Mitsos API 312 / 990 Build Thread

[mitsos]

I wan't going to mention (or am I doing that right now) proper shielding, as I know Ben has successfully built lots of gear, but here are my thoughts on the subject:

1.  IEC ground pin connected to chassis.
2.  Should have continuity between IEC ground and every panel of the chassis. Check by touching your DMM to the threads of different mounting holes, because, obviously, the paint is not conductive. If any panel does not have continuity, you will need to scrape away some paint where it touches the next panel or your case will not work as a shield.
3.  EVERY XLR Pin 1 connected to chassis (some may want to stick a switch between output XLR pin 1 and chassis to allow lifting the ground... your input XLR pin 1 MUST be connected to the case or your phantom won't work). 
4. If you use shielded cables for your I/O, connect the shield to the XLR pin 1 ONLY. DO NOT connect it to pin 1 AND your PCB.
5. Methods of grounding the circuit may vary a bit, but generally you will either take everything individually to one ground point (the so-called star ground) and then connect this star point to the main chassis ground through a switch of some sort OR you can run everything individually to your PSU and connect the PSU ground to chassis ground, again, through some sort of switch so you can break the connection if you ever need to.

Unless I'm forgetting something, that should do it.

Unshielded cables will not make or break a well grounded/shielded system. Unless your PSU trafo radiates noise into the circuit.  To give some perspective, a while back I bought a DIY312 which had NO shielded cables on I/O, even the DI. It did have an external PSU, but in any case, it never had any problems with noise.

Assuming your grounding/shielding is OK, as I assume it is, and your opamps are known to be working, I would look for problems where the signal is balanced. Basically check from the XLR input up to and including the input trafo, and output trafo through XLR out.  Check DCR on your trafo windings to make sure you don't have an open winding. Check the input pad resistors and switch, as well as the whole daughterboard for cold solder joints. And don't forget the 5-pin molex that connects to the smaller board.

Hopefully you'll find it quickly and it is something simple to fix.  If not, then inject a signal into the circuit and probe around with your scope or a signal probe (if you don't have one of these, here is a link to DIY instructions by a fellow member here, Mark Burnley:
http://www.diyfactory.com/data/mbsignaltracing.htm

good luck!

 

[tmuikku]

-- I wan't going to mention (or am I doing that right now) proper shielding, as I know Ben has successfully built lots of gear ...

Sorry if I sounded too arrogant Ben! I find it difficult to properly express myself on the Internets sometimes. What I ment is:
I'm positive it is something rather simple since I got mine running with relatively limited knowledge. Hopefully you get your preamps running good!

 

[benlindell]

I'm pretty sure I figured out what's causing this, I forgot to solder on my DI Jack  : I figured this out randomly in my head during a session today, I'll put that on tomorrow and recheck it, hopefully it's as simple as that.

 

[mitsos]

oops!  Yes, if you did not install anything in there (switching DI jack or jumper if not using a DI) the pre wont work as the mic signal goes through the DI to get to the opamp. Post back after you test this!
cheers!

 

[benlindell]

Sweet, works perfectly!! super quiet too considering it's just on my bench. Time to finish the other 15!!

 

[grantlack]

haha this made me realize that i never jumped mine either.  well played ;D.  now i just need to find the time to get these things powered up  :'(

 

[mitsos]

great news Ben!  (and grant too!)

let us know when you get further with them!

cheers!

 

[baadc0de]

Sweet, works perfectly!! super quiet too considering it's just on my bench. Time to finish the other 15!!

Mine got working with phantom, microphones, leds, DI and the works yesterday too. I even have an output attenuator and some mods that were suggested by ppa. And, yes, now to finish the other 7...

 

[mitsos]

people slowly coming out of the woodwork! cool! 

I too have to get back on track with the 8-pack... right now there are too many things to fix in the new house... the other DIY..

Curious about the mods you mention too.. care to share? 

cheers!

 

[baadc0de]

Curious about the mods you mention too.. care to share? 

I've been wanting to ask about it anyway.. the mod is a 470k resistor from DI out to gnd, implemented by having c11=link and r13=470k.

The other one is R21=22Ohm instead of 47Ohm, but I haven't tried that one I think. This is done to accomodate (protect?) ppa's germanium opamps of which I have a pair.

This is WIP, but still, I like pictures..
http://farm6.static.flickr.com/5126/5341722357_010156a1ef_b.jpg
http://farm6.static.flickr.com/5242/5341722175_9a75665468_b.jpg

Sorry about the blur. They should all come together after 1-2 days.. and then to link them up with 8ch soundskulptor tape simulator and mic some drumkits!!

 

[mitsos]

I've tried 1M in there, but didn't find it to make much difference. ( I actually stuck the resistor on the DI jack but I think it is better closer to the opamp.  I'll try 470K next

Haven't tried 22R after the opamp either. Just 47R or 0R. 

I like the case!  I can't remember... was that you who had Grandmasteraudio do one up?

thanks!

 

[baadc0de]

Haven't tried 22R after the opamp either. Just 47R or 0R. 

I like the case!  I can't remember... was that you who had Grandmasteraudio do one up?

I will be trying 22R tomorrow I think. Yes, Pete slaved about this one for some time .. I'm sure he still has nightmares about it, we must have exchanged over 50 emails though in the end the results were worth it. I love the compact, unassuming badassery that this is. With coloured LEDs no less! I'll be sure to post some better pictures when the project is complete.

 

[mitsos]

Here is a question from the BM, thought I'd post here cuz the BM threads get deleted after a while.

Those blue switches will work fine, if they are MBB.

Bob's spreadsheet is great. It is made to calculate individual gain steps like you see in his 312 schematic, only one resistor is in circuit at each step. To wire this on an offboard switch, you need to wire each resistor to one switch contact and join them all together on the other side and run a wire from here to one of the pot pads. Connect the switch pole to the other pad. Do not wire the resistors between contacts the way you see most switches wired here (I found this out the hard way).

The way he has it makes it easier to get exact gain steps but it's a bit harder to solder to a round switch.

I can email you a modified version of this file (to make a switch that resembles a rev log pot - resistors get soldered between the switch contacts) but you'll have to double(triple) check the math.

It would be great if you would email the spreadsheet and maybe if you have some, a picture of your switch with the resistors in place.

Thanks guys.

Cheers
Soren

well, my internet sucks atm.. can't seem to upload attachments to my email.. but I did find the spreadsheet so I should hopefully be able to get it to you tomorrow. Here are the values I used on the switch in front of me. I ended up changing to a custom dual pot, 1K and 10K I think,kinda like JH uses in the M1, mainly cuz I could get them and wanted to try.  I modified the output pad switch to switch between the two gangs of the pot for high and low gain and it works very well. If anyone is feeling adventurous, I can shoot some pics whenever I get them back from my friend (remind me end of the month).  But here goes the switch resistors... and a pic attached. The wires need to go to the pole and to the wire. ah, yeah, the wire that the resistors are soldered to is some leftover 1mm solid wire which I made into a pretty awful circle.

6 10027.60338 10000
8.2 6347.821887 6190
10.4 4306.548657 4320
12.6 3042.041472 3010
14.8 2204.491877 2210
17 1624.883544 1620
19.2 1211.510669 1210
21.4 910.3212243 909
23.6 687.4330937 681
25.8 520.5859528 523
28 394.6132219 392
30.2 298.884002 301
32.4 225.779123 226
34.6 169.741873 169
36.8 126.6639159 127
39 93.47505663 93.1
41.2 67.86160163 68.1
43.4 48.06849711 47.5
45.6 32.75762199 32.4
47.8 20.90466534 21
50 11.72309384 11.8
52.2 4.607493396 4.5
54.4 -0.909015696 0

 

[benlindell]

Quick silly question, the Altran Transformers from the groupbuy have a channel frame on them, is it safe to remove those so they fit better into each channel?

 

[delphi]

Yes - its safe to remove the channel frame. It does not affect the integrity of the transformer at all.

 

[Hank Dussen]

Simple question:
Does Bob Schwenklers spreadsheet also work with 990 opamps?
In other words: is there a difference in gain between the 2050 and the 990? 

 

[mitsos]

It will work yes, the gain circuitry works the same, you just need to enter the right numbers for each.  For a typical 60dB gain, the API circuit has about 20dB gain from transformers (depending on the trafos you use) and about 40dB from the opamp, vs the M1 type circuit which normally has 6dB of trafo gain and 54dB of opamp gain. Of course, these numbers are basic starting points, you can tweak each to your own purposes.

 

[grantlack]

Yes - its safe to remove the channel frame. It does not affect the integrity of the transformer at all.

easier said than done, i'm afraid.  out of the 7 i've tried so far, i've managed to completely separate ONE frame from the lams... they're all stuck to the varnish    considering i have to separate 16, i need a better method.

any suggestions other than brute force?  it's getting to the point where i'm worried about damaging the laminations and/or slipping up and breaking leads.

EDIT: nevermind.  got a method down, it was just a pain in the ass.  one step closer!

 

[mitsos]

Yes, channel frames are normally varnished together with the transformer to prevent mechanical vibration noises.  Depending on the varnish type some can be more difficult to remove.  Usually I use a flathear screwdriver to carefully pry up the little tabs holding the thing together, and, if I can't pry the sides off by hand, use the same screwdriver for leverage. It's not good to dig into the lams with the screwdriver so you should be careful and go slow, but there isn't much you can do to destroy the transformer unless you slip and dig the screwdriver into the windings...

btw, I've gotten some emails asking for help, and at some point soon I will take a day and compile all the info to paste here for everyone's benefit.

Anyway, I'm happy people are building up their preamps. I love mine so far, and have even used one prototype board for makeup gain for a pultec. With a couple of jumpers and maybe a trace cut or two it can be set up just like the original solid state pultec amp. I think I posted this earlier, and it's one more piece of info I've been meaning to post details about.

Good luck with this everyone!

cheers!

 

[grantlack]

that's what i ended up doing more or less, using screwdrivers through the mounting holes gives enough leverage to at least get one side to pop off.  just had to hone the craft  8).
should have these up and running soon enough!