Great River MP-2 , info, schematics & discussions

[tonedude]

What is the difference in impedance to the circuit? that loading maybe one thing to consider...

It's mostly hooked up to my MH 2882 interface, that's 10k. I'm not sure if I have any 600ohm stuff, maybe my ADL1600 is 600ohm (I guess it is... ).

 

[dmp]

So I looked at my problem (buzz) with the scope and did not figure it out yet. I see the buzz on the scope on the power rails. 60 Hz, sawtooth, about ~200 mV.
Tried a bunch of things, with the load disconnected, the sawtooth is gone from the power rails, but with the GR boards connected it comes back.
I think the rectifier board I made might have a bad layout.  The schematic is the same as Dan Kennedy's design.
Any tips on a recitifer layout? Right now the audio and chassis ground connections come off one side of the board, while the transformer connects on the other side. Should I move the ground connections directly to where the transformer connects?

 

[okgb]

I use a psu pcb from Dan , no problem , it was in production ,
All grounds go to the xlr bussbar,  psu straight there not the preamp board , any preamp
ground goes to the bussbar , I have to open mine up to see but,  I may have ignored the chasis ground
you may have to play with that.

I was also referring to the loading of the preamp with the output xfmr .

 

[tonedude]

I see the buzz on the scope on the power rails. 60 Hz, sawtooth, about ~200 mV.
Tried a bunch of things, with the load disconnected, the sawtooth is gone from the power rails, but with the GR boards connected it comes back.

Have you tried messuring the PSU with another load?

 

[dmp]

Have you tried messuring the PSU with another load?

No - that might be a good test.

 

[dmp]

I tried the rectifier board with 4.7k resistors as load and did not see any noise on the scope. I think the next test will be to hook up a alternate power source to the GR boards.

 

[tonedude]

Searching the web for inspiration about zobel and load to use on the LL1538 in my MP2 (I will use the xl version), and this is what I came up with:

- Soundskulptor MP12 recomending the use of a 33k load and no zobel for the 1538xl (giving a 1320 ohm input impedance)

- Focusrite ISA 110 using a load of 34k in parallel with a 220k resistor for the LL1538 witch is about 29,5k as load on secondary (approximately 1200 ohm input impedance), and a zobel in parallel with load consisting of a 5,6k resistor and a 220pf cap.


Googling info about the ISA 110 gives me the info that the zobel and load for the LL1538 in the ISA 110 is designed by Neve himself, that can't be to bad, can it?    ( http://old.focusrite.com/products/mic_pres/isa_two/key_features/ )

Questions:
- I have no scope to do the measurements and testing for a zobel myself, but does somebody have an opinion on the use of the "ISA 110 Zobel" and load in an MP2?

- And, any comment or thoughts on using a 30k load with the MP2 for a 1200ohm input impedance?


... picture below showing the ISA 110 input trans incl load and zobel...

 

[okgb]

I believe the Zobal is not needed with the 1538 , the boards have or were designed to use one for the jensen
When I suggested to Dan that the Millennia pre was a touch airer the my MP-2 with the  1538XL [ Which I really like ]
he recommended to remove the cap & resistor [ I'd have to check the parts number , should be obvious ]
check the Lundhal website but I believe that's correct.

from the Lundhal site

Optimum termination for best square-wave
response (Connection 1:5, source imp. 200Ω )
No termination necessary

 

[tonedude]

Greg, I guess you are right about that. I just started to "obsess" about what loading to use on the secondary of the transformer, to adjust the input impedance of the pre, and found out that a zobel is used in all the Focusrites using the LL1538.

When trying to read in on the topic I stumbled upon this thread, Are we loading our mics properly?. When reading it I found out that the impedance on different preamps spans from hundreds of ohms to thousands of ohms (Millennia Media HV-3B is 6.2k), and the GR MP2 with jensen input has about 1600. So just based on that I guess a pre with an input impedance of 6.2k is probably gonna sound airier than one with 1.6k. 

And in this article Shure SM57 Impedance Modificationl, it's noted that loading the microphone is fundamental to how it's gonna sound, especially if it's a dynamic microphone or some other kind of mic with a transformer output.

I'm gonna try different values for the loading of the LL1538xl i my MP2 and se how it affects the sound of the preamp. Maybe I'm gonna make the loading resistor switchable to make the mic input adaptable to different sources, say 600 and 1600. Right now it's 1200 and I'll try that for some time to se how I like it.

 

[tonedude]

The transformer loading is now on a switch, I ditched the pad switch in favour for a load switch, so now I can chose between a 18500 and 39000 as a load resistor on the LL1538 secondary, giving me the option to chose between a 500 or 1600 input impedance. The pre sounds beautiful!

below is a picture of my MP2 build, almost done. External PSU, shared with another pre I have built. I have to find knobs that fit those long shafts... but its on it's way.

 

[okgb]

Let us know if you end up using one load position more than the other, nice
to see you went with unbalanced outs as well, may it serve you well !

 

[dmp]

Well, I'm still having a buzz on my pair.
I tried a different psu (mnats) and the buzz was even a little worse. So that eliminates the recitifier board.
I triple checked the grounding and there doesn't seem to be anything wrong.
I found that if only one board is plugged in, the buzz goes away. As soon as I plug both boards in it starts up.
And when I unplug the power cord, the buzz dissappears for the 2 seconds as the caps discharge (while the preamp is still working).
So, ground loop or what?
The psu is internal - could the psu transformer be sending off bad vibes under the load of both boards?

 

[dmp]

After much trial and error I got this figured out.
I had a bad ground connection at the star point to the case (dumb!)
Then the Jensen input grounding was just strange. With the black and white tied together to the case I had a buzz that would not go away. Putting the black to the audio ground was the only way to make it quiet. 
Wires to the DI are very sensitive and moving around would hear some buzz. Might put shielded wire there but I'm going to use this guy a little first.

 

[G-Sun]

Grounding scheme
As recommended by Dan 
Run a heavy bus wire along the back of the i/o connectors - in Canada & U.S. the copper house wire fits perfectly into the XLR connector
Have the Chasis & PSU grounds meet at this point
Run the Preamp board ground [ including the psu ] to this point
Run the Input Xfmr screen & ground as well as any shielded i/o wire to this point , preferably between their respective i/o connectors .
Do not connect any of these grounds [ i/o & Xfmrs ] to the input side of the board

regards Greg

Do I ground the output unbalanced jack ground to the bus-ground-wire?

 

[dmp]

Do I ground the output unbalanced jack ground to the bus-ground-wire?

The unbalanced ground should go to the J7 connector along with the positive, i.e. 1/2 go to the transformer, 3/4 go to the unbalanced output.

 

[G-Sun]

Do I ground the output unbalanced jack ground to the bus-ground-wire?

The unbalanced ground should go to the J7 connector along with the positive, i.e. 1/2 go to the transformer, 3/4 go to the unbalanced output.

Ok, so I don't connect unbalanced ground to anything but the pcb then. Thanks!

 

[G-Sun]

For the JLM ACDC

Do I connect case(in) to ground-bus (xlr-pin1), 0v(out) to ground-bus or both?

 

[G-Sun]

Ok, I'm soon ready to fire up

Finding the value for R46 [ with fet ]

R46 is determined by putting a FET in the circuit, then adjusting
for minimum output offset. This value drifts some with temperature, so
I try to adjust for <.5mv when the circuit is warmed up.

You are selecting r46 for minimum output offset voltage at output
[ before output transformer of course  , if you installed one ]

just power the board up, it already has the feedback resistor, and put
a 2K or so pot in for R46, adjust for minimum, let things warm up some
more and readjust. After 15 minutes or so take the pot out of the
circuit, measure and replace with a fixed resistor of the closest 1%
value and then retest for warmed up minimum offset.

It takes a little while, but it's something you can walk away from in
between tweaks.

Further refinement [ and quote ] from Dan on slecting the resister

When adjusting for R46, I first power the boards up, and see that the
offset is less than about 15mV, and frankly I forget which polarity.

This proves that the FETs are a reasonable match.

Then insert the trimpot in for R46, adjust for zero offset, which you can't
do because just breathing on the board will move it a little, but you can
get it trimmed into less than 1mV pretty easily.

Remove and measure the trimpots, then replace with the closest 1%
resistor you can get.

It's not super critical, the offset won't get any worse, and anything under
5 or 10mV's of offset isn't going to cause any problems.

What fet is this? Any fet? Is it on the boom. Is it on the pcb?
.. I guess it's two gain-fets on the pcb/boom,
and that I'll just now balance + and - side, right?
Ok, so need to order a 2k pot then

 

[G-Sun]

The bom for resistors has quite some esoteric values,
with prices accordingly..

 

[dmp]

What fet is this? Any fet? Is it on the boom. Is it on the pcb?
.. I guess it's two gain-fets on the pcb/boom,
and that I'll just now balance + and - side, right?
Ok, so need to order a 2k pot then

Take a look at the schematic. There's only two FETs in the whole circuit and they are right above R46. Q1: 2sk-389
The small trimming using R46 balances the FETs to calibrate the amp.