Improving amp

[moriendi]

Hi fellas,

I have some Behringer Truth 2031A active monitors and I'm looking forward on improving amps a little bit.

First I have replaced in both amps bulging main caps , blown caps in bridge(red wimas) and added heatsinks to 7915 and 7815:

front: http://i48.tinypic.com/jq29gz.jpg
back: http://i50.tinypic.com/oswy9j.jpg

Lot of NJM4580 opamps, input volume adjustment, switches for muting woofer or tweeter, eq adjustements etc.

First, maybe changing these caps:
http://i48.tinypic.com/ixgzdh.jpg

Amp is full of them, any better at this size?

Thanks for thoughts and ideas.

 

[Sebastiaan]

Dear Moriendi,

Interesting compact design. I had one Truth 2031A here around too, so after your post I opened it. They paralleled two LM3886's for the woofer.

Funny side note, to see that many people on this forum seem to have problem with paralleling two LM3886's while Behringer did the trick cheap and dirty (no matched resistors, 1% ratings, no decoupling, no Zobel, no servo) and it works!

Like all Behringer equipment it is full with NJM4580's. The sound really bad. I suggest to replace them for OPA2134's if you like a darker/warmer sound, or for LM4562's if you like a very detailed and clean but somewhat lifeless sound. Above is of course subjective. C11,C3 and C9 are in series with a 1K resistor to ground forms the feedback-loop in al three amplifiers. Those capacitors have a high influence for signal quality. Replace them with better. For the other caps in the signal flow, you have to search yourself, since I am to lazy now to investigate further.

Another note, All LM3886's chips are not decoupled. Is strongly suggest you solder good quality (wima MKS) 100nF caps from V+ (pin 1+5) to ground (pin8) and from V- (pin4) to ground underneath the board.

If you do above, replace the opamps, replace those critical capacitors, and add the decoupling, I think this monitor has a very decent and good sounding amplifier stage.

With kind regards,
Bas

 

[jackand08]

As most of the components are surface mount, it might be difficult to change the parts. Anyway I'll carry on.

You could try replacing the capacitors in the signal path with some better quality ones from a respected manufacturer such as Panasonic's or Nichicon's. This should improve the sound.

Also, try replacing the opamps for some higher quality ones. It's up to you as to how much you spend, just have a look around the forum for people's suggestions and opinions and then decide based upon that. It may be worthwhile to solder sockets in instead so you can test different opamps. A socket may be quite difficult to solder to a surface mount board though.

You could also try moving the power supply so that it was located off-board. This would allow you to use bigger filter caps a larger, more capable rectifier and possibly a higher VA transformer if you really wanted to .

That's pretty much all from me, but I'm sure that they should help to improve sound quality.

 

[ginetto61]

Hi fellas,
I have some Behringer Truth 2031A active monitors and I'm looking forward on improving amps a little bit.... .

Good evening !
I would be interested to know your opinion of the drivers used, particularly the woofer
I had the opportunity to see one monitor open and felt the quality of the woofer debatable, more than the amps
Thank you very much and kind regards,

gino

 

[moriendi]

Good evening !
I would be interested to know your opinion of the drivers used, particularly the woofer
I had the opportunity to see one monitor open and felt the quality of the woofer debatable, more than the amps
Thank you very much and kind regards,

gino

Salve Gino,

tweeter is better quality than the woofer even it has pretty bright sound due to aluminium diaphragm. Woofer is really muddy and lacks details in midrange, however, I have some parameters for finding suitable replacement:
4 ohm
90db sensitivity
21,5cm exactly hole spacing
~130w power handling
crossover at 1000hz

There is a lot of suitable Seas, Scanspeak etc. replacements.

For others, I was going to put some Nichicon Muse ES(green) 10uF and 47uF caps as replacement and I didn't had steady hand so I accidently damaged pcb(read: fried my finger on solder tip and went with solder tip through hole on pcb) so now I have constant "broom, broom" when I turn them on. I guess I'll need to bring them to someone skilled to fix my idiotic mistake.

 

[abraxalito]

Improving the grounding would be my first effort. The underside of the board shows a rather half-hearted attempt at star point grounding - for no more effort they could have gone all the way. Also the balanced input looks like it might just have a pin1 problem but there's a shadow in that corner of the board. Its not recommended to connect XLR pin1 (ground) to the circuit board 0V but rather direct to the chassis.

I'd also check the grounding to the analog supply regs is single pointed too. Don't do what Bas has said in decoupling the 3886's, rather decouple them either V+ to V- or give the decouplers their own dedicated ground direct from the star point. A cap from V+ or V- to signal ground will only corrupt the local ground and result in a gritty sound.

 

[Sebastiaan]

Improving the grounding would be my first effort. The underside of the board shows a rather half-hearted attempt at star point grounding - for no more effort they could have gone all the way. Also the balanced input looks like it might just have a pin1 problem but there's a shadow in that corner of the board. Its not recommended to connect XLR pin1 (ground) to the circuit board 0V but rather direct to the chassis.

I'd also check the grounding to the analog supply regs is single pointed too. Don't do what Bas has said in decoupling the 3886's, rather decouple them either V+ to V- or give the decouplers their own dedicated ground direct from the star point. A cap from V+ or V- to signal ground will only corrupt the local ground and result in a gritty sound.

Dear abraxalito,

Sorry but I am disagree. I have the board here in front of me. The visible ground line you see, only goes direct to the ground pins of the chips, which only use them for mute reference) there are no other parts connected to this particular ground tie. It is not a problem to connect de-coupling capacitors to this ground, because the other ground (feedback, input enz.) run on separate tracks individual back to the star point. For fun I measure the outputs of the amplifiers. The seem to deliver more power then Behringer promise. Second there are small oscillations around 500Khz. But it doesn't say this happened also when connected to the drivers. The outputs have no zobels or inductors, so the wires from the amplifier to my dummy load can cause this oscillations. However after soldering 100nF from supply pins to the ground this oscillations almost disappear.

You are right about the "pin 1" problem. Disappointing, because I know on other Behringer electronics they treat the pin 1 the right way.

To solve this, Cut the track of the middle pin of X3, and solder a (thick) wire directly on 1 of the XLR and connect it to the chassis as close as possible to the connector.

With kind regards,
Bas

 

[abraxalito]

Dear abraxalito,

Sorry but I am disagree. I have the board here in front of me. The visible ground line you see, only goes direct to the ground pins of the chips, which only use them for mute reference) there are no other parts connected to this particular ground tie.

I defer to your better knowledge of the board, I was just guessing that that ground was also being used as a signal ground. Thanks for correcting me

Second there are small oscillations around 500Khz. But it doesn't say this happened also when connected to the drivers. The outputs have no zobels or inductors, so the wires from the amplifier to my dummy load can cause this oscillations. However after soldering 100nF from supply pins to the ground this oscillations almost disappear.

I'd guess and say that a zobel might well correct those HF oscillations. But its only a guess - generally the kind of oscillations a zobel is intended to prevent are parasitics in the output devices, hence considerably higher frequencies (>10MHz or so).

I also notice they've got ground fill around the opamp circuitry - are they decoupling to this ground fill or just using it as a signal reference? Its not going to sound as good as having individual ground returns.

 

[Sebastiaan]

I defer to your better knowledge of the board, I was just guessing that that ground was also being used as a signal ground. Thanks for correcting me



I'd guess and say that a zobel might well correct those HF oscillations. But its only a guess - generally the kind of oscillations a zobel is intended to prevent are parasitics in the output devices, hence considerably higher frequencies (>10MHz or so).

I also notice they've got ground fill around the opamp circuitry - are they decoupling to this ground fill or just using it as a signal reference? Its not going to sound as good as having individual ground returns.

Dear Abraxalito,

yes it is very hard to see on images. There are many "vias" for the very small ground tracks on the board. Was only able to see the tracks by keeping the board under bright light, so I could see through the board.

I am sudden not so sure anymore. I open one of my other "Truth" (what a name ) monitors, but they have a different revision board, with a changed grounding scheme, and a slightly changed layout. So it is kind of confusing..... That also explains why one monitor had more His then the other.

Btw. slightly off topic. The new Behringer B3030A are a bargain for the money and do sound much better then the first generation. The ribbon tweeter is actually very good in my opinion, and the woofer is improved, and doesn't sound "muddy" anymore.I think with some clever mod's you can make this a very good monitor. For the low price it is worth the experiment

BEHRINGER: B3030A

With kind regards,
Bas

 

[Sebastiaan]

I also notice they've got ground fill around the opamp circuitry - are they decoupling to this ground fill or just using it as a signal reference? Its not going to sound as good as having individual ground returns.

I took a quick look with multimeter in hand

Yes, They mix ground-plane and star ground. The ground-fill for the line level stage is eventually connected to the star-ground point, but as far as I could see this plane is only reference for signal related components for the filters. (Hence non of the opamps have any power supply decoupling ) The regulators and additional capacitors are grounded with a individual track to the star-ground and not connected (directly) to the ground-plane.

To be honest, In almost all active studio monitors I've seen, they are using only one ground-plane for everything, the line level and PSU decoupling. So yes in general all types of current are mixed up. I think Behringer did a good job for the minimum money the design may cost. But the Pin1 error is unforgiving after the many many AES papers and tutorials given about it.

With kind regards,
Bas

 

[abraxalito]

Thanks for the recommendation - that's USD240 per speaker, so $480 for a pair? I agree, a lot of kit for the money. Still quite a lot more though than my (lower-tech) Fayou XPEs which are around $185 for a pair. Equally far off topic, I've been playing with a very small pair of actives - QMS Q4 this week ($280 a pair). At first I wasn't impressed with the sound, couldn't decide if it was the tweeter that was harsh. After fixing the grounding (which was more royally screwed up than your terribly-named Behringer - I'd be tempted to rename them 'Pravda' to emphasise the irony ), I found its really sweet and airy. So call me a recent convert to ribbon tweeters

 

[Sebastiaan]

Thanks for the recommendation - that's USD240 per speaker, so $480 for a pair? I agree, a lot of kit for the money. Still quite a lot more though than my (lower-tech) Fayou XPEs which are around $185 for a pair. Equally far off topic, I've been playing with a very small pair of actives - QMS Q4 this week ($280 a pair). At first I wasn't impressed with the sound, couldn't decide if it was the tweeter that was harsh. After fixing the grounding (which was more royally screwed up than your terribly-named Behringer - I'd be tempted to rename them 'Pravda' to emphasise the irony ), I found its really sweet and airy. So call me a recent convert to ribbon tweeters

Wow! Thank you for your recommendation! This is really interesting, and amazingly cheap. What kind of amplifiers they use in the QMS Q4?

Hifi get so much more fun those days, since technology get so much cheaper

With kind regards,
Bas

 

[Sebastiaan]

Thanks for the recommendation - that's USD240 per speaker, so $480 for a pair?

Price is for a pair. But with some smart shopping I saw them already for $339 a pair.

with kind regards,
Bas

 

[abraxalito]

Wow! Thank you for your recommendation!

Do you mean the Fayou's or the QMS? I'd recommend the QMS if you wanted a really small one, otherwise the Fayous are better bang for the buck. But I'm not sure if they have any distribution beyond China. With enough interest though, they might be persuaded Trouble is that the price tends to double (as a minimum) when that happens as the western distributor and retailer take their slices of the pie.

This is really interesting, and amazingly cheap.

I've put some links on my blog as well as details of the mods. Welcome any feedback you have, naturally!

What kind of amplifiers they use in the QMS Q4?

TDA7265

Hifi get so much more fun those days, since technology get so much cheaper

Yes, chip amps rule

 

[Pocoyo]

how big diameter of caps supply ?
if 35mm change that with elna for audio
i hope theres a lot improve

change the diode with Mur 860
change the wima with vishay roederstein 1837

remove terminal input & output, make direct solder to the board
(change the output cable with good speaker cable)

 

[moriendi]

how big diameter of caps supply ?
if 35mm change that with elna for audio
i hope theres a lot improve

change the diode with Mur 860
change the wima with vishay roederstein 1837

remove terminal input & output, make direct solder to the board
(change the output cable with good speaker cable)

Supply caps ars 25mm diameter, how to fit mur860 due to leg spacing, they're pretty narrow, do they require heatsinks?

Sorry, if questions are dumb, but I'm pretty new to all of this.

However, I now have bigger problem, it seems that my pcb is irreversably damaged, how to get a new one? There is no Behringer service center in my country. Can you guys help me with that? To check at your countries, I'll pay to someone via paypal or something else of course.

 

[Pocoyo]

Hi

how to fit mur860 due to leg spacing
just bent the lead until left & right feet same big as the old diode

 

[Pocoyo]

For others, I was going to put some Nichicon Muse ES(green) 10uF and 47uF caps as replacement and I didn't had steady hand so I accidently damaged pcb(read: fried my finger on solder tip and went with solder tip through hole on pcb) so now I have constant "broom, broom" when I turn them on. I guess I'll need to bring them to someone skilled to fix my idiotic mistake.

1
try to look slowly ... all you need in here just work slowly
are there any solder gone to the other rails

2
Take a picture once again after your board was broooom ... broooom
take the up and under the pcb, i'm sure a lot of people will help you

best regards, jeffry

 

[moriendi]

Bumping this topic.

Got a new board from Fullcompass ( Behringer Q05 25204 00674 PCB,MAIN B2031 | Full Compass )and ordered new LM3886TF for that board, found them on ebay.

I don't understand why, but this board which is basically the same except it's Xunda's caps everywhere and on old were Samxon and LM3886's are isolated versions... Has a completely different sound. Lacks bass, highs are extremely detailed than before and somewhat harsh.

Now I'm in trouble, I'm left with two completely different sounding speakers. Maybe they need some burn-in time?

 

[rfenergy]

Dear abraxalito,


You are right about the "pin 1" problem. Disappointing, because I know on other Behringer electronics they treat the pin 1 the right way.

To solve this, Cut the track of the middle pin of X3, and solder a (thick) wire directly on 1 of the XLR and connect it to the chassis as close as possible to the connector.

With kind regards,
Bas

Hello Sebastiaan,

I have the rev.1 B2031.

Will this improve the sound?
What (or where) is the middle pin of "X3"?
Pin 1 is shield from the xlr, should this be done to the TRS jack as well or are they tied at input?

Thank You,
Robert