Inexepensive SDC on blowout 33€ (Avantone CK1/MXL 603)

[Khron]

Just a tip - it might be worth / helpful to raise the minimum level of the graphs, so that whatever differences there are, are easier to see

Ie. the minimum in that graph is "115dB"; just zoom in on the vertical axis and pan the graph so the traces are "in frame", and raise that minimum to maybe 140 or so (since there's not much relevant data under 40Hz anyway)

 

[MagnetoSound]

It's whole area between 500hz and 1k that is boosted. That peak stands out, but it is present in C1 as well. I am pretty sure it's my room or speaker as it' present even with ECM8000 measurement mic.

Could be pattern and capsule type diference.

Subjectively i dont hear it poking through.

Thanks. The levelling off above 3k, is that the result of your top cut mod?  Without the mod, is the response closer to the other mics in your test?

 

[kingkorg]

Yes. High end is stock roughly in between  modded and C1.

I am really sorry for graphs looking like this, it's very, very, rough way of just presenting what is going on. I simply do not have time to move speakers, change the room, it would be too time consuming.

However i think i found the ultimate, and possibly easiest way to flatten out capsule response.

I tried varying almost all the elements of one of the capsules today. Closed holes, used thicker spacers, even put thin mylar dot between membrane and backplate to simulate center termination,  I thought i could widen the holes, but just before that i found an element that directly impacts response of high end, with little to no effect on polar pattern.

It's these copper washers between metal and plastic backplates that determine volume of the air chamber. There is two of them. Remove one of them, and you get instant hi end boost and crazy 7K peak. I added one washer made of thinnest paper  i could find today, everything up to 5k boosted for about 5db, i got -3db bell cut at 7KHz, and rest of high end remained at the same spot. That would say i got -5db of very linear HF reduction with no roloff, and additional cut at very harsh sounding spot.  And yes, it was actually boost, so capsule sensitivity improved. I guess membrane is too dampened by stock chamber? Just guessing.

By the way, this is all in a flat, stock mic, without modded circuit.

However i think that is too much, as the capsule isn't so bright. This paper washer brought it into Oktava MK319 without HF disk world.  I just want to make it as flat as possible, and flatten out that 7K peak it seems to have. I will experiment, and give you feedback. Shame i dont have a micrometer.

It seems that this backplate is exactly what Marek was tweaking. Just by looking at the picture of his mod.

Cool thing is that backplates can be twisted by any degree in regards to each other, super cool way of manipulating polar pattern. A kind of thing Khron was doing with k67. And it's so easy with this capsule. For both of these tweaks you dont need even to disassemble the whole thing, but just unscrew  lockring, central screw and remove only plastic backplate. This way you avoid contaminating membrane and metal backplate space. 

Tuning frequency of this diaphragm on this example is about 1500Hz

Here is the thing i am talking about.

 

[kingkorg]

Here is best picture of Marik's capsule i could find.

 

[jarvis]

Hey there,
firstly I want to thank you for the tip and for keeping us posted with all your measurements and findings, it's really interesting.

I'm just getting started DIYing and I've been lurking for a few months, just picking up some bits and pieces of information here and there, and when I stumbled upon this thread I thought this mic would be a perfect candidate for me to get into mic modding. I read up on the "standard" mods and I'm keen on trying a few things out.

So I received it today recorded a few samples to get a feel of the mic and now I'd to take a look at the guts but I can't for the life of me figure out how to open that thing up. I unscrewed the 3 screws at the bottom and thought the tube would slide right up but I guess I was mistaken. Any hints? ???

Cheers
Jannis

 

[Khron]

Have you removed the capsule first, just in case? And what about removing the caps of the switches? I'm guessing the latter are (still) preventing your disassembly...

https://groupdiy.com/index.php?topic=67447.msg856730#msg856730

 

[jarvis]

Hey Khron,
thanks for the fast reply!

Haha, yes, I did remove the capsule beforehand.  I also suspected the switches but I can't really figure out how to get the black caps off of them, since I think that these are the culprits, correct?

Cheers
Jannis

 

[Khron]

Lever them out with a pair of tweezers, or a narrow-enough-bladed screwdriver? There's a good chance they're just glued onto the "shaft" / arm of the switches. They were, on my t.bone SC140's.

 

[kingkorg]

Forgot about that. Just turn silver part where the XLR is counter clockwise and switch caps will fall of the switches. However, good luck putting caps back into place. I gave up, i left switches without caps. I also think about removing switches as i think LF cut is at wrong point in the circuit. But i might be wrong. And i think PAD switch is causing stray capacitance. I might use them for something else.

Some say screws should be screwed all the way in to open it, however i unscrewed them and removed. didn't work the other way around for me.

 

[kingkorg]

By the way. Standard mod with changing caps with better ones and larger values, removing output caps didn't do anything for me! I think NP0/C0G are already in there, but i might be wrong.

I did null test with modded and stock mic, same spot played through speakers, pink noise, drum loops, vocals, different levels. Modded and stock cancelled perfectly out.  A bit of low end rumble got through modded one, but that was in noise floor.

Using this mod with bigger 1uF C3, C4 (in original MXL schematic c40, c90) caps, will change EQ curve! I used it with stock ones for results i measured.
http://www.audioimprov.com/AudioImprov/Mics/Entries/2014/2/8_EQ_Pt.3_-_Transformerless_Mics.html

I am sure though fet biasing can be done, but i am still learning there.

 

[Khron]

Ironic that the HPF is "causing stray capacitance", since 1) it looks pretty much "in line" with the signal flow, and 2) it switches a bigger capacitor in parallel with a smaller one (small one alone "creating" the high-pass)

The pad switch, though, now THAT might cause some stray capacitance indeed ;D

 

[jarvis]

Thank you both for the answers,
I just levered them out with my screwdriver as Khron suggested and I dont see myself ever putting them back on again either.  ;D
But even after pulling them off I still felt a significant bit of resistance. Turns out a small part of the mesh behind the cutout was glued to the plastic of the capsule connector part.  A bit of glue must have dripped down when they glued the mesh to the tube. With a bit of force I was able to get it off in the end so It's all good now.

But considering I read a lot of comments on how it's important to keep the PCB clean after soldering, I'm surprised that thing isn't noisy since it has a lot of flux residue on it.

The mod not making any or not much of a difference for you is kind of a bummer to hear, but I already ordered a few caps to try the mod out myself and I guess I'll have to see if I get the same results. I ordered some 0.22uF as well as 0,47uF and 1uF caps for C3/4 to mess around a bit. Since people have been reporting a better bass response with upping the values of these caps I'd like to know how the eq circuit works. Is this an RC being formed by R60 and C90 / R70 and C40 respectively?

Cheers
Jannis

 

[kingkorg]

Ironic that the HPF is "causing stray capacitance", since 1) it looks pretty much "in line" with the signal flow, and 2) it switches a bigger capacitor in parallel with a smaller one (small one alone "creating" the high-pass)

The pad switch, though, now THAT might cause some stray capacitance indeed ;D

I am getting tired, just edited that. Yes i was thinking about PAD. Sorry.

The mod not making any or not much of a difference for you is kind of a bummer to hear, but I already ordered a few caps to try the mod out myself and I guess I'll have to see if I get the same results. I ordered some 0.22uF as well as 0,47uF and 1uF caps for C3/4 to mess around a bit. Since people have been reporting a better bass response with upping the values of these caps I'd like to know how the eq circuit works. Is this an RC being formed by R60 and C90 / R70 and C40 respectively?

Cheers
Jannis

I might be wrong, you might have more luck, or better caps, dont know. Just please be sure you are not fooling yourself. People play guitar in front of mics, move, mod them, spend tons of energy and time, come back after a week and play guitar again, different spot, different mood, strings can change over a week if played a lot, and claim they get better low end response. I don't trust my own ears and brain that much to fool my self.  Make fair testing setup, and dont even look which take is which after modding.  I prefer to record pink noise and drums before and after any change i make. Seems to be most revealing. Same spot, levels, preamp gain... Normalize, and null test.

I don't think either this mic has low end problem, as long as there is no cut there is enough room to work with. My frequency sweep of the circuit shows its flat from 40Hz.  It's that 7.5K peak that might be troubling me on some sources.  But capsule mod seems to address this exact point. 

The irony is that i plan to use this guy on toms, and that peak could be useful there.

 

[Khron]

http://audioimprov.com/AudioImprov/Mics/Entries/2014/2/7_EQ_Pt.2_-_Transformerless_Mics.html

The caps coming from the JFET's source and drain and the base-to-collector resistors from the PNP output transistors do indeed form an RC high-pass. The high-pass switch only switches the two larger-value caps in parallel with the ones mentioned above (or not), to achieve the "flat" response.

 

[jarvis]

First off, the site both of you posted is a goldmine, thank you guys for the hint!

I might be wrong, you might have more luck, or better caps, dont know. Just please be sure you are not fooling yourself. People play guitar in front of mics, move, mod them, spend tons of energy and time, come back after a week and play guitar again, different spot, different mood, strings can change over a week if played a lot, and claim they get better low end response. I don't trust my own ears and brain that much to fool my self.  Make fair testing setup, and dont even look which take is which after modding.  I prefer to record pink noise and drums before and after any change i make. Seems to be most revealing. Same spot, levels, preamp gain... Normalize, and null test.

Yes, I'll keep that in mind when I'll do any before and after tests.

The caps coming from the JFET's source and drain and the base-to-collector resistors from the PNP output transistors do indeed form an RC high-pass. The high-pass switch only switches the two larger-value caps in parallel with the ones mentioned above (or not), to achieve the "flat" response.

So I'm assuming that C3 in conjunction with C14 and C4/15 respectively are switched in series when the highpass switch position is not engaged and C14/15 are switched out when it is?
So are the values for C40/90 in the MXL 603s schematics even the same as the ones for C3/4 in in this circuit?

Cheers
Jannis

 

[Khron]

See where they're connected, they only have two legs, that Schoeps circuit is used all over the place in cheap mics. Component designations don't mean much; their ROLE does

So I'm assuming that C3 in conjunction with C14 and C4/15 respectively are switched in series when the highpass switch position is not engaged and C14/15 are switched out when it is?
So are the values for C40/90 in the MXL 603s schematics even the same as the ones for C3/4 in in this circuit?

Cheers
Jannis

 

[jarvis]

See where they're connected, they only have two legs, that Schoeps circuit is used all over the place in cheap mics. Component designations don't mean much; their ROLE does

Yes, I think I got that (but probably not).

I was under the impression, that they do play, although not vastly, different roles, since the MXL 603s circuit doesn't look like it has a HPF-switch, while this mic does. So I suppose, the MXL circuit has higher values for the capacitors at the same position since the "flat-mode" with this mic, as I understand, switches the capacitors in series. So the values for the capacitors in this circuit at the same position should be lower for the fc of the HPF to be higher when the switch is engaged. Which would make changing the capacitors in C3/4 kind of pointless, since it would pretty much only really render the HPF less effective. Or am I wrong here (also very probable  ;D )?

 

[Khron]

In the 603, yes, the caps are most likely the higher value, in order to achieve a flat low-end response.

On the other hand, i'm pretty sure that in this mic (like in my Alctron HSMC001), the "default" (unswitched) caps are the smaller values that achieve the high-pass effect, and the switch only connects the larger values in parallel with those. You DO know / remember that's how capacitor values add up, right? ;D

From what i can understand from the photos kingkorg posted, the white and black wires on the underside of the board, are from the JFET drain and source.

If we ignore the switch, the signals go through C4 and whichever its "mirror twin" is on the other side of the board, and on to the PNP's, right?

In most switches, the middle leg is usually the "common", and depending on the position of the switch (or lever, in the case of toggle switches), it gets connected to one of the two other legs of the switch. You with me so far?
http://www.zseries.in/electronics%20lab/switches/slide/pictures/construction%20of%20spdt%20slide%20switch.png

If we look over here:
https://groupdiy.com/index.php?action=dlattach;topic=67447.0;attach=54957;image

We see that the right-most pair of legs of the switch (towards the XLR) isn't even soldered in. On the other hand, we can see that the right-most legs of both pairs of caps are on the same trace, so they're connected together.

The signal coming through the black / white wires, in addition to going through the (i'm quite sure) smaller-value caps, also goes to the left-most legs of the switch. Looking at the outside of the mic, the "towards XLR" position of the switch is the "high-pass on"

If you slide the switch the other way, the midle pair of switch legs gets connected to the left-most pair, and as such, C14 and C15 are getting connected in parallel with C4 and the other one, respectively

 

[kingkorg]

Khron, you are totally right. I omitted those caps totally.

Jarvis, you can check numbers on those  and use  this to figure out values.
http://www.electronics2000.co.uk/calc/capacitor-code-calculator.php

I don't have the mics with me right now. Could be they add up to 1uf, or close, and that is why i didn't have luck with mod.

As far as i can see on pictures on my phone, c14 and c15 are 470nF + value of c4 and c3, and we dont get a lot by upgrading to 1uF.  Stock MXL caps are 220nF. Additionally they seem as higher quality caps than MXL's already. But i might be wrong.

These guys from China follow mods on the internet and improve designs. Alctron has upgraded all of Neve rip offs to tantalums when GAP started advertising more ''advanced'' versions of GAP pres and comps. 

 

[jarvis]

First off, thanks for taking the time writing such a thorough response, Khron. I appreciate it.

In the 603, yes, the caps are most likely the higher value, in order to achieve a flat low-end response.

On the other hand, i'm pretty sure that in this mic (like in my Alctron HSMC001), the "default" (unswitched) caps are the smaller values that achieve the high-pass effect, and the switch only connects the larger values in parallel with those. You DO know / remember that's how capacitor values add up, right? ;D

Yeah, that's what I meant when I wrote: "So the values for the capacitors in this circuit at the same position should be lower for the fc of the HPF to be higher when the switch is engaged."

From what i can understand from the photos kingkorg posted, the white and black wires on the underside of the board, are from the JFET drain and source.

Yup, correct.

If we ignore the switch, the signals go through C4 and whichever its "mirror twin" is on the other side of the board, and on to the PNP's, right?

In most switches, the middle leg is usually the "common", and depending on the position of the switch (or lever, in the case of toggle switches), it gets connected to one of the two other legs of the switch. You with me so far?
http://www.zseries.in/electronics%20lab/switches/slide/pictures/construction%20of%20spdt%20slide%20switch.png

If we look over here:
https://groupdiy.com/index.php?action=dlattach;topic=67447.0;attach=54957;image

We see that the right-most pair of legs of the switch (towards the XLR) isn't even soldered in. On the other hand, we can see that the right-most legs of both pairs of caps are on the same trace, so they're connected together.

The signal coming through the black / white wires, in addition to going through the (i'm quite sure) smaller-value caps, also goes to the left-most legs of the switch. Looking at the outside of the mic, the "towards XLR" position of the switch is the "high-pass on"

If you slide the switch the other way, the midle pair of switch legs gets connected to the left-most pair, and as such, C14 and C15 are getting connected in parallel with C4 and the other one, respectively

Yes, that's what I figured, when I looked at the board. Ok, so my assumption was correct!

So replacing C3 and C4 with higher value caps, lowers the cut-off frequency of the HPF, rendering it pretty much useless. Thats why, in order to mod the mic with film caps while retaining the function of the HPF, I'll need to find out the values of the lower value caps and then pick a value that is ≥ 0.22uF for the caps that are switched in series when the switch is not engaged.

Furthermore, for the mod with better quality film caps to have an audible effect (that others claimed to get), you'd need to switch both the ceramics before the PNPs  to film caps, so not only the C3/4 but also C14/15. Is that what you did, kingkorg?

Cheers
Jannis

Edit:
Ah, my bad. I didn't see your response until after I hit send.
Thanks for the link and your insight, that helps!
I just looked it up and C3/4 are 10nF and C14/15 are, as you said, 470nF.
I guess I'll have to order some lower value films as well just to check it out.