DI/Reamp Theory Questions

[MrG]

Hey folks,

Forgive my noob-ness, but I am trying to make respectable sense of building DIs and Reamp boxes. Here are a couple questions/scenarios which will hopefully illustrate my confusion:

-I have a UTC A-10 wired backwards - instrument goes in 50k, comes out 200 ohm - and this works when plugged into a mic pre and makes sense to me, in that it changes an instrument’s pickups impedances from high Z to lo Z mic level, however, does this actually make the signal (voltage?) quieter?

-I know that a DI backwards is _not the same as a reamp_, but given that some folks like Jakob E. have mentioned simply turning signals down instead of using a true reamp process, I wanted to see how much it bumped line level signal down and so I tried this. The signal was extremely hot - unusably hot to go into an amp. However, when I plugged the Line level signal into the 50k “input” of the DI, it dropped the signal appropriately and sounded great through the amp... SO,

-I then inserted it onto the track from my DAW - line in to 200 ohm side, and 50k out into the DAW - and it did not change the signal level at all, only sounded like good harmonic distortion from a transformer at a 1:1 level type of sound... so then I switched it to being inserted as a DI (50k in, 200 out) and it lowered the signal significantly...

So my question is, what am I misunderstanding here? I hope I have presented some useful thoughts for my confusion.

One huge unexpected observation I’m having is that this backwards transformer DI works well for _both_ instrument in to mic preamp _and_ for reamping when used the same direction as the DI... confused!

Thanks in advance for all the help.

MG

 

[abbey road d enfer]

it changes an instrument’s pickups impedances from high Z to lo Z mic level, however, does this actually make the signal (voltage?) quieter?

Yes it does.

-I then inserted it onto the track from my DAW - line in to 200 ohm side, and 50k out into the DAW - and it did not change the signal level at all, only sounded like good harmonic distortion from a transformer at a 1:1 level type of sound...

The reason is that you load the secondary with a medium impedance, which creates attenuation.

One huge unexpected observation I’m having is that this backwards transformer DI works well for _both_ instrument in to mic preamp _and_ for reamping when used the same direction as the DI... confused!

Actually, there's no reason for being confused. As a DI, it provides impedance adaptation between the high impedance of the guitar and the low impedance of the mic preamp. As a "reamp", it provides the attenuation necessary to prevent overloading the input of the guitar amp.

 

[gridcurrent]

Abbey's comments:  +1

the patent implies the original transformer was a UTC O-10 with a resistive network.
https://patents.google.com/patent/US6005950A/en

A UTC A-24 into a 1000 ohm potentiometer works superbly.

There exists considerable confusion about using a DI backwards or even a 1:1.

 

[Bo Deadly]

DI-ing and reamp-ing is all about matching impedances. You need to know what the source impedance of the guitar is and what the input impedance of the amp is.

The guitar source impedance is about 30k but it can be a lot more if the volume knob is not max because you have the impedance of the pot. That can be an extra 50k or so. Also, the cable has capacitance that could be as high as a couple hundred pF which defines the impedance at high frequencies (low which means high freq. atten).

The amp input impedance is going to be quite high. The standard Fender style tube amp input is 1M (or ~136k for the low gain input). The tube also has potentially quite high input capacitance of maybe 400p but after 32k in series (or 68k for the low gain input).

So a good DI / reamp needs to replicate these two interfaces to make the guitar behave like it's connected to the amp when it's not and the amp behave as if it were connected to a guitar when it's not. For output to an amp (reamp) this could be as simple as a 33k resistor and possibly a capacitor to ground depending on what cable is used to connect the reamp box to the amp. For the input from the guitar (DI) you want 1M to ground but then also 33k in series with probably a fet input and possibly also add a capacitor to ground.

Then you have the issue of taking the resulting guitar signal and outputting balanced and taking balanced signal and converting to single-ended for interfacing with the amp.

Of course I don't think most people think about this stuff that much. They might think it's not necessary. But I would argue it really is. If you can visualize the guitar -> cable -> input, that's two rounds of ~30k / 400p which both have a roll-off of ~13kHz. That's quite a bit of high cut. I would be curious to see a schematic of one of the Radial DI boxes to see if they have the series resistance or if it's just a transformer. Transformers are popular in DIs because they provide isolation and because they can be made passive which means no issues with power and the ground noise that might go with that. But in practice it's really not necessary to use a transformer. It would work equally well or better if the circuit were active. It just has to be designed properly so as to minimize ground loop noise.

In fact, for reamping, I would not use the transformer you describe in reverse. You don't want to send a really low level (noise) and then load the driver with 200R (distortion). And the output impedance will not be well defined. I would actually just use a simple passive balanced attenuator, 30k series resistor + 400p cap and leave the shield unconnected at one end (amp end I'm thinking). When it comes to guitar, it's almost never a bad idea to roll off the highs. Guitar is only putting out maybe 1kHz. A Jensen P12Q crushes everything above 7kHz. Personally I would roll off guitar highs at every opportunity.

 

[john12ax7]

I do a lot of reamping,  and don't really like boosting / attenuating the guitar signal.  Keeping things roughly 1:1 works well.

I would argue against  using a passive  DI / reamp setup.  The main thing is having a large input impedance on the DI part,  1M is good.  For the reamp part you actually don't even need anything,  converter to amp works fine,  though a transformer box can be useful for long cables runs,  gnd lift,  tone,  etc.

 

[MrG]

Yes it does.
The reason is that you load the secondary with a medium impedance, which creates attenuation.
Actually, there's no reason for being confused. As a DI, it provides impedance adaptation between the high impedance of the guitar and the low impedance of the mic preamp. As a "reamp", it provides the attenuation necessary to prevent overloading the input of the guitar amp.

Thank you very much Abbey Road. I think this all helps some issues of my understanding.

I’m very detail-oriented but also very practical minded in my understanding, so to clarify and potentially help others in my situation, is the following basically an accurate understanding of the theories of impedance matching?:

For example:

- take bass (30-80k?) > di (50k in, 200 out) > mic pre in (150?). The hi z instrument sees approximately the same impedance, thus minimally overdrives, then the transformer attenuates the level of the instrument significantly while/via lowering it’s impedance, and then the mic pre sees matching impedance so minimal gain/attenuation changes happen?

-take converter line level (600?) > di backwards (200 in, 50k out) > converter  line in. It stays near unity with some added harmonic distortion bc line at 600 overdrives the in of di backwards at 200, then gets gained up significantly via the step to 50k, then gained back down significantly by the step down to 600?

-take converter line out (600?) > di (50k in, 200 out) > converter line in. Line out at 600 does not overdrive input of 50k, is significantly attenuated by step down to 200, and then does not overdrive the input of line in at 600?

-take converter line out (600?) > di backwards (200 in, 50k out) > 66 fender showman in (1M?). Line out at 600 overdrives input of di backwards at 200, is significantly boosted by step up to 50k, and then does not overdrive the input to fender via impedance, but is already far too loud for the circuit?

Btw, the above are my best practical explanations of what _actually happened_ in terms of level changes when I tested. I welcome any correction and hope this isn’t pedantic.

This all perhaps suffices to ask, does input/primary impedance mainly determine overdriving vs. not overdriving or does it boost/attenuate according to mismatch? And transformer ratio/imp determines boosting/attenuation, then secondary impedance matching determines overdriving and boosting/attenuation to match input impedance?

@squarewave - I need to understand more principles before responding to you intelligently, and I’m working on filling my gaps in knowledge asap. But one note - you mention the impedance and capacitance mismatching affecting high end above 13k, and also mention rolling off highs on egt - couldn’t agree more. I think I enjoy this more primitive approach due to, first, what my ears are telling me is simple yet gorgeous, and also due to the rolloffs in unnecessary highs, as well as interesting inconsistencies in the frequency spectrum.

Also, to clarify I am somewhat baffled because my Di backwards has the amp seeing 50k, yet is so incredibly high level that the amp plays distorted signal even with volume at 0! The DI _forwards_ (same direction as using instrument into mic pre) is the only way I was able to use it in a reamp function, and it sounded awesome.

I do a lot of reamping,  and don't really like boosting / attenuating the guitar signal.  Keeping things roughly 1:1 works well.

I would argue against  using a passive  DI / reamp setup.  The main thing is having a large input impedance on the DI part,  1M is good.  For the reamp part you actually don't even need anything,  converter to amp works fine,  though a transformer box can be useful for long cables runs,  gnd lift,  tone,  etc.

Agreed and good to know. I am thoroughly enjoying the tone of my chosen transformers, which is my main reason for pursuing this all in the first place, but I’ll keep this in mind - thanks!


Thanks all for the advice,
MG

 

[john12ax7]

Are you adding a lot of gain with the mic pre ? That could be the reason you are subsequently overdriving the amp.  Try it without a mic pre, straight into the converter. Then see if the levels and noise are ok with the amp.

 

[Bo Deadly]

Also, to clarify I am somewhat baffled because my Di backwards has the amp seeing 50k, yet is so incredibly high level that the amp plays distorted signal even with volume at 0! The DI _forwards_ (same direction as using instrument into mic pre) is the only way I was able to use it in a reamp function, and it sounded awesome.

The impedance of a transformer is actually not specific values but depends instead on what the other end is connected to. So it's more like the gears on a bike. What really matters is the ratio of the two windings. The formula to remember is that the impedances are the squares of the ratios. Or you could also say the ratios are the square roots of the impedances.

So in your 50k:200 transformer, the voltage ratio is sqrt(50k):sqrt(200) = 224:14 = 16:1 = -24dB.

Correction: The impedance ratio is 50k/200 = 250 (or voltage ratio squared). So I have corrected the various examples below.

But again, I think a point that you're missing is that the impedance depends on what is at each end. Your transformer is only 50k if the other end is 200. And visa versa. If you do guitar (30k) > transformer stepping down 50k:200 > Fender (1M), then the impedance "seen" by the guitar is actually going to be 250 * 1M or 250M. And the impedance seen by the amp is going to be 30k / 250 = 120 ohms. So if the guitar sees super high impedance, you will get every immaculate detail at all frequencies (including probably Norwegian radio broadcasts). And because you attenuate 24dB, you have to crank the amp which is always a good idea.

Just to drive home the point, regarding one of your other examples:

> take converter line level (600?) > di backwards (200 in, 50k out) > converter  line in

The converter line level out would "see" the impedance of the converter line in (600?) / 250 = 2.4 ohms (although converter line in is probably more like 10k so 10k / 250 = 40 ohms). And the converter line in would see 0-50 ohms * 250 = 12.5k. Needless to say, this is a horrible arrangement as you're loading the line out pretty hard.

 

[abbey road d enfer]

So a good DI / reamp needs to replicate these two interfaces to make the guitar behave like it's connected to the amp when it's not and the amp behave as if it were connected to a guitar when it's not.

Actually, only one of theses conditions need to be met. When the guitar is played into the amp, only one interface effect exists.
In the process of reamping, since connecting the guitar directly to a mic input is a receipe for disaster, that's where the impedance conversion is mandatory. For outputting a balanced line level signal to a guitar amp, there may not be a need for a gobo. Many coverter outputs have floating balanced outputs that can be directly connected to an unbalanced input, and proper level staging can solve any level issues. Indeed a transformer would take care of ground isolation.
The big question is where do we actually produce the "interface effect"?
For a guitar into a xfmr DI does not react at all like it does when plugged into an amp. An active DI with 1Meg input Z would be closer, but still may not provide the same interaction. It leaves a position for a tube DI there.
Or the "interface effect" could be done at reamping, with a reamp box presenting the right source impedance to the guitar amp. That would suppose DI'ing the guitar as neutral as possible, i.e. probably via a SS active DI.
Anyway it shows that reamping cannot sound exactly like a direct-to-amp connection. But pretty close if one understands what happens.

If you can visualize the guitar -> cable -> input, that's two rounds of ~30k / 400p which both have a roll-off of ~13kHz.

That will be swamped by the fact the amp's speakers roll off drastically much earlier (4-6 kHz).

  I would be curious to see a schematic of one of the Radial DI boxes to see if they have the series resistance or if it's just a transformer.

I don't know the radial specifically, but there are many reamp boxes around that include an RLC network supposed to simulate a pick-up impedance.

 

[ruffrecords]

Once you have recorded a guitar via a DI, you have captured the guitar signal and raised it to line level. So when you play it back  for reamping I would have thought all you needed to do was connect it straight to the guitar amp or have I missed something.

Cheers

ian

 

[volker]

Most reamp boxes use a transformer to avoid ground loops between PC/DAC and the guitar amp. A volume control makes a lot of sense too, considering the level difference between line and the typical guitar output. Another thing some boxes do is raise the output impedance to put it more in the realm of a guitar pickups. Some go as far as using an inductor or half a transformer in series to make it more inductive, or emulate the whole tone network within guitars.

 

[ruffrecords]

Most reamp boxes use a transformer to avoid ground loops between PC/DAC and the guitar amp. A volume control makes a lot of sense too, considering the level difference between line and the typical guitar output. Another thing some boxes do is raise the output impedance to put it more in the realm of a guitar pickups. Some go as far as using an inductor or half a transformer in series to make it more inductive, or emulate the whole tone network within guitars.

Ah, I see. I come from the days when all pro outputs had transformers in them already so there was no need to add one. I can see the point of raising the output impedance so the guitar acts as it would if a guitar was plugged in. Not sure about the inductor because that has already been captured by the DI.

Cheers

ian

 

[abbey road d enfer]

Not sure about the inductor because that has already been captured by the DI.

It may or may not. If using a passive DI, the interaction between guitar and DI is very different than between guitar and amp. Times and times again it has shoiwn its shortcomings.
Now an active DI may not react with the guitar in the same way as an amp, particularly a triode input imparts non-linearities a SS DI can't. Also, DI'ing a guitar very often uses a shorter cable.

 

[ruffrecords]

It may or may not. If using a passive DI, the interaction between guitar and DI is very different than between guitar and amp. Times and times again it has shoiwn its shortcomings.

Agreed

Now an active DI may not react with the guitar in the same way as an amp, particularly a triode input imparts non-linearities a SS DI can't.

But the amp will do this when reamping. Surely the point is to capture the guitar signal and leave the coice of amp and it input foibles till later?

Cheers

Ian

 

[abbey road d enfer]

But the amp will do this when reamping.

It will only if the source impedance is correct. If it's the output of a step-down xfmr, it won't happen. I'm talking about gain non-linearities that actually modulate the Miller capacitance.

 

[ruffrecords]

It will only if the source impedance is correct. If it's the output of a step-down xfmr, it won't happen. I'm talking about gain non-linearities that actually modulate the Miller capacitance.

I understand that and I agree emulating the guitar source impedance is important when reamping but we were talking about an active DI.

So summarise it is important that the DI captures the guitar signal unadulterated.

It is important that the reamp emulates the guitar source resistance..


Cheers

ian

 

[abbey road d enfer]

I understand that and I agree emulating the guitar source impedance is important when reamping but we were talking about an active DI.

So summarise it is important that the DI captures the guitar signal unadulterated.

It is important that the reamp emulates the guitar source resistance..

It seems we both agree...

 

[MrG]

Thanks for your direct input (pun intended) all.

There are a few residual things I’m thinking and trying to make sure I relate clearly:

-I understand now the concept of the turns ratio and how it affects the impedance each side is seeing.

-To be clear, as a mixer, I’m making many decisions based on years of honing my taste, and my ear-photographs of sounds which I consider the apex of recording history for each sound. So clean/unadultered is not my interest. In my personal DI-ing, for example,  I’m mainly using it as my only source for bass or egt, or my main source with a blend of an amp. For reamp-ing, it is not necessarily a guitar or bass - could be anything I’ve recorded. I do it old school - I mix to tape and don’t leave myself every option, making decisions as I go based on the vision of the record.

-Regarding the DI, I believe I understand why it works well now. And sonically, I think it does not display neither a bandwidth- nor dynamically-limited picture of the bass, in a way I consider detrimental. The best passive transformer DIs I’m building actually have a euphonic and very balanced signal process, sounding almost like a tube comp plus pultec post-process...

-Reamping is where this gets confusing to me, and I’m still trying to make sure I’m clear. In my attempt to run line level out of my DAC to my transfo DI backwards (200 > 50k) it sounds like I’m doing something you all think is wise, in that I’m raising the impedance of the source drastically, and making it resemble a pickup, effectively. However, the signal is unusably loud... How would one ever raise the impedance that much without boosting the line level signal through the roof?


-When the above reamp is inserted in line level - DAC > 200/50k > ADC, it does not affect the gain at all, sounds 1:1/unity practically, maybe .2 dB difference. Sounds great and rich. Is this potentially damaging either side of the connection based on the loads squarewave was describing?

-and when I insert the DI _forwards_ from DAC to amp, it sounds _great_, full-spectrum, great dynamics, and does not overload... ?

Could someone explain if/how this “reamp” behavior makes sense? To be clear, at the present time, any lay person may consider this DI both a DI and a Reamp all in one go, connected the same exact way, just due to the practical findings.

Again, I hope I’m being clear and very much appreciate you all’s expertise.

Best,
MG

 

[abbey road d enfer]

In my attempt to run line level out of my DAC to my transfo DI backwards (200 > 50k) it sounds like I’m doing something you all think is wise, in that I’m raising the impedance of the source drastically, and making it resemble a pickup, effectively.

No, we haven't recommended that. The high-Z side should be receiving the line level, and the low-Z side go to the amp, which takes care of the attenuation needed to avoid overdriving the amp's input stage. In doing that, the amp is presented with a very low impedance, which is not right for the purpose of recreating the guitar-amp interaction. In order to do that, an additional LRC circuit should be added. However, the main part of using an amp is the loudspeaker/air/room interaction, that is not lost; many SE's are satisfied with that and do not go the extra mile of simulating the impedance of the guitar.

  However, the signal is unusably loud...

That's why we don't recommend it.

How would one ever raise the impedance that much without boosting the line level signal through the roof?

It can be done easily by inserting a resistor of a few dozens kiloohms in the output. Alternatively, you can install a 500k potentiometer there, which will also allow level trimming.
Actually, there are many possibilities in this respect, like adding the aforementioned LRC circuit.

 

[Bo Deadly]

I'm talking about gain non-linearities that actually modulate the Miller capacitance.

I'm not familiar with the "gain non-linearities" you speak of but most tube amp inputs are going to have a significant grid stopper (Fender 5F6-A is 68k||68k). So if series resistance with the grid is already 34k, is adding another 30k for the guitar going to matter to these non-linearities?

The only effect of the guitar source impedance that I'm aware of is it's interaction with the capacitance of the cable. I just measured two guitar cables and got 670pF and 358pF. So if the guitar is 30k that's 7.9kHz and 14.8kHz respectively.

And of course the source impedance of guitar is almost certainly not a flat 30k at all frequencies. I suppose one could feed signal into the output through a series resistor that yields 6dB attenuation across it. Then you could also do a spectra to see what frequencies are attenuated more / less. Might even be able to play with the knobs and see the effect on the response in real time. Mmm, I sense an experiment.