Summing issue. (solved)

[rhythminmind]

EDIT: Solved! - not an impedance issue. Read on towards the end.


Is it possible to lower a line outputs impedance, other then the use of transformers?
I have a mixing network/makeup gain (fulcrum+JCF LEVR) that doesn't seem happy when fed anything over 50ohm.  I get very high THD otherwise.

 

[abbey road d enfer]

Weird. Please elaborate escription of problem.

 

[Harpo]

You already exchanged the balanced connection wire between fulcrum and your zero-impedance active summing device that relies on the series and shunt resistor values inside the fulcrum ? A shorted XLR-pin1/2 or pin1/3 might cause this behaviour. An above 50 ohm output impedance of the sending device into the fulcrum will only decrease level by a probably hardly to notice degree. Just an idea.

 

[rhythminmind]

Yes I've swapped out the cables between the fulcrum & LEVRs. Didn't cause any change. Would cable length matter? These are some details I posted elsewhere.

I'm getting really high distortion (THD) & a rather curved response with the LEVR. The distortion amount changes a lot depending on what feeds the fulcrum. I can't figure this one out. Your guys reviews & the description make it seem that these should be very clean.
FYI - Everything works really well if I use a mic pre (312 for example). Any ideas?

THD via fulcrum > LEVR = 3.371%

THD via 1968 > fulcrum > LEVR = 0.4%

THD via fulcrum > API 312 = 0.0060%

-----

All these devices are producing the same level test signal, But with drastically different results?

I've tried to be very thorough. Everything going thru the same wiring method. (No change in results wiring direct)
All test measurements are with a 1k tone @ +10 dBu.
Here are my findings.
NVision DA4040>fulcrum>LEVR>A/D = Low THD???? This is great but why? (1k= -14 & 1st harmonic = -92)
TC SK48>fulcrum>LEVR>A/D = High THD (1k= -14 & 1st harmonic = -45)
Lynx L22>fulcrum>LEVR>A/D = High THD (1k= -14 & 1st harmonic = -75)
Aphex 141>fulcrum>LEVR>A/D = High THD (1k= -14 & 1st harmonic = -46)
TC m3000>fulcrum>LEVR>A/D = High THD (1k= -14 & 1st harmonic = -62)


Ok for some reason it's happy with the NVision and not the other D/A's. Next I added a device (without processing) after the D/A.

NVision DA4040>DBX165a>fulcrum>LEVR>A/D = High THD (1k= -14 & 1st harmonic = -46)
NVision DA4040>Overstayer>fulcrum>LEVR>A/D = High THD (1k= -14 & 1st harmonic = -46)
NVision DA4040>1968me(in bypass)>fulcrum>LEVR>A/D = Low THD (1k= -14 & 1st harmonic = -82)
TC SK48>1968me(in bypass)>fulcrum>LEVR>A/D = Low THD??? (1k= -14 & 1st harmonic = -81)
TC SK48>Eventide H3500 100% dry>fulcrum>LEVR>A/D = Low THD??? (1k= -14 & 1st harmonic = -81)
TC SK48>TC m3000(analog i/o)>fulcrum>LEVR>A/D = High THD (1k= -14 & 1st harmonic = -62)
NVision DA4040>TC m3000(analog i/o)>fulcrum>LEVR>A/D = High THD (1k= -14 & 1st harmonic = -62)

The effects are even more exaggerated the hotter the signal. Any ideas why the LEVR is picky about source device? As I said earlier everything functions as expected with a Mic pre.

---

After looking over the results seems like if the Fulcrum via LEVR is fed anything with an output impedance over 50ohm it will have very high THD.
Hmmm. Not cool in my situation with majority of my gear being over 50ohm.
..


---

 

[rhythminmind]

A shorted XLR-pin1/2 or pin1/3 might cause this behaviour.

Being that the LEVR turns the fulcrum in a virtual earth summing stage, could the shunt resistor be causing the problem?

 

[abbey road d enfer]

Looks to me like the inputs are not properly balanced and some equipment might not like it.

 

[rhythminmind]

Looks to me like the inputs are not properly balanced and some equipment might not like it.

Don't have a schematic, & mine have transformer outputs, but It's supposedly a 4-transistor Push-Pull Active Summing Amp, that is inspired by the AMPEX 440 REPRO/Playback amplifier design.

Bad transistor matching? Both modules perform the same so unlikely.

 

[okgb]

It's not a problem with unbalancing or unbalanced outputs ?
some like to be floating whiles others need to be grounded

 

[rhythminmind]

Found something new... Whenever the LEVR is connected to the output of the Fulcrom, & a device is plugged into the input of the Fulcrom I see resistance across pins 1 & 2 or 1 & 3 of the fulcrom outputs. This does not happen if I use a mic pre. The resistance is also lower with the "high distortion" devices. This would explain the higher THD levels.
Now the question is, how is signal leaking into the ground of the LEVR input?

 

[keefaz]

Maybe compare the output of devices that work with LEVR (low thd) and the other devices ?

For example, it seems the Aphex 141 has impedance balanced output (signal only in hot)
http://www.bayviewproaudio.com/141-p-23159.html

 

[rhythminmind]

Maybe compare the output of devices that work with LEVR (low thd) and the other devices ?

For example, it seems the Aphex 141 has impedance balanced output (signal only in hot)
http://www.bayviewproaudio.com/141-p-23159.html

Interesting, would a virtual earth summing amp have problems with impedance balanced?

 

[abbey road d enfer]

Found something new... Whenever the LEVR is connected to the output of the Fulcrom, & a device is plugged into the input of the Fulcrom I see resistance across pins 1 & 2 or 1 & 3 of the fulcrom outputs. This does not happen if I use a mic pre. The resistance is also lower with the "high distortion" devices. This would explain the higher THD levels.
Now the question is, how is signal leaking into the ground of the LEVR input?

I think it's normal; the LVR should have input resistors. In addition, whatever you connect at the inputs reflects somewhat at the output.

Looking at the PCB, my understanding is that it delivers two outputs (one for each leg of the bus) that need to be connected to a balanced input for proper operation.
How do you do that?

EDIT: Litt clearly states that the output is non-floating. Are you sure you don't ground one leg?

 

[rhythminmind]

Looking at the PCB, my understanding is that it delivers two outputs (one for each leg of the bus) that need to be connected to a balanced input for proper operation.
How do you do that?

Not sure what your asking exactly. The setup requires two LEVRs for stereo operation. Each module has up to Three inputs & One output.

EDIT: Litt clearly states that the output is non-floating. Are you sure you don't ground one leg?

Audio Ground/Shield is connected to Power ground

I have transformer outputs fitted on mine. Currently the shield is attached on the LEVR input connecting it to the fulcrom/mixing network (short direct XLR cable). Without the input shield, gain/levels goes crazy. I drop LEVR output shield on input to the Patchbay/ADC.
The BAE Rack connects input & output shield to the 500 slot audio ground. It also connects chassis & power ground.
The LEVR connects Audio Ground & Power Ground.
Audio, Power, & Chassis are all tied.

FYI - I just got an email for JCF saying they will answer my questions shorty.

 

[rhythminmind]

@keefaz
Well you seemed to have guessed right.  that was a few days of unneeded wiring trouble shooting.

Brought home my D-scope & tested between Impedance vs Differential balancing. The LEVR is not happy with Impedance balanced outputs. Makes it's pretty unusable with my DAC converters & most of my outboard.
Anyone have any recommendations? Buy a crap-load of transformers?  :-\

 

[abbey road d enfer]

Anyone have any recommendations?

Return the LVR and get something that's not as fussy.

 

[rhythminmind]

Solved!
Well I got a definitive update thanks to help & minds of the Jensen company, Explains the situation.

The inputs to the Folcrom RMS 216 must not only be impedance balanced but signal symmetric (equal but opposite polarity signal swings on + and -) as well. Sources having no signal on one input pin will create rather high common-mode voltage at the output. But the common-mode voltage can be eliminated by a simple 1:1 mic isolation transformer like our model JT-MB-CA, at the output.

My mic pres have transformer inputs & in-turn cleared the problem in the past. So for the LEVR you need a transformer (a way to create signal opposite polarity signal swings on + and - on input or output)

So I now have to wonder how many people with passive mixing networks are unknowingly experiencing this with similar transformerless input stages? Scary.

 

[keefaz]

Technically the Levr is an active summing stage, not passive 

 

[rhythminmind]

FYI - Seems the problem I had originally stemmed from how the LEVR is setup with the output transformer. I removed the trannies, added the 100 ohm isolation & 100k drains to ground. It now performs as per the published distortion SPEC. Only moderately worse with non-symmetrical sources. Something with the transformer configuration really kills it's common-mode rejection, I informed JCF.
I also received two Jensen input transformers today & it brings the THD/IMD down even more. From 0.03 to 0.0012 with previously troublesome gear. The transformers do cause a high end roll-off starting at 10k, -5db or so once it hit's 20k. Well my frustrating adventures in LEVR land did help me learn a thing or two. I am finally enjoying them, & mixing/creating once again.

 

[Ack]

.....I removed the trannies, added the 100 ohm isolation & 100k drains to ground. It now performs as per the published distortion SPEC.

Hi,

I'm building a 24-channel passive  mixer with Mute/Polarity/L-C-R Pan to use with the JCF LEVRs. Could you share some hand-writing or info on these resistors? Are the isolation 100Rs placed between JCF outputs and pin 2/3? And the drain resistors 100K connect the output pin 2/3 to Ground? Any special notes on Grounding scheme before the JCF LEVRs?

Thanks a lot!