Group Buy Interest?

[antichef]

On the more aggressive side, here's a mix my daughter and I have been working on, passed through the ACMP-81s, with some EQ. I had tried this pre-mod (and with an earlier version of the song) and posted to the MP3 clinic -- I think the 81s are much better post mod.

First, without the 81s:

and then with:

I ran them a little hot, and so got some clipping, but for now let's say that was intentional

 

[nuemes]

ACMP73 Bass DI settings

Try this with your ACMP73:

Front single coil jazz bass w/ flatwounds, treble rolled slightly off on bass, thru ACMP73 DI

From left to right:
(Nob 1) Gain: 7 o'clock
(Nob 2) Low Rolloff: 50Hz
(Nob 3) Low Shelf Cut level: 11 o'clock
(Nob 4) Low Shelf: 35Hz
(Nob 5) Para EQ Cut level: 10 o'clock
(Nob 6) Para EQ: 0.36 KHz
(Nob 7) 12KHz at 7 o'clock (rolled off completely)
(Nob 8) Output: set at around 12 o'clock (set to ensure you can play hard without any high peaks on your DAW)

Instant John Paul Jones! Beautiful sound and the level is tamed by the output so no compression needed - great for taming sloppy players.

 

[vicenzajay]

Now...

Try this with your ACMP73:

Front single coil jazz bass w/ flatwounds, treble rolled slightly off on bass, thru ACMP73 DI

From left to right:
(Nob 1) Gain: 7 o'clock
(Nob 2) Low Rolloff: 50Hz
(Nob 3) Low Shelf Cut level: 11 o'clock
(Nob 4) Low Shelf: 35Hz
(Nob 5) Para EQ Cut level: 10 o'clock
(Nob 6) Para EQ: 0.36 KHz
(Nob 7) 12KHz at 7 o'clock (rolled off completely)
(Nob 8) Output: set at around 12 o'clock (set to ensure you can play hard without any high peaks on your DAW)

Instant John Paul Jones! Beautiful sound and the level is tamed by the output so no compression needed - great for taming sloppy players.

I KNOW there's something wrong with at least one of my ACMP 73's DI circuits. A few pages back you'll find my impressions of Bass DI'd through the 73. There was some discussion on how I was setting ouput gain for headroom, but if I'd set up the unit this way, I would have had a completely unusable (barely registering) signal at the output.

Bummer...not only do my 81's need fixing, now at least one of the 73's needs major work as well.

 

[nuemes]

I KNOW there's something wrong with at least one of my ACMP 73's DI circuits... if I'd set up the unit this way, I would have had a completely unusable (barely registering) signal at the output.

It's a shame to hear this. There is no excuse for differences in quality control between units like this.

I'm currently reading Will Durant's book The Age of Lious XIV and came across this yesterday; this is how they handled manufacturing defects in France in the 17th Century, you guys might get a kick out of this too:

Boards were established in all town halls to check defects in the output of local crafts and factories. Specimens of faulty workmanship were publicly exposed with the name of the worker or manager attached. If the offender repeated the offense, he was censured at a meeting of the guild; if he offended a third time he was tied to a post or public exhibition and disgrace.

 

[antichef]

I'll try those settings -- I notice you have the 12k completely gone. That region (and north from there) seem to be where most my ACMP73 trouble is, so it would make sense for me to just turn it off. There's got to be a way to make it better, though - the 73 and the 81 have the same preamp circuitry, right? My 81s have much nicer HF.

 

[crazydoc]

Well, I finally feel like I've made some progress on the noise problems with the 1081 I've been working on.

I think the most important point is that the problems with these are multifactorial. I've found three with this first unit (besides the switch pop), and dgatwood and others have found others.

1) The output connector (from the power amp board at the output gain pot) was loose (partially disconnected) - this was the easy fix.

2) There was oscillation in the eq section (see pic below). In this particular unit I finally ID'd it at 70MHz - about the limit of my scope - and about 8mv as measured at the output of board 5. This was present on all the eq boards and on the +24v rail that powered the preamp and eq sections. I don't know why this affected the audio signal, but it caused buzz at the audio output when the eq button was engaged, despite all the eq section switches being in the off position. The problem was worse when eq sections were engaged, especially the two inductive ones.

I swapped out the Q4 and Q5 transistors in board 5 with a 2N2222 and a 2N2907 (as I had these around, they have similar characteristics to the original Neve BC441 and BC461, and are a hell of a lot cheaper) and this reduced the magnitude of the oscillation by over 50%. I switched the same parts on board 4, and the oscillation disappeared completely.

Now with all the eq switches off, but the eq pushbutton switch engaged, there is no noise. However, engaging the eq of board 5, and to a lesser degree board 4, does insert some 120Hz and harmonics noise into the output, augmented by the boost control, but diminished by cut.

3) I next removed the bolt from the toroid power transformer, and moved it around in different configurations. Removing it as far as I could from the eq boards (only a few inches, unfortunately,) made a significant difference in the loudness of the noise, and twisting it this way and that changed both the intensity and the harmonic content of the noise.

So my next step will be to either remove the transformer and make it external to the preamp, or get some Mumetal shielding to shield the transformer, or the eq inductors, or both. This will hopefully give me a fully functional 1081, with only the gain switch "pop" problem, but I've seen that is being worked on also.

 

[dgatwood]

Okay, my hum is back with a vengeance all of a sudden. I seem to have blown an output cap on the supply board. The output has even more hum than before (when not attached to the boards; the caps on the actual audio boards seem to reduce this to a negligible level when measured from there). I have no idea if this is related, though. I'm starting to think the power supply fixes were mostly red herrings.

That said, I have confirmed that the hum appears to be entirely induced. I hooked cables up between two units, and even with a very limited ability to move (due to short cables), I was able to eliminate almost every trace of the hum through reorientation of the unit whose power supply I was using. I'm confident that if I had cables about three inches longer, I could completely get rid of it, but I'm not going to make custom miniature Molex cables just to prove a point.

I don't know why we're getting the amount of induction we're getting, nor do I know where it is entering the audio path. The two things that I'm pretty sure about are A. that the power supply in these units sucks in a very fundamental way, and B. that the hum is being induced into some part of the circuit due to being in close physical proximity with that massive EM field generator that the manufacturer calls a transformer....

A quick measurement of the current drain from this device (a brief inrush spike notwithstanding) shows the total load at around 10 or 12 W, according to my UPS. Given that most of this is roughly a 30V circuit, we're talking 300-350mA total. I can't imagine why they designed the supply the way they did.

I'm thinking that perhaps I should scrap the entire power supply circuit and start over. Use the 30VAC secondary to provide 48VDC with this circuit:

http://sound.westhost.com/project96.htm

and also use it to provide 24VDC through a regulator circuit (one side of the current supply circuit). Use the center tap to provide a ground reference for +/- 15VAC going into caps followed by +/-12VAC regulators.

With such a change, that huge toroidal transformer gets replaced by a single tapped 30VAC secondary. Digikey has standard transformers as cheap as $14 that would do the job.

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=237-1005-ND

Now doesn't that look a whole lot more inviting than a toroidal monster the size of your fist?

 

[mshilarious]

Now doesn't that look a whole lot more inviting than a toroidal monster the size of your fist?

I don't think it's quite that easy:

it is extermely important that the transformer is not used to power other supplies or equipment. The centre-tap must not be connected to anything, and needs to be insulated to prevent contact. The supply circuit uses the full 30V AC in a floating configuration, and connection to another supply or rectifier will cause a short on the winding.

If you use the center tap as a ground reference for the opamp supply, that ground reference is probably shared with all the other signal grounds on the board, which means that your phantom is no longer floating.

 

[wmtunate]

Did anybody ever check to see what the unregulated DC voltage going into the 24V regulators (LM317) was? Sorry if I missed it back there somewhere. According to the schematics, I'm seeing a 25VAC tap from the torroid feeding the 24V section of the PS. That would give you about 23VDC unregulated at the output of the rectifiers, which isn't enough to feed a 317 to produce a regulated 24VDC, especially under load.
If that is the case, then a torroid swap may be in order. Or adding a smaller secondary power transformer just for the 24V section.
I'd check myself, but I'm still waiting for the shipping invoice on my 81s.

 

[crazydoc]

Did anybody ever check to see what the unregulated DC voltage going into the 24V regulators (LM317) was? Sorry if I missed it back there somewhere. According to the schematics, I'm seeing a 25VAC tap from the torroid feeding the 24V section of the PS. That would give you about 23VDC unregulated at the output of the rectifiers, which isn't enough to feed a 317 to produce a regulated 24VDC, especially under load.
If that is the case, then a torroid swap may be in order. Or adding a smaller secondary power transformer just for the 24V section.
I'd check myself, but I'm still waiting for the shipping invoice on my 81s.

I didn't measure it, but if the secondary is rated at 25v RMS, that would easily give 30v no load prior to the regulator (and the load on this circuit is minimal, anyway -a few hundred ma .) And my regulated output shows 24v with no observable ripple at the highest gain setting of my scope (5mv/div.)

 

[mshilarious]

I didn't measure it, but if the secondary is rated at 25v RMS, that would easily give 30v no load prior to the regulator (and the load on this circuit is minimal, anyway -a few hundred ma .) And my regulated output shows 24v with no observable ripple at the highest gain setting of my scope (5mv/div.)

Also, since dgatwood has noticeable a bad cap on the supply board. Unsmoothed rectified DC is not going to be something with which a regulator can cope that well . . .

 

[P. Glaser]

Try this with your ACMP73:

Front single coil jazz bass w/ flatwounds, treble rolled slightly off on bass, thru ACMP73 DI

From left to right:
(Nob 1) Gain: 7 o'clock
(Nob 2) Low Rolloff: 50Hz
(Nob 3) Low Shelf Cut level: 11 o'clock
(Nob 4) Low Shelf: 35Hz
(Nob 5) Para EQ Cut level: 10 o'clock
(Nob 6) Para EQ: 0.36 KHz
(Nob 7) 12KHz at 7 o'clock (rolled off completely)
(Nob 8) Output: set at around 12 o'clock (set to ensure you can play hard without any high peaks on your DAW)

Instant John Paul Jones! Beautiful sound and the level is tamed by the output so no compression needed - great for taming sloppy players.

Thanks for this, nuemes. I'll give it a try.

 

[dgatwood]

Also, since dgatwood has noticeable a bad cap on the supply board. Unsmoothed rectified DC is not going to be something with which a regulator can cope that well . . .

The cap is post-regulator (at least in my reconfiguration). It's not one of the big supply caps. I think what I'm seeing is just switching noise from the regulator, but it's possible that I need to add some smaller caps pre-regulator to smooth out some of the high frequency ringing. It looks roughly like a high frequency sine or square wave though (haven't zoomed in to see which), so I'm guessing it is switching noise.

 

[cannabis]

The cap is post-regulator (at least in my reconfiguration). It's not one of the big supply caps. I think what I'm seeing is just switching noise from the regulator, but it's possible that I need to add some smaller caps pre-regulator to smooth out some of the high frequency ringing. It looks roughly like a high frequency sine or square wave though (haven't zoomed in to see which), so I'm guessing it is switching noise.

the 317 and 78XX regulators

are linear regulators

so there's no switching involved



i'd guess the oscillation

is generated on the eq boards

and somehow finds its way

to the power supply



have you done

the transistor swap yet?

 

[rodabod]

i'd guess the oscillation

is generated on the eq boards

and somehow finds its way

to the power supply

Very much possible.

Has anyone else checked if the loacl supply decoupling is decent? I seem to remember looking at the schematic and thinking it would probably be sufficient, but I can't see a board to look at the ground path.

Oh, and as I mentioned before, you could just run the supply with a dummy load and measure the ripple then. Logical tests will help you figure out what is going on.

Roddy

 

[crazydoc]

Did anybody ever check to see what the unregulated DC voltage going into the 24V regulators (LM317) was? Sorry if I missed it back there somewhere. According to the schematics, I'm seeing a 25VAC tap from the torroid feeding the 24V section of the PS. That would give you about 23VDC unregulated at the output of the rectifiers, which isn't enough to feed a 317 to produce a regulated 24VDC, especially under load.
If that is the case, then a torroid swap may be in order. Or adding a smaller secondary power transformer just for the 24V section.
I'd check myself, but I'm still waiting for the shipping invoice on my 81s.

The voltage is 33v give or take 0.2v at the rectifier output (regulator input - measured on C9 and C14) of each 24v supply. The ripple is a sawtooth 120 Hz of o.2v on C9 and 0.1v on C14. Output ripple is an unmeasurable <1mv on each.

Before replacing the output transistors on boards 4 and 5 (the ones with inductors), there was a 70MHz oscillation of about 2mv in the regulated output of the PS on the eq boards. Removing this oscillation quieted the output considerably, with only 120 Hz noise left when boards 4 or 5 are engaged, presuming at this point it is induced EMI from the PS toroid.

 

[dgatwood]

the 317 and 78XX regulators

are linear regulators

so there's no switching involved



i'd guess the oscillation

is generated on the eq boards

and somehow finds its way

to the power supply

I'm seeing the noise only when the supply is unloaded. The part in my reworked supply isn't a 7824. It's an NTE972. In theory, that's supposed to be an exact substitute for the 7824, but reality often differs from theory when part substitution comes into play. The 972 might be linear, but the specs don't say. All the other parts in that range explicitly are described as linear except for that one and a couple of others, so my guess is that they've updated the part specs to use a switching regulator, but I don't know that for certain. If it's not a switching regulator, then I have no idea what's going on. Maybe one of the new diodes is bad or maybe one of the big supply caps is also bad. *shrugs*

Either way, the caps on the EQ boards seem to be pretty much cleaning that up, so I doubt that further power supply fixes are going to help unless they involve moving the transformer at least a foot outside the unit or replacing it. That doesn't mean I'm not going to try, but I'm not holding my breath. I'm probably going to switch back to the original 317-based circuit, but with the upgraded diodes.

 

[mshilarious]

I'm seeing the noise only when the supply is unloaded. The part in my reworked supply isn't a 7824. It's an NTE972. In theory, that's supposed to be an exact substitute for the 7824, but reality often differs from theory when part substitution comes into play. The 972 might be linear, but the specs don't say.

Switching regulators usually either require external parts or are bigger than a TO-220 (and if they needed external parts, they would have more than three pins). I'd give a near 100% chance that's a linear regulator.

 

[stevelindsay]

73 pre and 1200 mic FS

Well, the wait's worn me out... if anyone would like to buy a 73 pre and an ACM1200 mic at the original cost please pm me. These are still with Chance, I'm in Australia... the red no's win.......

 

[redddog]

Well, the wait's worn me out... if anyone would like to buy a 73 pre and an ACM1200 mic at the original cost please pm me. These are still with Chance, I'm in Australia... the red no's win.......

Steve, we're still what appears to be a long way from these being serviceable. So in a way, none of us really have the pres yet. I'd wait it out if I were you. You really haven't missed anything.