[ACMP investiupgradifications] All things PREAMP

[maxwall]

Boy, I need to check with Cinemag. They can do pot core inductors. They have the ability to encase them in Mu Metal cans just like the Pultec Cm150's they offer. The metal can is a option that adds a additonal $20-30 to the end price each. But may help in this case.

Looks like these inductors may have been wound poorly, very poorly.

Unfortunately my units are shelved for now until I can find time to get to them, but I want to fix the annoying gain poping and overloading first , just can't stand this issue will all my units.

 

[alexc]

Just been testing the modified ACMP81 in my main rack.

Placing it in my rack next to other units has added about 5-10dBu of noise, so my figures have become :

Line Mode, Unity Gain, Bal In/Out
IP Gain at Position 7, Output Gain 11 o'clock

EQ In, Filters Out, Mid-Lo EQ Band  On, Max + Gain -> Overall Noise Floor -85dBu, individual components < -90dBu 

EQ In, Filters Out, All EQ Bands On, Max + Gain -    > Overall Noise Floor  -72dBu, individual components < -76dBu 

That's a bash test, pretty unlikely and so, my basic worst case test.

So, here's where the shielding and externalising would  improve things, I think.

 

[alexc]

Now for the fun stuff ..

Playing my Tele Thinline into a clean, hi quality DI and into the ACMP81 Mic In with 48V on.

Sounds really good. No noise that I can hear. The EQs are pretty powerful. Its sounds great on guitar  ;D ;D
As far as guitar sounds go, I'm pretty fussy. I'd have no worries using this to record clean guitars.

Bass - fabulous. Hofner heaven  and JBass paradise  8) 8)

Acoustic Guitar jim dandy! The eq really works well on this. Crisp highs all right.

I'm going to have a go with the ACM1200 now. Start burning it up.

At last, some good soul-injecting going on

 

[okgb]

For as heavy as everyone says they are , it sounds like a steel case
so putting one on top of another makes them more noisy in a
grounding way ? or ?

does one rack space of blank make a difference or enough ?

Of course now Alex , as if you haven't worked hard enough
we want to hear your gtr playing  , lol !

 

[maxwall]

AlexC

As I am gathering
Seems the only way to fix the problems is re-design the boards, which is'nt an likely endevour for a lot of owners here. It also seems that if the boards can be re-worked the noise floor will not be eliminated like we hope for, only reduced. But this does not guarantee we're left with a good recording tool.

Am I exaggerating or is this on target with the price tag. Because the amount of effort going into making these useful equates to just making one from scratch using tried and true design techniques which would result in a better sounding preamp.

Looks to me , we've been had.

 

[crazydoc]

AlexC

As I am gathering
Seems the only way to fix the problems is re-design the boards, which is'nt an likely endevour for a lot of owners here. It also seems that if the boards can be re-worked the noise floor will not be eliminated like we hope for, only reduced. But this does not guarantee we're left with a good recording tool.

Am I exaggerating or is this on target with the price tag. Because the amount of effort going into making these useful equates to just making one from scratch using tried and true design techniques which would result in a better sounding preamp.

Looks to me , we've been had.

So far, the fixes to get me 80% of the way to a no-issues unit are easy and cheap - rotate the toroid, replace the output transistors on boards 4 and 5 (I used 2N2222's and 2N2907's), and shield the toroid and the inductors. That's an hour or two's work (mostly for cutting and forming the shields), and $20 of hardware.  My only remaining issues are residual hum (though greatly diminished) from board 5, and general circuit noise at max gain from the preamp board, and to a lesser extent from the eq boards in general.

I still have more fixes to try: replace small signal transistors with lower noise units, possible grounding issues, and of course, the input gain "pop" (though there is a workaround for this.)
So the story isn't finished yet. And though some of us are spending some considerable time and effort on this, I think in the end the sum total both in time and $ of the fixes for individual units will be minimal.

So will we be left with a "good recording tool"? That will depend on individual needs.

 

[okgb]

Some seem to have units that are totally fine , [ or are they ? ]
if that's the case then it's fixable

maybe a combination of things
bad inudctor with board layout

but it you have to replace the whole thing , hey
maybe we can get the next version on warrenty
[ not bloodly likely , how about a deal & less fixes , ha ! ]

 

[alexc]

I think that the performance is varying for different people. Some people's work well, some have grounding issues
and some have other issues.

Shielding of various kinds will certainly make a difference. As crazydoc said, it is relatively easy to do.

It's not uncommon for additional noise to be added when the unit is in a rack. I see it all the time
in my units. It can add several dBu easy.

My experiences with the unit show me that there is definately an issue with the inductor.
It resonates beautifully at frequencies at which it operates ie. round the area of 50Hz harmonics.
The problem followed the inductor in all my tests. Nothing else made a significant difference.
Shielding never showed the potential of fixing the inductor noise.

I am preparing a set of analyser screen shots showing the performance I'm getting.

I do think the unit is worth persevering with. When it's not buzzing like a bastard, its sounds great.
For me the cost so far is around 20aud for transistors and 30aud for an inductort to get it where it is now - at the minimum level of good performance.

As to being had - I'm OK with it. But then I luv DIY and I didn't buy lots of units.

If I wasn't an EE type and expected a working unit out of the box, I'd be mightily bummed.

If it had of arrived performing like it is now, I'd have had no problem at all with it.

That said, there is a reason why a real one costs 6Kaud and this one cost me 350aud - not much more than an
effects pedal.

Overall - stick with it. It's a really good sounding unit.

Yes, I think I will post some guitar noodlings thru the unit. I was really impressed with the sound.
Played thru yesterday for hours, which is a great sign that it is doing some unexpectedly nice.

 

[rodabod]

I think I owe you a few pints Alex. Thanks for continuing with this.

I'm interested in how using the Carnhill inductor helped; I just don't exactly see why. As far as I understand, these inductors are simply a coil wrapped around a core - why would the Chinese model perform worse with respect to EMI (I'm assuming the problem is inducted EMI)?

Cheers,

Roddy

 

[alexc]

No worries Roddy - I like to be part of a community and DIY is my favourite place.

I'd be lost without it.

I started in EE when I was a kid to learn how to build  a guitar amp cause I couldn't afford one on my McDonalds wage.
Then I built those dodgy kit mixer kits and stuff only to find they were pretty awful, so I gave up diy audio.
Focussed on my day job to afford commercial gear and left DIY behind.

Then a bunch of years ago, a friend told me about this place. Now this is *real* audio diy.
I rediscovered the love of EE - back to my roots, you might say.
7 units later and 15 more in the works!

But I'm still mystified by the black arts of grounding, noise and impedance matching!
And  also neat wiring!

----------------------

As far as why the inductor may be dodgy? I don't know. The other one is very similar in construction and
its board is very similar. Yet it doesn't buzz. Mainly because it operates at frequencies not near 50Hz harmonics.
On the dodgy one, when you operate it at the higher frequencies, the noise is not there.
It's that 10% care that makes 70% difference.

The Carnhill reduces it a lot. More than 20dBu on the 50Hz harmonics.
If I was inclined to go all the way, I'd get a shielded inductor from Sowter at 100aud.
But overall, I'm happy with the Carnhill for the cost.

This unit is like Eliza Doolittle. You can clean it up and it will be pretty and passably mannered..
Buut it will never be an aristocrat.

Which is fair enough for the cost.

Would I recommend it to a non-DIY friend. Nope. Glad I didn't. I'd be up for reworking theirs too!

On the upside -  As a guitar/bass/acoustic box, you cant beat it (with mods) at the price.


-------------------

What mystifies me is why I can't see any improvements when improving grounding, as others have reported.
It's clearly a problem, yet when I regrounded and upped the filter caps - nothing. I watched on the analyser as I did it.

In fact on that unit, 3 of the EQ boards mysteriously died within 5 minutes of each other, after testing for 1hr. I still haven't identified the
failure. Nothing obvious. Changed Q1, Q2 and Q3 with no fix. I don't know if it's related, but only that unit has had any failure.
And  2days of constant messing around with no issues before I regrounded them.

---------------------------

Bottom line is that with the mods, it's basically  OK * for me*.
Your mileage may vary.
Others are reporting good results with shielding and regrounding alone.

Me - I will try keep the Carnhill, maybe add some shielding.

Then I'll focus my energy on my 1081 builds.

 

[wmtunate]

Check this out:

http://homerecording.com/bbs/showpost.php?p=3092577&postcount=13

 

[jdbakker]

Check this out:

http://homerecording.com/bbs/showpost.php?p=3092577&postcount=13

Running extra ground wires to "nearby ground screws" as well as the PSU ground screw sounds like a recipe for lots and lots of ground loops.

(Hm, maybe this should go to the HR thread)

JDB.
[no ACMPre here yet, so no way to do anything other than being an armchair EE]

 

[maxwall]

jdbakker ,

I agree , ground loops galore.

These units are going to look more like a EE project than anything which is'nt all bad for the learning experience ( in favor of that) but when talented minds are sratching there heads over this , its like AlexC said not for the weak handed.

Rodabod,
Your inductor question raises some valid questions
for example, like I learned with Pultec cores, the chinese pot cores may be a cheaper recipe
than what is used with Carnhill pot core recipes. But how do you check for inductive EMI
compared to Non-inductive EMI ?

 

[rodabod]

Could it be that the Chinese inductors have too much resistance in the winding? Ie. currents induced forming greater voltage than with a low resistance coil?

Alex, could you compare with the Carnhill possibly (unless I am talking nonsense!)?

Roddy

 

[alexc]

Yes I can compare. When I fix the bad EQ boards, I'll capture and add to my big file of analyser screenshots.

There was absolutely no doubt that in my case, the inductor was  major noiser compared with the Carnhill.
It was clear as my shouts of joy when I replaced it.

I tell you, I spend half my life looking at buzzes, ground loops and so on.
In my home studio, all I did for the first 6months was measure it and check for good combinations.
It was a standing joke with my professional studio owner friend!

I guess it is pot core material, winding guage, purity of the copper (or other!) wire.

The Carnhill taps DC resistances are 1036, 840, 611, 500 ohms
The 'soul-injecter buzz-o-matic' inductor shows  455, 227, 176, 82 ohms

That's a fair difference. I reckon if you set the other one up to operate boost in the 50Hz harmonic region,
it too would buzz.

Also the pot core seemed pretty conductive and no pot core clip ground as there is on the Carnhill.
Maybe significant or not. It certainly buzzed more when you touch it.

The Carnhill is just thrown in there with 5 wires hanging around. No special construction, it's quite
small and ordinary. Yet it instantly and repeatably removed 2/3rds of the noise.

 

[alexc]

Still doing more. It shows the unit + transistor mods + Carnhill.
Rest is stock.

 

[maxwall]

AlexC,

I'm trying to get a macro view of the problem before I narrow my sights on something.

Has anyone tried removing all the different board sections out of the power ground signal buss. Then starting with the preamp board ,listen, and add the eq/filter boards in one at a time until the origin of the noise gets isolated. I not sure if this preamp has a modular capability or all boards must be in place to pass signals. I hav'nt tried this yet, just brainstorming.

If its even correct to call it a buss its that backplane that runs along the backside of all the boards.
seems to carry power supply voltages and grounds.

Is it possible that we have a floating ground bleeding over into a power ground or vice versa. Its then feeds the unit a ground loop that gets filtered thru the inductor and comes out sounding high pitched only on certain eq switch settings. Then disappears as the eq is switched to other settings. becasue we know that certain buzz hum noises are not found in a majority of eq settings. just particular ones.

zzzzz time for sleep.

 

[alexc]

It is definitely a bus. Parallel lines all the way across the unit.

It's a little difficult because the power is distributed in one direction - right to left while the signal goes in the other direction.
You could do it by jumpering the power but that's 5 rails to jumper across lots of connectors.

I'm not sure that would clarify things.

--------------------------------

What I have seen is very little contribution to noise from anywhere except the 2 mid bands. And the mid-hi doesn't contribute to
worst noise, that is 50Hz harmonics, because it is not boosting in that part of the frequency spectrum.  ie. 700Hz and below.

As far as different frequencies giving different noise - yes that is correct for the mid-lo band. The closer you select to 200-700Hz
and boost, the more pronounced the noise.

The lo band, of course does boost that region, but it has no inductor to pick up the noise.

Certainly switching frequencies causes all manner of pops, farts and squeals. No surprises there.

And I also have seen some 'twitching' in the noise floor - the spontaneous 'clicks' that have been reported.
I think that is from the power supply - didn't see it in the other unit with replaced psu but I haven't focused on it yet.

Until that mid-lo mains noise is tamed, nothing else matters.

In my rack, the unit is 3 rack spaces from  an amp below. Turning on the amp alone adds 5dBu of additional mains noise.
Rack unit 3 above, a little less. Without the case lid, you get about 2.5dB worse.

This inductor is an exceptional antenna. The Carnhill significantly less so. I expect a shielded Sowter less again.
If you then remove the toroid externally, you'll get a couple more dB. Rotating the toroid is good for 2dB.

To summarise, in order of significance the road map as I see it.


Out of Box ----> Embarrassing Crap ----> Change Transistors ---> Mostly Acceptable  sans mid-lo --->

                ----->  Change Mid-Lo Inductor  ---> Minimum Good Performer  ----> Shield mid-lo Inductor  ---->
               
                -----> Getting Semi-Pro  ------> Change and/or shield mid-hi Inductor  --->  Almost Studio Ready    ---->

                -----> externalise toroid and/or psu  ---> Sounding Good!  --->  rework grounding

Achieve full potential to become a full-fledged project-studio box with a big sound

Oh yeah - and do the input gain switch mod so as not to punch a hole in your ears or speakers. Repeatedly.

 

[okgb]

I'd think some of the random noise to be additive ?
but that star grounding the individual boards
[  cutting their ground bus traces ] shouldn't be risky
6 or 7 lines back to the input or psu ground point

 

[TomWaterman]

Alex this is superb work.

I have found some of the Chameleons pick up buzz on the MF inductor only when it is set to resonate at freqs approaching mains A.C.

Remote PSU and a new inductor sounds like a good plan - however this is really turning into a lot of work if having to do multiple units.

To be fair here, Chance claimed all pres were heard by a tech and they were good to go before leaving china, and a batch of 1081s had been pulled as they didn't pass QC specs.

I'm starting to wonder if I made an all mighty mistake getting involved here with a large order for friends... based on some internet statements which appear to be BS.

-T