Aguilar DB750 overcurrent

[Swampy]

Holdup. I repeated the experiment but disconnected the connector to the driver board and the rails have climbed up to + and - 60 volts, with 0v DC on the output, lovely dim glow from the current limiter. Back to the drawing board I go...

[J M Fahey]

Holdup. I repeated the experiment but disconnected the connector to the driver board and the rails have climbed up to + and - 60 volts, with 0v DC on the output, lovely dim glow from the current limiter. Back to the drawing board I go...

Yes, In point (2) I had specified separating the MosBlock from the driver board for independent testing, but you figured it out.
The Mos Block is behaving normally, that is with true balanced "signal" (in this case DC) its output stays close to 0V too. Gooodd !
And with 0V bias, they do not pull excessive current.
I´m satisfied the (expensive) MosBlock is fine, back to the driver board, which we will fix *without* connecting it to the MosBlock for now.
Now I'm starting my workday, but at night will reconnect with some suggestions.
I think this Dragon can be slain.
Somehow the driver block is overbiasing it (although those +.18 and -.78V you measured *looked* very tame) or the Mos driver transistors are lossy or fail under stress and pass too much current *under normal rail voltages*.
We´ll see, won't leave a stone unturned.

[Swampy]

Lifting R521 sees a minor drop in current and rails up to about + and - 10v. Reconnecting and moving on...

I don't believe the driver transistors are at fault, lifting R522 / R514 / R524 sees no drop in current consumption. Unless C508 is leaky...doubt it.

Lifting R505 and R506 sees a very large drop in current consumption and the rails lifts back up to + and - 40 volts approx. I'm not familiar with this branch of the circuit however, it looks like a compenstation circuit that supplies a potential to the differential pairs emitters??

Chur

[Swampy]

Hey everyone,

I've managed to get the voltage rails back to normal on this beast by way of replacing Q503 and Q504. I still have a dc offset of .42 v and the dc offset / balance pot neither effects this value nor the voltages on the diff pair emitters. The bias pot also seems to be non operational, I've already replaced Q510 as I believed there to be a base to emitter short earlier in the troubleshooting process.

Thoughts would be appreciated, this thing is almost back to full strength.

Chur

aguilar_driver_for_web_5.pdf

[Swampy]

Yo,

I put this sucker aside for a bit. Here's where I'm at, can someone explain why the voltages at the nodes of D507 / D508 and D509 / D510 are the opposite of what I was expecting. Also when I measure them they fluctuate + and - approximately 0.30 v with a period of around 10 hz (just watching my multimeter swing), I get a similar + and - 2 mV on the output also?

These measurements were taken with the amp biased to cutoff and drawing 1.2 A from the mains, I don't want to bias it to active region until the imbalance between the + and - mosfet gate voltages are ('drive') are amended.

Any advice? Go for it?

Chur

[mozwell]

Apologies if i am off the mark here
We are just testing the driver, without the mosfet block connected ?
If so, we dont have negative feedback.... connect NFB temporarily to +DRIVE or -DRIVE (doesnt really matter which).
When you apply an AC signal, you shoudl get a big AC output signal on the drive pins, approx 40 times the voltage of the input signal.
Dont worry about the voltages at nodes of those diodes, just measure the +DRIVE or -DRIVE signal.

[Swampy]

Apologies if i am off the mark here
We are just testing the driver, without the mosfet block connected ?
If so, we dont have negative feedback.... connect NFB temporarily to +DRIVE or -DRIVE (doesnt really matter which).
When you apply an AC signal, you shoudl get a big AC output signal on the drive pins, approx 40 times the voltage of the input signal.
Dont worry about the voltages at nodes of those diodes, just measure the +DRIVE or -DRIVE signal.

Mozwell,

The mosfet block is connected.

I've measured 212mV RMS AC at the input on the driver board corresponding to 9.54 V RMS AC at the +DRIVE node.

9.54/0.212 = 45.

and I get a dirty looking measly 15.5 mV RMS at the output?

I tried bringing the voltage diff over the bias resistor up but once I read over 5 amps current draw from the mains and felt how warm the heatsink was getting I thought it was time to let it cool off.

Apparently this amp draws 1800 watts from the mains though, does that mean around 7 amps from the mains (240 V ) in idle??

Also, I'm measuring a potential difference of 1.7 V between the drive - and + nodes...still in cutoff (this will need to be around 7 V before the mosfets enter active region right?)

Chur

[Steve Conner]

Apparently this amp draws 1800 watts from the mains though, does that mean around 7 amps from the mains (240 V ) in idle??

No, the 1800W figure applies when it's flat out delivering full power to the speaker. I'd expect idle current to be 1-2A at 240V depending on details of the design. You should be able to get it down to almost 0 by turning the bias down.

[J M Fahey]

In post 17 I asked you to

I´m satisfied the (expensive) MosBlock is fine, back to the driver board, which we will fix *without* connecting it to the MosBlock for now.

Now in post 22 you tell us

The mosfet block is connected.

[Swampy]

In post 17 I asked you to
Now in post 22 you tell us

J M,

Yes, my apologies as I didn't communicate that I reconnected the + / - drive with the mosfet gate resistors as of this post.

Swampy

Hey everyone,

I've managed to get the voltage rails back to normal on this beast by way of replacing Q503 and Q504. I still have a dc offset of .42 v and the dc offset / balance pot neither effects this value nor the voltages on the diff pair emitters. The bias pot also seems to be non operational, I've already replaced Q510 as I believed there to be a base to emitter short earlier in the troubleshooting process.

Thoughts would be appreciated, this thing is almost back to full strength.

Chur


This morning I have disconnected them again and taken the following measurements.

Aguilar_03_04_12.pdf

There's a negative voltage on the base of Q12, Q511 Q510 and Q509 have been replaced along with C506, I suppose Q512 is the next in line??

Swamp

[mozwell]

Swampy
Lets disconnect the main mos power block & verify the drivers are ok, especially if you are getting 9VAC on the +driver line & 15mV on the output. This would suggest to me the mos block is not ok, but lets verify step by step first.
The mos power stage is a voltage follower, with a voltage gain of slightly less than 1. The sources of the mosfets will follow the voltages on their gates, as these are driven up & down by the driver transistors Q509 & Q511.
In order to make sure these drivers are working properly, we need to do some checks.

Disconnect the mos power block & connect NFB to the +driver signal
With 200mV AC signal on the input, do you get approx 8VAC on the driver output, if so, the driver stage is ok, & the problem is with the mos power stage.
Measure drain to source of all mosfets, they should read high ohms, check the zener diodes D101, D102, D103, D104 are ok, check all the gate resistors are 470 ohm, check R118 & R119 are 0.5 ohm

If this is a 750W amp, the a single 5W 0.5 ohm resistor for R118 & R119 are undersized. 750W into 4 ohm is 14 amp, 14A on 0.5 ohm is 7W, even at 50% duty this is 3.5W on each resistor. It might be worht replacing R118 & R119 with 2x 1 ohm in parallel, in each position.

[Swampy]

Swampy
Lets disconnect the main mos power block & verify the drivers are ok, especially if you are getting 9VAC on the +driver line & 15mV on the output. This would suggest to me the mos block is not ok, but lets verify step by step first.
The mos power stage is a voltage follower, with a voltage gain of slightly less than 1. The sources of the mosfets will follow the voltages on their gates, as these are driven up & down by the driver transistors Q509 & Q511.
In order to make sure these drivers are working properly, we need to do some checks.

Disconnect the mos power block & connect NFB to the +driver signal
With 200mV AC signal on the input, do you get approx 8VAC on the driver output, if so, the driver stage is ok, & the problem is with the mos power stage.
Measure drain to source of all mosfets, they should read high ohms, check the zener diodes D101, D102, D103, D104 are ok, check all the gate resistors are 470 ohm, check R118 & R119 are 0.5 ohm

If this is a 750W amp, the a single 5W 0.5 ohm resistor for R118 & R119 are undersized. 750W into 4 ohm is 14 amp, 14A on 0.5 ohm is 7W, even at 50% duty this is 3.5W on each resistor. It might be worht replacing R118 & R119 with 2x 1 ohm in parallel, in each position.

Mozwell,

I'm 99.9 percent confident the power mosfet section is fine as the section was tested and discussed, concluding in this post.

"The Mos Block is behaving normally, that is with true balanced "signal" (in this case DC) its output stays close to 0V too. Gooodd !
And with 0V bias, they do not pull excessive current.
I´m satisfied the (expensive) MosBlock is fine, back to the driver board, which we will fix *without* connecting it to the MosBlock for now."


Furthermore, with the drive + and - connectors lifted the DC conditions are off and when I fired her up today (with the drive connectors still lifted) R514 went up in a huff...D507 - 510 were replaced prior to this happening, I suspect it's the driver transistors Q512 and Q513 that are shitting the bed.


Can you suggest any other tests that can verify this?

R117/118 are mounted to the power mosfet's heatsink and rated @ 100 W each - http://www.caddock.com/Online_catalog/Mrktg_Lit/MP9000_Series.pdf

Chur
Swamp

[basstard]

Hi everyone!
I'm having a problem with one of these beasts.
Basically it fried a couple of regulators [a 7815 and a 7915], a couple of caps and a bridge rectifier on the PSU board [which got badly damaged].
Changed said parts but when I turn it on it'll give a NASTY noise thru the cab [can't really keep it on - also checked and there seems to be no DC on the output] and the new regulators will get VERY hot in a matter of seconds.
I can't find any schematic around the web and the ones attached here in the 4th post are empty.

Anyone willing to help with schems or suggestions on what to check?

Thanx in advance!

[Jazz P Bass]

What are the voltage regulators 'feeding'?
Opamps?
See if any are getting hot.

[Jazz P Bass]

[QUOTE=basstard;299264]I can't find any schematic around the web and the ones attached here in the 4th post are empty. QUOTE]
That is really weird.
I distinctly remember posting the pdf files (the originals were GIF or something)
Yet when you try to open them, they are blank.
If you search Google Images "Aguilar DB750 Schematic" the pdf files show up as a thumbnail but you cannot view the original.(blank file)
Moderators, any clue to what's up?
Copyright infringement?