Topic: Poorly done modifications to R4... progress report

[poorlyconditioned]

OK, I've started hacking my R4!

So far I've done the following:
- replaced NJM2068 opamps on Ch1/2 with AD8672.
- replaced some (*) capacitors on Ch1/2 with poly

About the opamps.  It did not help noise floor.  In fact, noise went up a bit!  Not much, but a bit.  It appears there are are two opamps for each channel, one at the mic/line inputs, using NJM2068, and a second NJM2100, driving the ADC (AKM4528VG).  I'm guessing that all the gain is being done at the first stage, and it may be hard to reduce the noise floor.  Did Busman or Oade succeed at this?  I'm not sure.  I noticed that the max gain is something like 60dB, so I'm not expecting miracles.  Anyway, I'm hoping (!) that this opamp improves the sound at least.  Results/opinions to follow.

By the way, there appears to be some funky stuff going on with Ch1/2, using a HC4052 MUX chip.  I don't understand this, and I'm not sure if it is on Ch3/4 or not.  I'm wondering if it is the mic/line selection circuitry.  I don't have a clue about this!  I would really like to understand the gain structure.

About the capacitors.  For XLR input, there are two electrolytics (10u/63v), one for each of "hot" and "cold" and they are positive at mic end where phantom power is applied through a 6k8 resistor.  This makes sense for phantom power (the caps will be correctly polarized), but not for ac input (they will not be polarized).  There are also two lytics for TRS in, this time negative end at input.  So, I did the following hack: I removed the lytic corresponding to the "hot" input on the TRS and put a 2u2 film cap there as well.  When phantom is off the input impedance is 10k.  So, here is the plan: When using non-phantom mics, put *unbalanced* in a TS input, and I will only see the film cap.  For phantom mics, put in the XLR input, and we'll see two caps, but they will be correctly polarized.  Yeah, this makes sense to me.

OK, I'm going to run some comparsions and then probably modify the other two channels.

I'm still undecided about the ADC drivers, but I'm considering something like OP264.  I need a good sounding single-sided opamp.

Whew!
  Richard

[china_rider]

Hey now... You're supposed to be spending all your time on the AKG actives.  Whats up? ;-}

T+ for experimenting

[bgalizio]

T+ for experimenting

[Shawn]

I'm guessing that all the gain is being done at the first stage, and it may be hard to reduce the noise floor.  Did Busman or Oade succeed at this? 

Doug claims that his transparent mod "dramatically reduces hiss" I assume that is the same thing you are talking about.

+t for experimenting.

[RebelRebel]

haha your posts say LEET

OK, I've started hacking my R4!

So far I've done the following:
- replaced NJM2068 opamps on Ch1/2 with AD8672.
- replaced some (*) capacitors on Ch1/2 with poly

About the opamps.  It did not help noise floor.  In fact, noise went up a bit!  Not much, but a bit.  It appears there are are two opamps for each channel, one at the mic/line inputs, using NJM2068, and a second NJM2100, driving the ADC (AKM4528VG).  I'm guessing that all the gain is being done at the first stage, and it may be hard to reduce the noise floor.  Did Busman or Oade succeed at this?  I'm not sure.  I noticed that the max gain is something like 60dB, so I'm not expecting miracles.  Anyway, I'm hoping (!) that this opamp improves the sound at least.  Results/opinions to follow.

By the way, there appears to be some funky stuff going on with Ch1/2, using a HC4052 MUX chip.  I don't understand this, and I'm not sure if it is on Ch3/4 or not.  I'm wondering if it is the mic/line selection circuitry.  I don't have a clue about this!  I would really like to understand the gain structure.

About the capacitors.  For XLR input, there are two electrolytics (10u/63v), one for each of "hot" and "cold" and they are positive at mic end where phantom power is applied through a 6k8 resistor.  This makes sense for phantom power (the caps will be correctly polarized), but not for ac input (they will not be polarized).  There are also two lytics for TRS in, this time negative end at input.  So, I did the following hack: I removed the lytic corresponding to the "hot" input on the TRS and put a 2u2 film cap there as well.  When phantom is off the input impedance is 10k.  So, here is the plan: When using non-phantom mics, put *unbalanced* in a TS input, and I will only see the film cap.  For phantom mics, put in the XLR input, and we'll see two caps, but they will be correctly polarized.  Yeah, this makes sense to me.

OK, I'm going to run some comparsions and then probably modify the other two channels.

I'm still undecided about the ADC drivers, but I'm considering something like OP264.  I need a good sounding single-sided opamp.

Whew!
  Richard

[Shawn]

Assuming you come up with a mod you like do you have any plans to offer it to the community?

[it-goes-to-eleven]

Assuming you come up with a mod you like do you have any plans to offer it to the community?

Would a blackboard schematic be sufficient? 

[Church-Audio]

One trick I use to figure out gain on each stage, is to inject a 1k tone on the input and measure the voltage on the outputs on each stage. One can assume for most stages the impedance will be 600 ohm for balanced and 10k for unbalanced. Then you will know what stage does what. I have always felt in order to get your noise down you need to have two stages, one quiet one with say +10 to +20 DB and a second +40 to +50. I hope that helps. It would be pretty cool to develope a logic type probe with a VU meter and selectable input impedance for doing these measurements.


Chris Church

OK, I've started hacking my R4!

So far I've done the following:
- replaced NJM2068 opamps on Ch1/2 with AD8672.
- replaced some (*) capacitors on Ch1/2 with poly

About the opamps.  It did not help noise floor.  In fact, noise went up a bit!  Not much, but a bit.  It appears there are are two opamps for each channel, one at the mic/line inputs, using NJM2068, and a second NJM2100, driving the ADC (AKM4528VG).  I'm guessing that all the gain is being done at the first stage, and it may be hard to reduce the noise floor.  Did Busman or Oade succeed at this?  I'm not sure.  I noticed that the max gain is something like 60dB, so I'm not expecting miracles.  Anyway, I'm hoping (!) that this opamp improves the sound at least.  Results/opinions to follow.

By the way, there appears to be some funky stuff going on with Ch1/2, using a HC4052 MUX chip.  I don't understand this, and I'm not sure if it is on Ch3/4 or not.  I'm wondering if it is the mic/line selection circuitry.  I don't have a clue about this!  I would really like to understand the gain structure.

About the capacitors.  For XLR input, there are two electrolytics (10u/63v), one for each of "hot" and "cold" and they are positive at mic end where phantom power is applied through a 6k8 resistor.  This makes sense for phantom power (the caps will be correctly polarized), but not for ac input (they will not be polarized).  There are also two lytics for TRS in, this time negative end at input.  So, I did the following hack: I removed the lytic corresponding to the "hot" input on the TRS and put a 2u2 film cap there as well.  When phantom is off the input impedance is 10k.  So, here is the plan: When using non-phantom mics, put *unbalanced* in a TS input, and I will only see the film cap.  For phantom mics, put in the XLR input, and we'll see two caps, but they will be correctly polarized.  Yeah, this makes sense to me.

OK, I'm going to run some comparsions and then probably modify the other two channels.

I'm still undecided about the ADC drivers, but I'm considering something like OP264.  I need a good sounding single-sided opamp.

Whew!
  Richard

[poorlyconditioned]

OK, I think I'm finished hacking for now.  Until I can learn more.

I replaced the capacitors on TRS in, two caps each channel, with film caps 2u2/100V.

I haven't done any more than replacing the initial 2068 opamps.  I used AD8672 I think.

I'm stuck with the rest though.  In particular, I don't know if the remaining 2068s on that board are part of the input circuit or not.  I can't trace this sucker, it is really complicated.  I see a 2100 (dual single supply amp) is used to drive each channel of the AKM chip.  But as I said, I can't trace whether those extra opamps are used.  If so, it is much more work to replace.

Hey, Busman, I could use some help if you're willing...

Thanks!

  Richard

[china_rider]

Get back to the active 480s :-} T+

Busman rocks!

[Daryan]

What brand of capacitors are you using as replacements?

[poorlyconditioned]

What brand of capacitors are you using as replacements?

I'm using standard polypropyline film caps.  I think any kind of *nonpolarized* cap is fine, whether polyester, polypropyline, or whatever.  My point is that *unless polarized correctly* electrolytics are suspect.  I'm not too worried if they are polarized correctly.  This is confirmed by some professional gear, like the GP DMIC-20, which uses standard 'lytics, but has a switch to short them out if you're not using phantom power.

  Richard

[poorlyconditioned]

Get back to the active 480s :-} T+

Busman rocks!

Yes, I will get back to the 480's.  I've got to take a break and rethink some of the *mechanical* issues.

As far as Busman's mods.  Yes, they are great.  But I want to do my own work.  Why?  Well, I don't want to pay (and ship) my unit, I am by nature curious, and finally I believe that all DIY types should have access to this info.  I see no problem with a free/open mod info coexisting with some great for-profit mods.  Believe me, with the number of R4's I'm seeing, there will be lots of business for Busman, and anyone else who is willing to do mods for others.

  Richard

[willndmb]

Get back to the active 480s :-} T+

Busman rocks!

Yes, I will get back to the 480's.  I've got to take a break and rethink some of the *mechanical* issues.

As far as Busman's mods.  Yes, they are great.  But I want to do my own work.  Why?  Well, I don't want to pay (and ship) my unit, I am by nature curious, and finally I believe that all DIY types should have access to this info.  I see no problem with a free/open mod info coexisting with some great for-profit mods.  Believe me, with the number of R4's I'm seeing, there will be lots of business for Busman, and anyone else who is willing to do mods for others.

  Richard

+t
both you guys have given me great gear/info in one way or another

[Daryan]

Why do you think electrolytics are suspect?  I use them all the time and find them 100% reliable.  Of course, I am using blackgates, jensen 4-poles, and panasonic fc series so I am spending a lot of money on them too, but the improvement in sound quality is far from subtle.

[poorlyconditioned]

Why do you think electrolytics are suspect?  I use them all the time and find them 100% reliable.  Of course, I am using blackgates, jensen 4-poles, and panasonic fc series so I am spending a lot of money on them too, but the improvement in sound quality is far from subtle.

I am saying that electrolytics are suspect *if they are not polarized* with a positive voltage.

A "good" scenario would be a blocking cap with a voltage of +48VDC (phantom power)  +/- 0.1VAC (mic signal).

A "bad" scenario would be a blocking cap with just an AC mic signal on it (eg., mic or line, but without phantom power).  In this case the voltage would fluctuate from a small positive to a small negative value.

Since electrolytics only "work" with positive voltage across them, I would be worried.  But to be honest, I don't know exactly what the effect of putting a negative voltage is.  Ideas welcome on this.

  Richard

[Daryan]

I see and understand now.  Mine are always used in polarized situations, so I misunderstood.  Do you have a picture of the internals by any chance?  I would like to see inside if I could and possibly make some suggestions.