my poor power amp... :-(

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi everybody,
a few months ago i was thinking about building a power amp but i bought one for less than all the parts cost me so... you know how it goes.

Anyway, came home today from a party switched on the amp to hear a nice reasonably loud humming. Well actually its about music listening volume. I turned it off straight away and went and had a word with my brother and he said he hadnt had it up loud and i believe him because he payed halves with me.

Ive had the lid off and done some testing and the humming/buzzing is independent of the two level controls on the front, you can still play the music sort of behind it, my multimeter is showing 108Hz on the speaker outputs.

Im thinking power supply because of all the above.

I'll do some more testing.

I can also supply pictures.

any ideas?

Thanks,
Tim
 
yeah, it sure is, measured DC voltages accross the two psu caps (all dissconected and out of the amp etc) and one of them had no volt and the other had 60V ish

What do you think could have killed it?

it also has no voltage on it so im not sure about ordering another one.
it says....

6800uF 80WV

any ideas? could it happen again?

Would i get any gains out of upgrading them while im here?
 
Measure the AC output of the transformer. Multiply that by 1.414. You need capacitors to withstand that sort of voltage. From the sound of it, you need 80-100V capacitors (the 100V ones are probably going to be the easiest to get).

What make/model amplifier is it ? Any chance of some internal pictures ?
 
yeah wacked a 4700uF 80v cap that i had lying around in and it seems fine. doesnt get warm and the hum has gone. I will buy a replacement that is the same as the previous one.

Yeah its a ProSound 1000w peak into 8 ohm
in a 2u rack mount case weighs a ton
i wasnt expecting to much from it but it has surprisingly good sound

its got a 1000VA transformer (wouldnt want to have to replace that!), 25A bridge rectifier some protection circuits, then on a separate board there is the amp, 16 large transistors on a fan cooled heatsink. level controls on the front and clip indicators etc. speakon outputs etc.

When i switch it on my computer UPS battery backup supply complains about the drop in power for about 1/2 a second on the ring main....

i'll get some pics for you when my brothers finshed with the camera

The fan is good for when you have a party but is annoying when you want to listen to music quietly so i made a controller circuit for it. The fan is off until a certain temperature then it kicks in when its needed. i still want to make a thermal cut-out with a couple of thermal switches so i cant blow it up.

tim
 
apparently its meant to be

2x400Wrms into 4 ohm but i dont really care too much, you dont need that sort of power in a house.

What has surprised me the most is i can play it so much louder than my old stereo and not even notice because it still sounds clear.

On my old stereo it actually sounded "loud" because it couldnt cope with it.
 
Might be best to replace both of them if one has popped - could be that the originals are from a faulty batch. 6800uF is a bit weedy for such a high power amplifier though, I would have expected at least 2x6800uF per rail.

I would check the bridge rectifiers. If there's a 1000VA transformer in there, i would hope there are two! (one per rail). They should be big 35A jobs bolted to the case. Make sure none of the diodes in the bridge are short.
 
yeah the bridge rectifier is a big metal one with 4 6.3mm (?) push on connectors. 2 are marked AC and the others are +ve and -ve. bolted to the case, not found it to get even warm yet though.
yeah i double checked it and its fine

no the caps were fine, we did some playing at college connecting caps up the wrong way round and over-voltaging them by about 2 times.... if you put them in a metal tube you can get some distance on them....

i was thinking last night... could i make a separate PCB with lots of caps in parrallel and then link it in with wires or would this introduce lots of noise?

I gather the capacitors smooth out the rectified AC but do they also supply that extra bit of power (quickly discharging) for the loud low frequency sounds? Or do they just have to suply enough power and not leave too much ripple?

Tim
 
You sure can build a capacitor bank and wire it in if theres enough room in the case. Keep the leads as short as possible, and use good thick ones, 18AWG or so.

Ideally you want 2-3 parallel caps per rail, bypassed by say 100nF film caps. I would go with 3x4700uF if you have the room, if not, 2x6800uF (this is per rail, so 6 or 4 capacitors total)

http://www.tnt-audio.com/clinica/ssps1_e.html is a good article about amplifier power supplies
 
switched on my amp today and guess what....

the other power supply cap had gone, completely, same as before

replaced it its fine. this looks to me like old age :)

and while im here, here is an image of the inside of the amp.
 

Attachments

  • dscn5084.jpg
    dscn5084.jpg
    93.6 KB · Views: 311
yeah, they are the caps i replaced the broken ones with. They are 4700uF i think :)

the old ones were 6800uF and much bigger cases (dont worry the voltages are fine)

but would say, 10000uF capacitors make any difference to the sound? The bass isnt sounding as nice as it did but im not sure whether its just this song.

Tim
 
yes,

now, if i'm correct, the capacitor is used in an AC/DC power supply for two main reasons. First, to filter AC ripple, which what you have now does that.

The other reason is to keep up with power requirements of an amp. When a deep bass sound goes through an amp, many amps are drawn from the power supply. The power supply cannot keep up with it, so thats what the capacitor does. When the amp isn't drawing the full power of the power supply, the caps take that time to charge back up. Then when it needs it, the instantaneous power is much higher than with low value caps. So, higher mf rating, the more bass you can push through an amp without distortion.
 
soundNERD is correct, decreased bass-performance is the result of substituting the capacitors with a smaller capacity. Replace them with good quality items of at least 10,000 uF each.

I took the time to read the article about power amp power supplies that jaycee posted and learned a great deal from it.

Now that I know a little more, it seems incredible that your amp has a big toroidal transformer and only some puny capacitors... And it's not only the capacity that makes a difference, also the quality (slew rate) of the capacitor.

For comparison:
- Marantz PM40SE (2 x 55/75 W into 8/4 Ohms): 2 x 10,000 uF,
- Marantz PM80SE (2 x 120/180 W into 8/4 Ohm) 2 x 18,000 uF.
In both cases the capacitors are Elna "for audio".
 
timfishy said:

I gather the capacitors smooth out the rectified AC but do they also supply that extra bit of power (quickly discharging) for the loud low frequency sounds? Or do they just have to suply enough power and not leave too much ripple?

Tim


It is as simple as the fact that the energy only comes from the line. Capacitors don't make energy they just store it. The amount of charge (coulombs) that is stored is a fuction of the capacitance value and voltage. When you first power the amp up, the rails have to build voltage. Think of this voltage as the water level, and the capacitors are reseviour tanks. The inrush of water is at a level set by the peak secondary voltage of the transformer and has to fill the tanks also. This is why the lights dim for an instant when you charge them up. Once they are charged and the tanks filled, the circuit will take current(water) from the tanks(caps) in bursts, depending on the output taken. The reason that larger caps are better in bass amps is because the frequency of bass is lower than the refresh rate from the rectifier diodes to the caps. Have you ever noticed that the capacitor value for SMPS's are much smaller? The refresh rate is much much faster. The caps may have to hold a large current through several cycles of the rectifier. If you ask me, the larger the caps or storage tanks, the better the performance and power supply rejecton and less ripple voltage on the rails. This is why the peak output current is much higher than the transformer current rating, but only for a short time. The only limit is the current surge rating of the rectifier and transformer during initial turn on.

Capacitors made today are much smaller physically than older models because of better matierials and construction. They should be used according to the values printed on them. Paralleling caps is no problem as long as the voltage rating is not exceeded. It is sometimes cheaper to use many small caps as opposed to one huge cap, works just as good but looks a little bit more messy.
 
The transformer may cycle 60 times (or 50 times) per second, but how much of each cycle can effectively be used depends on the recovery speed of the rectifier. A slower rectifier will take longer to recover, therefore less of the cycle can be rectified and put to use.

However, if you read the article on http://www.tnt-audio.com/clinica/ssps1_e.html, you will find out that ultrafast diodes will not neccessarily lead to better sound (when used in current gain stages) than good old bridge rectifiers.
 
Administrator
Joined 2004
Paid Member
Actually, with a bridge rectifier, the pulses are 120Hz in North America. The switching time of the diodes is of no importance here. What is important is that you obtain all of your energy in the very brief time the crest of the AC wave is a diode drop above your instantaneous DC level. Thus, you peak current pulses are very high. Also, the inductance of your filter caps prevents them from storing as much energy as they math indicates. Therefore, you want capacitors with low series inductance. In other words, a bunch of small caps in parallel.
I think the worst supply I've seen was a Bedini where the filter caps were huge. The AC ripple was incredible. I can't remember the value, but the hum was more than audible. These were custom caps. So use common sense and parallel some reasonable caps, bypassing them with smaller values may help the sound too.
-Chris
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.