Replacement Toroidal Transformer

[NoisyIndividual]

The diagram says 15vac to meter lamps but that diagram has a mistake. You can't have the - side of the bridge and the center tap of the xfmer going to ground. It's a short through the bridge rect.

The center tap should go to ground and the 15vac comes off the blue winding. Then the lights get AC.

You might want to review that.
You can't ground the centre tap.

The centre tap is the 15vAC supply for the meters, the return for this is either of the secondaries which aren't the centre tap, marked Y or BLK I think.

It is this return that is not designated in the schematic.

 

[RishiRishiRishi]

Yeah so I don't think the centre tap goes to ground - it almost certainly goes to the meter as a floating ground, as per schematic. I believe the mistake on the schematic is describing that as 15VAC, perhaps this is to just give an idea of replacement transformers.

After discussing with someone else who owned the unit, the original mixer did in fact contain a 2 x 15 VAC transformer at 15VA, despite the caps being rated for a lower voltage. I have a 2 x 15 VAC 50VA transformer that I can test out in there and see if it gets things going

 

[NoisyIndividual]

The 15v AC supply is either of the 2 secondaries.
It's as simple as that.
The schematic only shows one side, that's the only error.

Just don't ground any of them, you are correct to keep them "floating".

 

[NoisyIndividual]

Can you tell if the caps are original? I wouldn't be surprised to find Allen & Heath knew they'd be just fine.
I wouldn't do that with any Electrolytic you can buy today mind.

Perhaps the only "error" was using a poor little overworked transformer that wasn't able to keep up with the rest of the components.

 

[RishiRishiRishi]

Agreed - yeah it looks like the caps are original - they almost certainly haven't been changed and are in keeping with the caps on the rest of the board, by way of branding etc. Do older caps just tolerate more?

Also doesn't look like anyone has made any changes to the board at all. My theory is the previous owner thought it was wired pin 2 hot (this has been carved into the underside of the chassis by someone and underlined) and assumed it didn't work and then cannibalised the Tx. You'd imagine if the previous Tx failed, then at least one of the fuses would blow?

 

[NoisyIndividual]

I suspect in the old days the cap manufacturers would conservatively rate their product. Maybe they thought "we'll mark our 50v caps 40v, because we know what those naughty boys like to do with them".
Or maybe they just let the manufacturers know there was a certain tolerance built in.

Dunno, I'm making stuff up.
Today I would think the bean counters demand you just don't do things like that any more or we'd be out of business.

 

[RishiRishiRishi]

Apparently that is the case - they'd have -20%/+80% tolerance - maybe down to shoddy manufacturing, or anticipation for people overloading them - who knows!

 

[Brian Roth]

FWIW....the -20%/+80% tolerance is for the capacitance (microfarads) and not voltage rating <g>.

Bri

 

[Mr. Moscode]

You might want to review that.
You can't ground the centre tap.

The centre tap is the 15vAC supply for the meters, the return for this is either of the secondaries which aren't the centre tap, marked Y or BLK I think.

It is this return that is not designated in the schematic.

On second look you're right. The CT should float and only connect to the lamps. I was thinking a bipolar supply. My mistake.

If it was wired as the diagram indicates there would be short circuit of the xfmer though the negative side of the bridge. My X marks the line to be erased to fix the mistake.

 

[Mr. Moscode]

Can you tell if the caps are original? I wouldn't be surprised to find Allen & Heath knew they'd be just fine.
I wouldn't do that with any Electrolytic you can buy today mind.

Perhaps the only "error" was using a poor little overworked transformer that wasn't able to keep up with the rest of the components.

The "overworked xfmer" may have been chosen with the intention of including the resistance of the windings to keep the voltage on the first cap below 40volts. Also it's cheaper. That said, it may be why the OP had no xfmer in the first place as it may have burned out. Certainly if the input cap off the + of the bridge shorted. There's a good chance that the cap swelled or leaked.
Anytime you over-voltage a cap you're living on the edge!

Putting in a larger (50va) xfmr with lower resistance will probably raise the raw DC a few volts. One may want to take that into account. You could raise the 10 ohm R by a few ohms to compensate.

OK, enough from me. I have a warm day. Time to do some stuff outdoors fighting entropy.
Good luck.

 

[MidnightArrakis]

The CT should float and only connect to the lamps. My mistake.

[The CT should float and only connect to the lamps] -- Wouldn't it still be better if the "Signal GND" was isolated from the "Earth GND"? -- Since "Earth GND" is famously known to be rather "dirty", besides being an electrical hazard IMHO (i.e., accidental mains voltage coming in contact with "Earth GND" and instantly going > POOF!!! <), I would think that you would desire to separate those two GND's, no???



[My mistake] -- GO!!! ..... And live your life in peace, My Son!!! You have been absolved of all your sins!!!.....(joke).....

/

 

[Mr. Moscode]

[The CT should float and only connect to the lamps] -- Wouldn't it still be better if the "Signal GND" was isolated from the "Earth GND"? -- Since "Earth GND" is famously known to be rather "dirty", besides being an electrical hazard IMHO (i.e., accidental mains voltage coming in contact with "Earth GND" and instantly going > POOF!!! <), I would think that you would desire to separate those two GND's, no???

View attachment 122147

[My mistake] -- GO!!! ..... And live your life in peace, My Son!!! You have been absolved of all your sins!!!.....(joke).....

/

Earth and signal grounds is a different argument. I don't know how the xfmer is mounted and what the chances are for a short from AC to anything are. So if you have a primary to secondary short in the xfmer you'll have AC in your circuit.

If you use a split-bobbin xfmer the odds of that become negligible.

You can put some parallel 6A diode with the polarity reversed to the other and 10k R in parallel connected between earth and signal grounds with them to give you safety/ bleed off leakage and avoid a ground loop.

The lamp returns should go to the minus post on the bridge rectifier to reduce ground currents.

Perhaps we're over engineering this.

 

[NoisyIndividual]

The lamp returns should go to the minus post on the bridge rectifier to reduce ground currents.

No it shouldn't, you've done it again.

 

[NoisyIndividual]

I would suggest that since this desk still has the original capacitors some 40 years or so after it was made, there was nothing wrong with their design in this respect, they knew the capability of those caps.

If you are going to replace these capacitors it would be prudent to update the voltage rating, as modern caps are not a like for like replacement.

 

[Mr. Moscode]

No it shouldn't, you've done it again.

It could go across a winding but since I don't see a return on the schem I assume the lamps went to ground somewhere.

Don't get snarky with me. >"You've done it again." Just be civil.

 

[JohnRoberts]

Play nice guys....

JR

 

[NoisyIndividual]

OK, I have just noticed an ambiguity in the schematic that Mr Moscode highlighted.

The Red Cross purports to show there is a connection between ground and centre tap, but actually there isn't one.
All connections between traces are designated by a black blob.
There is no black blob between ground and the centre tap, but unfortunately the schematic has extended the ground wire left, so it does actually touch the drawn trace from the centre tap, so looks at first glance as if they are connected.

There is NO connection between centre tap and ground, and definitely shouldn't be.

That includes connecting the lamps between centre tap and ground, as Mr Moscode suggested in the second of his posts I corrected.

As I have said several times now, the lamps need to go across one of the secondaries. Period.
It really is that simple.

 

[RishiRishiRishi]

So I replaced the transformer with the 2 x 15 VAC, 50VA Tx, and wired it up as per the diagrams I attached and it is powering, and working! Thank you for everyones help with this!

I now have two new issues:

1. The mixer is very noisy through the headphone jack (haven't tested the main outs yet), and is very noisy even when nothing is plugged in, all channel faders are down, and so are the gain knobs.
2. Channel 2 is essentially dead, and only audible at max gain but is super distorted, and channel 5 is super quiet, but no distortion.

To fix this I was thinking of carrying out a recap of the electrolytics, and also replace all of the ICs. I really like the sound of the unit so want to stick with 741's. Do you think I should replace the transistors too? I'm sure this should sort out the majority of issues - what do you think?

 

[MidnightArrakis]

1. The mixer is very noisy through the headphone jack

What does the noise sound like? Hissy? Scratchy? Motor-boating scratchy? What???

2. I really like the sound of the unit so want to stick with 741's.

Those 741's are really "old-school" slow-speed, bandwidth-limited opamps. Are you sure you wouldn't to upgrade to something even a wee-bit more modern?

/

 

[RishiRishiRishi]

I would say the noise is very scratchy white noise, and a crackling even at low levels with respect to the monitor knob. I know most peoples thoughts on 741's as being negative, but I think they're cool and ideally would want to fix the noise issues/ dead channels.

I've just checked the main outs and they're a lot less noisy, compared to headphones out - but the individual channels are noisy

I can record a bit of the noise and upload if useful?