| Arc |
Hi I am new to the whole DIY Chipamp thing, and I have a couple questions. Mainly about Toroidal transformers and what are recommended. I know there are a lot of these threads, and I just didn't see any that answered my questions directly. If there are some please direct me and I will look again. Thank you for bearing with me here.
I am planning on making the Dual Mono kit from chipamp.com. I am planning on keeping the left/right/power all completely separate. And the chassis will be completely aluminum, with copper plates for the amp boards to sit on.
The speakers I will run off of this will most likely be around 4ohms with slight spikes to 8ish. I know this comes to a factor of what V I need on the transformer. After looking at Vikash's designs that he made I would like to get a 300va 20/20 because this seems to be the best option for me.
The question I have is would it be best to get one transformer or two smaller ones? I have searched and seen many designs with two. The only benefit I can see from this is complete separation of channels. Any others? If I go with two should I just divide the 300 by two to 150Va for each toroid?
Also, any good (reasonably priced) sources to get these custom wound?
Thank you for all the help, any other comments or suggestions on this?
-Alex |
|
|
| sek |
Hello Alex,
| quote: | | The question I have is would it be best to get one transformer or two smaller ones? |
This has been discussed many times before, the consensus is that true monoblocks benefit from better load regulation behaviour of the transformers.
But using a shared transformer and completely independent rectifier pairs, capacitor banks and distribution wiring is almost as good. The higher load influence on the transformer can be compensated for by higher transformer part quality! :)
| quote: | | If I go with two should I just divide the 300 by two to 150Va for each toroid? |
In theory, yes. But practical transformers aren't ideal, the behaviour under load will be a little better if you keep the power rating higher. But all in all it depends on your needs. You won't use a GC for sound reinforcement in a movie theater, so what? :D
OTOH, I use a 120VA transformer to power two LM3886 and I power my main speakers (8Ohm load) in my living room with them (four LM3886 in active speakers total)! You see, such discussions can quickly get a matter of taste.
I can't help out with custom winders, sorry. ;)
Cheers,
Sebastian. |
|
|
| Buhl |
If Your speakers are 4 ohm, I would defenetly go for paralling the 3886 chip - it looses a few % in the mid/high region, but gains 100% in the bottom end, mostly in low bass definition. What you do is you tie the in and outputs of two channels together, inserting a 0,1ohm powerresistor in series with each output - se app note an1192 from national - very, very good reading, even for more "experienced" diy'ers ;-)
Power in parallel @24 volt trafo(34/35volt after rectrification) is 50watt in 8 ohm, 100 in 4 and 140 in 2 if I recall. because of the parallel configurations better abilities to handle current in low impedances, 24 volts are a good choice here, but read the app note, and consider to build the amp on "labboard" i.e. hardwire it - it can be done even with a small experience on diy building.
Regarding the single/double transformer question, I would say that the most important thing is seperate retrifiers for each channel - under load the bridge can drop a few milivolts, and I think that can be heard - double trannies is the best, but I would use the money on good retrifier diodes and good caps, in my opinion this gives best performance for the $$$ - regarding size - take the poweroutput @ half the nominel speaker load, and multiply it with two, and multiply it with the number of channels. Ex. - a stereo PA100 amp delivers 140 watt in 2 ohm(assuming 4 ohm speakers), x2 x2 = 560 watt - so I use a single 500 VA trannie.
I have build single 3875, 3886, zen, aleph and more, but for everyday use I use a parallel 3886, also called PA100 - its really good if build properly.
Good luck !
Hans |
|
|
| Arc |
Sebastian,
Thank you for your reply. Although I am a little fuzzy on the terminology used with making amps I will learn before I continue much further.
It appears (to my untrained mind) that for a "true" dual monoblock, I would use two transformers. But with the dual power supplies, it is almost as good to use just one. So this seems to be a matter of cost and preference?
If cost permits I would like to do a dual transformer, might as well do it right the first time, but with this I wouldn't need dual secondaries correct? So if I went with two, would 250Va be a good fit. And then I would only need 20V instead of 20/20, which should help cost..a little.
This will only be used for periodic music sessions when alone. And to brag and show off the craftsmanship.
Thanks agian,
Alex |
|
|
| Arc |
| quote: | Originally posted by Buhl
If Your speakers are 4 ohm, I would defenetly go for paralling the 3886 chip - it looses a few % in the mid/high region, but gains 100% in the bottom end, mostly in low bass definition. What you do is you tie the in and outputs of two channels together, inserting a 0,1ohm powerresistor in series with each output - se app note an1192 from national - very, very good reading, even for more "experienced" diy'ers ;-)
Power in parallel @24 volt trafo(34/35volt after rectrification) is 50watt in 8 ohm, 100 in 4 and 140 in 2 if I recall. because of the parallel configurations better abilities to handle current in low impedances, 24 volts are a good choice here, but read the app note, and consider to build the amp on "labboard" i.e. hardwire it - it can be done even with a small experience on diy building.
I have build single 3875, 3886, zen, aleph and more, but for everyday use I use a parallel 3886, also called PA100 - its really good if build properly.
Good luck !
Hans |
Thank you for your reply. If I am understanding you correctly, you are saying to use the LM3886 dual mono kit to drive one channel? In other terms taking two chips to one channel? Sorry new to this.
If this is your intention, this was a plan that I did have for the future. But for now I just would like to get my feet wet, and since the only thing I really need to figure out with this kit is the Transformer, I feel it a good kiddie pool for me.
I intend to read through the spec sheet form National when I get some more free time. I will also do some research into the "hardwired" technique you speak of.
Thanks,
Alex |
|
|
| Buhl |
You are rigth about the dual mono channel - two chips in parallel pulling each speaker. A way to wet your feet slowly, is to make sure you can build two identical stereospeakers - build each with one transformer and one set of bridges - keeping in mind that you are gonne use each stereo amp as a monoblock in the future - make a switch that couples the inputs together, and add a 0R1 resistor in series with each output - this way all you have to do is to flip the switch and put a piece of wire between the amp positive outputs - now its a paralle amp.
Read this: http://www.national.com/an/AN/AN-1192.pdf
Absolutly "must read" that i'm sure will teach you alot (it did for me...)
Cheers !
Hans |
|
|
| sek |
Hi again,
| quote: | | If cost permits I would like to do a dual transformer, might as well do it right the first time, but with this I wouldn't need dual secondaries correct? |
If cost permits, You should probably do it. :)
Let's jump into terminology right here: Power amplifiers of the kind we discuss (e.g. chipamps) need a positive and a negative supply voltage, as they have to resemble both the positive and negative component of the eletrical music signal. You can't get around supplying two voltages of equal magnitude and opposite sign (e.g. +/- 20V).
Those two voltages can be derived out of one single secondary winding, but it has to have a so called center tap, a reference point exactly in the middle of the winding - a third wire. Technically, this would make it a center-tapped 40V secondary winding (denoted as 40Vct or +/- 20Vct).
But as windings on transformers are nonideal, the ct is never exactly in the middle - neither physically, nor electrically. For that reason, a dual secondary transformer is preferable. It can resemble a center-tabbed transformer by connecting the windings together in series (which is why it is always at least as good as a ct one).
The improved performance shows when the two secodary windings get rectified and smoothed independently. The absolute voltage levels might vary under load (as with any real transformer), but the secondaries never 'fight' each other, shifting the amplifiers ground level actually up and down with varying loads. The ground reference point is derived by joining the split supply halves in series at their outputs (instead of at their inputs, so to speak).
This very advantage is not so important with only one single amplifier connected (one could compensate for the drawbacks elsewhere). But two amplifier modules inside the same enclosure actually require a stable ground, as one channel should never be influenced by the signal and load situation of the other.
So the alternatives are: Either two regular ct (or dual secondary, doesn't really matter) transformers for true monoblocks - or one big, quality, dual secondary transformer for a common transformer supply.
Regarding parralleling channels: this doubles the current capability of the resulting amp channel, and thus increases it's power capability into lower load impedances (Ohm's law applies ;)). E.g. twice the power into 4Ohm, but not into 8Ohm, let alone 16Ohm. Why not (and how to get it done anyway) can be read in National's application note AN-1192.
Hope this helps,
Sebastian. |
|
|
| Arc |
| quote: | Originally posted by sek
...
So the alternatives are: Either two regular ct (or dual secondary, doesn't really matter) transformers for true monoblocks - or one big, quality, dual secondary transformer for a common transformer supply.
Hope this helps,
Sebastian. |
Thank you again. I think I am going to go with dual trans. I can understand why to a point. I will start looking for high quality toroids. I think the only thing I need now is the Va rating for each. I am thinking about going to 200-250, with 18/18v or 22/22v dual secondaries because I seem to see a lot of those, maybe get a pair of 250va 20/20's made.
I am not going to parallel them, if this turns out like I want, I will think about making another one and just bi-amp the components. This why I can stick to 4ohm mids, and then build the other to match the ohm's of the tweeter. Maybe find a way to be able to exchange the trans. out for others depending on the resistance of the drivers. |
|
|
| ianpengelly |
Just to help seal the deal, it is worth looking at my thread on "Got Chipamp Hum? Possible Solution".
I have sound that using the dual mono kits with a single transformer results in hum and the solutions are to either use a single rectifier board for both channels (which is a compromise) or the best solution is to go with two transformers.
All the best!
Ian |
|
|
| sek |
Ian has a point about hum and his findings, as those are the boards you intend to use. :)
| quote: | | maybe get a pair of 250va 20/20's made |
Don't bother. 20V vs. 22V won't make any difference, but getting them custom made will heavily increase their price.
A good toroid has some properties as: low output voltage sagging under load (a good manufacturer specifies that), shield winding (preferably copper foil) to help safety and mains noise suppression, a good vacuum resin cast (to highly damp the copper winding from vibrating with line voltage fluctuations), and no thermal protection inside (as this works destructive and can't be reset).
If you want a quick set of properties to look for, choose these two: power rating (for the load stability) and resin cast (for the noise damping). The rest can be skimped upon. Good transformers don't neccessarily have to be expensive while bad ones don't need to be cheap, either. ;)
As for power: the datasheet states that for your around 30V DC (from around 20-22V AC) the maximum power - even with extremely ugly sound due to a variety of protection circuits at work - will be no more than 80W per chip. That's why people tend to recommend 150VA transformer ratings in such a setup. Anything above 200VA per channel will be oversized by a factor of 4, unneccessary in my opinion.
Sebastian. |
|
|
| Arc |
So you think that 2 150-200va, 22/22 v would work nicely? If so that is great, I am starting to get a good list of things I need. This is the last thing I am lookning for.
Do you have any brands to recommend/avoid?
Thanks for all the help thus far.
Alex |
|
|
| sek |
Looks expensive to me (although reasonable), but very good. Plitron is renowned even among Gaincloners. E.g., look at these from Chech Republic. With electrical and magnetical shielding, though. But same size, same quality of load regulation and also going for the price Plitron quotes. Thus, it can't hurt to order the Plitron. If in doubt, here's more to look for.
I can't give an alternative for you, as all companies I know are European. :)
That's a local business oportunity. Come on, fellow Northern Americans. :D
Sebastian. |
|
|
| BWRX |
| Avel Lindberg toroids are good quality and quite a bit cheaper than Plitron toroids. I (and I'm sure many others on here) can highly recommend Parts Express as well. |
|
|
| Arc |
| I love PE. I shop there often. I think I will go with the PE Avel Toroids. I think those will meet my needs nicely. I just splice the original AC in between the two toroids..me=noob at amps. |
|
|
| sek |
This model would be the right size and type, but it's of the average quality kind. I'm using an equivalent from Talema for cheapness. :D
It will work very well, but it will hum slightly. If you choose this one, consider a mounting solution that damps the toroid and decouples it from wood, metal, etc.
The AC goes from input receptor to mains switch and then to a suitable distribution block. You then fuse the two toroids separately, because a single fuse for both wouldn't survive the power-up current draw. It's likely hat the circuit breaker of your wall outlet will trip when switching on two such toroids at once, in which case you'll need an 'inrush current limiter' circuit. Try it and see... ;)
Cheers,
Sebastian. |
|
|
| sek |
| Hammond Manufacturing also has a model that's been used by some DIYers on the forums. |
|
|
| Arc |
I work with car audio so I understand seperate fusing and distrobution blocks for car, but what type of "suitable" distrobution are we talking about? I was thinking about dual fuses.
I am planing on doing a turn on switch, this would go between the ac in and the toroid, correct? Also this is before the fuse also right. Sorry if these are very easy questions, I just don't want to get anything wrong on this.
Thanks for you help thus far. |
|
|
| sek |
Hi,
I just wanted to point out that transformer primaries don't belong crimped together at the mains switch. The term for 'screw terminal blocks' escapes me, but you'll know how to do it. ;)
The rest of what you wrote is spot on. Some pointers on recommended power supply projects would be a widely known TNT project and Carlos' chipamp PSU.
Have fun. |
|
|
| Arc |
Ok, I was in Radio Shack today and I saw exactly what you were talking about. They were called ground screw terminal or something. It was just a block of metal with 6 holes and screw terminals. So if this is something for the trans. primaries, let me map this out so I can think of this.
AC In--> Switch-->Power Block-->Fuses-->Toroids-->PS |
|
|
| sek |
I was actually just talking about something like in the attached picture. Make sure that vibration and lifting/carrying can't harm the installation wiring inside the amp enclosure. you know that from car audio, but in 110V/220V devices, touching the leads is more dangerous than in a car, y'know...
But take care to bolt the incoming PE (earth) wire directly to the metal chassis, i.e. as shown in this example (scrool down to the trafo picture). Then, connect every single metal part (that could be touched from the outside) to this PE bolt for safety reasons (just like shown in the example). One connection must then exist between each of your PSU grounds and PE (and only one, which also means that any input terminals must be isolated from chassis metal).
Your connection sequence is right (all this is also explained in detail in the example projects I linked to in my previous posts). ;) |
|
|
| Arc |
Sebastian,
I just want to say thank you for all your help with this. You have been a great source of information and help. You have made this project seem a lot more completable to me, thanks.
I just want to reword what you said to make sure I am correct. I create one point on the chassis and connect the earth from the AC in to this point, and then any exterior pannel will also be connected to this.
I also need to take care that my inputs are not connected or grounded to the chassis. Should I also try to keep the actual amp boards off of the aluminum of the chassis? I am mounting the chip and board to .25" copper for a heat sink, should this be seperated from the aluminum?
Thanks agian. |
|
|
| sek |
Hi Alex,
you're very welcome. :cool:
Basically you're exactly right about 'earthing' the panels. But in practice it would be overkill for every single additional panel that already has an electrical connection to an earthed one. It depends on the connection between the panels - and a safety agency would measure and test. ;)
Good practice (and my recommendation) would be to do a PE bolt on a central place with a short PE wire directly from the inlet, and an additional PE wire from the heatsink to that PE bolt. Then, measure resistance between each panel and the PE bolt. If it's more than a couple milliohms (which I doubt), give it an own PE wire to the bolt. Just account for ageing, alluminium corrodes in free air... :)
Wether your interconnects should be grounded, earthed (note the difference) or left floating depends on your interconnects. ;)
The usual way is the one described in chipamp.com's manual. For standard ('unbalanced') hifi connections like cinch, mount the connectors insulated from the panels. Cinch ground (signal ground) from the cinch socket then goes to the amp board with the (hot) signal wire.
The mentioned manual describes it on page 12. Don't forget the connection from CHG to the PE bolt from each PSU PCB. Page 13 also mentions the safety earthing story.
The amp boards must of course be inulated from any potential (voltage). The boards themselves isolate the mounting holes from the copper. The chip must be mounted on the heatsink with an insulation layer between chip and sink.
Your options for the boards are plastic or metal standoffs or bolts. Both are good.
As for the heatsinking, there are silicone pads, mica washers and ceramic plates availabl. Silicone is the easiest, mica is classic, ceramic looks best. :cool:
If you - as per Brian's and Peter's recommendation - use the LM3886TF, you don't have to care for that part, as the TF type is plastic insulated in itself. Just use some silicone grease as heat transfer agent, like on CPU coolers.
The copper heatsink would then not have to be separated from the aluminium. The more metal you have earthed, the better. Just measure the resistances between metal parts after mounting. For conductivity it should be a few milliohms maximum, for insulated parts it should be many many megaohms minimum. ;)
Sebastian. |
|
|
| Arc |
So far so good. I am gathering parts. The amp itself should be here in 2 days. I am wondering about hook up wires inside. What is everybody using?
Would this suffice? http://www.gepco.com/products/proav...ual_xband_M.htm
Also, any good brands of Pots. to recommend for a Dual Mono set up? |
|
|
| sek |
As noone else seems to answer... ;)
The cable you link to looks expensive :(
And it's for balanced interconnects, something you don't seem to use in this project. But "mono" microphone cable of the same quality should give optimal results.
People use all kinds of cable for internal wiring, a search should come up with a lot of discussion about the topic.
For power wiring, refer to recommended values per given current requirement. PSU wiring should be thick and flexible. :D
Same for potentiometers, a lot of discussion has been done about this. ALPS (blue series) and Panasonic (plastic tape) are recommended frequently. Just look for something 'encapsuled' with a solid mechanical feel. ;)
Cheers. |
|
|
| teemuk |
| quote: | Originally posted by sek
A good toroid has some properties as:... ...and no thermal protection inside (as this works destructive and can't be reset). |
Care to elaborate this? I understand the "can't be reset" part but not the "works destructive" part. |
|
|
| sek |
A temperature protection device in a transformer is usually a 'thermo-fuse', an element that opens on overtemparature. It then stays open (in order to prevent a fire or other hazard) and the transformer has to be exchanged.
It's bad because of two things: You wreck a transformer by overheating it (which is the least problem, as such high winding temperatures should never happen) - and the variable impedance of the thermal fuse might (possibly maybe) affect the sound.
I've never heard it, though, as I've never overheated a transformer myself. ;) |
|
|
| teemuk |
| Ah, I see. I always thought they use a bimetal switch or similar protection method in those (heating opens the switch - switch gets closed when the transformer cools down). |
|
|
| sek |
Well, there actually are transformers with auto-reset thermal switches, and I'd say then they are good. ;)
My inital vote against a temperature protected trafo was with regard to the non-resetting (destructive) type only. |
|
|
| samsagaz |
| quote: | Originally posted by sek
Hi again,
If cost permits, You should probably do it. :)
Let's jump into terminology right here: Power amplifiers of the kind we discuss (e.g. chipamps) need a positive and a negative supply voltage, as they have to resemble both the positive and negative component of the eletrical music signal. You can't get around supplying two voltages of equal magnitude and opposite sign (e.g. +/- 20V).
Those two voltages can be derived out of one single secondary winding, but it has to have a so called center tap, a reference point exactly in the middle of the winding - a third wire. Technically, this would make it a center-tapped 40V secondary winding (denoted as 40Vct or +/- 20Vct).
But as windings on transformers are nonideal, the ct is never exactly in the middle - neither physically, nor electrically. For that reason, a dual secondary transformer is preferable. It can resemble a center-tabbed transformer by connecting the windings together in series (which is why it is always at least as good as a ct one).
The improved performance shows when the two secodary windings get rectified and smoothed independently. The absolute voltage levels might vary under load (as with any real transformer), but the secondaries never 'fight' each other, shifting the amplifiers ground level actually up and down with varying loads. The ground reference point is derived by joining the split supply halves in series at their outputs (instead of at their inputs, so to speak).
This very advantage is not so important with only one single amplifier connected (one could compensate for the drawbacks elsewhere). But two amplifier modules inside the same enclosure actually require a stable ground, as one channel should never be influenced by the signal and load situation of the other.
So the alternatives are: Either two regular ct (or dual secondary, doesn't really matter) transformers for true monoblocks - or one big, quality, dual secondary transformer for a common transformer supply.
Regarding parralleling channels: this doubles the current capability of the resulting amp channel, and thus increases it's power capability into lower load impedances (Ohm's law applies ;)). E.g. twice the power into 4Ohm, but not into 8Ohm, let alone 16Ohm. Why not (and how to get it done anyway) can be read in National's application note AN-1192.
Hope this helps,
Sebastian. |
Let me ask something i really dont understand very well how i need to order the transformers :S
i need the transformers to run an 5 channels AMP. 5 x lm3886.
i was thinking to use 30v on each rail, i read in datasheet that maximum are 35v, so i think that will be ok. i want to run it as high as possible because my Speakers (FR125s) are just 84db/1w. are 30v ok? can i use 32v too?
i want to use 2 transformers, each one need to have just 2 wires right? 2x 0v-30v 200-250VA will be ok?
Thanks in advance. |
|
|
| pinkmouse |
| AC voltage is measured by RMS, basically an average level. When rectified you need to multiply it by 1.41 to get the DC level, so for your 32V rails, (ideal for 8Ohm speakers), you will need 32/1.41 = 22.7V secondaries. |
|
|
| facundonu |
shouldn't you add the 1,4v for the bridge rectifier before dividing by 1,41?
i own 30+30 VAC (with no load) transformer, 250VA. i guess (i'll measure it today) it will give something like 27+27 VAC loaded. after rectifying:
30*1.41-1.4=41
28*1.41-1.4=38
it should vary within the +-41VDC to +-38VDC range.
pretty tight for a LM3886, right?
anybody tryied this chip so close to the 42V limit? any suggestions besides avoiding it?
thanks. |
|
|
| sek |
Hi,
| quote: | | i was thinking to use 30v on each rail, i read in datasheet that maximum are 35v, so i think that will be ok. i want to run it as high as possible because my Speakers (FR125s) are just 84db/1w. are 30v ok? can i use 32v too? |
The absolute maximum supply voltage |V(+)| + |V(-)| is stated as 84V. This gives 42V absolute maximum per supply line. Bute the diagrams only specify sensible and safe function up to 70V (i.e. datasheet p. 8).
On p. 9 THD+N is specified for both +/-28V and +/-35V, so anything in between gives results as rated. Anything beyond gets worse. The Power Dissipation vs. Output Power graphs on p. 11 indicate a maximum power loss of 70W when driving 4Ohm from +/-35V. Sinking this heat off is difficult to say the least! For an 8Ohm load and +/-40V the graph indicates a loss of 50W, still a lot.
And it suggests the conclusion that driving 4Ohm from +/-40V would strike the amp with 90W of losses, nothing a chipamp could handle! As you can't guarantee any load impedance to be only well above 8Ohm, this scenario becomes a no-no. Sound-wise, the internal protection circuitry would long have kicked in and ruined the sound. This is also the answer to the actual question: it would work, but that it would sound good at higher volumes is unlikely.
| quote: | | i want to use 2 transformers, each one need to have just 2 wires right? 2x 0v-30v 200-250VA will be ok? |
There's no need to have two transformers, one beefier one would be smaller and less expensive. But it seems to be an interesting idea. ;)
Although, 2x 0v-30v don't conform to 2 wires, so I probably don't get what you mean. :cannotbe:
What you need ist 2x 0v-30v to each amp, how you supply it is up to you. ;)
From one transformer 2x 0v-30v means exactly three secondary wires.
From two transformers with one secondary each it means two secondary wires per transformer, so four secondary wires total. This is probably what you meant.
As Pinkmouse and Facundonu correctly pointed out, 30V secondary doesn't mean 30V DC supply to the amp.
I recommend +/-35Vmax DC for 4Ohm speakers, +/-40Vmax DC for above 8Ohm. This would lead to 2x 0V-24.75V and 2x 0V-28.3V secondaries, respectively (not accounting for rectifier losses).
| quote: | | shouldn't you add the 1,4v for the bridge rectifier before dividing by 1,41? |
Rectifier losses vary with rectifier type. The value 1.4V fits a silicium diode bridge rectifier when cold and loaded with higher currents. In this regard I would recommend to only design for safety - and thus not increase the AC voltage to compensate for the rectifier loss.
Two reasons: first, a Schottky rectifier has less than half this loss, so you never know what everyone wants to use. Second, 1V more output voltage only means 10Wmax more output power in the region of a +/-40V supply. Sound-wise, this gain is negligible, but from a safety standpoint some volts too many are some volts too many. :cool:
Cheers,
Sebastian. |
|
|
| samsagaz |
Thanks too much, i read in the forum that using 2 transformers 0-28 for example, will give better sound that just use one 28-0-28. Thats why i ask abt use 2 instead just one, but one will be cheaper of course, so i think that will use just one.
400VA will be ok for an 5 channels amp? i dont want an bigger trafo, and i cant purchase toroidals, because in my country are too expensive.
So i will goto the store and just ask for 28-0-28 400VA transformer. thats ok right? :) |
|
|
| sek |
This actually only depends on the power you require. You wrote that you use FR125s with just 84db/1w, that's not a lot. :cannotbe:
At 10W, they could deliver approx. 94dB SPL (@1m), for 100dB you'd need at least 40W. The amps would then still have a headroom (dynamic reserve) of 3dB, accounting for power compression and safety margin.
You probably want to max out the dynamic range of the chips, but with almost +/-40V DC this is a total power consumption of 120Wmax per channel. That's 600Wmax per 5 channels.
So for regular casual and occasional party or home cinema listening, a 400W transformer would probably be enough for good sound all the way. It's definitely enough regarding safety, provided you fuse accordingly. ;)
To squeeze the very last out of the chips, a 1000W transformer and active heat sinking would be required, though. But in case it's a 5.1 channel home theater system, this will not be required, because there won't be many moments when all channels get driven hard simultaneously. :whazzat:
In any way, with +/-40V supply lines, the usual small heatsinks are definitely not sufficient any more, especially if you mount five of them into one enclosure. You'll have to cool the thing like a big one (because it actually will be). :hot:
Hope this helps,
Sebastian.
PS: Take care! You need heat sinking and capacitor buffering for five amplifiers putting out anything between 2W and 120W each, thats between 10W steady and 600W short term power total. |
|
|
| Thomo |
Hi guys. I hope you don't mind me coming in on this thread but it seems better than to start another new one...
I have built LM3886 amp and am impressed by what I have heard so far. But from other peoples comments I know it's capable of more.
I have attached diagrams of what I have. Could anyone be so kind as to give some suggestions for improvement.
I realise the tx isn't up to scratch and that the local decoupling caps need to be improved but i'm unsure about snubber values etc....
Cheers, Lee. |
|
|
| sek |
Hi,
I still prefer TNT's approach and what Carlos used to make out of it - as mentioned above - even though it's been a while ago already.
Just make sure you read both topics to their last pages.
And 2x30V is way too high for the transformer, especially at 120W for two channels!
On how to calculate it and what values I recommend, please re-read the whole thread. In my postings above I detailed a lot, I cannot repeat all of it.
:2c:
Sebastian ;) |
|
|
| sek |
Excuse me, there has been a post by 'juergenk' in this place, which is gone! Any explanation?
He might have deleted it, but I don't think so, as I read it about half an hour after he postet it. |
|
|
| pinkmouse |
| He did indeed delete it himself. |
|
|
| sek |
And I guess you stole my Cop, didn't you? :D
Happy evening,
Sebastian. |
|
|
| pinkmouse |
| Yup, impersonating a police officer is a very serious offense! :) |
|
|
|
