Schematic Request: Behringer EP1500

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Another two important advantages of this "grounded output" topology is the absence of the classic VAS (voltage amplifier stage) and its non-linearities, and the fact that the differential input amplifier is powered with +/-15V regulated rails thus allowing to achieve better performance.

Concerning the QSC versus Behringer dilemma, I agree with sakis. There are plenty of factories in eastern countries working both for "first order" and "second order" brands, and the latter usually end up selling what the former has discontinued and replaced with newer stuff.
 
EP1500 Circuit - grounded topology

Now had a look at the QSC circuit - the main voltage gain is just in the last stage (the combined driver transistor and 'upside down' emitter follower quads). One still has a VAS, it's just handling a bit more power and presumably working in class AB ??. I'm not too keen on the design - but I bought my EP1500 before I knew and it works fine and sounds alright.

There was another 'grounded topology' style of circuit used in the UK C-Audio ST400 or ST600 Amps, except they used MOSFETs rather than transistors.
The C-Audio amps basically consist of an NE5534 driving the 3-4 MOSFETs directly in a grounded source configuration - thats it !
I think that is asking a bit much of an Op-Amp.
I rather get the impression it is done for economy in these - it means one does not need a large drive voltage for the output devices as they function as voltage amps rather than just source followers

The problem is that the open loop gain rises hugely when one disconnects the speaker - and if one is unlucky it can oscillate at RF.
I don't think the Behringer Amps are subject to this problem (and I eventually sorted the C-Audio amp after a bit of tweaking), but its made me wary of this configuration.
 
Eva said:
Another two important advantages of this "grounded output" topology is the absence of the classic VAS (voltage amplifier stage) and its non-linearities, and the fact that the differential input amplifier is powered with +/-15V regulated rails thus allowing to achieve better performance.

I wouldn't exactly call the op-amp supply "regulated" in theQSC amps. A couple of zeners and boot strap caps - and the op-amp rails do drop as the base current into the OPS goes up. It's actually part of the current limiting scheme. Get rid of this, add another current gain stage, and run the op amp off a truly regulated supply and the SQ would be improved.


Concerning the QSC versus Behringer dilemma, I agree with sakis. There are plenty of factories in eastern countries working both for "first order" and "second order" brands, and the latter usually end up selling what the former has discontinued and replaced with newer stuff.

The circuit is just dirt simple to copy and (almost) foolproof. The RMX is not discontinued. I have a brand spanking new RMX5050 (sub duty in my second system). I wonder if anyone is cloning those yet?
 
this is basically how that output stage works..... nothing really complicated about it, just turns your brain inside out the first time you see it...... all that's really been done here is the ground is moved to the other side of the load and the power supply center point floats with the signal instead of the amplifier center point floating with the signal. this also means that if you were to scope the rails on a working amp, the rail voltages have audio impressed on them as well.
 

Attachments

  • qscout.jpg
    qscout.jpg
    26.2 KB · Views: 526
practice also

show that amps like that are solid rock ..... meaning that cause of the simlicity of the design is not much that can fail inside the amp .....

some days ago i finished a couple boards of a moded clone of the hill audio chameleon that is based on this kind of simplicity

like 15+15 opamp class AB +B sziklai idle 120ma rails 63+63 and out almost 180w@8R.....with 4 pairs of 1943+5200 per board

the amp was finished some 30 days ago and used so far about 60 hours on full power at various PA application....

cliping is very soft distortion figures are fine compaired to the output power and amp is very happy to drive any small PA duty like monitors or subs .at 4 or 8 ohms ...

the quality is nothing really impressive but generally the amp works really very fine ....

( in one set up i had to play with a couple of monitors 15"+ huge piezo made from a company called Dibeisi its the cr*piest speaker you can imagine costs about 80 euro the all thing ( box woofer piezo )

so i said to my shelf i am going to seriously push them to hard cliping and if any fail ( speaker or amp ) i will blame it on the speakers..... ( belonged to the costumer he he he )

well compaired to the trafo 1KW 46+46ac, the banks 33.000+33000, and the 180w boards i presumed that all that power was delivered to the speaker heavilly ......

it was producing a totaly distorted sound for all night since clip led was most of the times constantly on .....

nothing happend though ....amp was fairly hot ( sinks +vents ) but the speaker was still working by the end of the night ....

final conclusion : is that for a simle cheap to make amp is a hell of performer and also cheap power might never produce any quality but it can be very stable if designed correctly


happy chameleon sakis
 
More QSC Info

Someone asked what the "QSC configuration" was called. They call it a "grounded collector" configuration and it is true as stated that it allows the transistor's heat spreader to be tightly coupled to the heatsink for better thermal performance.

There's also another design feature that is subtle but important: The speakers are effective AC coupled to the amplifier via the filter capacitors. This is the part that blew my mind when I first started working on QSC amps. The importance of this design feature is that in the unlikely case that you short an output transistor, it is not possible to pass DC to the connected speaker which usually roasts the voice coil quickly or causes mechanical damage.

As a side-note: Peavey for years used a crowbar circuit in the CS series amplifiers that would effectively short circuit the amplifiers output terminals in the event that it sensed DC on the output. They did this to protect the speaker when an output transistor shorted as they did not use this sort of AC coupling. After repairing at least a hundred of those CS series amplifiers I suspect that the crowbar may have been false triggered in many cases by clipping distortion or by oscillation due instability when driving low impedances and led to a lot of blown transistors that would have otherwise been perfectly okay.
 
"The speakers are effective AC coupled to the amplifier via the filter capacitors."

Only on the smaller QSC, the larger ones were all DC coupled with the speaker hooked to the center-tap of the power transformer (just like the Behringer EP2500 too).

"After repairing at least a hundred of those CS series amplifiers I suspect that the crowbar may have been false triggered in many cases by clipping distortion or by oscillation due instability when driving low impedances and led to a lot of blown transistors that would have otherwise been perfectly okay."

Not so (the Behringer EP2500 has a crowbar too, as does the QSC RMX2450).

The crowbar has an integrator and takes quite a while to trip, even with DC present. The trigger SBS was selected for about 8V~10V.

If you would turn off the DDT protection and drive the amp hard into clipping with asymetrical waveforms it will produce enough infrasonic energy (40V below 5hz) to be sensed by the crowbar.

Previous repair attempts generally lead to shorter mean time before failures.

One frequent cause was replacing the crowbar'd fuse with the same value glass type, and turning it on again. The new fuse would blow right away as there was something wrong in the first place (causing the crowbar to activate).

The amplifier will now be extensively damaged because of the glass fuse (the original being a ceramic type). In addition to outputs, we will now have emitter resistors to replace, and the crowbar triac. If the repair tech missed an emitter resistor, the amplifier will work , but the power must now be carried by fewer outputs.

The foldback current limiter transistors may be open too, the problem here being that the amplifier will work after being 'repaired', but not have current limiting. These amps had their 1/4" output jacks used (or the same on the speakers), if a cable got tripped over and pulled out with the amp on, no short circuit protection.

Replacing only the outputs that were blown this time. That's fine if the amp is new, after it's a couple of years old all the outputs are stressed and need to be replaced.

Re-using the rubber output insulators. They're not re-useable, and early non-filled ones are too high of a thermal resistance and they should all be replaced anyway.

Really old CS800s develop leaky transistors on the drive boards, and may cause the crowbar to trip on power-up. If you disconnect the crowbar the amp seems to work fine. Replacement driver boards are inexpensive, less money than the time you spend looking for the bad parts.
 
another thing with the CS800 i saw frequently was the wiring harness, where it passed under the output transistor board. the bundle of wires is twisted together and there is barely enough room for it, as a matter of fact the board actually puts a bit of pressure on it. with the rough handling road amps get, the board pinches the wires, and eventually two of them will short, or there will be a short between one of the wires and the board or ground. result...a molten pool of charred insulation and wire, and sometimes blown output devices or a charred section of output board. i; ve begun fanning the wires out of the bundle before it goes under the board to prevent this from happening. maybe this adds a little 2nd harmonic to the output through magnetic coupking, but it prevents the charred mess i have seen in several of these amps so far.

another thing about these amps is theat after replacing the bias diodes, fill the hole the diode is in with heatsink grease, even if there was none there before.

somewhere i have seen a chart that cross references peavey transistor part numbers to common 2N MJE, and MJ devices.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.