the shipping cost exceeds the cost of capacitors? If so then it might pay you to order even more expensive capacitors - like OSCONs for example. They're the gold standard but they don't come in very high values, you'd need to parallel smaller values.
The thing to do is experiment with what sounds best, learn as you go. There's no 'right answer'.
> power brick 19v .... > 25V ..... > better to have a 12v
Wait a second. You are riding off in all directions.
How much power do you need? Will there be a test?
In home speech/music systems, you need at least a Watt to play "loud". With low sensitivity speakers you would like considerably more. Many old-time speaker makers said "minimum 20 Watts". They would certainly play with less; they just didn't want to get blamed if the speakers didn't impress with flea-size amplifiers. There is no upper limit except your budget, and how careless you might be on your volume.
For Power you need *both* Voltage and Current. There is math. Let's skip this for now.
And 12V 19V 25V, while different, are not "very" different. "Much bigger" would be like 3 times the voltage (10X the power). Which exceeds the Maximum for this amplifier.
Your 19V supply, in this type amplifier, will make almost 20 Watts in 8 Ohms. Many many hi-fi fanatics have been happy with 20-watt jobs. At this level you need about 3 Amps, your brick has that and a bit more. You are in much better shape for power than many first-time builders.
And the laptop brick in your hand BEATS buying another supply for very little "improvement". Sure 25V is more than 19V, but I would not drive to town to get a 25V 4A supply. The difference in audio level isn't worth it.
And any amplifier YOU build will be "the best in the world", in your eyes. It is a very intoxicating thing to bend-up some wire and stuff, find electric life in it, and make sweet music. (Even if, as in my first self-design, it didn't get "sweet" until a 20 minute hard run every morning.)
Since you do have the 19V supply, those 15V caps have to go. As the chip is rated to 30V, and you could change supply later (get it working first!), 35V caps would be nice. But if 25V caps are in your hand (or in the mail), use those.
Build it. Don't agonize "could this be better?"
Wait a second. You are riding off in all directions.
How much power do you need? Will there be a test?
In home speech/music systems, you need at least a Watt to play "loud". With low sensitivity speakers you would like considerably more. Many old-time speaker makers said "minimum 20 Watts". They would certainly play with less; they just didn't want to get blamed if the speakers didn't impress with flea-size amplifiers. There is no upper limit except your budget, and how careless you might be on your volume.
For Power you need *both* Voltage and Current. There is math. Let's skip this for now.
And 12V 19V 25V, while different, are not "very" different. "Much bigger" would be like 3 times the voltage (10X the power). Which exceeds the Maximum for this amplifier.
Your 19V supply, in this type amplifier, will make almost 20 Watts in 8 Ohms. Many many hi-fi fanatics have been happy with 20-watt jobs. At this level you need about 3 Amps, your brick has that and a bit more. You are in much better shape for power than many first-time builders.
And the laptop brick in your hand BEATS buying another supply for very little "improvement". Sure 25V is more than 19V, but I would not drive to town to get a 25V 4A supply. The difference in audio level isn't worth it.
And any amplifier YOU build will be "the best in the world", in your eyes. It is a very intoxicating thing to bend-up some wire and stuff, find electric life in it, and make sweet music. (Even if, as in my first self-design, it didn't get "sweet" until a 20 minute hard run every morning.)
Since you do have the 19V supply, those 15V caps have to go. As the chip is rated to 30V, and you could change supply later (get it working first!), 35V caps would be nice. But if 25V caps are in your hand (or in the mail), use those.
Build it. Don't agonize "could this be better?"
>
Build it. Don't agonize "could this be better?"
Thanks for the advice. I know I'm all over the place I'm not 100% sure what I'm doing but I really want to learn. I have a pair of MB42Xs in this setup and I guess I wanted to get all I could out of them but you're right they do sound great. I will try to get some good caps and will swap them tomorrow if I can.
As others have touched on the amp is rated for 15 v for a reason, although the chip can be used with higher voltages this has been limited in this amp mainly so Smsl can use the same chip in a more expensive amp with higher rated caps etc enabling a higher power rating.
I initially used mine with 12 v 5 a ( i don't trust cheap " in the box psu s ) and it sounded fine then in my ignorance tried a 24 v 4.5 a supply and fried it taking out the tweeters on my new JBL s at the same time.
Whilst i understand your enthusiasm for modding just make sure you understand exactly what you are doing or it will end in tears![Smile :) :)](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
I initially used mine with 12 v 5 a ( i don't trust cheap " in the box psu s ) and it sounded fine then in my ignorance tried a 24 v 4.5 a supply and fried it taking out the tweeters on my new JBL s at the same time.
Whilst i understand your enthusiasm for modding just make sure you understand exactly what you are doing or it will end in tears
tried a 24 v 4.5 a supply and fried it taking out the tweeters on my new JBL s at the same time.
So out of curiosity did you also have 16v caps and this is why you fried it? I mean I see what you are trying to tell me but if I do get better caps, what could the problem be? What should I be watching out for? I can always try the amp with some old speakers first as a "security measure".
That's why I was wondering about using a lower voltage, higher current amp. I've seen people mention that using astron psus made a big difference in sound. I currently only have a 12v 2A PSU laying around but I feel it sounds tiny. Actually, I feel the one I'm using now sounds "tiny" or like too "tight"? Don't know the exact words.
Problem might be that not just the two big electrolytics could be rated below 25V, I don't know.
Your Teapo are general purpose series, low impedance or ultra low impedance/esr series should be better. Teapo have bad name because they tend to fail in stressfull positions in computers, here they could be fine, again I don't know. If only above 2 were available I would take the Teapo.
Your Teapo are general purpose series, low impedance or ultra low impedance/esr series should be better. Teapo have bad name because they tend to fail in stressfull positions in computers, here they could be fine, again I don't know. If only above 2 were available I would take the Teapo.
So I did some calculations based on what I read on this thread and found out that:
Vpeak=sqrt(32w*8R)=16
Ipeak=16/8=2A
Cbulk=ipeak*td*dmax/vripple
=2A*1/250khz*0.95 / 0.1v = 76uf
cbulk 2 channels = 152uf
esrmax= 0.1/2 = 50mR
So I need at least 152uf for capacitance and a maximum esr of 50mR.
These OSCONs are 390uf and 14uR esr. So if I can fit 3 in there it would be the best option correct? I'm not completely sure how to set them up in parallel if they currently are in series but if they have the minimum capacitance there shouldn't be any problem.
Otherwise I could get a couple of these. and maybe just 1 OSCON.
Vpeak=sqrt(32w*8R)=16
Ipeak=16/8=2A
Cbulk=ipeak*td*dmax/vripple
=2A*1/250khz*0.95 / 0.1v = 76uf
cbulk 2 channels = 152uf
esrmax= 0.1/2 = 50mR
So I need at least 152uf for capacitance and a maximum esr of 50mR.
These OSCONs are 390uf and 14uR esr. So if I can fit 3 in there it would be the best option correct? I'm not completely sure how to set them up in parallel if they currently are in series but if they have the minimum capacitance there shouldn't be any problem.
Otherwise I could get a couple of these. and maybe just 1 OSCON.
So I read up un how to solder them in parallel, basically I just need to put them through the same hole. I found These which are 20v 680uf and since my psu is 19v I guess that would be ok? I don't think I could fit 3 but my guess is 2 should do the trick in parallel. Am I going in the right direction?
I get tired again thinking about then LOL Advice still remains the same as then, two large high frequency low impedance Chemicon electrolytics that fit comfortably and one solid polymer. If you can only get Teapo easily, go for Teapo. If you only can get Teapo 50uF solid polymer easily, add that value. Shipping can get ridiculous.
But what happened above in post #11 does make me wonder
But what happened above in post #11 does make me wonder
two large high frequency low impedance Chemicon electrolytics that fit comfortably and one solid polymer.
Ok this seems doable. Just to clarify, the solid polymer one MUST be the same voltage as the others but it doesn't really matter if the capacitance is low as long as ESR is low.
Yeah it's 30usd to Mexico but thankfully I know someone in the US so I can ship it to them and then pick it up. I'm a bit eager to do this but I guess I can wait if I can get something better.
But what happened above in post #11 does make me wonder.
I know it scared me a bit but I'm guessing I should be ok if I do it correctly
Last edited:
Just for clarity i didn't make any changes my amp was as supplied , i just tried 24 v because i read that the chip could take 26 and assumed that Smsl were just limiting output by using / rating it at 15 v for marketing purposes .Obviously there are other components that play a part it made a loud squealing/cry for mercy then went silent. Next time i wired the Seakers up to a different amp tweeters were dead .
Live and learn as the saying goes![Smile :) :)](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
Live and learn as the saying goes
I made a 3118 board with 16V Chemicon electrolytics and 16V Nichicon polymers and powered it with 24V powersupply and it started to pop sometimes while playing music, that surprised me a little till I figured it out
Ampboard still seems to sounds the same as identical others, with a 15V PSU, the figuring what was going wrong maybe took one minute max. But the polymers probably are damaged inside.
> Vpeak=sqrt(32w*8R)=16
You have computed V RMS, not V Peak.
Peak of a Sine wave is 1.414 times higher. 22.6V.
In conventional single amplifiers, the power supply must cover the peak-to-peak, 45.2V.
These little class D modules give you *two* amplifiers per speaker. Each one can run on 22.6V (plus losses) and deliver the 45V p-to-p output.
Which is why it wants 25V DC supply to reach its 32W claim.
At 12.5V supply it would be 8W. At 18V about 16 Watts.
Remember that Double Watts is nowhere near Double Loud!!
But it does increase chance of break-down.
You have computed V RMS, not V Peak.
Peak of a Sine wave is 1.414 times higher. 22.6V.
In conventional single amplifiers, the power supply must cover the peak-to-peak, 45.2V.
These little class D modules give you *two* amplifiers per speaker. Each one can run on 22.6V (plus losses) and deliver the 45V p-to-p output.
Which is why it wants 25V DC supply to reach its 32W claim.
At 12.5V supply it would be 8W. At 18V about 16 Watts.
Remember that Double Watts is nowhere near Double Loud!!
But it does increase chance of break-down.
- Home
- Amplifiers
- Chip Amps
- Modding the SMSL SA-36A Pro