Someone did mention keeping the impedance at 600ohms for that point.. can't remember who. We're already 6 pages deep lol. But do you mean on the secondary as well? Or just going into the transformer? (I hope I used secondary correctly).
I'm no transformer expert, but I think it is the magnetizing current that matters, the part of the current that flows into the primary but does not come out transformed on the other side. That current depends on the magnetizing inductance of the transformer and the time integral of the voltage that is applied to it.
Yes the transformer load is on the secondary, the output. Its that load that the source sees and its that load and the voltage that determines the current. Another way to saturate the core is some DC current. This might be easiest.
FYI someone did some testing on 111C distortion. Here is the link
https://www.google.com/url?sa=t&sou...kQFnoECAgQAQ&usg=AOvVaw2sMOmdiqP_f4eVuXJLWA0D
Just search 111C bc they tested multiple units.
https://www.google.com/url?sa=t&sou...kQFnoECAgQAQ&usg=AOvVaw2sMOmdiqP_f4eVuXJLWA0D
Just search 111C bc they tested multiple units.
Just proves the point again: wrong xfmr for "color".
Clearly. I just hate to see someone do everything right except for one thing and miss achieving their goal entirely. But you go ahead and do whatever you want.
I tested a 111C today myself. Hit it with stepped levels from +26dBu to +36dBu and did a distortion sweep. It's very difficult to get that transformer into saturation. Really, nothing happens at all, even at +36, except for very low frequencies, below 30Hz. It's basically colorless.
The document shows 0.1 % distortion at 30 Hz at 0 dBu, 0.05 % at 100 Hz at +17 dBu, nothing about levels above 17 dBu or about high signal levels combined with frequencies below 100 Hz. It does confirm post #122.
Last edited:
Well I can hear the cliche rounding of top end and some midrange that you swear is absent, even at typical mix bus levels, so I'll be interested to hear what it does with a little more gain. I'm also listening in a treated room with very high grade studio monitors and some room correction for the last 5db of correction, so that might be something as well. Granted that means the average listener with average gear won't hear it, or care, but I'm doing this for me. I love the experimentation.
I think he's saying he did his own testing. He also has these units. The document doesn't go above like +26bc the test gear couldn't go higher.
One thing I want to mention is there is a difference between slamming a transformer with an electric guitar to make it sound crunchy and running a completed mix through it to tame some of the top end and midrange peaks. Engineers that use this on the mix bus the way I am are not looking for thick saturation, they're looking for subtle effects that make the final track even more easy to listen to, and I believe these transformers are perfect for that. My whole thing is I want to know what the max effect is and the min affect so I know how much to use and when.
Are you sure it isn't just frequency response abberations you want, which you will also get at lower levels? The distortion is essentially nothing except at very low frequencies.
On a side note, I was reading up on power amp input sensitivity and I'll be curious to see how much level I can send to the input with it distorting. There was no max input spec so will have to test. It's rated to provide max output at +4 but my tracks can easily hit +24 or +25 out of the converters, even with the max output of the interface being +20.
The song I've been testing measures at +5dbfs peak before it goes out of the interface into the compressor/transformer, which would be +25dbu. I guess the converter distortion wasn't audible enough for me to care. I may have to reduce the signal going out to the amp by quite a bit if I can hear distortion, then turn up the sensitivity on the amp.
The song I've been testing measures at +5dbfs peak before it goes out of the interface into the compressor/transformer, which would be +25dbu. I guess the converter distortion wasn't audible enough for me to care. I may have to reduce the signal going out to the amp by quite a bit if I can hear distortion, then turn up the sensitivity on the amp.
I've never been after distortion. Actually if I hear distortion, which obviously happens at low frequencies first, I back off the gain. I'm looking for anything that makes the brittle top end of digital music, less harsh. That can include some of the upper mids as well. While this transformer might not "color" the sound, I do believe it eats up some of the harsh peaks of the top end in a way that digital compressors and clippers just can't do. Granted the peak attenuation I'm talking about would be like 1db or less, that's all you're really doing at that stage of a mix. Generally you shouldnt want or need a lot of gain reduction at that stage because you would have already taken care of the big problems on the individual tracks themselves. The mix bus is for minor alterations.
This time I don't understand you. Eating up peaks sounds like non-linear distortion to me, so that's inconsistent with the first sentence. Besides, you need to have enough volt-seconds to get the transformer to behave non-linearly, which you won't have at high audio frequencies. Maybe I just misinterpret what you wrote.
If you're "reading up", try reading the manual. The amp has input level controls. With them full clockwise +4dBu=full output power. When turned CCW they attenuate the input signal up to 70dB.
You're confusing figures here. Anything stated in dB is relative to a reference, the level of 0dB. If your interface maxes at +20dBu and your tracks peak at +25, the question you need to ask is +25 relative to what? If 0dBu, then you're clipping your interface quite badly. If 0dBFS, same thing, you're clipping your converters horribly.
Um...is this a professional studio situation? Because peaks at +5dBFS is simply wrong. Not just bad practice, complete amateur hour. You'll clip ever converter in the world, on every play device.
This is now comic. Someone who can't near distortion at +5dBFS (that's around 12% THD @1KHz) claims to hear transformer color of a 111C not even driven to saturation (less than .1% @1KHz). 12% is intollerable. Anyone can hear it. .1% is inaudible.
Houston, we have a problem.
Read the amp manual then turn down the amp input controls.
If you want to get rid of the brittle top end reduce your +5dBFS peak levels to something below +0dBFS. -1dBFS would be a good start. It'll be 100% effective in a way no processing or transformer can ever do.
I'd consider posting the graphs but this thread has seriously gone off the rails. The OP is on his own with unrealizable goals and, apparently, without knowledge of how to properly set record levels. I just don't think any further information will help.
THD at 20Hz exceeded 1% at +36dBu, low Z source and 600 ohms termination. At all other higher frequencies it was well below 1%. Raising the source Z isn't the answer, lowering the termination might be, but again, there's no point in discussing this further. What we have is the very definition of fanaticism: redoubling efforts once the goal has been lost.
Pity you didn't measure with a 600 ohm source, the distortion is always lower at low source impedance.
This doesn't make sense to me.
The whole project was a bit weird from the start, but that's not unusual at diyaudio.
I appreciate the help you've provided. Your expertise is in electrical engineering and mine is in audio engineering. Thanks again.