Sony FH909 amp common fault?

I have read some posts and www info around this but never seen a certain fix
On pressing the on button for the first time, the volume lights remain off and there is no sound, nor can any of the functions or settings be used.
Switch off and wait 5-10 seconds and turn on again. Then I get the volume leds light up and functions work and there is audio.

I have read that it is one of the relays but checking voltages there shows the same when it works to when it doesn’t, though people who said it was the relay had general audio issues but didn’t fully describe whether their issue were the same as mine exactly. The issue seems more like a capacitor issue by the symptoms described.

Can anyone add their precise symptoms and fixes please because it is hard to find the exact relay replacement. If it is the relay, could someone assist with which to replace the Matsushita hb2a-yd-dc24v-s with please or if not the relay which caps you would target first? IC952 and associated caps?

I guess i’ll try to replace a few caps and that ic regulator.

Does anyone know equivalents for these relays please?

DEC Type DH1U 12VDC TV-5

and

Matsushita HB2a-YD-DC24V-S

Thanks

I've replaced many of the caps, no change.
I removed the amp board relay and jumped it to determine if the fault was still there and it was.
Anyone else suggest where to look next?

Is the relay coil getting power? Because if it’s not, putting in another relay will make it nice and shiny but won’t make it functionally better.

This unit suffers from many cold joints, only the relay with the see through plastic needed replacing on mine, my engineer had the relay from china Dave beyond that I can't help, yours has a slightly different issue than mine as mine always powered up fine lad but needed to warm up to stop the sound from cutting out....i have 2 now and the other has the same issue but no engineer

he also replaced the main amp but im not sure why

good luck..ohh my relay is different from the original as he couldn't find that

i think you're issue could be on the power supply board or the black relay but im only guessing Dave, the other relay is for sound alone

Thanks Guys, I've tested without the relay on the amp chip board by jumping it and same issue was there so I'm assuming that relay is good.

There is definitely voltage across this amp relay and it does click regardless of whether the actual audio kicks in or not.

I have 3x 100uf 50v Caps to source and replace and a few of the smaller ones. I'll try to replace these last few before moving onward.

I agree Reno, this issue seems different in that the volume and functions don't work initially then wait 10 seconds and turn on and most of the time it turns on properly. I'll try do a video.


Here is a visual if the issue.

https://youtu.be/LZnclhI0wIE

My advice for you is to monitor the power supply rail(s) voltage(s) so you can see what it is doing during manifestation of the symptoms.

Thanks Norm, i’ll take some readings at the connection to the amp board.

All rail voltages to the amp board remain identical when both working and not working.
-11v rail is -11.44
+11v rail is 11.26
-34v rail is -33
+34v rail is 32.5

Doesn’t appear to be anything unusual here.

Question... does the power switch always toggle working/non-working? Does it ever working/working or not-working/not-working?

Ok, looks like this unit has a lot of pcbs. You should monitor the voltages in all of them to rule out issue with the interconnects. Also check voltages on IC102 when working/not-working.

It seems like a random outcome but I have noticed that switching on and off say within a few seconds it will never result in successful power up but waiting around 10 seconds after a non working switch on it will almost always switch on properly.
Though the actual answer is that there is no pattern, I’ve seen times it switch on successfully a couple of repeated times but mostly power on is unsuccessful.

Tested voltages at ic102 and there is a discrepancy at pins 2,3,4,5 between working and not working.
When working all read 0v as expected.
When not working all read 8.7v.
Going to investigate where they go!

Hmm, ic303 pins 22, 23, 24, 25 have voltages like above when not working and expected 0v when working.

I’m not sure why the 11v rail is appearing on those points to make it not work and vice versa.

Isolating R352 connecting the 11v rail to the mic amp results in the unit never successfully powering on properly. It still powers up but no functionality as shown in the video.

So, isolate means what, taking R352 out of circuit? Why did you do that?

I desoldered one leg because I wanted to see if it would alter the issue and perhaps tell me the issue was being caused in that section. Not sure whats occurring.

Ok, I don't understand the logic in that approach. If you disconnect something under the hood of your car, and it becomes broke 100% of the time, is it the same reason why when it's reconnected, it's only happening 50% of the time? For example, if your car ignition is partially failing and therefore only starts your car 1/2 the time, but you disconnected the fuse to your fuel pump and now your car won't start 100% of the time, could it be related? Who knows? Maybe it tells us something. Maybe it tells us nothing. Maybe you just confuse yourself.

If you did something and it fixed the problem, that might tell us something and be helpful. But randomly taking things out of circuit until it doesn't work anymore.... not sure that approach really tells us anything since the circuit is not designed to have parts pulled. Pulling that leg is useful in this case simply to measure the resistance in that resistor.

I looked at that mic amp, which doesn't just amplify the mic, also seems to buffer the input from IC102 as well. But I don't see any obvious reasons why the system won't work with R352 disconnected. But you only disconnected one resistor. R353 which runs the negative rail is still connected, and the negative rail is not ground, keep that in mind. While interesting, this discovery doesn't give me any aha moments because I don't have confidence that maybe some other condition was occurring at the same time.

So are you saying you measured voltages at IC102 and the only changes observed between (working/non-working) is to the output pins? Also, are you saying you measured all pins on IC303 and all pins are behaving the same except pins 22-25? The rest are all the same voltage? I'm just trying to get a feel for what is happening.

Other observations:
Did you check all the components in the Q307 reset switch circuitry? Seems like it powers on at least 1 of the controllers.

There are a bunch of controllers there. Did you check to see if they all have proper voltages? I'm sure it's a pain in the butt, but to get a complete picture....

Many of the controllers share a common data/clock signal. If the clock signal is not present, they will behave like a frozen computer (hint).

Several ceramic crystal oscillators or oscillator-circuits are present. Seems like at least 3 of them. You may want to check and see if they are functioning properly. No oscillator, means no clock, means the MPU's no work.

So many MPU's this is a very complicated circuit. You aren't going to be able to decipher all of the functions. For example IC951 looks like it might be a programmable MPU in which case, it's anyone's guess what or how it really works. Not even sure you can just stick another chip in there as it might need programming first to function. In cases like this, probably would only work if a factory correct replacement from Sony is procured. Not saying it's bad, I'm merely making an observation that if it has failed.... you get the picture.

There are a lot of grounds. Analog ground and digital grounds are separate. It's also important that when you check voltages, to also check and make sure that the grounds are not switched or broken as well. Many boomboxes will show voltages everywhere when the system is powered off because the power switch disconnects on the ground side.

Q307 circuit checks all good.

I’ve checked ic303 in full and something weird occurs.
Other than the differences below between working and non-working, when in a working state, checking volts at pin 20 turns from a working state to a non working state and has to be powered off and on to get back working.

Changes in working and non-working, * denotes discrepancies.
1-6 identical as manual
* 7-10 working 0v, not working 5v
11 working/not working 3.4v
* 12 working 5v, not working 0v
13-16 working/not working 5.4v
* 17 working/not working 1.5v (should be zero)
* 18 working/not working 1.54v (should be 5.5v)
* 19 working 2.7v, not working slowly rising to 5v
* 20 working 2.17v, not working 2.7v
21 working/not working 0v
* 22-25 working 0v, not working 10.08v
26-27 working/not working 0v
* 28 working/not working 3.4v (should be zero)
29 working/not working 3.4v
30 working/not working 5.4v

Dave I would inspect the power switch itself.

Hey Floyd, i gave it a good clean and it seems to be functioning with clean clicks in the small microswitch and it does power the unit on.

Dave I would inspect the power switch itself.

Click to expand...

Floyd, it would be great if this was a simple power switch, but this unit is an always ON unit with standby. So the power switch doesn't really power on the set, it really just lets the MPU (micro computer) know that Dave wants the set to turn on.

I’ve checked ic303 in full and something weird occurs.
Other than the differences below between working and non-working, when in a working state, checking volts at pin 20 turns from a working state to a non working state and has to be powered off and on to get back working.

Click to expand...

So, does your meter have a 20 meg Ω input? Well, anyhow, if you paid attention to what I said in my previous post about the set having multiple data/clock oscillators, you'll also note that I mentioned that without a clock signal, the system will freeze, or how a computer might crash if you will. Almost all computers including micro ones require this clock signal to function. So what is so special about pins 19 and 20 of this controller? Pin 19 provides the power out to the crystal oscillator circuit, and pin 20 is the clock-in lead. It could be that your multimeter imparts too much of a load and drags down the oscillator, thereby killing the clock signal and crashes the MPU. A 20 meg Ω input usually is enough to prevent such circuit loading but not always. An oscilloscope with a 100:1 probe through a cap might allow you to actually see the clock signal.

This is why I told you to check your oscillator circuits. Some MPU's require a single X-tal. Others require an X-tal circuit. Controller IC303 requires an oscillator circuit which is comprised of 2 150p caps and 2 resistors, and a ceramic crystal oscillator. 5 components total. If in doubt, you can go ahead and change those 2 caps and resistors. The crystal can also fail too so I'm not sure if this one is weak, or if your meter loaded down the circuit too much.

Anyway, I'm just throwing things out there, might be helpful, maybe not. I wouldn't want to be you right now, haha. But that's all I got.

Thanks Norm, i'll check those resistors first see how we go.
If you were going to replace the crystal oscillator, what part would you go for? I guess that might be the million dollar question and would need to put a scope on it.

Dave the service manual is on stg, someone posted it on there when i had issues, i'll try and find it for you lad