Possible leaking output mosfets in Hafler XL600

Help needed quickly​

I have an XL600. When I put it in stereo, then the right plays just fine, but the left seems to clip at 300mv input. When I run it in BST (mono), the unit clips at about 150mv, and it kicks some kind of DC to the ground. I had the 10 electrolytic caps replaced, and the problem persists. I took it to the shop again, and now they are telling me that the output mosfets need to be replaced on one of the channels. They told me they checked everything. They told me that they even swapped the driver boards. Something sounds kind of fishy. Mosfets usually don't leak current when driven, do they?

I need to know if I should pull the unit or let them replace the mosfets. Please help. They are getting back to me with prices supposedly tomorrow.

Sorry about this. I am new here, and I reached out to Mondialfan who is on a trip. He suggested I reach out to the forum.

Thanks in advance,
Peter
 
Moderator
Joined 2011
If they claimed that replacing the capacitors would fix it, they owe you a free fix.
If you just asked them to replace the capacitors, hoping that would fix the problem,
that was a mistake.

You are putting a lot of money into an old, so-so amplifier. Maybe it's been fed enough
cash. There are newer, cheaper, better ones around.
 
Last edited:
No - I asked them to fix it, and they replaced the caps. Now they are telling me the mosfets are leaking. From what I have been told they usually don't leak like that. They usually blow and take a fuse with them. They owe me a fix, definitely, but why do I feel like I am about to be robbed of my hitachi mosfets?
 
Moderator
Joined 2011
"They owe me a fix, definitely, but why do I feel like I am about to be robbed of my hitachi mosfets?"

Because you are.


Why did they give the amp back to you the first time, if it was not fixed?
Sounds like a sleazy place. Cut your losses now, before you get in even deeper.
How much was the first repair charge? Did they even use high quality capacitors?

If I were extremely cynical (and I am), I would say they want to replace the Hitachi
parts with some random transistors, and sell the Hitachis on ebay.
 
Last edited:
Member
Joined 2005
Paid Member
Even better put a Cordell DHC 200 front end driver board. competes with the best. Leaky fet mmm not so common, they either work or don’t usually. The problem is these guys don’t know what their doing (replacing the old caps is a good idea,, but it wasn’t what was wrong with the amp). And it seems you don’t have the experience to work on it, so your stuck with their nonsense or move it elsewhere
 
Last edited:
I had to inform the tech that it will go down to two ohms in BTL and put out 1500 watts. He was disagreeing until I told him to look at the specs. That is when he stopped disagreeing with me. Now I suspect they want to harvest the caps. I guess the fundamental question is whether they can leak like that. It seems odd to me that they could leak. The cap replacement was $700, and though it is Manhattan, I feel like I got screwed.

Ticknpop - sorry - I want to keep this one stock. It is one of a pair that I like to run in BTL.
 
I do not know of anyone using the Cordell DH-220C driver in a XL600. XL600 uses a separate boosted supply and iirc has fuses on the driver.
Also it does not support btl either
Make sure the shop returns all components to you. If mosfets get changed make sure they use new exicon devices and not old hitachi
 
Last edited:
So - an update. I took everyone’s advice, and I picked the unit up. They didn’t change anything. I am planning to take it to Audio Classics in Vestal NY. I am very grateful to everyone for their advice. It makes no sense to me that the mosfets “leak”. Either they work or they don’t. It sounds fine at 70 watts, but clips at 80? I suspect that it is more likely that the resistors or caps on the mosfets are more likely the culprit.
 
huge room for improving XL600

[sorry to say] the "stock PC40" drivers cards are marginal sounding, at best. I have (4) PC-40 boards, harvested from amps, they all measure differently. I suspect the (unobtainable) front end FETS to be suffering from matching or increasingly out of tolerance over their life cycle. With the absence of DC offset adjustment, I struggle to even consider recapping / rebuilding / selling them.

Having that powerful of an output stage... the driver stage electrolytics need to be much larger to maintain adequate drive on the MOSFETS gates / not clip... not just a meager 100uf. (image 1 after re-work) (image 2 - XL600 as I bought it)

I have come to suspect the 30-40 year old MOSFETS are also showing their age. I am about to pull the vintage MOSFETS and install (tightly matched) single die Exicons. But I am waiting to make sure the amp is burn in.

The single power supply (even with *dual rectifiers) is anemic, on both the 500 and *600 amps. Having the L/R driver front ends driven from a shared mono power supply, was one of the disappointing things that I have seen in design, done on the cheap... I have been experimenting with several advanced P/S designs, doubling and even quadrupling over stock, there is so much room to refine it's design and performance.

(image 3, Franken'amp) comparing 500 to 600 power modules. I have built and studied linearly over numerous builds, to establish a hierarchy of importance of what "will improve" and lend to meaningful differences, as shared above.
 

Attachments

  • XL600 top.jpeg
    XL600 top.jpeg
    510.4 KB · Views: 89
  • XL-600 ad.jpg
    XL-600 ad.jpg
    151.4 KB · Views: 93
  • 600 : 500_5252 sm.jpeg
    600 : 500_5252 sm.jpeg
    621.6 KB · Views: 92
shared front end supply (is just GOOFY)

I do not know of anyone using the Cordell DH-220C driver in a XL600. XL600 uses a separate boosted supply and iirc has fuses on the driver.
Also it does not support btl either
Make sure the shop returns all components to you. If mosfets get changed make sure they use new exicon devices and not old hitachi


Personally, I would opt for front end driver card improvements, over maintaining a funky design PS implementation ~and~ OMIT the front end shared power supply altogether that was poorly implemented in both XL280 and XL600.

*I realize the original post(er) mentioned, keeping it stock.
I am not ignoring that...
 
I had experienced gate leaking in one of my DIY hitachi power amps in the 80-ies. These MOSFETs are extremely rugged in the d-s path, but sensitive at the gates. Due to some oscillation there was some gate overvoltage that destructed the gate. You can easy measure this: faulty Mosfets measured only several 100 ohms gate-to-source instead of infinite.

Btw - 700US Dollar for the "repair" is incredible greedy imho.
 
So - an update. I took everyone’s advice, and I picked the unit up. They didn’t change anything. I am planning to take it to Audio Classics in Vestal NY. I am very grateful to everyone for their advice. It makes no sense to me that the mosfets “leak”. Either they work or they don’t. It sounds fine at 70 watts, but clips at 80? I suspect that it is more likely that the resistors or caps on the mosfets are more likely the culprit.

I sold a lateral mosfet amplifier pcb a week back.
The guy who bought it got in touch and the dc offset was wandering all over the place. He checked every component on the pcb and they were fine.
He had used "scrap amp pulls" for the mosfets.
There were no shorts on them but one was showing a bit of leakage when measuring it with DMM. So he replaced it and the amp works fine now.
So he is thinking the mosfet was leaking a bit.
I must admit I have never seen a faulty transistor that wasnt shorted.
The only one that was odd was a bipolar transistor with Hfe of 1 !
I replaced it and the amp was good after that.