Hi All,
This topic HAS come up in the past, however i'm looking for advice on this situation 🙂
There is 2 "main" problems i'm looking to alleviate!
1. The amp sounds "bad" on main speakers, it is a "cheap" pro audio amp and mostly built well, but the pre-amp stage (i'm guessing) really sucks badly!!! It goes from Cannon's into a board (that looks like it was bought from a $2 shop!) with NE5532's (2 of them) to drive the preamp and obvious gain adjustments go to the front panel... I understand the NE5532's aren't that bad and when setup right are quite good, however the sound is SO stale and dry that it's not actually listenable!!! I use it bridged to a sub and it sounds very average... From the NE5532's it goes into 2N5551's and 2N5401's (pairs obviously and there's 2 lots of them) Then into 2SD669 and 2SB649 into 2SA1942 and 2SC5200 (there's 5 pairs of them in each side (channel A 5 pairs, channel B 5 pairs).... There is a 1500VA (output) transformer hooked up to this which has +-78V on the DC side (after 10,000uf filter caps (2 of them))... The amp supposedly does "1200w" into bridged 4 ohm... Now, the capacitance of the filter caps needs to be increased in my books as it's not enough... However would that affect sound quality? Sorry to post such a long post, however, would the pre-amp section (i can post a picture if that helps?) affect the output more? Like i said the whole preamp could have come from a $2 shop to be honest, it has the cheapest of cheapest parts on it and they don't look too excellent!!! What would be the first port of call? I'd really like to modify this!!!
2. Power output - into 4 ohm bridged i've measured 89v across the outputs (it was obviously clipping!!!) It does get hot quite fast (15 seconds it's getting very hot but heatsink is still touchable), it is fan forced (which i will be upgrading too) but looking at possibilities of adding another torroid and having one p/s per channel - realistically i will only load it down to 4 ohm bridged, however the headroom there is very important as it's for home cinema use and needs to develop at LEAST 1200wrms continuous (preferably!)... I don't mind about heat, heat is easily enough fixed with fans and most amps in home cinema don't get continuous output, more pulses so it will cool enough to not have an issue... Can anyone help me with this? I've looked into adding more outputs, this is not easily done (no room and obvious problem of doubling up BJT's - not easy and potential to blow a lot of chips!)... I'm really after more power and more stable, i feel the power supply isn't holding up it's end of the bargin to start with - even the internal wiring gets quite warm (i think it's 14ga - each channel has 154v DC on the chips (to ground) so it has the voltage!!!) Wiring is an upgrade and i think the capacitance needs to go up by 4x and maybe add another torroid? Can someone help me with some ideas for stability? Thanks very much in advance!!!
Aaron
This topic HAS come up in the past, however i'm looking for advice on this situation 🙂
There is 2 "main" problems i'm looking to alleviate!
1. The amp sounds "bad" on main speakers, it is a "cheap" pro audio amp and mostly built well, but the pre-amp stage (i'm guessing) really sucks badly!!! It goes from Cannon's into a board (that looks like it was bought from a $2 shop!) with NE5532's (2 of them) to drive the preamp and obvious gain adjustments go to the front panel... I understand the NE5532's aren't that bad and when setup right are quite good, however the sound is SO stale and dry that it's not actually listenable!!! I use it bridged to a sub and it sounds very average... From the NE5532's it goes into 2N5551's and 2N5401's (pairs obviously and there's 2 lots of them) Then into 2SD669 and 2SB649 into 2SA1942 and 2SC5200 (there's 5 pairs of them in each side (channel A 5 pairs, channel B 5 pairs).... There is a 1500VA (output) transformer hooked up to this which has +-78V on the DC side (after 10,000uf filter caps (2 of them))... The amp supposedly does "1200w" into bridged 4 ohm... Now, the capacitance of the filter caps needs to be increased in my books as it's not enough... However would that affect sound quality? Sorry to post such a long post, however, would the pre-amp section (i can post a picture if that helps?) affect the output more? Like i said the whole preamp could have come from a $2 shop to be honest, it has the cheapest of cheapest parts on it and they don't look too excellent!!! What would be the first port of call? I'd really like to modify this!!!
2. Power output - into 4 ohm bridged i've measured 89v across the outputs (it was obviously clipping!!!) It does get hot quite fast (15 seconds it's getting very hot but heatsink is still touchable), it is fan forced (which i will be upgrading too) but looking at possibilities of adding another torroid and having one p/s per channel - realistically i will only load it down to 4 ohm bridged, however the headroom there is very important as it's for home cinema use and needs to develop at LEAST 1200wrms continuous (preferably!)... I don't mind about heat, heat is easily enough fixed with fans and most amps in home cinema don't get continuous output, more pulses so it will cool enough to not have an issue... Can anyone help me with this? I've looked into adding more outputs, this is not easily done (no room and obvious problem of doubling up BJT's - not easy and potential to blow a lot of chips!)... I'm really after more power and more stable, i feel the power supply isn't holding up it's end of the bargin to start with - even the internal wiring gets quite warm (i think it's 14ga - each channel has 154v DC on the chips (to ground) so it has the voltage!!!) Wiring is an upgrade and i think the capacitance needs to go up by 4x and maybe add another torroid? Can someone help me with some ideas for stability? Thanks very much in advance!!!
Aaron
I would say you are well on the right track. You estimate for the PSU caps sounds about right. Maybe just a 2kVA toroid in place of the existing one rather than adding another one.
I wouldn't worry about how naff a board looks, it's the sound that matters. If it's got NE5532s on then it's a good start. Maybe you could change some signal path caps for better ones, but I doubt you will notice much to be honest.
Maybe have a look at the bias for the amp. It could be set really low.
I wouldn't worry about how naff a board looks, it's the sound that matters. If it's got NE5532s on then it's a good start. Maybe you could change some signal path caps for better ones, but I doubt you will notice much to be honest.
Maybe have a look at the bias for the amp. It could be set really low.
my oppinions on things like this, and i deal with a lot of pa gear and some REALY cheap nasty stuff at that, is
A, driven within limits any amp should sound fine if its the corect one for the job, given speakers that are up to the quality you want to listen at.
B, yes up the caps to at least 15 to 20 thou.
C, upgrade the pre amp components, all the resistors and caps, resistors should be .5 watt metal film, maplins do a bag of 1500 varied values for under £20. Caps the choice is yours, there are plenty of poly caps designed for audio, but that choice is another thread all together.
D, do you need a pre amp? is there a conection for pre out? and pre in? try making a conection between pre and power amp, and feed in line signal, that will show if its the pre or power amp thats sounding bad.
E, bridging should give double voltage swing, (i think!) so even 70 volts should give 140 before clipping, id check the pre amp!!!!
this is just a few thoughts, and hopefully they will help, see you later, steve.. ..
A, driven within limits any amp should sound fine if its the corect one for the job, given speakers that are up to the quality you want to listen at.
B, yes up the caps to at least 15 to 20 thou.
C, upgrade the pre amp components, all the resistors and caps, resistors should be .5 watt metal film, maplins do a bag of 1500 varied values for under £20. Caps the choice is yours, there are plenty of poly caps designed for audio, but that choice is another thread all together.
D, do you need a pre amp? is there a conection for pre out? and pre in? try making a conection between pre and power amp, and feed in line signal, that will show if its the pre or power amp thats sounding bad.
E, bridging should give double voltage swing, (i think!) so even 70 volts should give 140 before clipping, id check the pre amp!!!!
this is just a few thoughts, and hopefully they will help, see you later, steve.. ..
Hi
1200w into 4 ohms bridged is putting a lot of load on the output stage and it's drivers.
It is equivalent to 600w into 2 ohms. The voltages on the rails indicate you should have about 200w to 250w into 8 ohms. So test each amp into 4 ohms and see how close to power doubling you get. Check temperatures. Usings temps and currents check the output stage SOA against manufacturers data. If all this is OK then proceed with very big heatsinks and try testing into 2 ohms going through the checks again. If your power supply, smoothing and heatsinks are upto scratch then you may get 600w or 700w into 2 ohms.
good luck
Andrew T.
1200w into 4 ohms bridged is putting a lot of load on the output stage and it's drivers.
It is equivalent to 600w into 2 ohms. The voltages on the rails indicate you should have about 200w to 250w into 8 ohms. So test each amp into 4 ohms and see how close to power doubling you get. Check temperatures. Usings temps and currents check the output stage SOA against manufacturers data. If all this is OK then proceed with very big heatsinks and try testing into 2 ohms going through the checks again. If your power supply, smoothing and heatsinks are upto scratch then you may get 600w or 700w into 2 ohms.
good luck
Andrew T.
I think the comlementary of the 2SC5200 is the 2SA1943, not the 2SA1942.
The 5200/1943 have a maximum dissipation of 150W/ 15A, a bandwidth of 30mHz., not bad for a PA amplifier i think.
With 5 pairs you have 1500 maximum dissipation per channel, bridged 3000 watts.
Why not through out all the PCB's, add another transformer and go for a decent front end design ?
The 5200/1943 have a maximum dissipation of 150W/ 15A, a bandwidth of 30mHz., not bad for a PA amplifier i think.
With 5 pairs you have 1500 maximum dissipation per channel, bridged 3000 watts.
Why not through out all the PCB's, add another transformer and go for a decent front end design ?
Hi NUTTTR,
I think there are serious driver problems here. There is no way a 2SD669A/2SB649A will drive that output stage into a 4 ohm load at those power levels. Bridged is out of the question. By chance is there an output pair acting as drivers for four paralleled output pairs? You should also check the standing bias current. The amp may not have bias given your description of the sound.
Any way you can post a schematic, or make & model?
-Chris
I think there are serious driver problems here. There is no way a 2SD669A/2SB649A will drive that output stage into a 4 ohm load at those power levels. Bridged is out of the question. By chance is there an output pair acting as drivers for four paralleled output pairs? You should also check the standing bias current. The amp may not have bias given your description of the sound.
Any way you can post a schematic, or make & model?
-Chris
Pro gear may well have protective circuitryat both the input and output has an adverse impact on sound quality. It's there because on the road prospects of "helpful" folks plugging the wrong thing into the wrong socket must be considered. There may also be compression, peak limiting, or "soft clip" cicuits in there that are not likely to help with audiophile goals. If you could dig up a schematic, it may be possible that someone here can take a look at the possability of just bypassing some features that useless to you a=or worse.
BTW, the transistors you mentioned are pretty good (as in almost nothing available to "upgrade to"), so I think you should leave those alone unless you can get your hands on the schematic. You can do Plug-n-play with the opamp but don't expect too much - there even some who swear the NE5532 has yet to be improved on -m arguable but it's still pretty good.
BTW, the transistors you mentioned are pretty good (as in almost nothing available to "upgrade to"), so I think you should leave those alone unless you can get your hands on the schematic. You can do Plug-n-play with the opamp but don't expect too much - there even some who swear the NE5532 has yet to be improved on -m arguable but it's still pretty good.
I've worked in live sound performances for some time and my experience is that even badly designed PA amplifiers are far more perfect than loudspeakers or rooms. Your problem is probably related to these two last items. Don't mess with amplifiers prematurely
*whoops* thanks muchly for picking up my typo, it is a 1943 🙂
And anatech - sorry - i have also missed something else too (it was late!) 2SA1941 2SC5198 is inbetween the D669 and the outputs 🙂 Sorry, that was kinda cruial!!! 🙂
The PCB's themselves seem ok quality (the amp part anyway) and i have added wiring to the back of them (from memory, i've done this on several amps...) and it's all 12ga on the board + tracks so it can handle current now! 🙂
As for brand - it was an "OEM" style unit made by soundking and it is an SKAA1200E version... I'll have to post photo's, schematics (AFAIK) i can't get hold of and no one has!! Very very simple design on the output section!
http://nutter.caraudioaustralia.com/gallery/thumbnails.php?album=20
There's pictures in there!!! 🙂 If you move the mouse over the pictures it will have the file name "big" is ~300-500kb each "SM" is about 100-200kb each... Can someone look there and give me some advise on it? Would appreciate it greatly!!!
1500w per side? Excellent! That's what i wanted to hear!! 🙂
Potentially there may be no bias? not sure i've tried measuring it, unless my multimeter is stupid, i have yet to get a measurement!!!
There is only protection on the output in the form of a relay on either side... That's it, no input protection at all 🙁
Speakers and room are great they normally run off a Peavey PV-8.5C which they sound unreal on and it sounds just as good as my H/K home theatre amp! However it's almost not listenable with the other amp 🙁
Thanks for the great help folks!
Aaron
P.S. The big resistors in those pics are 0.47ohm 5w with 3 legs (centre is common)... i'll go ahead and try and measure bias again!
And anatech - sorry - i have also missed something else too (it was late!) 2SA1941 2SC5198 is inbetween the D669 and the outputs 🙂 Sorry, that was kinda cruial!!! 🙂
The PCB's themselves seem ok quality (the amp part anyway) and i have added wiring to the back of them (from memory, i've done this on several amps...) and it's all 12ga on the board + tracks so it can handle current now! 🙂
As for brand - it was an "OEM" style unit made by soundking and it is an SKAA1200E version... I'll have to post photo's, schematics (AFAIK) i can't get hold of and no one has!! Very very simple design on the output section!
http://nutter.caraudioaustralia.com/gallery/thumbnails.php?album=20
There's pictures in there!!! 🙂 If you move the mouse over the pictures it will have the file name "big" is ~300-500kb each "SM" is about 100-200kb each... Can someone look there and give me some advise on it? Would appreciate it greatly!!!
1500w per side? Excellent! That's what i wanted to hear!! 🙂
Potentially there may be no bias? not sure i've tried measuring it, unless my multimeter is stupid, i have yet to get a measurement!!!
There is only protection on the output in the form of a relay on either side... That's it, no input protection at all 🙁
Speakers and room are great they normally run off a Peavey PV-8.5C which they sound unreal on and it sounds just as good as my H/K home theatre amp! However it's almost not listenable with the other amp 🙁
Thanks for the great help folks!
Aaron
P.S. The big resistors in those pics are 0.47ohm 5w with 3 legs (centre is common)... i'll go ahead and try and measure bias again!
Ok - I'm stuck measuring bias!!! I've tested everywhere and the only point i get a reading is between the ground (heatsink ground - earthed) and the emitter resistor... that gives me 78mv on one side (left) and right i get nothing, 0mv... Somethings not right here!!! There is differences between the 2 channels all the way back, when i pull it apart, i guess i'll be able to fix that then, for now, it still works, so anyone got any help if i'm measuring bias wrong? I'm measuring between the small blue 1/4w resistor near the big white ones in the picture 🙂 Thanks heaps!!!
Neither heatsink gets hot at idle, they stay pretty darn cold (compared to my peavey that puts out HEAT (!) at idle)
Thanks!
Neither heatsink gets hot at idle, they stay pretty darn cold (compared to my peavey that puts out HEAT (!) at idle)
Thanks!
I dont see the really cheap parts you mentioned.
If that toroid rates 1500VA it is not a cheap buggar.
The PCB's look rather good, as do the white block double emitter R's.
Using a single centreplaced vent for such a high power amp seems odd.
The 5532 used at entrance are dual opamps, unlikely that they are only there for inverting the signal to bridge the channels.
Using 5532 is not cheap, far cheaper dual opamps could have been used at that location.
Running the right channel signal cable so close to the toroid could have been avoided.
I would have opted for rotating the powersupply 180 degrees to keep the toroid at distance from the entrance and the front end.
Is there a noticable difference in sound between left/ righ channel and bridged ?
Silly me, why does one of the lines of the entrance cable of the left channel look as cut ?
If that toroid rates 1500VA it is not a cheap buggar.
The PCB's look rather good, as do the white block double emitter R's.
Using a single centreplaced vent for such a high power amp seems odd.
The 5532 used at entrance are dual opamps, unlikely that they are only there for inverting the signal to bridge the channels.
Using 5532 is not cheap, far cheaper dual opamps could have been used at that location.
Running the right channel signal cable so close to the toroid could have been avoided.
I would have opted for rotating the powersupply 180 degrees to keep the toroid at distance from the entrance and the front end.
Is there a noticable difference in sound between left/ righ channel and bridged ?
Silly me, why does one of the lines of the entrance cable of the left channel look as cut ?
NUTTTR said:
No, AndrewT said 600 or 700 into 2 ohms per side. That means 1200 to 1400 into bridged 4 ohms.
NUTTTR said:
You are measuring wrong. Remember those big white 0.47 ohm 3 terminal resistors you mentioned earlier? You should be measuring the voltage from the centre pin to one of the end pins.
Jacco said 1500/side worth of dissapation - wouldn't that be equal to the output - or is it half? (Ok, so i'm a bit rusty on the basic stuff 🙁 )
Ok - i figured i wasn't meant to measure there because across that there's ~3mv on one side of the resistor (side closest to the side of the amp) and that's what i'd guess "noise" through the outputs? The other side (of the ones i took pics of) shows 0mv across it (from outside pin to centre pin)... Same thing on the other side of the amp, also the side closest to the outside wall (opposite to the right side (right side = one in photo's)) and same thing on the other leg... 0mv (so going by resistor |<-3mv|output|<-0mv other side (of amp) |<-0mv|output|<-3mv if that makes sense?) Anyway, that says to me there is no bias at all - which i'd guess WHY it sounds so bad?! Now that i know this - i WILL need help adjusting the bias (if someone would be so kind 🙂) and how i go about it... Assuming i'm meant to measure at idle with no input? 🙂 Thanks heaps 🙂
Most of the components in the output, transformer, etc aren't cheap, but the pre-amp board has the cheapest of cheapest components on it (granted, not the NE5532, but it's the parts around it that's the problem in my book!) it's mono each side, that's why the extra wire has been cut off (it's a bit dodgy like that, thats what i meant by poorly built, they simply cut the extra wire not needed off, rather than not having it in the first place i'd guess?
I'm going to DEFINATELY upgrade the wiring, meaning power wire, coax wire, etc, i've got some high quality stuff that will allow me more length to properly route it!!!
The sub it's driving is dual 2 ohm so no, i've not noticed any difference between running 2 channels and bridged...
The silver zener diodes near the audio input (from preamp) get HOT HOT HOTTTTTT one side is 90 the other side (the part of the amp in photo's) gets 97deg!!!! at idle... which is surprising and needs attention ASAP in my books!!! I'm definately going to look at what i can substitute it with, have it sit up slightly higher and have fans blowing through the heatsinks 🙂
Thanks for the help so far it's been great!!
Aaron
Ok - i figured i wasn't meant to measure there because across that there's ~3mv on one side of the resistor (side closest to the side of the amp) and that's what i'd guess "noise" through the outputs? The other side (of the ones i took pics of) shows 0mv across it (from outside pin to centre pin)... Same thing on the other side of the amp, also the side closest to the outside wall (opposite to the right side (right side = one in photo's)) and same thing on the other leg... 0mv (so going by resistor |<-3mv|output|<-0mv other side (of amp) |<-0mv|output|<-3mv if that makes sense?) Anyway, that says to me there is no bias at all - which i'd guess WHY it sounds so bad?! Now that i know this - i WILL need help adjusting the bias (if someone would be so kind 🙂) and how i go about it... Assuming i'm meant to measure at idle with no input? 🙂 Thanks heaps 🙂
Most of the components in the output, transformer, etc aren't cheap, but the pre-amp board has the cheapest of cheapest components on it (granted, not the NE5532, but it's the parts around it that's the problem in my book!) it's mono each side, that's why the extra wire has been cut off (it's a bit dodgy like that, thats what i meant by poorly built, they simply cut the extra wire not needed off, rather than not having it in the first place i'd guess?
I'm going to DEFINATELY upgrade the wiring, meaning power wire, coax wire, etc, i've got some high quality stuff that will allow me more length to properly route it!!!
The sub it's driving is dual 2 ohm so no, i've not noticed any difference between running 2 channels and bridged...
The silver zener diodes near the audio input (from preamp) get HOT HOT HOTTTTTT one side is 90 the other side (the part of the amp in photo's) gets 97deg!!!! at idle... which is surprising and needs attention ASAP in my books!!! I'm definately going to look at what i can substitute it with, have it sit up slightly higher and have fans blowing through the heatsinks 🙂
Thanks for the help so far it's been great!!
Aaron
NUTTTR said:
As far as I could see, Jacco just said it had a 1500 VA transformer which would not be cheap. Nothing about per side or dissipation.
NUTTTR said:
Correct.
NUTTTR said:
Yes, you do measure with no input signal and no load connected. However, I suspect that there is no facility to adjust bias as the output stage is a pure class-b design. It's the price you pay for a budget 'pro' amp. If you did manage to hack the thing to implement a bias adjustment, the chances are the resultant increase increase in heat dissipation may push things over the limit. IMO you are pushing it running at 2 ohms per side as it is and by upgrading the PSU etc. you only serve to stress the amp more.
Upgrading the PSU is fine if you run at lower power as you will benefit from a stiffer supply, but if you are doing it to squeeze more out of the amp, you are treading the path to quicksand IMO.
sam9 said:
I agree
you must try to deconnect the protective circuitry
Trying different PSU caps ( of physically bigger size ) is also very important
And check that the op amp pins+/- are correctly decoupled to grd ( for example 100uF electrlytic//100nF ceramic )
alain
jacco vermeulen said:
That's what i was meaning 🙂
Thanks for the comments richie, i was ultimately out to get more out of it, yes, however even keeping it making rated power is good too... More stable PSU is needed (IMHO) and class b is why it'd sound like it does!!! I didn't even think twice about that, the literature that came with it claimed it was Class A/B from memory, but no matter...
So obviously the difference between class A/B and B is simply bias? Class B must have a fair bit of crossover distortion too i'd guess (hence the sound)....
I COULD hack it - however i'm going through upgrading everything anyway so i'd be happy to give adding bias a go - can anyone provide me with some idea on what to do and where to start? I'm new at something like this (i've never played with class b before!)... No wonder it stays ice cold the whole time!! Heat shouldn't be an issue, so long as it doesn't thermal runaway!! ROFL, that'd be my luck!!
So in adding bias, EVEN just a bit to make it more a/b rather than just b should affect the sound a fair bit, as class a/b sounds about 50,000,000 times better in my books (My car stereo has 4.5K worth of amps alone! I like the quality!).....
Thanks for the help folks 🙂
Aaron
rha61 said:
Excellent i'm glad it was my thought too 🙂 I've got to look at that and make sure everything is right in it, bigger caps are on the books so to speak 🙂 Is there anywhere over there (assuming US?) that i can get bigger caps, no one in Aus really specialises in that and when they do they charge stupid amounts for them!
Thanks 🙂
Aaron
NUTTTR said:
My apologies, I missed that bit. However, I must point out that Jacco made 2 errors. Firstly 5 pairs per side would be 5x 150W whereas Jacco has used 10x 150W to give 1500W. But even this is wrong. The power of an output stage is not just a straight multiplication of the device rated power. You have to take the device Safe Operating Area (SOA), thermal resistance and temperature into account. This can mean that you may only get 75W out of a 150W rated device in some applications.
NUTTTR said:
Yup.
NUTTTR said:
It may be possible, or it may not. It depends on what components are already in place. If the components are there, then it's a case of swapping some out and/or piggybacking a bias generator/Vbe multiplier in there. At the other end it could mean cutting tracks and piggybacking parts on, which may not be possible due to space constraints. IMO as a newbie I wouldn't be attempting this just yet. You need to learn up a fair bit on how amp output stages work before you can begin to analyse the amp and ask how to mod it.
My guess is that the output stage could be something like Rod Elliot's P68 subwoofer amp.
As for protection circuitry, IMO as this is a budget amp they will have kept it to a bare minimum. This means just current limiting the output stage. This is simple to remove, just one small transistor for each half of the output stage. Look at Rod Elliot's P27 guitar amp for how it's implemented.
Thanks richie - i'm not a newbie at it as such, but i'm not familier with class b or playing god with bias 😉
I've been looking for stuff to read up on it, so thanks heaps!!
It's fairly easy to hack/mod this amp due to the design, i mean there's nearly enough space to drill holes in the boards to mount extra parts 😉
Thanks
Aaron
I've been looking for stuff to read up on it, so thanks heaps!!
It's fairly easy to hack/mod this amp due to the design, i mean there's nearly enough space to drill holes in the boards to mount extra parts 😉
Thanks
Aaron
- Status
- Not open for further replies.