STK1080ii Devices

Morning Mooly -- Thanks for that --
I was just measuring between pins 2 and 3, and then between pins 4 and 5, so wouldn't you think that this should find R7 and R8 as on the pdf? It appears that the emitter leads from the output dies are spot welded to pads which then go directly to the output terminals, so that would eliminate anything under those devices.
And, all the STK 1080ii devices I have here measure the same -- Interesting, when R7 and R8 are shown and numbered on the pdf. I have attached the original Sanyo Datasheet for some of the genuine parts, which could be of interest for You.
In my Replacement, I'll stick to those .22 Ohm values and closely monitor what happens.
Can You tell me the purpose of that D10 Diode which is connected to the Collector of Limiting Transistor TR14?
Thanks Again for all!https://www.diyaudio.com/forums/attachment.php?attachmentid=733896&stc=1&d=1549315819
 

Attachments

  • STK Data 2.pdf
    416.2 KB · Views: 208
Just a followup thought: Maybe when this Akai AM-U61 was built, the devices would have been the Genuine STKs which would have contained those Emitter Resistors --
Maybe since then, (with it's recent history of unreliability and blowing up STKs), that these new STKs that have been fitted lack those resistors? - As the STK I removed definitely does not have any that I can find/measure.
From What I read on the 'net these Akai AM-U61 are very reliable. -- Interesting.
Thanks for all --
 

PRR

Member
Joined 2003
Paid Member
> purpose of that D10 Diode which is connected to the Collector of Limiting Transistor TR14?

And its twin.

The output is a Darlington. Needs 1.2V to turn it on.

Clamp transistor can pull-down to <0.1V. That's a hard turn-off.

Clamp plus diode pulls down to 0.6V. This is sufficient to to turn-off the Darlington without getting so hard-off that it recovers badly.
 
Administrator
Joined 2007
Paid Member
The original STK's have to have the 0.22ohms somewhere. There is no other way to sense the output stage current.

Try passing a fixed current of say 1 amp between pins 2 and 3 and then measure the volt drop between those pins. If you don't see a meaningful volt drop then I'd have to suggest all the devices you have perhaps aren't genuine.

The two diodes are as PRR describes.
 
Hi Mooly
Of Course, You are absolutely right -- how can the Limiters sense output Device Current if there is no Emitter Resistance; Looks like what You said is correct; these devices I have are ALL virtually a Short Circuit between Pins 2 and 3, and 4 and 5. This then explains the recent unreliability of this Amp.
Did what You suggested, and there is practically NO Voltage drop.
In that Datasheet I sent You, Sanyo state that there should be .22 Ohms in each case, so I will go with that value in my replacement circuit. I fiddled with those Vbe Multiplier Base Values in order to use standard values of resistors for the final build; the Original STK Device looks like it would have run between about 10 to 15 Ma IQ, so I'll shoot for that at the end. Small problem is that it will probably need a Trimpot in the Base/emitter side of the circuit to be able to set it up to that; I decided if I put it there, and it goes Open Circuit in the future, that the IQ will only drop and therefore protect the rest.
Once again, Thanks for All the help!
 
Administrator
Joined 2007
Paid Member
That all sounds like a plan :up: and its interesting that the devices you have seem to be missing this resistance... which might explain a lot actually. Its not just current limiting that is affected but perhaps more importantly the thermal stability of the stage and likelihood of thermal runaway.
 
STK1080ii Device Replacement

Thanks for that --

I have now completed the Pc Boards and are now assembling them; see attachment;
I Need some advice; Due to the fact that I don't know how this amp is going to be treated, Do You think it could be an idea to increase the values of those Emitter Resistors to say .33 of an Ohm to make sure that the Current Limiting Circuit of TR14 and TR15 comes into play a bit earlier -- Or is it just best to leave them at .22 of an Ohm?

Thanks for all, and Kindest Regards --
 

Attachments

  • DSC00701.JPG
    DSC00701.JPG
    294.6 KB · Views: 298
Administrator
Joined 2007
Paid Member
I can't answer that one I'm afraid... because many current limiting schemes would not protect an amp from major abuse. Thermal issues would still come into play under sustained overload and probably see the module off.

All you can do is make a judgement call. Try 0.33 (or 0.47) and see how loud you can play the amp and then decide if it will be OK for the customer. Try adding a fixed a 8 ohm load across the speaker to make things a bit tougher and see if it still plays OK without noticeable clipping.
 
STK1080 Devices

Well, I seem to have had significant success.
I mounted the Output Devices onto a piece of 5 mm Thick Aluminum which I have attached to the Amp's Heatsink with Thermal Paste and screws.
After powering it up on the Variac, I slowly Increased the supply voltage up to our normal 230 Vac while monitoring the Quiescent Current, which I have set to 15 Ma in both channels. I have also attached the Vbe Multipliers to the heatsink and these appear to be doing what they should. After watching it for a while to see where things were going, I connected up the Oscilloscope and lo and behold, it seems to be Rock Stable, which is good considering there is probably some extra wire length in there between the mother board and my additions. Then I connected a set of speakers while still monitoring the outputs, all good. Then, a Signal Source to the input; and it sounds great.
Tomorrow I will give it a Power Output Test into my Fan Cooled Resistor Stack and see what is what. In the Final Build, I used .33 of an Ohm Emitter Resistors as I believe this might help the Thermal Stability and Current Limiting to activate a little earlier. I have attached some shots of the exercise. Thanks to All who have given help!
 

Attachments

  • One Channel.JPG
    One Channel.JPG
    287.4 KB · Views: 290
  • Ready for Install.JPG
    Ready for Install.JPG
    325.3 KB · Views: 297
  • Installation in Akai.JPG
    Installation in Akai.JPG
    346.3 KB · Views: 292
Morning All --
While poking around with the Dvm, I have discovered something else that does not look quite right, and so before I do an all-up Performance Test, I would rather wait for advice, as I do not want to risk any of my work up to this point.
There is an Op-amp, IC3 on the drawing, which looks like it may help with DC restoration of the Output Rail. - (I could be wrong on this.)
On one Channel, the Output voltage of this is .444 Volts, which agrees perfectly with the Akai Schematic. On the other Channel, the Output voltage is only .058 Volts which seems a little low.
Looking at the Collector Output voltages of the Driver Devices on that Channel, TR9=+4.21v and TR10=minus 4.22v.
The Other 'Good' Channel measures TR9=+4.22v and TR10=minus 4.24v, so nothing really amiss here. The 'bad' Channel input voltages to the Boards I made are pin 12=+1.58v and Pin 10=minus 1.58v, so even, great. The Input Voltges to the 'good' Channel on the boards I made are pin12=+1.58v and pin 10= minus 1.61v?
Thinking that the TA75558s Op-amp may have been damaged in the previous history, I carefully replaced it with the correct new one, but those output voltages remain exactly the same.
i don't really want to go poking around too much in the High Gain Front End, as there is always a chance of slipping with the probe and damaging something down the line.
If necessary I could solder in some test points to alleviate this risk --
Does this point to something out of calibration in the Front End?
Both Channel Output Rail Voltages hover around plus/minus 2 or 3 millivolts, so I can't see much wrong here, but of course the DC Gain of that Op-amp is probably high as there is no DC Feedback path around it, only AC Feedback C28, which I have replaced in both channels. I have also checked the Values of those 4 resistors around the input to the Op-amp, all good. Power supply rails to the Op-amp measure at plus and minus 15.2 Volts, with no apparent noise on them. When Powered down with NO Psu Voltages, everything I measure comparing the two channels looks practically identical, so this is a head scratcher --
I just thought I could field some advice before going any further --
Thanks So much for any and all help so far!
 
Administrator
Joined 2007
Paid Member
The opamp is used as a DC servo and the fact your DC offset at the main amp terminals is just a couple of millivolts is proof enough its all working correctly.

The differing opamp output voltages are simply the correction voltages needed to bias the amplifier to zero volts DC offset. Any imbalance in the amplifier itself will need higher correction voltages. When the opamp runs out of range (hits the rails) the correction stops.

Yours is fine :)
 
Hi Mooly
Thanks for that. I was just concerned that something might have been amiss because all the other voltages are so alike to the Akai Schematic. That was a Good Explanation of how it's supposed to work. I have Corrected that Akai Schematic re the connection of that TR10 Collector and it's voltages, which I have attached to be of help to others.
OK, now I can go ahead and do some performance testing --
Many Thanks for All!
 

Attachments

  • Corrected Akai AM-U60 Drawing.pdf
    966.6 KB · Views: 137
Hello, All --
Did an all up Performance Test after playing it for a couple of hours.
The result was 75 V p-p into 8 Ohms, which equates to 87.5 Watts RMS, measured with 1 Khz Input; that is just short of Akai's 88 Watts. That was with only One channel Driven.
No sign of any sort of instability, so that's good.
if Any One wants the Artwork for my PC boards, just ask.
Thanks for all help received; I'm really thankful for that.
Attached is the Output, just short of clipping, as viewed on the 'Scope.
 

Attachments

  • 75v Pk-Pk @ 1 Khz Sine Wave.JPG
    75v Pk-Pk @ 1 Khz Sine Wave.JPG
    331.6 KB · Views: 261
  • The Fan Cooled Load Resistors.JPG
    The Fan Cooled Load Resistors.JPG
    346.5 KB · Views: 263