Cinema Engineering Passive EQ - Pot Behavior Reversing with DI in chain

[emrr]

How do you know that? He has not stated that he tried that. If the +- symbol next to the terminals indicates polarity, then it was incorrectly wired yes?


Again, I don't think this contradicts any of your observations. It is possible that the problem is that DI ring is floating as you say.

You have not read my test descriptions with comprehension.  I've tried it.  Same circuit.  I've looked at the Cinema version, the Altec version, the Langevin version, and the Hycor version.  They will all behave similarly. 

 

[PRR]

> You have not read my test descriptions with comprehension.

 

[emrr]

nohatnoswim, it's great you got these.  They are THE top-shelf classic American program EQ, and it boggles the mind that people generally don't know them.  Variants were made for 20+ years by all those different companies, all but Hycor related to Art Davis's presence.  Several of the early US SS channel strip EQ's mimic the operation through other circuit methods. 

One make-up approach I've tested, but not committed to, is using something like an Edcor 600:10K transformer on the output.  Matches a typical 10K input, provides makeup gain pretty close to what's needed. 

 

[abbey road d enfer]

I seriously doubt a bridged-T EQ can operate significantly outside its nominal impedance range. The DI connected at the output must be reflecting several 10's of kiloohms, which is clearly a mismatch.
Instead of continuing to guess, I think it is quite simple to load the EQ's output with a 600r load and see what happens.

 

[emrr]

I seriously doubt a bridged-T EQ can operate significantly outside its nominal impedance range. The DI connected at the output must be reflecting several 10's of kiloohms, which is clearly a mismatch.
Instead of continuing to guess, I think it is quite simple to load the EQ's output with a 600r load and see what happens.

Also not reading with comprehension. 

You guys don't want to bother reading the paper, or the actual test results, yet you feel obliged to speculate. 

Just amazing. 

 

[abbey road d enfer]

Also not reading with comprehension. 

You guys don't want to bother reading the paper, or the actual test results,

What paper,? Is an Altec the same as a Cinema?

yet you feel obliged to speculate.   

I'm not speculating. I'm proposing a troubleshooting method that proceeds by elimination.
What do you propose, in practice? Basically you are telling the OP that you can't reproduce his problem. So what? Does it move one inch in the direction of solving his problem?

 

[mjrippe]

What paper,? Is an Altec the same as a Cinema?
I'm not speculating. I'm proposing a troubleshooting method that proceeds by elimination.
What do you propose, in practice? Basically you are telling the OP that you can't reproduce his problem. So what? Does it move one inch in the direction of solving his problem?

No, Abbey, Doug DOES reproduce the problem AND solves it.  The "paper" he is referring to is his thoroughly detailed post. 

I know folks are high-strung right now, but yelling at someone who is actually doing the work and providing the solution is counterproductive.  Let's all take a deep breath and start fresh.

 

[emrr]

This ^

Speculation does not help the OP.  I've run through the possibilities and replicated the problem. 

The paper I'm referring to as a starting point on loading is Altec Technical Reference paper #192.  It's been online for years.  Addresses the EXACT circuit. 

On another note, I changed a Hycor example to 1dB steps for a mastering house some years ago, and in looking at Rick Chinn's drawing in the tech docs, it's pretty much that with fairly minor R value differences here and there.  The Hycor at least is made originally with Cinema Engineering wire wound resistors.

 

[abbey road d enfer]

No, Abbey, Doug DOES reproduce the problem AND solves it.  The "paper" he is referring to is his thoroughly detailed post. 

I know folks are high-strung right now, but yelling at someone who is actually doing the work and providing the solution is counterproductive.  Let's all take a deep breath and start fresh.

OK, got it. I must say the "paper" goes to length explaining why this should not happen, and a the turn of a phrase, the solution is presented, in a not so clear way. To be clear: I didn't want to read the "paper" to the end because it was hard to read.

 

[nohatnoswim]

GOT IT.  Pin 3 or ring into your DI is floating.  The signal path is incomplete.  Ground it, you get proper action and much greater control range.
The transformer coupled DI left ring open.
The JLM shorts ring to sleeve. 

Into 10K, grounding the - path boosts gain +0.5dB at max boost while changing overall gain at 1K a little less than -3dB.  Into the much higher Z of the DI’s, control range is greater by 1.5-2dB and the treble peak boosts broaden out.

Thanks - that sounds right - I only soldered hot and ground to the bantam patch for all the unbalanced inputs - so if I were to add the cold in there that should solve some issues ?

I'm still struggling to understand how I'm loosing around 26 db round trip on the insert via I/O on my DAW though when according to Art davis' article it should be around 14db loss. Any thoughts on why that is so dramatic ?


Not sure how to approach the polarity thing yet though - the above is more of a job to deal with as it involves pulling the patchbay out of the rack as opposed to a few units in the other rack.

 

[abbey road d enfer]

Thanks - that sounds right - I only soldered hot and ground to the bantam patch for all the unbalanced inputs - so if I were to add the cold in there that should solve some issues ?

Simplest fix is to tie cold and ground (R to S). However best practice is to wire all connections as if they were balanced.

I'm still struggling to understand how I'm loosing around 26 db round trip on the insert via I/O on my DAW though when according to Art davis' article it should be around 14db loss. Any thoughts on why that is so dramatic ?

You didn't mention that before. Define "round trip"; scenario A or B, or...

Not sure how to approach the polarity thing yet though - the above is more of a job to deal with as it involves pulling the patchbay out of the rack as opposed to a few units in the other rack.

Testing polarity is quite simple. You don't need to take apart the whole studio.
Create a mult of a signal (your patchbay should allow that), run this signal straight to a mixer channel, and to another channel via the device you want to check for polarity. Make sure the levels from each channel is the same (or about). When you open both channels, the resulting sound should increase, if not, the channel under test is in reverse polarity (commonly but improperly called "out-of-phase"). Particularly noticeable on bass and midrange.

 

[emrr]

Thanks - that sounds right - I only soldered hot and ground to the bantam patch for all the unbalanced inputs - so if I were to add the cold in there that should solve some issues ?

I'm still struggling to understand how I'm loosing around 26 db round trip on the insert via I/O on my DAW though when according to Art davis' article it should be around 14db loss. Any thoughts on why that is so dramatic ?

What he said ^, just be sure cold and ground are tied for unbalanced use.  If the DI doesn't do that, adding to the patchbay won't help. 

You'll notice all of the connection and loading conditions result in varying amounts of loss.  The load resistor is one additional loss. 

 

[nohatnoswim]

You didn't mention that before. Define "round trip"; scenario A or B, or...

Touched on it briefly in reply 14.

Scenario A - Have the EQ setup as a hardware insert in the DAW  through the SSL - insert it in the chain and i'm loosing around 26db on the  journey out and back into the DAW.

It is indeed possible that the DI inputs are not balanced wired at the input themselves inside the box.  I assumed the unbalanced in wasn't balanced wired so that's why I didn't bother to connect the cold onto the patchbay, but that was an assumption on my part - I need to check the other end there.

 

[nohatnoswim]

nohatnoswim, it's great you got these.  They are THE top-shelf classic American program EQ, and it boggles the mind that people generally don't know them. 

Thanks - I'm aware of what they are as they've been part of my life for around 15 years.  It's a sad story tbh - my best mate and I worked together on music for years and we were both individually collecting gear, him substantially more than me. We eventually ended up in different countries , got married etc.  Unfortunately he passed away real young in late 2018 at 44 years old, and I bought as much of his gear as I could afford to from his widow to keep the dream alive.  There were 4 CE filters aswell in the collection but I just couldn't afford them with all the other gear I had bought - pained me to split them up but had to let them go as I valued the EQ's more than the filters when pushed to it.  We used to joke a lot that he had probably the largest Cinema Engineering collection on the planet - and it might have actually been true - he certainly aimed for it !  I'm far from as technical as he was -  he maintained all the gear himself, built tube eq's and built the DI racks I have here now- one for me a few years ago, and then I bought his one eventually - both of which are subject of this wretched chain !  I'm just trying to get the whole thing up and running as I had it all shipped to the other side of the planet to me here and I'm trying to get past the techie madness so I can focus on music .  It's been a ridiculous amount of gear to integrate in a short space of time frankly, but I'm getting pretty close to having it all functional.


Appreciate all the advice and help here folks - stay well - mad times for us all right now.

 

[abbey road d enfer]

Scenario A - Have the EQ setup as a hardware insert in the DAW  through the SSL - insert it in the chain and i'm loosing around 26db on the  journey out and back into the DAW.

Your previous description of scenario A does not include the SSL. What is exactly the signal flow?

 

[nohatnoswim]

Your previous description of scenario A does not include the SSL. What is exactly the signal flow?

First post image 4 shows Scenario A - top of the picture is the SSL . 

 

[abbey road d enfer]

First post image 4 shows Scenario A - top of the picture is the SSL .

Aah! OK, SSL interface. So you're comparing a signal that comes from one interface output, goes to a mult, then goes back directly to one interface input, compare it with the same mult going to the input of the EQ, then the EQ output goes back to an other interface input, right? And it's in the DAW you see a difference of 28dB? Is it constant with frequency. I mean have you checked at 100, 500, 2k, 4k,8k?

 

[nohatnoswim]

SSL Channel Out => Patchbay => Into the EQ's => Out of EQ's => Patchbay => SSL Signal In , 

Cable all wired balanced in and out of the patchbay through the whole route.

Though it's a pair I'm using at the same time and not just one channel.  Both channels behave exactly the same.

It's a flat gain drop across the board and not frequency specific .

 

[abbey road d enfer]

SSL Channel Out => Patchbay => Into the EQ's => Out of EQ's => Patchbay => SSL Signal In , 

Cable all wired balanced in and out of the patchbay through the whole route.

Though it's a pair I'm using at the same time and not just one channel.  Both channels behave exactly the same.

It's a flat gain drop across the board and not frequency specific .

Can you make a measurement using a shorter signal path?
Source (can be an output of the SSSL)=>calibrated mic preamp with built-in meter-> establish reference level
then:
Source=>EQ in, EQout=> preamp -> evaluate gain difference using mic pre attenuator
This is assuming you don't have an audio VOM.

 

[nohatnoswim]

OK - Finally managed to get some time to look at this again.
I've flipped the hot & cold on the inputs, and removed the resistor entirely.

It's now about 9db signal loss on the round trip, and boosting either the hi or low pots fully brings that back up to zero.  No need for a DI at all ! ( unless cutting of course )

Will pull the other 3 units out and fix them aswell