Martin

OK, I understand that you're focusing on the MPSH54 as the solution.

I checked my parts box, and I have 6 Brand-new NOS MPSH55, which is listed on the same MPSH54 original datasheet from Motorola, showing it as a low noise RF transistor. The only difference between the 54 and 55 is shown as max HFe.

These have the gold leads.

I checked the HFe, and it reads 39. The minimum HFe per the datasheet is 30.

 

I installed it in the left channel, and the Relay went into protection.

I removed it, and reinstalled the MPSA55 (Hfe 160), and the relay did not activate - but still reading 94mV at TP111.

 

So, perhaps my MPSH55 HFe is too low at 39.

Can you check the HFe's of your MPSH54's (if that is your shop)?

 

My batch of MPSH54 show an HFE of 62-63 FYI Menahem. Not certain at all it’s relevant though.

Have both a nice New Year’s Eve!

Menahem Yachad wrote:

MPSH55...
I installed it in the left channel, and the Relay went into protection.

Sure you got the pinout right? They have reverse E/C from most other TO-92.

Martin

Pinout correct - double-checked before installation! But thanks.

More progress:

I'm concentrating in the area around the Diff Amp.

I replaced TR112/113 BC546B HFe 300 with 2x BC546C with HFe 477. Down to 69mV.

So HFe is very important here - drop of 25mV with Higher HFe.

This is the best progress so far - I'm not intending to change these again.

 

Removed TR115 MPSA55, replaced with BC560C HFe 450 – 68mV – not significant

Replaced C140 5pF with 6.8pF – 63mV - drop of 5mV. The measured values are not exact anyway, and expecting drift. I was looking for a favorable change in mV, and that’s what I got.

Checked (very carefully) all the semis for correct readings under power.

Replaced D102/202 Z2.7V – both reading 1.8V, with NOS ITT Z2.7V (reading 2.66V) – I’ll guess that they were both damaged when DC caused the Relay to trip in.

L/R DC Offset now reading (using DBT) -29mV minus – I probably ignored the Minus sign in all my previous readings.

Just to see if there would be an improvement, I removed Diff Amp 2x BC546C, and installed 2x BC546B – No. DC Offset increased to -49mV.

Reinstalled the 2x BC546C again.

 

Today’s score is (net AC, without the DBT):

 TP111 Left = -26.7mV, 

TP211 Right = -28.4mV 

(now within the SM Spec of 0 +/- 50mV)

Speaker line (junction of 2x 0.18Ω Emitter resistors) 

Left = +15.6mV, 

Right = +13.8mV.

Bias is dead stable on both L&R at 13mV .

So, in summary, up to now, the components which made the most impact on reducing the high mV DC Offset are:

TR112/113 TR212/213 Differential Amp – Optimum = 2x BC546C (HFe 477), instead of BC546B

D102/202 Z2.7V

I have the MPSH54 on order – when it arrives, I’ll remove the 2x MPSA55’s (TR115/215), and see if it makes any difference – I’m very curious to see if it does.

 

Offset seems low enough now. I’ve had a look on my lounge 4400, Fluke says 13-14 mV.

But it was last adjusted many years ago!

Ha - you made me laugh!

I can just picture you reading my post, and then saying to yourself "Hm, I should get up and measure my BM4400, just to make sure!"

I saw a note somewhere about replacing R176/276 180Ω with a 220Ω. Is that something that any of you know about?

I'm certainly going to give it my best shot at getting it under 10mV.

I've invested a fortune of effort in this so far, I'd like to do the job as right as possible.

Menahem Yachad wrote:

Ha - you made me laugh!

Glad to have made your day then 😕

Believe this or not - I cleaned the BM4400 with compressed air, and the DC offset dropped to 12mV on both channels. It was not that dirty to begin with, but coming to the end of the job, I wanted to improve the look as much as possible. 

Today, I'll polish the Rosewood panels.

FYI - Another little trick I employed to protect the output stage semi's during this troubleshooting, is to remove the 4 emitter resistors, and replace them with 1Kohm 5W resistors. That drops the current to a couple of nA. Do NOT readjust the bias trimpots - leave them as they were before changing the resistors. This enables me to change components freely, without fear of blowing the transistors, while tracking the causes of excessive DC offset.

Of course, always use a DBT when poking around like this. It makes a slip of the probe a non-event.

The other change I made to the circuit, was to reverse the mounting position of the right-side resistor of each of the 2 emitter pairs, so that I can measure across both of them easily, from the top of the PCB, and measure each of them individually.

On both channels, I have identical Bias - left emitter 6.4mV, right emitter 6.9mV = 13.3mV total.

Last job will be to install the MPSH54's when they arrive.

I got the MPSH54's today.

First thing is power to the BM4400, to confirm the Baseline measurements.

Speaker Output L/R = +12mVDC

TP111 Left = -26.7mV, 

TP211 Right = -28.4mV 

 

Now, I installed the MPSH54's

 

Speaker Output L = +2.8mVDC

Speaker Output R = +0.23mVDC

TP111 Left = -35.9mV, 

TP211 Right = -37.9mV 

 

I had to readjust the Bias again, after installing the MPSH54's.

 

So, the MPSH54's are the icing on the cake.

In summary (after the full restoration), these were the troubleshooting steps which provided the most significant gains:

1. Replace Diff Amp TR112/113 and TR212/213 with 2x matched BC546C (not BC546B) - the SM specifies the C-grade, if the BC54x is used.

2. Replace D102/202 Z2.7V – both reading 1.8V, with NOS ITT Z2.7V (reading 2.66V) 

3. Replaced C140 5pF with 6.8pF

4. Replaced TR115/215 with MPSA55 first - that brought the 0mV rail down to 12mVDC, and afterwards the MPSH54 brought it down to 0.2mVDC.

 

My conclusions - the MPSH54 is ideal (I also don't know why), but if you can't get them, then the MPSA55 will also do an excellent job, to within 10-12mV. 

 

Restoration complete!