Noise in output from Orion 2500D

SQLGuy

Member
2009-01-11 5:18 am
Hi folks,

I've got a couple of 2500D's in for repair. Both had blown outputs. One was blown for one phase, and the other for two phases.

I've replaced all the blown outputs, and a bunch of gate drive resistors, and a couple of burned up gate pulldowns, and rebuilt a couple of the driver boards.

On the driver boards, there were a number of blown KTC1027's and KTA1023's as well as 1Ohm SMD resistors and a Zener.

I replaced the outputs with same parts, but I subbed the KT's with KSA1220A's and KSC2690A's, since I had some on hand, and I wasn't finding the KT's from my regular sources (also, they looked like better parts)...

Anyway, with them back together, both now run, but one of them has about .15V of DC offset and a decent amount of 8KHz or so AC noise. The other seems normal. The noise is there regardless of gain, and regardless of the setting of the low pass filter. Googling around, it seems like a few people reported similar noise issues with these amps, but either didn't find a cause, or had them swapped out by Orion. Anyone here familiar with such a problem? Worst case I'll be doing a lot of scope comparisons between the two amps, but I'm hoping someone's seen (and solved) this before.

Cheers,
Paul
 

SQLGuy

Member
2009-01-11 5:18 am
I just did a bit of comparison.

The main difference I found was a current draw difference between the sides of the drivers using the KT parts and the the ones using the KS (Fairchild) parts. The two sides of the driver boards each have a 330 Ohm resistor on their supply. In the good amp, each resistor drops about 10.8V. In the bad amp, the KT side is dropping about 6.9V and the KS side is dropping more like 11.5V.

At this point I'm wondering whether the Fairchilds are maybe not suitable subs, even though they meet or beat all characteristics of the originals. Either that or I need to use them on both sides of the driver boards.

Anyone know a source for small quantities of the KT (Korea Electronics) parts?

Thanks,
Paul
 

SQLGuy

Member
2009-01-11 5:18 am
OK. I didn't measure the voltage drop across the Zeners, just across the series resistors. I was thinking they were gate protectors because the original Zeners on the blown sides of the drivers were shorted. If they're just shunting current for regulation, I don't see what would have shorted them.

I'll probably take a closer look at this circuit this evening, but I still think the subbed transistors are the issue. Best I can find so far is ordering them from China through eBay. Someone mentioned having bought them from MCM, but I don't find them on their site now.
 

SQLGuy

Member
2009-01-11 5:18 am
I didn't have much time to look at the amp yesterday... maybe more today, but what i did see doesn't yet make sense to me. The "supply" side of the 330Ohm resistors seems to come from the end pins of the driver boards where they connect to the main board, and the other side feeds to collectors of a couple of the small transistors. Thing is, the "supply" voltages are about +/- 22V on the good amp, and the other sides are +/-33V. I am a bit confused as to why this looks backwards from what I would expect.

On the non-working amp, I still have -33V and -22V on the repaired side of the driver boards, but 36.8V and 30V on the side where I didn't find any damage. Maybe there is some damage there and the unrepaired side is actually the problem?

I think I'm going to need to draw this circuit out so I can better understand what's happening (unless someone has a schematic they could share?).

For the subs Sean suggested, unless these are other numbers for the exact same parts, I would still be concerned, as the KS's that I used should be a good sub too.
 

SQLGuy

Member
2009-01-11 5:18 am
I traced further into the circuit this evening. Here's what I'm seeing: rail voltage is fed through the Zeners (supposed to drop 12V) then through the 330Ohm load resistors, then NOT TO GROUND, but instead off the driver board and eventually to a 100uF cap in parallel with a couple of rectifiers, making a little power supply that's fed by a small winding off the main toroid whose output appears to be coupled by a 10uF cap.

I'm not really sure what they're trying to do here, though I get the impression that the power supply of this amp can run at different voltages depending on load. Is this system supposed to provide variable regulation for the driver stage as the power supply voltage changes? If so, this part of it is not working right. The rails for both sides are consistent at 45V. (Same as the working amp.)

Any suggestions from here?

Thanks,
Paul
 
FETs are commonly driven with 10v (Vgs) to ensure that they're switched on completely. They will turn on at about 3.5v but won't completely switch on until the drive voltage is a bit higher.

To get a 10v signal, they need 12v of supply (you lose some drive voltage across the transistors). If they have significantly more than 12v, it will cause the drivers to run hotter (and to some degree will have more trouble getting the FETs to switch off efficiently).

If the voltage is too low for the drivers, the FETs will typically run hot because they're not being switched on completely.

The power supply is likely regulated for the rail voltage. The other windings will produce an output that will vary with the pulse width and the 12v input voltage.

They use the resistors and the 12v Zeners to maintain a constant 12v from a not-so-stable voltage from those windings.

If this amp has windings dedicated to this purpose, that's why the voltage-dropping resistors have a much lower value than the JBL amp. Instead of dropping from rail, they're dropping from a much lower voltage.
 

SQLGuy

Member
2009-01-11 5:18 am
Hi Perry,

I think I didn't explain this clearly. The supply for the drivers is dropping FROM the rails, but it's not dropping to ground. The bottom side of the regulator circuit is to this cap-coupled dedicated supply.

On the working amp, the dedicated supply is providing about 22V or -22V on the bottom end of the 330Ohm resistors. On the non-working one it's providing 22V and -30V.

So, on the six working paths (counting both amps), voltage values are:
Rail: +/- 45V
Zener: Drops 12V to +/-33V
330hm: Drops 11V to +/- 22V
Dedicated supply: +/-22V

On the two non-working paths:
Rail: -45V
Zener: Drops 9V to -36V (since there's not enough current across them to regulate)
330Ohm: Drops 6V to -30V
Dedicated supply: -30V

Or maybe it is a bad Zener? They are effectively in parallel.... on the other hand, with the 330Ohm in there too, I would have expected to see some difference in the voltage drop of a bad one versus a good one; they are in fact dropping exactly the same voltages.

The biggest pain with this amp is not being able to run it without clamping everything down. Pretty tedious... Next step will probably be to reassemble it again, then scope out the A/C from the dedicated supplies... maybe that coupling cap or the filter cap are dry...
 

SQLGuy

Member
2009-01-11 5:18 am
A bit more digging this evening, but no solution:

1. I replaced the 10uF 50V caps that couple the transformer pulse to the rectifiers of the "ground" circuit for the 330Ohm resistors. No change.

2. I replaced the Zeners on the incorrect-voltage sides of the two driver boards. No change.

I also made some measurements. On the good side, the dedicated winding from the toroid is generating 16VAC. On the bad side it's 8VAC. All other outputs from both toroids match.

The windings on the bad side are loose as they wrap around the toroid core. On the good side they're lacquered down. Moving the loose windings around has no effect on the voltage they're producing. Seems odd, though, that the voltage is exactly half on the bad side. At this point I'm leaning towards the idea that it's actually a bad transformer, much as I think that's unlikely.

Don't suppose these transformers are available anywhere?

I may swap one from the good amp to verify whether that is in fact the issue.
 

SQLGuy

Member
2009-01-11 5:18 am
I scoped the two transformers and the square waves on the supply windings and those on the higher-voltage secondaries of both were identical. It was only the small winding that provides regulation for the driver circuits that was different. (20V p-p on the bad side, and closer to 40V p-p on the good side).

Still looking like a bad transformer....