Repair help requested - Creek 4240SE

[wutzacircuit]

A family member recently gave me a dead Creek 4240SE integrated amplifier. The unswitched outputs were inadvertently grounded together. I found the power supply fuses were blown when I opened up the case. When I replaced them, they quickly blew again (less than 2 seconds). I have contacted Creek to try to get a schematic (the sent me one for my old 4140 a few years ago), but I have not been able to get them to send me one for the 4240, so far.

I am pretty new to the DIY audio world (one Millet Max headphone amp build under my belt), and I don't have much of an electrical engineering background, so I am looking for any help I can get on diagnosing the problem(s) and repairing them. I am an engineer, however, and I have access to most of the tools I could imagine needing (DMM, O-scope, etc). This is not an urgent project, but I am hoping to learn a little about amplifier design in the process. If anyone has the schematic for this amp and can share it, or has any tips on where I should start with debug, I would appreciate the help.

Thanks in advance!

[wintermute]

I'd say it is likely that your output transistors have failed short circuit as a result of the original short on the output, before the PS fuses blew. Now you have most likely got a permanent short on the output which is why the fuses blew again.

Check the output transistors for shorts, as a first step There may be other collateral damage as well, but that's a starting point.

Tony.

[wutzacircuit]

Thanks for the quick response, wintermute.

Without a schematic, (OK, probably even with one) I'm a little lost on which are the output transistors. The board layout is not completely straight-forward to me. I will take a look at it and see what I can figure out. I'll report back with what I find.

[JoelS]

if you have access to a variac, that can be really useful when trying to troubleshoot a power amp with a short circuit. connect the variac output to the amps power connector, start with the voltage at zero and turn it up slowly while measuring the input current. you might only be able to turn up the voltage a little bit before approaching the current rating of the PS fuse. if you run it with low line voltage and keep the current low enough, you will have a chance to probe around and figure out what's happening without keep on blowing stuff up. I agree that it's likely the output transistors are blown, but if there's still something else wrong you might just replace all the blown output transformers, then turn it on and blow them all again because something else is still wrong. happened to me before.

-Joel

[alexcp]

PM diyAudio member x-pro, he was the developer of those amps and should be able to help

[wintermute]

without wanting to sound too funny I'm not familiar with the amp, Id say the big ones mounted on the main heatsink(s)

I just did a search and found an image of the guts of the amp on creeks website Creek Audio's Old Products :

the ones mounted to this heatsink in the image below.. it is very hard to tell from the picture if they are chips or transistors with some sort of heat spreader on them.

edit: ok I looked at the other pdf in the zip file and it says it has mosfet driver and output transistors, so they aint chips

Tony.

[eggontoast]

What usually happens with these amps is that the thermal compound dries up, then the output devices overheat and fail as the output stage idles quite hot on these.

This is the only schematic I have for the 4240

[alexcp]

Looks rather similar to 4330 (see attachment). In case you will be replacing the output MOSFETs, note they must be "logic level", e.g. Vgs(th) should 1..2V, not 4V as non-logic-level MOSFETs (such as IRF610) have. The ones used by Creek may not be available, but IRL540 seem to work well in 4330. There is no real replacement for ZVP3310, but those are available.

BTW, I believe 4330SE simply had a bigger power transformer than non-SE. The schematics were identical.

[wutzacircuit]

Thanks everyone for the help!

A few comments:

There was a definite shorting "event" on this amp that probably caused the failure. The thermal compound is a little dry, but not too bad. I will replace it when I replace the FETs.

There are 6 devices mounted to the heat sink. They are divided into left and right channels. 2 for each channel are TO220 package BUK555-100B logic level power FETs. I tried measuring resistance from source to drain and got the following:

(Right Channel)
Q111 - Starts off at around 125kOhm and levels off at around 42.5kOhm
Q112 - Starts off at around 146kOhm and levels off at 9.1 kOhm.

(Left Channel)
Q11 - Starts off at 0 Ohm and levels off at 18.6 Ohm
Q12 - Never really levels off, but is in the 15MOhm range.

The other two devices are TO-92 packages (Q8 and Q108). I can't read the markings on these because the flat surface of the package is against the heatsink which I haven't removed yet.

So with my novice level experience, I'm not sure if the transients I am seeing in the measurements are normal or not. I am just measuring with a cheap DMM at this point. However, from what I am seeing, I am going to guess that the Left Channel FETs (Q11 and Q12) are blown.

So now for more questions:

1) From my previous hybrid amp build, I understand that matching some transistor devices is important for sound quality. Is this important for this case? (I am assuming the answer is yes) If so, how do I go about matching them (what measurements should I take?) I expect I will replace all 4 FETs.

2) I understand that properly biasing the FETs is important. It would appear that the only adjustment here is via PP1 and that this should be adjusted so that the voltage drop across R22 is 11-12 mV based on this schematic. (Is this correct?) I have V1.3 board, so I probably need to research this a little more.

3) What else should I be testing before I make repairs?

Again, many thanks for answering my novice questions. If anyone needs any mechanical or thermal questions answered, I'll gladly return the favor!

[wutzacircuit]

Sorry, one additional detail, the comment alexcp made about the differences between the 4330 and the SE version seems to apply to the 4240 as well. Bigger output caps on the SE power supply (10000uF vs 6600uF). This could also be a version difference (1.0 vs. 1.3).