Technics SU-VX800 amplifier

JoeJester

Joined Apr 26, 2005
4,390
@MrChips Nice research.

It's certainly pointing towards that direction.

on edit ...

If there were an oscillation, lets say a 10 nS, 1 volt pulse @ 100khz rate, that would increase the 0.5v reading on the emitters of the Q503/Q505by a millivolt, something beyond the recognition of the multimeter. That's why a single trace scope test would be beneficial. If it triggered on a 2V signal that means the amplitude of the signal exceeded 1.5V (2V minus the base voltage of 0.5V). That's all we would know. Increasing the trigger level to where the scope no longer illustrates the trace would give an idea of the characteristics of the pulse.

The test can run as long as you want. Any noise impulses exceeding the trigger level will trigger the scope.

Right now, I'm still leaning towards the counterfeit transistors, from suspect to probable. The scope test will push me confidently into certainly.
 
Last edited:

Thread Starter

Marius83

Joined Dec 28, 2017
423
hehe. Not boring @JoeJester, i'm just not quite sure if i'm doing things right, check on my picture if the DSO is set up correct...


I have now just let the adjustment at TP401 been untouched since yesterday, i measured at power up and it stabilized around 26 mV, seems ok.

Then i turned VR403 to the same percentage as VR404, then i only got 0.1 mV at TP501.

Attached oscilloscope at Base Q507, i set it at 0.1 Volt setting, i have no 0.5 Volt setting, next would be 1 Volt.
I took a picture of the screen of it, amp has now idled for 20 minutes, i haven't seen any bigger spikes than you can see on the picture.
I also attached a DMM at Base Q509, it read 0.1 mV for a while, then dropped to 0.0 mV... and stays there.. if i raise the adjustment at VR403/TP501, a little voltage at Base Q509 shows up for a split second, then falls to 0.0 almost instantly.


Sorry for the worst video.
 

Attachments

Last edited:

Thread Starter

Marius83

Joined Dec 28, 2017
423
Just to follow up on fake Toshiba power transistors.

https://www.engon.de/audio/toshiba transistor fakes_en.htm

https://sound-au.com/counterfeit.htm

Yes, it would be wise to purchase your electronics components from a reputable distributor.
You can try looking for replacements such as 2SC5200N and 2SA1943N.
Good research.
But not all four installed at LEFT was from the same order...hypothetical, if two of them shorted hard, could they also take down the two others?
Just wondering..
 

Thread Starter

Marius83

Joined Dec 28, 2017
423
It has now idled for over one hour straight, without any peaks/changes in the voltage measured at Base Q509 or TP501.. depends if i'm measuring it right...
 

JoeJester

Joined Apr 26, 2005
4,390
@Marius83 Ok, we know it can get above 1 V and the normal looks to be about 0.8 or so volts. Increase the vertical scale to the next level.... run the test again, and if you get a trace post a picture. Take the scope out of auto and set it to single. Trigger level set to .2V under that condition, hopefully we won't see a trace at all.

Get some good transistors, enough to have spares if the right channel fails you soon.


on edit:

You notice that the scope showed a +0.8V dc level with noise on it. That noise had a peak of about 1V. Your DMM reads the average of all the noise it "saw" (within the meter AC bandwidth). 0.1 certainly seems good to me, and it could be lower if your meter has a +1 count on the accuracy.

Nice job. Killed two birds with one oscilloscope measurement. Confirming there were nothing preceding caused the problem leaving the "counterfeit" as the probably cause.
 
Last edited:

Thread Starter

Marius83

Joined Dec 28, 2017
423
Okey!
So no more testing at the moment..
Forget to mention, i affixed the transistors at RIGHT channel to the heatsink before i did start the testing.

I just ordered transistors from Digikey 2SA1934N and 2SC5200N, maybe it will work with them :)
 

JoeJester

Joined Apr 26, 2005
4,390
Let me tell you something about oscilloscope operations. When you select the single mode vice auto, the first time the monitored signal exceeds the trigger level, the trace starts. If you set the positive slope or negative slope, once that transition exceeds the trigger point, you get a trace.

Why am I saying this? Well, if you had an unknown impulse that was 10x the normal reading, you could set the trigger level to above 2 -5 times your normal reading and if a trace appeared, you can conclude the impulse was above that setting. If you had selected the auto mode, the trace re-triggers each time the time base restarts, so the trace is continuous.

This may not be a problem in your set. But it your trying to find a short time impulse, using the single trace is good to use. That type of test requires four tests .... 1. Positive x volts trigger, positive slope. 2. Positive x volts trigger, negative slope. 3. Negative x volts trigger, positive slope. and finally 4. Negative x volts trigger, negative slope.

Those four tests would prove an impulse if any one showed a trace. The trace may or may not show the impulse, but would definitely show whatever remained after the impulse till the end of the trace. If it triggered on a positive slop, you can decrease the time base and maybe you will see the impulse. This of course assumes the impulse can be seen as greater than 0.1 divisions on your time scale on the lowest scale available.


I also have high hopes the problem is solved. I'm glad you persevered with solving this problem.
 

Thread Starter

Marius83

Joined Dec 28, 2017
423
I think i will use the scope, monitoring when adjusting it next time with the new transistors in... if somethings happens...

I got a little inspired by the links about counterfeit transistors provided by MrChips.
So i inspected my own transistors just by curiosity, inside do look pretty similar, but i guess it could be other different things.

They do have some differences visually, the original ones have a shiny backside, and a little different print compared to the new ones i've bought.
Anyways, new ones from a reliable source is coming! :)

Original ones to the right and the new ones to the left in the pictures.

powertrans.jpgpowertrans2.jpg
 

Thread Starter

Marius83

Joined Dec 28, 2017
423
So, should i replace the power transistors in the RIGHT channel as well?
Or, should i just leave them at first, and continue with the LEFT channel?

toshiba.jpg
 

JoeJester

Joined Apr 26, 2005
4,390
Test the transistors before you install them. Keep the others stored in their ESD (Electro Static Discharge) container. I doubt the conditions exist up there where you can create a spark this time of year,

Yes, Insert Q507 and Q509. You should be able to achieve the lower limit (4 mV) easily and I think I would leave it there for this test.

Monitor the emitter of QQ503 with your Oscope and monitor tp501 with your DMM.

Any thing abnormal .... within the 30 minutes test, Make a note.

Test the transistors before you install them.

An overview of ESD


 
Last edited:

Thread Starter

Marius83

Joined Dec 28, 2017
423
So i tested the two new transistors.
Then installed new Q507 and Q509

i set the adustment at TP401 to 25 mV and monitored it for a while.ok
Then i hooked the oscilloscope to Q503 Emitter, monitored TP501 and adjusted.

4 mV was easily reached.
I monitored it and adjusted a little bit further down as it was slightly rising.
I stabilized it about 5 mV, and it idled for 30 minutes without anything happened.

A little slow downed flic of the oscilloscope.. got on the side..

 

JoeJester

Joined Apr 26, 2005
4,390
OK... do you see the spikes on your oscilloscope? You need to adjust the oscilloscope voltage to a higher scale 20V/cm (2V with a x10 probe) and set the trigger from trace to normal or single on a positive slope and adjust the trigger level positive of the ground reference.

As an aside test, put the oscilloscope on the right channel on the same test point, and let's see what's on that channel.
 
Last edited:

Thread Starter

Marius83

Joined Dec 28, 2017
423
It's just when i put the Oscilloscope in single or normal, the screen freeze like i put it on hold... only in auto it will show live...
i might be doing something wrong...
 

MrChips

Joined Oct 2, 2009
30,824
The oscilloscope screen is suppose to freeze in single or normal mode. That is how it works.

In single sweep mode, the oscilloscope will capture a single event and freeze. You need to rearm the scope to capture the next trigger event.

In normal mode, the oscilloscope will not update the screen until an new trigger event appears. The screen will freeze until the next trigger event appears.

It "unfreezes" when in auto mode.
 

Thread Starter

Marius83

Joined Dec 28, 2017
423
Aha, MrChips... a useful function i can understand.. hehe.. as you know i'm not good at the oscilloscope.

But, when a trigger event appear, i guess the green "Trigger" led at the Oscilloscop should light up of flash, right?
It did not in the measurement i performed over here... but it might be because my settings were not right..

I will put it on the right channel to check, but i need a little help to set it up correctly

@JoeJester says:
trigger from trace to normal or single on a positive slope and adjust the trigger level positive of the ground reference.

How do i select positive slope,and set the ground reference..?
 

JoeJester

Joined Apr 26, 2005
4,390
OK ... It looks like you will be using your parameter select button and the + and - buttons. Look at the trigger level indicator in the picture.

In the picture below, you will see that the trigger is set to AUTO and right next to it is the slope .... which starts and the top and moves to below the zero line and then continues to the right. That is the negative slope. The positive starts negative and moves above the centerline.

DSO-138.jpg
 
Last edited:

JoeJester

Joined Apr 26, 2005
4,390
In your movie, I see spikes exceeding plus or minus 4 cm. Your scope is set at 1 V/cm. You are using a x10 probe. I see a signal, at +5 volts DC with spikes exceeding 35V on the positive side and exceeding -45 volts on the negative side. I just realized your scope probe had to be in the x1 position as your scope is more consistent with the 0.5/0.6 V reading at the emitter. that would make the positive and negative spikes achieving up to plus and minus 4.0 volts.

The first step in the comparison is to set the vertical higher than 1V/cm so we can get an idea of what we are looking at. With a 2V/cm x10 probe, we should see the peak of the pulse which I assume will not exceed the 60V supplies unless it's on a differentiator circuit.

This is where the single trace will help. Once you get the trace, the time base can be changed ... without hitting the run button ... to a lower value to try and see the signal that triggered the scope. Right now we know that signal is greater than 40 volts, but for the purpose of the test, set the trigger level to 0.5cm to 1 cm above center or 10-20V at the 2V/cm scale. Ensure you set the trigger level to above 0.6 V when you do the single trace measurement. On the second reading, the negative slope, set the trigger level to about negative 0.5 or so.

once you get a stable reading on the left channel, snap a picture, then test the right channel and snap a picture. since this is a left channel problem that was "apparently" resolved once, we should see the same signal on the left channel. If we don't ... time for more troubleshooting.

We've already replaced those transistors a couple of times, so, we are being extra careful this time.

Hang in there. Your getting more experience with that scope.

Note: Edits are in italics.
 
Last edited:
Top