40 year old Technics Amp repair - troubleshooting circuit

Thread Starter

sparkyjf

Joined Feb 13, 2020
13
Hi all

I'm trying to repair an almost 40 year old Technics SU-V3 amplifier. It's a classic circuit design - almost completely made up of discrete components, with very few IC's. The fault is manifesting itself as the speaker protection relay not activating. This appears to be because there is a large negative DC voltage on the power amp stage, and the protection circuit is supposed to protect from DC on the speaker outputs (amongst other things).

I have a copy of the service manual which includes voltage measurements to take at various points in the circuit. However my own analog electronics is a bit rusty and I could do with some help. I've taken some measurements which certainly show some issues, but I'm not clear what parts I need to replace to resolve the issue.

Happy to post more details including marked up schematics - can anyone give me a hand with this?

Many thanks

James
 

Dodgydave

Joined Jun 22, 2012
11,284
If you can post the link to the service manual, ..
also have you measured the voltage at the protection relay speaker outputs, Tp1 to Tp4,, it should be Zero or near to..
 
Last edited:

MrChips

Joined Oct 2, 2009
30,706
Hello sparks, we'll be happy to lend a hand. You will need some basic tools, DMM, low wattage soldering iron, camera (e.g. smart phone).

We always begin with fundamentals:

1) History of the unit. Are you the first owner? Have you seen it working? Were you present when it stopped working?

2) Look, listen, smell, touch. Without disturbing anything give the unit a thorough inspection. Look for burnt, charred, discoloured components, capacitors leaking or bulging at the top. Sniff for evidence of burning.

We always do these basic things before applying power.

Next, we make photocopies of the circuit schematics pages if we anticipate that we're in it for the long haul. We will be marking up the copies and not the originals.
 

Thread Starter

sparkyjf

Joined Feb 13, 2020
13
Hi all!

First of all absolutely blown away by the responses - thanks so much for all your offers of help. I'll try and wrap up answers to all the questions I've been asked so far in this post, but if I've missed anything or anything doesn't make sense just let me know. Thanks for posting the service manual @MrChips - that matches the copy I'm working from (I found the copy I started working from here in case it helps anyone else out later on: https://www.manualslib.com/manual/690398/Technics-Su-V3.html)

Ok first off the history of the AMP - I picked it up from a charity shop about 6 months ago. Physically it's (sadly) a bit "well loved" but I powered it up and it worked from the moment I got it home. It has sounded amazing. Then a few days ago the right channel went first - it went all scratchy and I though it was either the volume control or input selector switch. However I noticed that the speaker protection relay was actively clicking in and out, and whether it clicked in or out seemed to depend upon the amplitude of the input signal. The left channel was perfect initially. However over the space of a few minutes these symptoms disappeared and now the relay won't connect the speakers up at all. There was no smell of burning, and I can't see any evidence of burning inside the case (I'll post photos later if it helps). The amp still powers on, I have +/-42V on the power supply which looks right, and the FL power meter and little "Class A" logo on the front still illuminate, so I'm presuming that power is good.

Initial investigation started with the two output inductors L301 and L302 - I would expect these to be near zero with no signal applied. They are resting at -40V on both channels, which just seems very wrong. I suspect this DC output is why the speaker protection isn't turning the output relay on (and rightly so!). The power amp transistors aren't hot so no real power seems to be flowing through them, but it appears that (picking on the right channel as that went first), Q334 is fully on, and Q332 is more or less turned off.

I haven't had time yet to probe all the voltages that are presented on the circuit diagram, but things do look very wrong around Q302, Q320 and Q304 (I'll attach a screenshot of my annotated PDF which shows my voltage measurements). I have focussed mostly on the right channel as that was first to go, but a handful of measurements from the left channel are showing almost identical readings (and indeed the left channel output is resting at -40V just as the right one is). So I fear whatever failure occurred to the right hand channel occurred to the left also.

More than happy to take more measurements, send photos, etc. I have access to most basic tools including a soldering iron, DMM, pocket scope and a very old (but still working) signal generator. Screenshots are attached of some of my measurements - it seems that Q320 (constant current source) might either be fully on or short circuited, and I imagine that by itself could throw out a lot of the circuit - however I don't want to randomly start removing and replacing components without a bit more understanding - don't want to cause more damage!

I think that answers all the questions I've been asked (more or less) - happy to take more readings and provide more details as required - just let me know.

Thanks again everyone!
 

Attachments

R!f@@

Joined Apr 2, 2009
9,918
Chips was kind enough to point me to this thread.
Don't mind if I step in.

Reading through I noticed that amp is having identical issues on both channel.
Regarding voltage and all, Right ?
 

Thread Starter

sparkyjf

Joined Feb 13, 2020
13
Chips was kind enough to point me to this thread.
Don't mind if I step in.

Reading through I noticed that amp is having identical issues on both channel.
Regarding voltage and all, Right ?
Yes - exactly that. Although the initial issue manifest itself on one channel initially, electrically both channels seem to be the same.
 

R!f@@

Joined Apr 2, 2009
9,918
OK.
If both are same, my approach would be to look at areas common to both.
Let's check IC301, AN6552, the amps DC servo OPAMP.
Through out the checking keep speakers disconnected, and all the internal PCB wire and PCB's connected intact. A disconnected ribbon cable will throw us out of the proper readings.

Check the voltage of the OPAMP and post it. Volume should be at MIN. No signal input
Always check the R.L amp outputs for DC voltage at L301 and L302. And always probe the inductors first each time to make sure there is DC at output.
 

Thread Starter

sparkyjf

Joined Feb 13, 2020
13
@R!f@@ I like the way you think - I had been a bit blinded by the seeming failure of the right channel first - confirmation bias at its best.

So, as suggested, AMP is intact (all ribbon cables attached). Speakers disconnected. No input signal (inputs are open circuit), input selector set to AUX. Volume set to min. Confirmed -40V DC still present at output inductors.

I have measured the voltage at each of the pins of IC301 - pins 4 & 8 look about right (supply voltage), but all op-amp inputs and outputs are pretty out. Screenshot attached of my readings - essentially all op-amp inputs are at approx -16.8V, and op-amp outputs are both at approx +15V.

So it seems both the inputs and outputs are pretty far out of spec (though I suppose the former could cause the latter). Where is a good place to look next?
 

Attachments

R!f@@

Joined Apr 2, 2009
9,918
We can say that ±15V of OPAMP supply is fine.
Before Jumping into conclusion there is another thing we need to make sure of.

When we take apart an Amp ( disconnecting the PCB from chassis ) results in disconnected grounds. The PCB holding screws + the RCA mounting screws are at 0V, chassis mounted. The PCB screws might have grounding plates soldered.
You need to make sure that the RCA input jacks are indeed connected to the filter capacitors ground.
Check for continuity between the filter cap 0V (GND) to the RCA jack grounds.
An open GND will lead to floating terminals resulting in DC at output.

My approach would be to replace the OPAMP with a similar one ( TYPE ) and check. But for this grounding needs to be good.

PS. Hope your DMM ground is the filter caps ground. Solder a wire there and use it as ur DMM common. Keep it long so it is out of the way always. Cause it tends to get in the way at the darnest time and touch the PCB area that might break ur heart.
 
Last edited:

R!f@@

Joined Apr 2, 2009
9,918
Before going for OPAMP there is also another thing we can do, but for this too all GND terminals need to be at 0V.
Check grounding and come back.
And confirm that filter caps are at ±42V or so.

It's midnight for me, I might not reply tonight but tomorrow I will.
 
Last edited:
What Rifaa mentioned: about having a ground wire soldered to o the PCB, and use it to attach the DMM or scope's common, this is a very valuable tip.

Not only is easier to measure around with a single probe, it also reduces the risk of a short in an inconvenient place. damaging the unit for good.
 

Thread Starter

sparkyjf

Joined Feb 13, 2020
13
Thanks @R!f@@ and @schmitt trigger. Yes regards ground I tacked on a wire at the point between the two power supply filter caps (C601 and C602). One thing I did find yesterday (and confirmed today) is that there is a big difference between chassis ground and the "0V" rail on the PCB (i.e. the mid-point between C601 and C602).

The amp is still physically assembled (thankfully Technics made it so you can access the entire PCB without removing it), and I've confirmed that the chassis, RCA jack grounds, ground terminal on the back and even the heatsinks for the power transistors are all commoned. However there is no connection (at least not resistive) between the chassis and the "0V" rail. If I measure between the chassis and the various test points, the difference in measurement is of the order of 20V (roughly).

Given that I haven't dismantled the amp, I assumed this was how it was designed, but of course I could be completely wrong about that. I wasn't clear from the schematic if there's supposed to be an electrical connection between 0V and chassis ground, and I wasn't sure if I could just short the two together, or whether that would cause problems elsewhere.

I won't take any further action until you advise. No worries about not replying tonight - grateful for your help. I expect this to take some time, and sooner or later I'll need to order parts which will incur further delay. Just happy to be working towards a potential diagnosis.
 

R!f@@

Joined Apr 2, 2009
9,918
I figured that you might not need to take it apart.
Still, from the schematic I can see that the main GND is connected to the RCA input jacks COM terminals. Which means GND is the same as Chassis. ( Unlike sometimes there are a small CAP in between the GND and Chassis , but not in this amp )
PS. Still I could be wrong as I never seen one personally.

I spent some time looking at the schema, not easy with a small laptop.

According to the schema the chassis is connected to Amp ground. Take a look. May be others can confirm.

Presuming ur DMM GND is the cap ground, take some DC voltage measurements respective to GND of C601 & C602.
Like I said, solder a wire at filter cap grounds and use it as ur DMM negative. (Main TX, T1, terminal 21 & 13, which is 0 volt line)

Measure DC voltage to :
1. Chassis.
2. Audio Input RCA Jack grounds or commons.
3. Connector J3, terminal no. 2
4. Connector J1, terminal no. 4
5. Connector J2, terminal no. 2 & 4
6. Connector J4, terminal no. 2 & 5


These are all Ground traces. It's routed in such a way that if terminated wire connection or PCB terminal termination gets disconnected or corroded, Amp develops ground issues.
Check them. Look at the schema, study the GND traces. It will help.
 
Last edited:

R!f@@

Joined Apr 2, 2009
9,918
Additional details needed :
I like to know how the ribbon cables / if any, are used for PCB inter connection.
If there are, how they are connected.
Are they soldered ?
Are they inserted into PCB connectors ?
If later, you need to take them out and check for oxidization.
 

Thread Starter

sparkyjf

Joined Feb 13, 2020
13
Hi @R!f@@

Thanks for your continued help. So this evening I've had time to probe the circuit again as requested. The behaviour of the grounds are consistent across the circuit (which I hope is a good thing)? My measurements are as follows (all with reference to the mid-point of C601 and C602):

chassis: -19.4V
RCA GND: -19.6V
GND terminal: -19.6V
J3 T2: -20.5V
J1 T4: -19.4V
J2 T2/4: -19.3V/-19.2V
J4 T2/5: -19.6V/-19.6V

All ribbon cables are connected using removable headers/connectors on the main PCB, but are all soldered directly into their daughter PCB's (e.g. volume control PCB, tone control PCB, etc).

I've reviewed the schematic and I agree with your assessment - I can trace the midpoint of C601/602 (my measurement reference) right through to the ground of the RCA jacks and chassis - and these are all mounted on the same PCB. According to the schematic, there's nothing in line with the the smoothing caps and the chassis - no resistor to blow, and no capacitor. Yet there's a massive voltage difference. The PCB layout isn't entirely clear to me how the ground flows - there isn't a continuous ground plane.

From what I can tell it's only single sided PCB, but there are a lot of wire links in it that connect across various tracks. I'm a little stumped at this stage as to why something that should be directly connected is behaving like it's not - but also the voltage readings are consistent at all points I have measured ground at. It seems to me unlikely that a wire link would go, but the behaviour is almost as if that is what we're seeing here.

Have you ever seen anything like this before?
 
With your multimeter, measure the resistance from the common of C601 & C602 to the chassis. Make sure the unit is un-powered.

In very old units, environmental conditions sometimes create a thin insulating film between a pair of joined metal surfaces.
 
Top