Alpine 3566 issue

[Kei]

I took my working 3566 and replaced all of the capacitors and the 1 & 10 ohm resistors with like for like replacements. (old caps were starting to leak and R765 & R760 were a tad cooked and measured around 1Mohm)

Pre test run checks indicates that the resistance between BATT and GND is 3.7ohms which is definitely not right. I've taken it back to bits and checked for shorts and not found any. The caps I put in were all good from what I could tell as were all the resistors. I've tested it on a bench supply with current limiting and as the current/voltage rose I got the expected got very toasty resistors R765, R760 & R773, but no catastrophes.

Any ideas what I have managed to do? I did the same work bar the resistors on my 3553 at the same time and that amp seems to be ok.

Images for reference Dale CPF 1% 1 & 10 ohm used to replace originals

An externally hosted image should be here but it was not working when we last tested it.

Main bulk changed for panasonic FC/FM/FR.

An externally hosted image should be here but it was not working when we last tested it.

 

[Kei]

I've been testing the D1669 inverters, Q715, Q719 & Q723 are shorted between base and emitter. The other half of the inverter, Q716, Q720, Q724 all tested good as did Q711 & 712. The Diodes D713 and D714 also tested good.

 

[Perry Babin]

Did they read as shorted out of the circuit?

 

[Kei]

Yes, I tested all components out of circuit.

 

[Perry Babin]

Those have been discontinued for many years. You need to be careful when buying replacements. Most all on ebay are going to be fakes.

 

[Kei]

Cricklewood in the UK is considered a reputable supplier from all I've read and they carry the original Sanyo D1669. (at a rather considerable £3.30 each) I've got about half a dozen somewhere for my 3552 that I bought 7 years ago but I can only find the ones that I removed that still tested good.

Out of interest, are there any suitable equivalents?

 

[Perry Babin]

I know of no exact sub but the D44VH10 with insulators and shoulder washers may work. If you found a reputable source, use the original part.

 

[Kei]

Presumably it would be sensible to replace the whole bank of 6 for both halves of the inverter even though the first set test good?

Also I assume it would be wise to test the output transistors for the channels powered by this half of the power supply? (It's split into 1/2/3/4 on one half and 5/6 on the other)

 

[Perry Babin]

When power supply transistors fail, it's mandatory to replace all operating in parallel for the best reliability. If something stressed the transistors to cause 1/2 to fail (not something like a broken lead), it's best to replace all in the failed power supply.

I wouldn't suggest removing the outputs for testing but a quick in-circuit check for shorted outputs would be a good idea.

 

[Kei]

I've replaced all of the blown components and all seems to measure ok. I've run it on my bench supply and with no remote signal it sits happily drawing no current after the initial cap charge spike. Turning it on using the remote, the current draw hit the limit I'd set at 5A without any signs of trouble from the amp. (some buzzing but no glowing components or smoke) Just how much current does this amplifier draw at idle? The service manual suggests 3.5A for no input for ch1/2/3/4 and 3.5A for 5/6 so I'm assuming that it is roughly 7A. My 3553 which is 4x35W seems to idle around 2.5A.

 

[Kei]

I've tried allowing more current all the way up to 9A and still sitting on the current limiter. (the inverter quietly squeals as the current rises) My bench supply only goes up as far as 10A so I've not got much more I can give it.

Is it really supposed to idle at such a high current with no input or load connected? Also should the parts Q711 & Q712 (2SC4024) be replaced too? They tested ok so I left them in.

 

[Perry Babin]

If all semiconductors are clamped to the heatsink, you can measure the voltage across the emitter resistors in the output stage. None should read more than about 2mv across the resistor legs. If you find one or more well above that, the bias may be set too high or there could be some other fault causing the channel to draw excessive current.

Also make sure that you didn't install any caps backwards.

 

[Kei]

Pretty certain all of the capacitors are all in the correct polarity. One thing I did change was the 0.47uF electrolytic capacitors over to wima metallized polyester film. (E113, 213, 313, 413, 513 & 613)

Measuring the emitter resistors (big 5W 0.1ohm parts?) needs to be done with the amp remote on? Assuming it can be done with the current limited via the bench supply.

 

[Perry Babin]

Yes, the white ceramic components.

Yes, amp on. Limit the current to about 5 amps.

 

[Kei]

I had a look at this earlier and access to the legs on those resistors is extremely tight on the main 4 channels. (blocked by the bus bars and nec driver IC's) Channels 5 and 6 should be easy to check. Bias resistors and transistors checked out ok on the meter as did the outputs. I will attempt to measure the voltage across the emitter resistors that I can get at tomorrow.

 

[Perry Babin]

The emitter resistors are directly connected to the emitter legs of the output transistors. You can place the probes on the corresponding emitter terminals for the two transistors in each channel.

 

[Kei]

Measured DC voltage across the emitters of the output transistors here are the results:

Q604/605 - 2.0mV
Q204/205 - 1.7mV
Q404/405 - 1.8mV
Q304/305 - 2.1mV
Q104/105 - 1.8mV
Q504/505 - 2.1mV

This is when the 5A current limiter is hit. Voltage from the supply falls to 7.1V. Channels 5 & 6 peak at upper mid 20mV's, the other channels all spike higher before the current limit is hit.

These are the peaks before the limiter is hit:

Q604/605 - 2.5mV
Q204/205 - 3.6mV
Q404/405 - 3.3mV
Q304/305 - 4.1mV
Q104/105 - 3.7mV
Q504/505 - 2.2mV

 

[Perry Babin]

You have channels 5 and 6 at less than 3mv. Where does the 20mv+ apply?

 

[Kei]

Sorry, my error. In my hurry to write the post before leaving for work I got muddled with my decimal points converting from volts to millivolts.

Here is how that should have read:

DC voltage across the emitters of the output transistors here are the results:

Q604/605 - 20mV
Q204/205 - 17mV
Q404/405 - 18mV
Q304/305 - 21mV
Q104/105 - 18mV
Q504/505 - 21mV

This is when the 5A current limiter is hit. Voltage from the supply falls to 7.1V. Channels 5 & 6 peak at upper mid 20mV's, the other channels all spike higher before the current limit is hit.

These are the peaks before the limiter is hit:

Q604/605 - 25mV
Q204/205 - 36mV
Q404/405 - 33mV
Q304/305 - 41mV
Q104/105 - 37mV
Q504/505 - 22mV

 

[Perry Babin]

If you have the service manual, look at the emitter voltage for the various output transistors. You'll see that one is positive and the other is negative. The total difference is what you should read. For example if it's 0.004 and negative 0.004, you should read 0.008 measuring as you did.