[BUILD] TB550A

[emrr]

I'd catalog a bunch of different potential interface conditions, and see if you can spot a pattern, or any particular change. 

 

[rickc]

Cataloging the different interfaces is a great idea, but we'll have to come up with a methodology because this stuff is usually not found in manuals today.

Even the manufacturers are loathe or unable to provide it, partly under the guise of 'secrecy' or because the tech-support person is too lame to recognize how important this tidbit of information is.

I discovered this the hard way when trying to interface a Control 24 in a friend's studio. We called Digi and asked, and the person responding didn't know, and would not make the effort to find out. They thought it was information that we did not have "need to know."

Here's what I think we need to know:

output topology:
transformer?
transformer with grounded center tap? (not likely today, but possible with vintage gear)
transformerless?
If transformerless, then which topology: chain of inverters or cross coupled, or impedance balanced?

There are now several different cross coupled topologies, and I believe at least one of them is tolerant of being treated like a chain of inverters output. Others require grounding the low-side at the source, and some are tolerant of grounding at the far-end of the interface (like a transformer would).

Knowing all of that is esssential to making an intelligent decision (as opposed to a SWAG) about how to interface with a given piece of gear.

Once you go thru all of this, it's easy to see why balanced inputs are preferred.

 

[emrr]

I was thinking closer to home, checking every likely connection at your place, see if there's any that do behave correctly, note the tech differences. 

 

[horvitz]

OK.. So Rick, you make a lot of good points on what potential items to look at.  I'd like to focus on two interesting items.  First with Doug, when he went from the OSA rack to the 51x rack, his -4 when in went away.  Second is on Robert's where he had his MOTU on both sides of the EQ and sees the same.

My testing environment here consists of Apogee converters and either a 51x rack or my bench setup which is the same thing just without the card cage.  This is very similar to Robert's setup.

I've been sitting here trying every possible combination I can of reversed pins, disconnected pins, grounded/not, etc.  The only somewhat interesting result I've had is that on the input side, if I float the low, the EQ magically makes -4.5db.  Inside the 550, this is grounded via pin 5.  If the low were not connected at all on the card edge, bypass wouldn't work.  This tells me that the only way to float the low and see the results we have would be with a grounding issue.

Another note is that if in your rack, pin 5 were not connected to ground, there is a backup path to pin 13 via the 10R on the bottom rear of the board.  If both were not connected, the unit wouldn't power on.  So here's my wacky proposal - pin 5 doesn't go anywhere AND there's some problem with the 10R, then when EQ is in low would float and when in bypass, low in would connect to low out as normal.

That's a REAL stretch but, it does seem to fit the symptoms anyway and causes the same result.  Robert or Doug, could I ask you guys to do a very quick test for me?  With the unit out of the rack, check DCR from 5 to 13.  Should read 10R.  If that's an open, then see if pin 5 in your rack goes anywhere.

So then if all these crazy IFs line up, then we roll back to Rick's question about some commonality of what might be interfacing the device's input and maybe actually draw a conclusion.  Of course if it's a grounding problem like that, the conclusion may be that the design of the thing is just inherently not tolerant to these sorts of issues.  That's a nice way of saying I could have done it better

 

[rickc]

That's well said.

I think if I were to do it, I'd have put in one of the THAT ingenius input chips and made a way to hard wire bypass it if someone were really stuck on having a unit that is identical to the original unit. Then those of us who prefer balanced inputs could have their cake, and those who want stock at whatever the cost can have theirs too.

Now having made that statement, I confess to not having looked to see if there's even room for another dip8 chip. When you're outa room, you're outa beer too.


--rick

 

[rob61]

I ran the test, but not sure this is what you wanted. I took the EQ out of the rack, and measured resistance between pin 5 and pin 13 on the EQ module. (tried to figure out if you were talking about measuring it on the module or of the rack itself, I went with the EQ module). By the way, I measured difference using a meter that shows numeric readout, and I get -4.5db reduction when switching EQs inline, exactly what you got. The MOTU is electronically balanced, and says it supports balanced and unbalanced operation, but my cables are all wired for balanced.

On both EQs, I get 10R between pins 5 and 13 no matter whether the hard wire bypass switch is up or down.

Does that tell you want you wanted? Or is there something else you need me to test here?

By the way, I've run cables from the MOTU 24I/O to the rack inputs/outputs. Before I had it in my patchbay but wanted to eliminate that variable so now its going straight in/out of the interface and not through the patchbay.

 

[horvitz]

On both EQs, I get 10R between pins 5 and 13 no matter whether the hard wire bypass switch is up or down.

Does that tell you want you wanted? Or is there something else you need me to test here?

That's exactly what I was looking for.  Thanks.  Unfortunately, it kind of blows my theory but it's good information.  We can eliminate that possibility.  I have one other thought that I'll check out today and will post my results if it goes anywhere.

 

[emrr]

DCR from 5 to 13 is 10R as expected.

Pin 5 in OSA rack goes to both XLR pin 1's, pin 1/chassis is not connected, but the MXLR shell connection is to chassis.  Weirdest thing is the output jumpers are swapped in terms of which jumps where, but no smoking gun I see there either.  I've detailed the OSA in another post for future generations. 
http://www.groupdiy.com/index.php?topic=48324.0

MOTU driving THAT coupled device driving 550 mounted in OSA, behaves as it should, MOTU driving 550 mounted in OSA, gain loss when engaged. 

MOTU maybe our smoking gun.  All works properly when driving in 51x.

I put in 620R loads and achieved gain match in my system.  Anyone doing this need be aware it must be at the transformer secondary pads, and not at the module output pins, as they reflect the state of the relay bypass. 

 

[horvitz]

Weirder.. but at least that MOTU seems to be common.  That makes me feel a little better.  Is it the 24I/O for you as well, Doug?

 

[emrr]

Yes. 

I'm still not seeing the critical difference between the two 500 racks, and my 51x works fine while Robert's doesn't. 

 

[emrr]

In my chain, I'd tried feeding my MOTU 24i/o  into the eq through the patchbay, then back into the 24i/o.

I've also tried going interface to Serpent 4001 compressor, to the 550a and back into my interface. Both exhibited this -4db drop.

The first situation is mostly like mine, does what mine did in the OSA rack, but not in the 51X rack.  Differs in that my output is feeding a balanced console input.

The second situation is like my THAT coupled experiment, up to the end point (console versus 24I/O).  The THAT device (SB4001 in your case) fixed my OSA issue, doesn't fix yours. 

 

[horvitz]

We've got some small differences here but I feel like we're real close to something.  Just like most problems, it's going to be something real trivial.  I'm trying to get my hands on a MOTU to borrow locally so I can test further.

  Brian

 

[rob61]

I have another MOTU box, the 828 MkII, which was also modded by Black Lion Audio. I thought I'd run the same test using that converter.

I get the same -4.5db drop in level when the EQ is switched in. I also tried flipping the pin1 to gnd/chassis in the 51X with no difference in the results.

 

[emrr]

These things are cool.  Actually used them on a mix bus, compared with several other EQ options, couldn't get anything like it with my other tools.  They have a definite sound at a lowly +2. 

 

[horvitz]

Neat!  Glad you're enjoying it  ;D

 

[Biasrocks]

Agreed, absolutely fantastic sound. I've pressed them into service over the past few days and I think they're my new favorite attitude EQ. Great on just about everything I've tried them on. Especially great on drums and guitars. I haven't tried them on the mix bus yet, a pair of Pultec's living there currently.

I may have to order another pair of these guys.

Thanks again Brian for a great project and the behind the scenes help and encouragement!

Regards,
Mark

 

[Biasrocks]

I think 2dB steps are an asset in a tracking EQ too.

Agreed, there are a few times where I wished for a 1db step, but in the end I made it work.

Regards,
Mark

 

[emrr]

You know better/worse right away with 2 dB, and can move along. 

 

[GregNey]

So, I recently finished building both of my units and I'm having problems with both of them. I went through tonight and checked for obvious causes (placement, solder bridges, cold joints, etc...) with no luck. Here is a sound clip that includes the following and demonstrates what's happening with both units:
-Unit 1 bypassed
-Unit 1 engaged (VERY LOUD)
-Unit 2 bypassed
-Unit 2 engaged (very quiet)

http://www.gregney.com/TB550Aproblems.wav

The chain is: Uher Mic > VP28 > TB550A > Echo Audiofire12 > Reaper. Both 500 units are in an EL500. Both slots are tested and working. All 4 GAR2520's (in both units) are tested and working.

I'm pretty good at putting things together but when things go wrong, I don't really know where to start... Maybe these behaviors are indicative of something common you guys have seen before? Any general troubleshooting tips are appreciated.

Thanks, Greg

 

[horvitz]

Hi Greg,

Sorry to hear things aren't working so well.  No worries, though.  We'll definitely get it all sorted out.  Unfortunately, troubleshooting these things is usually a lot of work to find one small problem, but we'll get to it.

Some questions first.  When you test, are the boost/cut knobs all at 0 and the Fltr switch down?  That's always the first place to start since it eliminates a large chunk of the circuit.

If that's how you're testing, and as you say the 2520s are all tested and known good, the next thing to do will be to inject a test tone and start tracing out the circuit to see what's what.  Do you have the ability to get this thing powered up outside the rack and do you own a scope or signal probe of some sort?

You will want to take it apart for this testing, too.  If the boost/cut are all set to 0, you can actually remove the filter board entirely and it will still pass a signal.  That will make the rest of it much easier to access.

  Brian