Great River MP-2 , info, schematics & discussions

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
hey guys,

anyone making progress with this project? I finally got time to start stuffing some boards. I have a few small questions which I don't believe have been answered in this thread (just read through it). None of these questions are deal breakers - just want to understand what I'm building.


-I've read about the process to select R46 but R15 and R18 are also marked as TBD on the schemo. On the parts list they are down as 221R. Do I go with the 221R or do they need to be selected in some fashion?

-If I skip the output transformer and just do an impedance balanced output what resistor should I add to "balance" things?

-Speaking of the output I went with 20ohn resitors for both R36 and R47. Dan mentions the difference this makes when driving an output transformer - am I correct in assuming it makes no real difference when not using an output transformer?

-Why are R30 to R33 and R36/R47 1 watt? It seems to me that the output current flows through R38 anyhow which is 1/4 watt.

-What is the function of the pair of RFI RXE-040 on the supply lines? Are they some form of current limiter. Do the they have a polarity? If so do we need to stuff them in a particular way?

-Final question!! I'm using Lundahl 1538 for input transformer and would like to leave out the zobel as per the advice from Dan. Am I correct that C20 and R7 are the zobel?


Thanks in advance for any info.

Cheers,
Ruairi
 
I'll alert Dan if he doesn't see this but
the unbal option is a great one , i just took that out off the board
and it sounds great , it wasn't until i made a quad that i added
output xfmrs [ the Jensens ] and it's a good sound too , a little thicker
for cleanest i'd go totally unbal [ i've driven 150 ' no problems ]

O.k. even if you do the impedance balenced , i'd strongly recomend
adding a 1/4" unbal out , can't hurt OR go wrong
[ doesn't massenburg like to run unbal ? ]

unless you already have a 1538 , the XL version is worth considering
no schematic on hand but yeah you can see on the board where the
cap & resister meet [ 200 pf ? ] leave em out , i think i noticed a slight
lift but couldn't tell you anymore

I started with the XL's and didn't want to mess with something that
sounded so good , the extra handling is in the lo freqs but i think it
makes it sound fuller [ brush ] overall



regards Greg
 
Hi Greg,

thanks for the answers. I'd rather do an impedance balanced arrangement than unbalanced if it's possible. Good to know the unbalanced output works well.

I already have the Lundahls, got a nice deal on a used set of 4 from a fellow lab member. I'm not too concerned about the extra level handling of the XLs, most of my work lately is acoustic and folk recording at lower levels.

Thanks for alerting Dan.

Cheers,
Ruairi
 
Hi Ruairi,

The 221's are what I ultimately selected for production runs years ago, they set the currents in the internal stages where I wanted them. I should've marked the schematic differently, but forgot.

The output impedance is pretty complex, due to the output stabilization network (R34, R35, C23, and L1) plus the active output impedance of the amplifier, plus the build-out resistor, be it 10 or 20 ohms. I'd be inclined to just match the build-out resistor and figure the mismatch isn't going to be too bad. Sloppy, but the math is beyond me. PRR???

The build-out resistor value isn't real critical, for a completely unbalanced system something in the 47 ohm range might be more appropriate. When you add the impedance balance resistor the output Z is going to be fairly close to nominal shielded twisted pair impedance, plus the 47 ohm 1W will operate into a short circuit all day long at just about any signal level.

True, the output current flows thru R35 (the quarter watt in parallel with L1), but only a portion, and only the highest of the high frequencies at that. So, unless you get the sucker oscillating at a few hundred kc's it'll never see any real load current. (In the schematic I'm looking at, R38 is part of the positive supply regulator).

C20 and R7 are indeed the zobel network. I think you'll want to change R6 to something bigger to provide the right load for the Lundahl. If I recall correctly that transformer wants to see a 100K load or so. Don't forget that R8 is in parallel with R6, so in this case a 121K for R6 ought to be about right.

You can leave the RXE-040's out, just replace with a jumper. I've never seen one trip, ever. The board is current limited by the LM-317's/337's to half an amp or one and a half depending on the regulator you're using.
They are a resettable fuse, no polarity.

Have fun,

Dan
 
Hey Dan,

[quote author="Dan Kennedy"]The 221's are what I ultimately selected for production runs years ago, they set the currents in the internal stages where I wanted them. I should've marked the schematic differently, but forgot.[/quote]

no worries on the 221s, I imagine you were building this thing in your sleep towards the end of its run! This thread updates everyone to the changes to the BOM and schemo so we're all clear.

The output impedance is pretty complex, due to the output stabilization network (R34, R35, C23, and L1) plus the active output impedance of the amplifier, plus the build-out resistor, be it 10 or 20 ohms. I'd be inclined to just match the build-out resistor and figure the mismatch isn't going to be too bad. Sloppy, but the math is beyond me. PRR???

I used the 20R for R47 so I'll match that.

True, the output current flows thru R35 (the quarter watt in parallel with L1), but only a portion, and only the highest of the high frequencies at that. So, unless you get the sucker oscillating at a few hundred kc's it'll never see any real load current. (In the schematic I'm looking at, R38 is part of the positive supply regulator).

Thanks for the info, the mention of R38 was caused by the couple of glasses of fine Italian wine I treated myself to after my nights soldering.

C20 and R7 are indeed the zobel network. I think you'll want to change R6 to something bigger to provide the right load for the Lundahl. If I recall correctly that transformer wants to see a 100K load or so. Don't forget that R8 is in parallel with R6, so in this case a 121K for R6 ought to be about right.

Not sure on this one ... according to the 1538 datasheet -
Optimum termination for best square-wave response (Connection 1:5, source imp. 200Ω ) - No termination necessary

I have no experience setting up a zobel so I'll defer to those who know.
Going from Greg's practical experience it probably makes little difference with this particular transformer.

You can leave the RXE-040's out, just replace with a jumper. I've never seen one trip, ever. The board is current limited by the LM-317's/337's to half an amp or one and a half depending on the regulator you're using. They are a resettable fuse, no polarity.

Thanks for this info dan, I already have them so I may as well install them unless they have any effect on the PSU quality. I was just wondering what they did. Being a DIY build they might come in useful to keep the smoke in.

Thanks for the great info and the great project, I'm really enjoying the build.

Cheers,
Ruairi
 
Hey All,

As usual, I am ~ a year late to the completion party. I am finally turning up some steam on building my backlog of projects (3 compressors coming soon). I actually got this working on the first power-up (thanks to everyone who had and commented on their mistakes/problems/build tips, particularly the power section wiring reversal on nrg's pdf document) but am getting enough hum to make me think there is a grounding issue. I'll post a photo of my build later tonight. A big ole thanks to everyone involved in putting this project together. This was a fun build (so far ;) )

2 quicky questions -

1) I built this using a JLM AC/DC powersupply - The output side has a section for the mains grounding and a 0v for audio. I have all my audio grounds in the MP2 being sent to a single, star mains ground connection on the case. Should anything be hooked up to the 0v on the powersupply? I assume no but I wanted to make sure.

2) Unrelated to ground and probably a ridiculously simple question - Since I am not using the fancypants powersupply used by AW Music and OneRoomStudios in their builds (thanks for sharing photos - gorgeous builds) which seem to handle the 48v indication light, should I just use a DPDT switch instead of the SPDT other folks are using & use it to control power to the indication led as well as the relay?

thanks!
 
sorry for the el stupido questions - all figger'd out. i'll get a chance to test it out in a session tomorrow afternoon. Thanks again for putting this project together!!!

inside.jpg
 
Guys i'm helping our mate Bluzzi (Jim) get his pair of GR done.

Now can someone point me or give me some direction in the lastest step of this? Calibration maybe?

Thanks.
 
Hey 3nity , other than getting the parts in right
selecting the fet resister is the only thing like
a calibration , r46 [ i think check back in this thread ,
first page ] use a trim pot & measure / adjust for lowest
offset on output, once you have checked back a couple of times
when it is warmed up it will settle down and you will have
a resistance value from the trim pot to measure and replace
with a fixed resister value .

check back in this thread to get the procedure as quoted from Dan ,
but that is basicaly it .

regards Greg
 
[quote author="Ptownkid"]I will be doing another run of kits for this project in the very near future.[/quote]

I am for sure on this one when you do. This was one of the first really great preamps I ever used, and still ranks at the very top after some years. I think if too many cool projects surface I am going to have to enter a 12 step program!
 
Hey Greg.
Will i do a horrible mistake if i use a CM75101 i have? as input?
This is the only quality input i have!

thanks?
 
What's the ratio of that one ?
a 1:5 or so is what's used a higher ratio may lose some headroom
, not really needed for the gain but no harm if you use one in the
meantime , if the ratio is close you should have no problems
Have Fun !
 
[quote author="Tungstengruvsten"]does anyone have available the small pcb's for the elma switches? I have 4 PCB style ones here leftover from my 33609 order, would be handy to put them to use with this project...[/quote]
That's interesting, I was looking for a way to turn Ptownkid's ELMAs into PCB mount ones for my 33609 build. Would you be willing to do a trade, either for ELMAs or money?

JDB.
 
JD,

how many Elmas do you need? I think I might have 2 of the pcb mount units for the 33609 to spare. Let me know if two is enough for you and I'll take a look for you.

Cheers,
Ruairi
 
Back
Top