Line Amp Capable of +35dbu

Hmmmm.

I'm still trying to wrap my head around using transformers myself, but I don't think 600R is ideal if the secondary is loaded with 600R. That would cause the low end response to be down -3dB I think. The source output impedance should be at least 10x lower(?). I'll let the pros weigh in on that one though.
From a practical standpoint I can tell you that when I currently run my audio out of the interface at 600ohm and into the transformer at 600ohm, then out of the transformer again at 600ohm and back into the interface which has a resistance of 10k, I do not observe any low end attenuation. I do notice an overall signal attenuation of about half a db, but that's actually what the 111c is rated for.
 
When you say "out of the transformer again at 600ohm", are you actually loading the secondary with a resistor? A transformer doesn't have an impedance of it's own. If the 10k of the interface is the only load on the secondary, then that's what source on the input sees.
 
When you say "out of the transformer again at 600ohm", are you actually loading the secondary with a resistor? A transformer doesn't have an impedance of it's own. If the 10k of the interface is the only load on the secondary, then that's what source on the input sees.
Yeah I guess I'm a bit out of my depth here. All I know is it's wired in the 1:1 600 ohms configuration, like in this picture, and that input impedance of my balanced interface is 10kohms. I also included the input specs of my interface.
 

Attachments

  • Screenshot_20220505-115008.png
    Screenshot_20220505-115008.png
    129 KB · Views: 62
  • Screenshot_20220505-115805~3.png
    Screenshot_20220505-115805~3.png
    50.3 KB · Views: 52
Last edited:
Yeah I guess I'm a bit out of my depth here. All I know is it's wired in the 1:1 600 ohms configuration, like in this picture
A 1:1 transformer just reflects impedance from input to output and vice versa, unchanged. There is no 600 ohms in the equation here unless that happens to be the output impedance of the compressor.

Just think of the load the source sees, and source impedance the load sees, as being the same as if the transformer wasn't even there.
 
A 1:1 transformer just reflects impedance from input to output and vice versa, unchanged. There is no 600 ohms in the equation here unless that happens to be the output impedance of the compressor.

Just think of the load the source sees, and source impedance the load sees, as being the same as if the transformer wasn't even there.
Exactly. Just to add...when a transformer is specified as having a primary or secondary at a particular impedance, that means it's optimized to work at that impedance, not that it IS that impedance.
 
A 1:1 transformer just reflects impedance from input to output and vice versa, unchanged. There is no 600 ohms in the equation here unless that happens to be the output impedance of the compressor.

Just think of the load the source sees, and source impedance the load sees, as being the same as if the transformer wasn't even there.
The balanced signal comes out of my interface at 600ohms, and I believe it remains that way out of the compressor.. I don't have specs on it because it was a clone built by a guy. All I know is that the output transformer of the SSL is a 1:1 JT-11-DMCF.

Also, I may not use the compressor once I have an amp that can drive the 111C, so it would just be coming out of the interface at 600ohms. But, I may end up mixing and matching. I have a patchbay so I can change the arrangement of these units on the fly.

One last thing is that I have -30db pads after the 111C before it goes back into the interface. The pads actually bring it down to 150ohms. I'm not opposed to changing those though. I only got them bc they were fast and cheap.
 
A 1:1 transformer just reflects impedance from input to output and vice versa, unchanged. There is no 600 ohms in the equation here unless that happens to be the output impedance of the compressor.

Just think of the load the source sees, and source impedance the load sees, as being the same as if the transformer wasn't even there.
So let's assume the output of the interface is 600ohms and goes into the transformer and then goes back into the interface that has 10k ohm input. Does that still mean the secondary of the transformer is 600ohm? Or does that become 10k because that's the impedance of the input it's going into? I was looking at ohms as a directional thing.. so because the input of the transformer was seeing 600, and it's wired for 1:1, that the output also is 600, and then gets converted to 10k after if goes into the interface, which then is converted to a digital signal...
 
One last thing is that I have -30db pads after the 111C before it goes back into the interface. The pads actually bring it down to 150ohms. I'm not opposed to changing those though. I only got them bc they were fast and cheap.
Right, I forgot about the pads. From what I can find online (I'm not familiar with this at all), it looks like the pad in parallel with the 10k of the interface would present a 4.55k load to the transformer.

EDIT: Scratch that; I don't know if a 150Ω mic pad is symmetrical in/out (H-pad).
 
Last edited:
It will be closer to 1 kohm than you think. The pad has to consist of at least three or four resistors and the 10 kohm is only in parallel with one or two of them.
You're right. I found the specs for the specific pads and it says they're H-pad type. Using an online calculator for 1k input and 150R output @ -30dB, and then simulating it with 10k in parallel with the output, I get 1.01k input impedance.
 
  • Like
Reactions: Obe1cannoli
You're right. I found the specs for the specific pads and it says they're H-pad type. Using an online calculator for 1k input and 150R output @ -30dB, and then simulating it with 10k in parallel with the output, I get 1.01k input impedance.
FYI I have a multimeter so if you want me to take any kind of measurements I can. I can use music or my music production software can generate any kind of tone/wave.
 
New thoughts, may have been mentioned. Transformer distortion results from core saturation which is caused by current not voltage. This means you need a large load (at least 600 ohms) to get lots of current and a source that can provide it. Once there you can probably adjust the amount of distortion by varying the load.