Pro amp to Home amp, mods advice please?

[matt09]

Hi there,

I've recently bought one of these amps use with a subwoofer with intention to mod it a fair bit. I've already replaced the fan and added a speed control, installed pro shielded cables on the input side of things, upgraded caps etc.

Would it be worth changing the opamps from the weird stock 'cool audio' ones, if so what would people reccomend considering I'm using this as a Sub amp? i.e LM6172 etc?

Also in light of the opamp upgrade would it be worth making a Jung Super Regulator for the upgrade as the stock uses cheap SMD mounted regualtors?

Thanks.

 

[dangus]

The only thing I'd suggest is checking where the low frequency starts to roll-off. If you have subs with response approaching single digit frequencies, you don't want your amp to be rolling off at 20 Hz.

I checked the specs for my best pro amp, an Audiopro AP4040: it has a 12 dB/octave roll-off starting at 20 Hz.

 

[Brett]

None of my Pro amps have unneccessary LF rolloff that would affect the majority of real life HT subs.. Check the specs though, and make sure the fan is variable speed as some can be too noisy for a domestic environmant.

Apart from that, use it as-is. No need to mod it for dedicated sub duty and I seriously doubt you'll hear any difference whatsoever in a competently designed and engineered amp from one of the major players.

Pro amps are usually excellent value for money in terms of power and reliability. Might amso be worth looking into the lead-sled slightly older versions secondhand and some usere are moving into amps that use SMPS, Class D etc, and weight should not be an issue in a home install.
I use one in my rig.

Crest CA9 and QSCs are very good and Mackie and Tapco get some good props on some of the Pro fora I peruse occaisionally.

 

[Khron]

From what i've read, some pro amps might be just a bit under-biased (for some thermal economy), so that might be a mod you might want to investigate.

 

[matt09]

Thanks for all the replys, I forgot to mention what the amp actually is, a behringer ep1500, 700 WPC @ 4ohms or supposidly and definitly a budget amp rather than a quality Crown or QMX.

Regarding the roll off from looking at tests done on these amps on AVS Forums, it is only very slight and wouldn't be very noticible in practice I don't think , something like 0.7. Also these amps are used everywhere for home subs so I don't think this is an issue.

I have already put in a fan controller and such as the stock setup was very noisy. The thermal issue won't be a problem most of the time as I won't be cranking it up too high but yes, I have read that the chasis can reach over 125'C

All advice points to leave it as it is then?

 

[carlmart]

All advice points to leave it as it is then?

Without an schematic to hint on what could be upgraded, there's little more than can be said, can't it?

 

[marchel]

How bout more or bigger caps?

 

[Zero Cool]

The Behringer amps are made by the same Asian company that makes the QSC RMX series of amps. QSC makes all but 2 of there lines in house in California. the RMX and one other series are purchased from an Asian company.

FYI, any CREST VS or CA series amplifier should have the power supply caps changed immediately. there was a bad run of caps and caps to close to there Working voltage limits. 90Vdc caps on 93vdc rails etc! and they leak and start fire. I have a half dozen of them stinking up my shop right now!


Back to the subject. Older Crest amps like the 4001, 5001 8001 are monsters that use Toshiba 2SD555 2SB600 output devices, and can be made to sound quite good. they will take having the bias cranked up a bit but you still have to deal with fan noise. but there is enough in the way of heatsinks that you could reduce the fan speed, and even with the bias cranked up for home duty the amp should stay well within limits.

I don't know that there is enough heatsink mass in the Behringer/QSC amps do do that with and get away with it. Those older amps used massive heatsinks and small fans. now days it seems MFG are using smaller heatsinks and bigger fans to reduce weight and cost.

For subwoofer duty. making sure you have a good enough AC line to power that amp i think will make a bigger difference then anything. Power line sag is a form of compression you can hear.


Zc

 

[matt09]

Hi there,

Schematic is here, you will see there are a few undesirable things about the design like multiple parallel connetions to the pre output but I didn't plan on totally canibalising the amp and disconnecting tracks for filters etc as it ruins any possibility of selling it in the future as a PA amp. This is why I came to the conclusion that the sensible things to upgrade would be the preamps and rails.

I have measured the voltage on the caps and found it to be about 80v with the cap limit of 100 so reasonably within the limits there. The output devices in these amps are 2SC5200 FETs and at normal duty they will run fine with no fan at all. When cranked up a fan becomes essential as you'd expect. The Behringer comes with a pretty hefty 7-8" toroid which does the job to the hard limits of the amp where by this would be the bottleneck before lack of power I believe.

Cheers for the advice so far.

 

[carlmart]

you will see there are a few undesirable things about the design like multiple parallel connetions to the pre output but I didn't plan on totally canibalising the amp and disconnecting tracks for filters etc as it ruins any possibility of selling it in the future as a PA amp. This is why I came to the conclusion that the sensible things to upgrade would be the preamps and rails.

I don't see so many problematic areas, and certainly not at the output.

The main area would certainly be the PS caps. Is that all there is: 6600uF on each rail or there's more outside?

Less than 30K on each rail for 80v seems like unthinkable to me for this circuit. But you may need some protection and power-up delay, which I don't know if the Behringer provides.

Capacitor quality is certainly important.

Next think I would try is hanging a zobel from each rail supply, like that suggested in TNT Audio PS articles.

The parts quality would come next, certainly upgrading the 4580 chip to something better. Perhaps a 2134 or something similar.

Filter capacitors, particularly if they are ceramics, should be upgraded, but this being a bass amp might not make that critical.

The next thing might be dealing with the heat, which may need looking at what is being used on the output transistor to transfer heat to the heatsink, which people tend to forget. Heat paste used in computing, like Arctic Silver, may improve on that, or using better isolators, like aluminium oxide.

For such a job you would not need to solder anything.

You can find silent fans now which might improve on what's on the unit, but they tend to run slower to lower noise. So check if they are doing their job.

Let's hope this helps.

 

[matt09]

Hi there,

You're right, I missed the obvious one that there is hardly any capacitance on the rails, it is just 6600 uf per rail, they look pretty generic too.
There is a delay on output I think but that's all.

I presume you mean a modification to the zobel network as there is already one on the output, I looked on TNT audio but wasn't able to find anything specific.

I will look at replacing the opamp and have already changed a few capacitors, presumably it would just be a drop in replacement and no other modifications would be required.

I have already replaced the fan and it seems to run fine on low rpm at mediumish volumes without getting too hot.

Thanks for the advice

 

[carlmart]

You're right, I missed the obvious one that there is hardly any capacitance on the rails, it is just 6600 uf per rail, they look pretty generic too.

Sometimes "generic" can be good. In any case is too little, I think.

There is a delay on output I think but that's all.

See if the delay is on the supply first stage or only on the speakers.

I presume you mean a modification to the zobel network as there is already one on the output, I looked on TNT audio but wasn't able to find anything specific.

No, I meant the PS rails, not the output. Look here:

http://www.tnt-audio.com/clinica/ssps3_e.html

They use 1R/17W in series with 680n, but you can try other values.

I will look at replacing the opamp and have already changed a few capacitors, presumably it would just be a drop in replacement and no other modifications would be required.

I have already replaced the fan and it seems to run fine on low rpm at mediumish volumes without getting too hot.

Yes, it would all be drop in replacements.

 

[richie00boy]

2SC5200 are bipolar transistors. How many per channel does this amp employ?

 

[Brett]

A Behringer. Allegedly a QSC circuit clone. Doesn't man the quality is the same. Dunno about the B and QSC sharing the same Chinese factory. Doesn't mean the parts or QC are the same. However, earlier I was referring to the US made models.

For sub only duty and assuming a <100Hz LR24 or similar, op-amps and bias aren't going to make much difference, if any at all that will actually be audible or measurable. A heftier PS bank and a soft start cct might as well as having a power lead of sufficient gauge to prevent sag as suggested earlier would be better places to spend the effort.

 

[matt09]

Hi there,

Cheers for the help, have everything sorted now I think, the amp has 8 output transistors per channel. I'll pick up some nice caps next time I go to a radio fair as the prices for the sort of voltage and capacitance new is going to amount to half the price I paid from the amp.

I'm a little stumped on the zobel across the supply though, 1R at 17w? and that was for a small supply. With such huge amounts of power are component or specifically wattage values going to be massive.

 

[richie00boy]

Well with only 4 output devices per channel (4 upper and 4 lower) each typically 125 watts dissipation, that gives you a net power output of 500 watts RMS. Maybe the 700 Behringer claim is peak?

The zobel parts wattage is related to the voltage on the supply rails not the amp power directly.

 

[matt09]

there are 8 per channel and the 2sc5200 are 100w each. Assuming 700w that would be 87.5w per chip

 

[AKN]

Hi,

Behringer EP-1500/2500 test:
http://www.avsforum.com/avs-vb/showthread.php?t=855865

 

[richie00boy]

there are 8 per channel and the 2sc5200 are 100w each. Assuming 700w that would be 87.5w per chip

See my post again, there are not 8 lots of dissipation per channel, only 4

 

[RyanW]

I agree with the general opinion of the thread, Supply rail caps are the key! Get whatever you can afford for now and add more later. I have an old Peavey amp; I upgraded the old dead 15 000uF per rail to 20 000uf. The amp is 30 years old and the supply caps must have been near dead. It made an incredible difference. Even the highs were substantially improved. I did not use a soft-start and it works fine (No sparks or light dimming)... Any more than 20 000uF per rail and I imagine it would be necessary.