LS28 power supply problem

[AudioIdiot]

Hi Forum,

After a long debate I bought a used LS28.
I was opting for a Mark Levinson but several tests showed me that I prefer tubes (The LS28 replaces a Copland CTA 305)

After a month of great use I lowered the gain from 18,7 dB to 12,6 dB according to an official ARC modification paper (changing 4 resistors) as the volume range was liited to the first 18 steps.
As I have access to an Audio Precision System 2 Cascade I ran a full set of tests and can only conclude that the unit not only sounds great but also measures extremely well. Very low distortion, I used 0,1% resistors for the modification and ended up with 0,005dB channel mismatch and a nice 200k bandwidth.
After the modification the noise and distortion was even lower (logical)

Then the next day I listened again to the unit and after an hour, the volume gradually went down to no output. I switched the unit off and on and the unit played again for about 10 minutes.

As the unit is straightforward, I opened it and found that after a while the PTC protection of the heater voltages was engaged and had blocked the heater current, so no output.
To investigate what the cause was I inserted a current meter in series with the AC voltage to the heater part and monitored the current. The AC current was 2,3A and that sounds about right (for DC the needed current is 1,7A DC, but as the voltage over the two heaters in series was 11,7V I expect that the current is a bit lower than nominal (1,5A)
The heaters run a non-stabilized voltage.

The amp ran for an hour, no problem, and then the fuse blew.
Investigating what happened I noticed that the core of the transformer was excessively hot to touch. (Even after half an hour, it was still warm)
A fuse replacement while removing every secondary output caused another fuse to blow. ...... the transformer is fried.
Measuring the different transformer winding showed a shorted primary.

After removing the transformer, I noticed that the PTC in series with the heater power was touching a metal stand-off of the transformer.
I think that earlier the heat of the transformer caused the PTC to trip (not a high current or a faulty PTC).

During the tests, I watched and probed the different heat-sinks to feel if something was drawing more current than it should but nothing was warm. Not the 5V, not the 180V, not the + or - 12,6V

But this means that the error was in the transformer to begin with.
Now the bad news.
ARC will only deliver the transformer trough an official repair center at a whopping €650, about 10 to 14 times the cost of such a transformer
As I am still not sure what caused the transformer to die I need to investigate the amp with 5 Lab power supplies to see if the unit is OK.
I think started in the transformer but I need to know for sure before I place an excessively expensive replacement transformer.

So I made a drawing of the power supply (quite straightforward with big troughs-hole components) and a lamp on one side of the board.
So the only thing that I need is what would be the nominal voltages the transformer supplies and what are the expected current.
There are 5 DC voltages that are made from 4 transformer secondary’s. Especially what is the voltage before the 180V regulator? In addition, what current is expected?
The 180V is the only voltage not protected by a PTC.

Is there a repair center reading this forum that could help me?
And does anybody have experienced such a strange error. Transformers hardly ever die by itself.

Thanks,
Peter

[tdelahanty]

have you replaced the tubes or tested them on a tube tester?

[AudioIdiot]

They work in the application. But I haven’t removed them. What tube error could cause this?
I will test the heater in isolation of the rest.
Than apply the +180V and -12.6 for the amp.

[tdelahanty]

there may be an inter-electrode short causing excess plate current. Substitute a new set of tubes or check the originals on a tester. There might be a bad cap in the p/s.

[AudioIdiot]

Well,

I powered the Amp today with 4 external lab supplies. I had an LS27 at hand that has a structure 90% identical.
The Amp works fine. All currents drawn are within what you would call “normal”.
Total power dissipation was about 40W so no problem.

Conclusion: I had a spontaneous transformer fault that fried itself.

Now I am curious how ARC will react to this as I send them a full report with all measurements.

Cheers

[tdelahanty]

I'm at a loss, power transformers rarely go bad ???

[AudioIdiot]

Originally Posted by tdelahanty

I'm at a loss, power transformers rarely go bad ???

I agree, but the primary windings are fried, all below 1 ohm
I guess sh*t happens

[piolaxo]

Originally Posted by AudioIdiot

Well,

I powered the Amp today with 4 external lab supplies. I had an LS27 at hand that has a structure 90% identical.
The Amp works fine. All currents drawn are within what you would call “normal”.
Total power dissipation was about 40W so no problem.

Conclusion: I had a spontaneous transformer fault that fried itself.

Now I am curious how ARC will react to this as I send them a full report with all measurements.

Cheers

That is very resourceful -- must have been exciting to power it with those
external supplies and working without issues. At least now you know that
it is the transformer and nothing else.

Sal

[AudioIdiot]

The transformer is not covered under warranty as I am not the first owner, bummer. A replacement transformer costs €650....pffff.

But in my research where I had to draw the complete schematic with the defect amp as the source I found two major design flaws and one curious choice. I have send my findings to Audio Research so they can comment but have not heard back from them.

I will share them with this forum if I do not get any feedback from ARC to explain them.

[W9TR]

For 650 euros I would take the failed transformer apart to see if you can find and fix the problem. It is a bit tedious and time consuming but not difficult.

From the LS-28 inside pictures it looks like a conventional E-I lam bobbin-wound transformer.

It is unusual for a transformer to fail. But they do fail - usually from external causes like moisture or lightning.

Tom