Passive summing mixer outputs

[hoyahelper]

Hey

Apologies if this has been discussed before, I couldn't seem to find anything on it though!

Im sure theres a straightforward answer to this but am fairly new so am not sure how it would work... anyway!

I am building a passive mixer, no make up gain (will be doing on an external set of preamps).

Design is slightly based on Ians DB25 info (thanks Ian!)

I am sending my summed signals into a grayhill switch to send to a few different pairs of transformers for different flavours. Then out to a set of xlrs

my question is, could I have multiple outs coming from the transformers (for maybe sending to two separate pre-amps for A/B etc) without affecting the signal?

Any advice would be greatly appreciated thanks!

 

[ruffrecords]

I am sure you realise the transformers are not necessary in a passive mixer and you have just included them for some added colour. Assuming also that you use 600:600 type transformers then you probably could feed each one to a couple of outputs. How little this would affect the signal depends a lot on what you are feeding them into. Usually you would feed the output of a passive summer into a mic pre and these have typical input impedances of about 1500 ohms. Feeding two of these from one output adds a pair of 1500 ohm loads on the transformer output; in other words a load of about 750 ohms.

Although I have assumed you are using 600:600 transformers, the output impedance of the summer will typically be in the region of 200 ohms and this will be reflected through the transformer. Bottom line is you have a 200 ohm source driving a 750 ohm load. Not ideal but not a disaster either. You will lose a couple of dB level when driving two outputs but other than that it should not affect the sound.

And if you use a sexy FET input mic pre with a 10K input impedance or more you won't have any trouble at all.

Cheers

ian

 

[hoyahelper]

I am sure you realise the transformers are not necessary in a passive mixer and you have just included them for some added colour. Assuming also that you use 600:600 type transformers then you probably could feed each one to a couple of outputs. How little this would affect the signal depends a lot on what you are feeding them into. Usually you would feed the output of a passive summer into a mic pre and these have typical input impedances of about 1500 ohms. Feeding two of these from one output adds a pair of 1500 ohm loads on the transformer output; in other words a load of about 750 ohms.

Although I have assumed you are using 600:600 transformers, the output impedance of the summer will typically be in the region of 200 ohms and this will be reflected through the transformer. Bottom line is you have a 200 ohm source driving a 750 ohm load. Not ideal but not a disaster either. You will lose a couple of dB level when driving two outputs but other than that it should not affect the sound.

And if you use a sexy FET input mic pre with a 10K input impedance or more you won't have any trouble at all.

Cheers

ian

Thanks Ian

I am gonna be using 4.7k resistors on stereo channels and was thinking of maybe adjusting the shunt resistor to accommodate for the transformer load?

The transformers are very much for colour yes but also able to bypass!

Was thinking of using either line input from a preamp - haven’t gotten to pre-amp stage yet and explored the options!

 

[abbey road d enfer]

I am sending my summed signals into a grayhill switch to send to a few different pairs of transformers for different flavours. Then out to a set of xlrs

my question is, could I have multiple outs coming from the transformers (for maybe sending to two separate pre-amps for A/B etc) without affecting the signal?

I'm not convinced the xfmrs will produce their "mojo" at such a low level there is at the summing point.

 

[hoyahelper]

I'm not convinced the xfmrs will produce their "mojo" at such a low level there is at the summing point.

Do you think less resistance on each of the input lines would help on that front?

 

[abbey road d enfer]

Do you think less resistance on each of the input lines would help on that front?

No. transformer saturation needs two things: a rather high source impedance and a rather high magnetization level. Decreasing the resistors would increase the latter but reduce the former.

 

[hoyahelper]

No. transformer saturation needs two things: a rather high source impedance and a rather high magnetization level. Decreasing the resistors would increase the latter but reduce the former.

ahah i see...

I guess this is why the make up gain circuit comes into play usually before the transformers right?

 

[hoyahelper]

No. transformer saturation needs two things: a rather high source impedance and a rather high magnetization level. Decreasing the resistors would increase the latter but reduce the former.

What do you think about changing the value of shunt resistors?

 

[user 37518]

No. transformer saturation needs two things: a rather high source impedance and a rather high magnetization level. Decreasing the resistors would increase the latter but reduce the former.

Yes and also important is the quality of the transformer, if you want "color" (AKA mud, grunge, distortion) the cheaper the better, a Jensen or Cinemag is probably not the go-to brands in this case, try something like the el-cheapo Triads like this one https://www.mouser.com/datasheet/2/410/TY_250P-1892894.pdf or you can try the Chinese ones on e-bay, 10 for $3 USD last time I checked. However, if you are going to use a Mic Pre to amplify the output, just use a pre with transformers?

 

[ruffrecords]

ahah i see...

I guess this is why the make up gain circuit comes into play usually before the transformers right?

The factors that give a transformer colour are core material, core size, signal level and its frequency. The type of transformer you would typically connect to the bus of a passive mixer would have a low distortion core and the signal level going through it would be rather small both of which mean very little colour is added.

One the output side though you have a much higher signal level, you typically use a plain steel core which produces more colour throughout the audio spectrum and if you make it smaller than would normally be necessary, you can achieve some nice colour below 100Hz.

Cheers

Ian

 

[Winston OBoogie]

The Carnhill modern equiv. of the old St. Ives 10468 might work to give a little mojo at this signal level,  it'll  do that when used as intended in Neve channel amps so why not here? 

In this case, I'd start off by wiring it as 1K2:1K2 rather than 1K2:4k8  -  the secondary in parallel rather than series. 
If your two external pre amps you're running it into are transformerless then the resistive input impedance of a couple of them strapped across the output will possibly be a decent termination for the 10468 transformer.

You might try putting a nominal high value resistor across the 10468 as a partial termination too?
Add a simple low pass filter cap across the secondary just as Neve did rather than fiddle with finding a zobel. In other words, let that sucker ring! 
Start with a value between 680pF and 1000pF, or not.

Just a thought. 

 

[My3gger]

You could also try OEP transformers, they are good for linear response but small enough to have a bit of sound when not driven hard. Cheap A262 series would be between suggested Triad and Carnhill.

 

[ruffrecords]

You could also try OEP transformers, they are good for linear response but small enough to have a bit of sound when not driven hard. Cheap A262 series would be between suggested Triad and Carnhill.

I agree, low cost OEPs or Edcors would be my first choices for colour.

Cheers

Ian

 

[Winston OBoogie]

Yep, I'll give a thumbs up that those would work too.

My main point really is that a smaller core mic input with 1:1 ratio would be better for colour in this application than your usual 600:600 type in my opinion.  But who knows?  Build something and adjust to taste I suppose.

 

[user 37518]

The Neutrik trannies may also work https://www.neutrik.com/en/neutrik/products/accessories/transformers

 

[hoyahelper]

Thanks all, very useful info!

I have a couple of 250 ohm 1:1 old Haufe Transformers, These Prahe 400 ohm 1:1 transformers (cant find any info about them if anyone knows anything let me know!) and a pair of Lundahl 1:1 600 ohms

I was thinking of 10 channels - 4k7 resistors so input resistance would be 9.8k - to get a 500 ohm load I would use approximately a 1k shunt resistor

Im very much looking to drive these to get the said "mojo" (not the lundahls though) and my preamps have between 6k-10k input impedance so they can handle the higher output impedance right?

 

[ruffrecords]

Thanks all, very useful info!

I have a couple of 250 ohm 1:1 old Haufe Transformers, These Prahe 400 ohm 1:1 transformers (cant find any info about them if anyone knows anything let me know!) and a pair of Lundahl 1:1 600 ohms

I was thinking of 10 channels - 4k7 resistors so input resistance would be 9.8k - to get a 500 ohm load I would use approximately a 1k shunt resistor

You don't necessarily want the mix to look like 500 ohms to feed 600:600 transformer because you will likely lose bass response. A lot depends on the primary inductance of the transformer at low frequencies. This is what determines the bass response. Unfortunately the inductance itself tends to be frequency dependent and the way it varies depends on the core material, but as a rule you tend to drive the  transformer from a lower impedance to get better bass response. Unfortunately, driving from a low source impedance tends to result in lower distortion == colour.

Cheers

Ian

 

[vari-mu]

I made an SE tube summing amplifier 10 years ago according to my circuit,
this still works today.
  The primary inductance was 25 Hy or about 3k at 20 Hz.
Perhaps 25 Hy is a lot, but I would not risk doing less than 10 Hy, low frequencies may suffer.
Summing resistors are used 1kΩ.
  Schematic :