Musical Fidelity A2 output offset after repair

[coolnose]

Hi All,

a friend came over and left his Musical Fidelity A2 with me after having started to butcher it, the kind of "beer job" one always appreciates...

After a quick look in vivo and and a google search, it was confirmed the diode bridge was shot, lifting the pads etc.

Replaced the diode bridge, powered up through a bulb tester, the bulb lights up on turn on, then off and on again as the the bias builds up.

I turned the bias down, same procedure, but the bulb stayed dim after start-up, so I started it up without the bulb and started resetting bias.

It went all over the place, the pots would dramatically change value just looking at them. So out they went, new multiturns in and bias conservatively set to 45mV over the 0.22R (200ish mA).

One channel has 10mV offset, the other has 60mV. Any hint for a remedy other than matching up a pair of input transistors ?

Documentation on these is scarce, so I worked with a drawn shematic someone made from the board (but missing all the helper circuits) and this russion link which was quite helpful.It's pretty close to what I have but mine is fitted with BC327/BC337 and looks "factory" to me.

Google Translate

 

[JonSnell Electronic]

60mV is not worth worrying about but the only cure is a matched pair for TR101 & 102, or a preset on R103 ground point to move the voltage just off zero is another way.

 

[coolnose]

Thanks for your quick reply, I knew it would be someone from the island being able to help me 😉

Matching TR101/TR102 is what I had in mind but not planning to do.. could you kindly elaborate on the TR103 preset to ground you are referring to ?

I am good at painting by numbers but not really an artist 😉

Thanks,

Max

 

[as_audio]

Using the third MPSA92 you can probably find a match.

 

[coolnose]

That's clever 😉

 

[as_audio]

.. and even the fourth from the other channel.

 

[Mooly]

That looks very very similar to the old Maplin Latfet amp of the 70/80s. That also had non identical bias currents for the input transistors and also suffered with DC offset as I recall.

 

[coolnose]

So you are saying the current imbalance on the input pair wold cause the offset ?

 

[Mooly]

It can do, yes. The amp design itself does not major on DC precision, it has no need to as offsets of 100mv or less were considered acceptable at that time.

 

[Mooly]

Here is the Maplin amp with equal and unequal resistors but remember the simulation has 100% perfect matching of transistors of the same type.

 

[coolnose]

Well all things considered, especially as it's working fine as is, I'll give it a clean and buckle the thing up again and call it a day.

I still would like to understand Jons remark about TR103 for my education, maybe he'll chime in later.

 

[as_audio]

He wrote about R103 and it was meant to add a small dc correction voltage to the input.

 

[coolnose]

Thanks for that info, I misread "TR103" and could not get my head around what that would do.... Now I also understand Mooly's post about the other amp...

So altering the value of R3, maybe substituting it by a trimmer for finding the correct value before final replacement, should help me get the DC offset close to 0.

I'll look into that!

 

[as_audio]

Easiest and quickest thing is to swap some of these transistors.
Changing R3 (103) will not change much.

 

[coolnose]

Alright folks, I thought I'd just close the lid, but doing some final measurements on the source resistors before doing so I noticed some currrent imbalance on a paralled pair of OT on one channel.

I think I will be pulling the source resistors all together and all the OT devices as well and measure them. Is it reasonable to assume "they" just slapped the transistors into the circuit (same batch) without matching them in any way ?

Can I fire up the amp without the OT and the relative source resistors to take care of the offset by replacing hte input transistors by matched pairs ?

Also I do hear the speaker relay clicking on startup (delay around 5 seconds) but the output offset is unchanged before/after relay trigger... The protection circuit is not shown in the schematic, I'll try to sketch that up from the board to maybe get more input from you guys.

Thanks,

Max

 

[Mooly]

I wouldn't worry over any small imbalance, without going to great lengths with device matching its just a normal occurrence.

I'm not following the bit about firing up without OT's and source resistors fitted. The feedback loop would be broken, nothing would work or be proved in any way.

Don't understand 🙂

 

[coolnose]

Mooly,

i followed your advice and set the bias across the highest source resistors readings on each channel, and let it run a couple of hours with music.

Controlled temp on the common heatsink, adjusted slighly and considered it was as good as it left the factory, lid closed now.

Thanks,

Max

 

[Mooly]

Pleased to hear its all OK... although I can't just recall mentioning source resistors as such 😉

 

[coolnose]

Mooly,

I was referring to your comment of slight imbalance with reference to the design goals of that period +/-25 years ago, etc.

I just considered that's the way it is (and was ...)and closed the lid...

Thanks for your assistance!

Max

 

[coolnose]

Actually I computed all the comments in the thread.. 😀