BA283 (summing) in Yamaha PM2000

Wanted to share this experiment and ask a couple questions. I've got it all wired up using the local power supply (PM2000 is +/-24V) So I 'm using one leg for one channel and the other for the other, reference to 0V (console bus ground)  These are for the main L/R summing amps.

The PM2000 uses 33K summing resistors. I have no transformer (input to the summing bus)and a 150ohm slugging resistor in front of the Ba283 right now just to see if this works. (Didn't have 200 ohm on hand, which is the value Ian recommended.) The channels in my console are very low output impedance into the 33K summing resistors, and there are a total of 38 channels going into the L/R buses.

So it works but I need to adjust gain to be higher in the BA283. I'm guessing between "T" and ground is where the best spot is? (First stage gain adjustment) I'm going by this literature http://www.celestial.com.au/~rosswood/data/images/NeveInfo2.PDF
It has a range of resistor values for this spot from 330 ohm down to 8 ohm.

I have a 10K fader between the first and second stages, so assuming I have the fader at "0" for setting unity gain,  I would then pick a resistance value between "T" and ground, theoretically should be in the ballpark of 40-56 ohm to provide around 30-33 dB gain from the BA283? That is assuming we're adjusting the first stage of the BA283?

With the Ba283, is the idea to set the first stage gain as high as I can without clipping and then increase the second stage to unity gain, or is it to set both stages about equally?

Another question. This is concerning the EA1166 output transformer. Here's a drawing from the Hairball site: http://library.hairballaudio.com/datasheets/1166-500.pdf

This is for the smaller 500 series transformer but I'm assuming the wire colors are the same for the larger (earlier) EA1166. Maybe my assumption is wrong.

It seems very confusing. In the original schematics, we have pin 1 of the LO1166 primary connected to "B" of the BA283, pin 3 is connected to "M" the +24DC. On the secondary side, pin 5 is in phase with pin 1 and pin 8 is in phase with pin 3, the 1.5K resistor is on the pin 5 side, and the .01uF cap is on the pin 8 side.

For the EA1166 on the page on the hairball site, we have pin 1 on the primary (brown) in phase with pin 5 (green) on the secondary, and pin 2 (red) in phase with pin 8 (grey) Okay, fine, looks the same except for some reason Ed or  (hairball?) decided to label what was pin 3 on the LO1166 as pin 2 on the EA1166  But wait. . .. maybe there's more that's different.

The way I read that at first was to wire "B" from the BA283 to 1 (brown) and  "M" to red (pin 2) Then on the secondary,  we would have the resistor connected to green (pin 5) and the cap to grey (pin8)

However, this yields much less gain than the Carnhill version (VTB1148)  in comparison. I couldn't figure this out until I swapped the primaries on the EA1166.

So, now it goes "B" to pin 2(red) "M" to pin 1 (brown) while the secondary stays the same, with the resistor connected to green (pin 5) and the cap to grey (pin 8)

How confusing. Seems like there will be a polarity flip, but I haven't checked that yet. The gain now matches the Carnhill though. So we're happy about that.

Okay, after all that blabber, my question is, do I appear to have the EA1166 wired up correctly now?

Final question: Where would be a good place to wire a "mute" switch for the BA283?

Here's some relevant info with the BA283 schematics: http://www.celestial.com.au/~rosswood/data/images/NeveInfo2.PDF



P.S. Should this go in the Lab or the Drawing Board sections? Seems like I never get replies if I'm asking about the PM1000 or PM2000. Like, people see Yamaha in the title and move on or something . . .

JW said:

With the Ba283, is the idea to set the first stage gain as high as I can without clipping and then increase the second stage to unity gain, or is it to set both stages about equally?

Click to expand...

The nominal internal operating level in a Neve console is -8dBu. So you should set the gain of the first stage of the BA283 to give you -8dBu at its output. The output stage gain is set to raise this to your normal operating level. Doing it this way will maximise your headroom.

Cheers

Ian

Thanks very much Ian. So, for -8dBu, I'm getting .308V.

Should the first stage be loaded with the second stage when I measure? And just a normal 1k tone feeding into an input strip?

JW said:

Thanks very much Ian. So, for -8dBu, I'm getting .308V.

Should the first stage be loaded with the second stage when I measure? And just a normal 1k tone feeding into an input strip?

Click to expand...

Load it with 10K and you should be fine. 1K tone is fine.

Cheers

ian

I am assuming you downloaded the ea1166 pdf?  I think I used Joe from JLM's Neve schematic to wire the EA1166.  I believe it made it understandable. Or keep bothering Mike at Hairball. He'll eventually spell it out for you like he had to me with the EA10468

He did get back on the white market thread, which spells it out.  I had it right in the beginning.  When I wire it like this it clips nasty like an opamp.  https://groupdiy.com/index.php?topic=28923.1080 I must have messed something up with the transformer (pulling on wires? I don't know. I need to at least ohm it out but I gotta get back into chassis to get at it. The other EA1166 is the same though. I have a JLM 111DC as well and that has no clipping but is lower in volume than the Carnhill, wired like JLM's 1290's suggest (1:2) Very weird to have this many problems with transformers (I even tried replacing the zobel cap) I thought for sure it was one of my channels (w/ the BA283 cards) but the problems follow the transformers, not the BA283 cards.  Carhill sounds great and can take a massive amount of level. Sounds right to me. Haven't sorted out the gain of the two stages yet though.

I'm getting 312mV at the output of stage one with a 7.2 ohm resistor to set gain. Looking at this literature again http://www.celestial.com.au/~rosswood/data/images/NeveInfo2.PDF it looks like a 8.2 ohm would be maxing out the gain of the first stage at 48dB. Can we call it 8.2 and still be okay I wonder. Okay, I'll set the second stage tomorrow and give a listen. I imagine this much is needed because of the 38 channels w/ 33K summing resistors. And Neve used less, so perhaps we're maxing out the first stage?

Update on this project. Setting the first gain stage for an output of .308V (-8dBu) results in an inability to reel it in in the second stage. What I have now is a 9 ohm resistor setting the gain of the first BA283 stage, and I've removed the resistor and cap between junction K and ground for a 15dB second gain stage. The output of the first stage isn't quite .308V, but it sure sounds good. We're talking the difference between a 7.2 ohm and 9 ohm resistor for the first stage. Very sensitive!

Also a couple of observations I want to share should they be concerning. . .

With Carnhill (VTB1148) the secondary has about 300mVf DC on it. Is this acceptable?

The EA1166 transformers clip. According to the way Mike at Hairball says to wire them up for  Neve output,  they clip nasty like an opamp. The primary is around 15ohm and the secondary is 53ohm. These are both series measurements. Does that sound right?

The output stage should easily be able to raise -8dBu to +4dBu. If it does not then there is something wrong with it.. Did you set the standing current in the output stage?

Cheers

Ian

Ian, the problem I was having was the output was too much even at minimum gain (15dB?) . So I had to lower the input slightly. I think you misunderstood or I didn't explain very well?

How do I set the standing current of the output may I ask?

Also, where would one wire a mute switch?

Seems confusing the way I'm explaining so I'll just do some bullet points.

-7.2 ohm resistor sets the first stage at .312V This proves to be too much gain over all. The only lowering of the second gain stage that I understand (from being fixed at 18dB) is to remove the 1.5K and 100uF cap between "K" and ground which makes it 15dB

-A 9 ohm resistor in the first stage results over all in a +4dB output, (with the second stage at 15dB) but the first stage output measurement becomes something like .250V

caveat: Could be that my Radio Shack meter is not accurately measuring the output of stage one at 1K.

But I'm pretty sure I'm close to maximizing the first stage (?)

JW said:

Seems confusing the way I'm explaining so I'll just do some bullet points.

-7.2 ohm resistor sets the first stage at .312V This proves to be too much gain over all. The only lowering of the second gain stage that I understand (from being fixed at 18dB) is to remove the 1.5K and 100uF cap between "K" and ground which makes it 15dB

-A 9 ohm resistor in the first stage results over all in a +4dB output, (with the second stage at 15dB) but the first stage output measurement becomes something like .250V

caveat: Could be that my Radio Shack meter is not accurately measuring the output of stage one at 1K.

But I'm pretty sure I'm close to maximizing the first stage (?)

Click to expand...

OK, now I get it - you have too much gain in the output stage. Don't try to reduce it, just place a 10K pot between the two stages and adjust for the correct levels.

I cannot remember the exact process for setting the output stage current but make sure J is connected to A and place a dc voltmeter between D and A. Adjust RV1 for about 2V on the meter.

Cheers

ian

Just wanted to reply again and say my levels for testing were faulty earlier when trying to set the first stage voltage.

After correcting, I actually ended up with about a 30 ohm resistor in the first stage, and a 620 ohm to get the second stage to a +4output, if anybody wants to try this. I'm pretty stoked at the sound of these amps. Everything going through the console now sounds more real, with seeming depth especially in the low mid area. Just my impression.

Many thanks to you Ian.

I guess I should ask once more about the mute switch, since I have one right there in the channel. I tried wiring one up to break the connection between the output of the first stage and the fader, but it's not absolutely silent when off. Hmmm.

Is there a better place for a mute? I'm having a hard time finding a schematic where a mute is implemented with this circuit.

Depends how you wired it. For best results you should not only break the connection but also ground the input to the fader. So you need a SPDT switch. Connect the wiper of the switch to the top of the fader. One switch position connection goes to the output of the first stage and the other goes to ground.

Cheers

Ian