Harman Kardon HK395i with issues

[runnin']

Hi, I posted in an existing thread in the solid state section, but on second thought figured I should start my own thread.

So the 395i, upon power up, pushes out the woofers of the connected speakers before they return to normal in about a second. Measuring dc offset at the speaker posts(all pots set properly) shows that the left channel starts off about -500mv DC before slowly climbing to between -58,-67mv where it moves around a little. The right channel stays close to -32mv. After 15 minutes, both channels are still at their above mentioned values.

I checked out the service manual on this unit from hifiengine, and it says that 32mv DC is the spec. Also, it shows where the dc offset trim pots are but adjusting them causes the left channel to bounce around but settle back at the 58-67 range. I did deoxit the pots beforehand, and then put them back at their original positions.

I'm reading that either input transistor pairs(can't find them) might be the culprits, or the main output transistors, which are on the heat sinks. Other than that, I can't find anything that looks like heat stress or otherwise damaged or failed components. I did reflow a few joints that looked like they might have gone cold, but with no success.

I've got a Harman PM 625 as well and it sounds really nice. The 395i plays fine but after I saw the woofers push out I'm scared to connect speakers to it. Does anyone have any thoughts as to what I might be looking at?

 

[runnin']

Okay, time for pictures. As you can see, there is no apparent "smoking gun" as it were!

 

[runnin']

And a few more.

 

[runnin']

After reading up on various threads here(Audiokarma is quite a help with the many Harman threads one can find), I'm going to go ahead and order some caps for the power and protection section. In photo "unnamed-4", you can see the main power caps, but there are at least 3 smaller ones behind it. I'll update once I get the caps and have them installed. Hoping for the best!

Edit: I found what is apparently a pretty good brand of capacitor for audio. Vishay BC.

http://www.digikey.com/product-detail/en/vishay-bc-components/MAL215831682E3/4688PHBK-ND/4967127

I'm sure they make many shapes and sizes, so make sure you have the right farads, volts, temp rating and that the base will solder directly onto your board before ordering!

 

[runnin']

I ordered the parts from Digikey.com Sunday night, and by Tuesday afternoon was wondering when I'd get an email saying they shipped the order. When I got home it was in my mailbox! Those people at Digi really run a tight ship. Unfortunately, no joy on the 5 capacitor install. The unit behaves exactly the same way, 500mv dc at the speaker leads left channel upon start up, and it settles down to 58-68mv at the 30 second mark, and it makes no difference after 15-20 minutes. The funny thing is, it measures under 10mv at TP1 and TP2, TP3 and TP4, which are the schematic points for measuring dc offset.

Anyway, I don't really know where to go from here, every thread I can find sort of says something different. Maybe I'll donate it to the electronics department at work, if they'll take it!

 

[Goldie99]

I'm not an HK expert whatsoever, but before anything else - you need to read up a bit on DC-Offset, and Idle-Current (or Idle Adjustment), they are completely different things, which you are confusing, with possibly dire consqequences (for the amp).....

When you measure voltages between the speaker connection terminals (under the right conditions), then you're measuring DC-Offset - there is a long, but very educational, thread from Echowars (and about a thousand others !) in the Solid State forum - at the following link. DC-Offset values should always be as close to 0mV as possible, but practically anything under ca. 50mV is generally acceptable - there may or may not be internal adjustment pots which can be used to adjust DC-offset. In your case, with the HK 395i I didn't see any in the manual - so there's likely no adjustment for DC-offsets.

http://audiokarma.org/forums/index.php?threads/amplifier-distortion-dc-offset-and-you.5634/

When you measure voltages between TP1 & TP2, or TP3 & TP4 (under the HK specified conditions), then you're measuring a voltage which corresponds to (and can be used to calculate) Idle current, which is also often called Bias. The HK manual for the 395i specifies a value of 36mV between TP1 & TP2 (or TP3 & TP4), after 15 mins heating up - it is usually very temperature sensitive (hence the 15 minutes to stabilise), and very sensitive to even small changes of the adjustment pots VR401 & VR402 respectively - the pots which I suspect you've been trying to use to adjust the DC-offset......

If you have been 'adjusting' VR401 & VR402 in an attempt to correct your DC-offset - I would suggest you adjust them back to the manual specified value ASAP - you run the very real risk of destroying your outputs if the idle is set too high (basically via thermal runaway).

 

[runnin']

Whoops, thank you. I thought you could adjust dc offset with the correct trim pots. I have marked the original settings so I can reset.

I did measure dc offset under the proper conditions(unless I'm missing something): input selector on an input with nothing connected, volume at 0, tone controls and balance at middle point and speaker selector on the outputs you are measuring. Like I said above, with speakers connected upon power up, the woofers push waaaaay out and take a second to return to normal. there is no sound during this occurrence. When I turn off, disconnect speakers, attach meter to same speaker outputs with everything set as above, I get 500mv of DC in the left channel, which slowly ramps down to between 58-68 mv of dc. The right channel ends up at around 28 mv.

In my uneducated view, I've never seen speakers do that on any equipment I've owned, therefore I think that something is wrong, but I can't find anything obvious on the pcb that is burnt up, etc.

I have tried reading that thread, but the first few pages have people chiming in on what their measurements are. I guess I'll have to plug on, maybe read 10 pages per night, hoping that they get around to something more educational!

 

[Goldie99]

I have marked the original settings so I can reset.

If you mean physically marked the pot positions, then that won't be accurate enough - you should use the procedure on pg. 7 of the manual to set the Idle Adjustments to 36mV - just the slightest touch of the pots will change the Idle currents.

I have tried reading that thread, but the first few pages have people chiming in on what their measurements are. I guess I'll have to plug on, maybe read 10 pages per night, hoping that they get around to something more educational!

I'd say the first few pages cover most of the key points - so maybe 5-10 pages...

If it's only the Left channel going up to 500mV, and the right is much better, then yes, it does sound like a potential problem developing - hopefully one the HK guys will be able to advise better on that though.