I was refering to my writing, not yours. Apologize if you understood i was talking about you.
I respect your point of view and understand it.
You're right, kind of. Some Portico, early Focusrite or old Neve products (not Ams-Neve) are effectively very expensive. Other works he made during 90's are much more affordable now.
I'm thinking about Amek 9098 derived series. Output stage have a nice transformer to drive the outside world!
I must confess those brands are 'standard' for me as i used to work in commercials studios for some years.
For me very expensive is much more mastering grade audio gears (and some Rupert Neve products enter this categorie but not all gear he made).
I don't know how many times I've said this, but after 3 pages and no mention I'll say it again. True professional interfaces are not only balanced but there 14 dbu hotter. So watch out for level mismatch. And the only truly balanced interface is a transformer. All electronic balancing circuits have some form of drawback.
Fully agree . But some thinks that transformers are still 50's design highly coloured. I think they're wrong but it sometimes start wars so i keep for me.
About level shifting, Myteks are truly professionnal and so give you choice for output voltage reference. It is stated in manual's quote in first page about alignement to -10dbv.
I agree with you about that too and it is the reason i talked about 1240 That buffers: one is -12db attenuation balanced to unbalanced line receiver.
If it still not enough attenuation a R2R ladder could be used in front to have constant impedance pad of the correct db amount. You gain in headroom too which is one of my obsession.
In some very particular case where EM/RFI could be problematic it has an advantage in keeping those low thanks to symmetric/balanced cable up to the buffer which is close enough to amp to be wired by a (very) short unbalanced cable. It has the advantage to represent really easy load to converters output too.
Yes.
You need to check what levels you intend to pass through and what effect the pre-amp will have on those levels.
It's called "design".
You could also check at what voltage level, distortion starts to climb for the single ended circuits in many pre-amps. Instead of measuring this, I adopt a rule that the maximum input/output signal be at least 10dB below the pre-amp clipping level.
This leads to a CD output level of 2.2Vac (6.2Vpp) requiring a single ended amplifier to have a clipping level greater than 18Vpp and thus requiring a supply exceeding 20Vdc.
The LED/LDR attenuators have no active gain, but they do increase their added distortion at higher signal levels. Might be worth looking at the datasheets to see if +0dB, or +10dB, ref your maximum signal could be doing to the LDRs.
Last edited:
Well it's becoming a bit technical for me now. I just plugged in with xlr to rca adapter. Audio volume comparison between CD direct to LDR pre and dac to pre does show volume is quite similar. Is this a sign it might work without a transformer conversion, or I am just an idiot and I understood nothing? Maybe levels where reduced a bit already at mytek trimmers?
Probably. Did you do some calibrating before?
Your not an idiot and we probably wasn't clear enough.
Transformer is not needed. It's just if you want to be sure to never gets into trouble about groundloop noise, the real definitive answer is a transformer and as a side it gives an accurate method to do balanced/unbalanced.
Note that electronic balancing unbalancing work perfectly well in many case but in some case only a transformer can withstand. This is one of the reason in pro audio gear you see them used often: environnement is (potentially) hostile in this
case.
In domestic environnement you (probably) should'nt have real benefit using a transformer.
To be clear about level shifting and transformers not all are doing this. In most case, input transformers are approximately 1:1 ration between primary and secondary, so they won't induce level shifting, an the impedance won't change neither.
Some others will 2:1 ratio (or other ratios) and in this case you'll have an attenuation and impedance seen by precedent stage (or source) will be upped by 4 if you compare to the stage it feeds. For example if your preamp has an input Z of 5k and you insert a 2:1 transformer between your pre and DAC ,your dac will now 'see' an impedance of 20k (5k * 4) and the voltage your dac output is now -6db at your preamp input (20log 0.5:-6). Hope this clarify a bit.
I will try to extend a bit on Andrew last post.
Reference levels are different in consumer and proaudio: this relate to -10dbv and +4dbu regularly stated on gear. I won't explain why there is difference between the two but for the end user this have some effects.
As stated by Cbdb and Andrew pro gear outlevel is 'hotter' than consumer hifi by 14db (approximately and given the pro gear is seriously engineered), and given the habits in pro field is to use balanced and consumer hifi is mainly unbalanced you'll have to treat that too.
So for interfacing pro and consummer you'll need to unbalance (in most cases) and take care of this level too.
For unbalancing we have already talked.
For level shifting you'll need to use a 'pad' in front of your preamp of approximately 14db attenuation. Or if your converter have an internal volume control be sure you always stay below -14db maxoutput and you should be safe operating area.
If you want to use a pad it's easy: you need to know the input impedance of your preamp and use a constant impedance pad of -14db using same impedance as your preamp. This way you are sure sound won't change in your signal path as the pad is seen by your converter as if it is the input z of your preamp so we can say the pad is 'invisible' and perform level shifting without inducing spectral content change (a symptom of bad impedance matching). Most pads use only some resistors so they should not introduce distortion or artefacts to signal (maybe at worst case some little bit of 'jonhson noise').
Reference levels are different in consumer and proaudio: this relate to -10dbv and +4dbu regularly stated on gear. I won't explain why there is difference between the two but for the end user this have some effects.
As stated by Cbdb and Andrew pro gear outlevel is 'hotter' than consumer hifi by 14db (approximately and given the pro gear is seriously engineered), and given the habits in pro field is to use balanced and consumer hifi is mainly unbalanced you'll have to treat that too.
So for interfacing pro and consummer you'll need to unbalance (in most cases) and take care of this level too.
For unbalancing we have already talked.
For level shifting you'll need to use a 'pad' in front of your preamp of approximately 14db attenuation. Or if your converter have an internal volume control be sure you always stay below -14db maxoutput and you should be safe operating area.
If you want to use a pad it's easy: you need to know the input impedance of your preamp and use a constant impedance pad of -14db using same impedance as your preamp. This way you are sure sound won't change in your signal path as the pad is seen by your converter as if it is the input z of your preamp so we can say the pad is 'invisible' and perform level shifting without inducing spectral content change (a symptom of bad impedance matching). Most pads use only some resistors so they should not introduce distortion or artefacts to signal (maybe at worst case some little bit of 'jonhson noise').
Last edited:
Krivium, you have been extremely kind to explain. Maybe the ideal situation is to use the cable, no transformer, but..... Check the mytek level and possibly decrease it if necessary. The guy that sold it to me said he never decreased the levels but he used in consumer system like me. Maybe the shop made it? No idea. Would it be to much hassle if I write in pm the mytek instructions to bring level from +4 to - 10? Then you could judge if a basic guy like me can do? Thanxs for all, all cases.
Very good idea. If i can help i will no problem. Could you tell which model it is exactly.
It should not be to difficult, probably some jumpers to move and probably check with a voltmeter which level is at output.
I don't know very much the stereo models but i used the 8channel one and remember it as an internal volume control. Can't remember if it's clearly stated attenuation in db but could easily be verified with voltmeter.
You're rght Andrew. Most sources like CD/SACD players and phono pre's today deliver between 1 and 2V I think - that should be between 0 and +6dBV.
Jan
And to the XLR to RCA connection: NEVER connect XLR pin 1 to any signal ground - it should be connected to chassis ONLY. Unless you like hum and ground loops .
A true balanced connection is pin 2 for hot pin 3 for cold. So pin 2 to RCA hot, pin 3 to RCA ground, pin 1 to chassis. If it doesn't work, it wasn't a true balanced connection.
Jan
.A true balanced connection is pin 2 for hot pin 3 for cold. So pin 2 to RCA hot, pin 3 to RCA ground, pin 1 to chassis. If it doesn't work, it wasn't a true balanced connection.
Jan
Yeah I know. As an example, there's still loads of people that think a balanced line requires two signals in opposite polarity and equal amplitude. Not realizing they unconsciously introduce again a reference to ground, which is what we try to get rid of in balanced connections in the first place....
Jan
Jan
No you haven't, my bad. I was trying to use a short cut to make things easier to understand from a non electronician point of view but i think we agree about general idea. And the value given are more inline with my point of view about gain stagging and nescessity to allow for 2/3db minimum headroom when possible.
I must confess i didn't checked your math neither, as i consider most (if not all) of the time your post about theory and explanation are spot on. edit: in fact i just reread the post we are talking about, and don't kow maybe i was thinking something else at the time but did'nt noticed some things... We came to more or less same conclusion about some things.
Maybe you are right i can't tell. I don't use consumer audio gear except digital player. I came from professional audio and use the standard in use in this field. As far as i can tell +4dbu= -18dbfs (with +/- 2db margin allowed depending on bradcast/recording- mixing studio/mastering facilities) is pretty standard and -10dbv still used in some 'low cost' hardware. In this i'm ok with what Art M is saying.
From what i've seen 2v is becoming more or less standard maximum rms out level on digital disk players converters and with some stand alone units (DAD units if i remember correctly).
So for odbfs: 2v rms which equal +8dbu. If you follow same principle as for +4dbu example given you'll see we are not far enough from -10dbu reference point. let's say -18dbfs should output 0.2449 vrms which equal -10dbu.
So -10dbu reference seems still valid to me.
Or you are facing a gear which use the now obsolete Pin3 hot schem... Still happen sometimes in studio. Introducing headache.
Be assured you've got mine Andrew.
Probably because the way it is usually presented give more focus to that part of the schem than for impedance.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.