[ACMP investiupgradifications] All things PREAMP

[alexc]

>Hey alex did you say removing the torriod still helps a little with the carnhill inductors ?

Haven't explicitly tested that. I think it would make a small difference in the 1KHz-6KHz
region perhaps 5dB or so. But I haven't tested that combination.
My tests were with ACMP81 Q4/Q5 mods and (3 feet away, DC connection) externally mounted Altronics toroid and JLM PSU.

In my case, I saw the overall noise benefit of replacement PSU and toroid onboard or offboard
as not being significant unless the major noiser inductors was elseways tamed. This test was done in the rack with
other units on 3RU above and 3RU below.

It was unequivocal for me - no mains noise benefit at all. Some mid-hash noise benefit across all the filtering bands.

However - your milage may vary.

 

[Steve Hogan]

My latest rev of the new transformer design looks like 1.4" max height and about 3.4" to 3.5' max diameter. Making a smaller OD is certainly possible, but not with the same good regulation and extremely low flux density even at 50 Hz.  It appears to be able to fit nicely into a 73 and the 84 chassis, but the 81 box appears to have the most real estate covered with circuit boards and the least amount of available extra space for the larger tranformer. I need to get an 81 in house to work on  I will comment on the regulator connection tomorrow, when I'm not so tired.

 

[peteb]

Both my preamps (73s) are instantly blowing fuses.
Haven't had them working at all yet.
Before I plugged both of them in I set em to 230v (For Australia) , tried 100mA then 125mA.
Any advice appreciated.
What's the max fuse I could try?
Cheers,
peteb

 

[Mylithra]

On the '73, I measured the current draw (with an RMS meter so I dont know the peak value) off the primary tap to be only .064A. The original fuse is a 250ma fuse so I think you should start there.
At this point, If I were you, Id take my meter to the power transformer and start looking for shorts.

Winding DCR
15v 14.2R , 7.5R to ct
26v 6R
48v 52.1R
115v 47.5R@115 188.1R@240

here are the DC resistences on the windings from the Transformer (taken from the one that had the least noise with an RMS meter). Id make sure you're at least in the ball park. (or cricket park as the case may be  ;D)

Hope it helps.

 

[Steve Hogan]

Start by measuring the DCR between the L and N blades on the EIC301 Mains Inlet with the transformer switched to 230.  It should read about 188 Ohms.  If so, the transformer primaries are series connected as they should be.  Intitial inrush current drawn by a toroidal transformer can be very high, depending on where the flux in the core was when it turned off and where the AC waveform is when it turns on.  Worst case surge is limited only by the DCR of the primary winding due to the complete saturation of the toroidal core during intial turn on.  In this case: 240V x 1.414 = 340 Peak Volts/188 Ohms = 1.8 Amps for first several cycles, which are 10 msec each instead of 8.3 msec per cycle since you are 50 Hz instead of 60 Hz.  If the small fuses you are using are fast-acting types, they will blow for sure. Slow blow is required Littlefuse 313 series or equivalent in whatever size fuse is being used in the AC inlet.  If you have access to a Variac you could try ramping up slowly to make sure that there are no transformer shorts or mis-wires, but it seems unlikely that 2 units would have the same problem. One other possibility is that the primaries may be wired in reverse polarity.  That would put a 188 Ohm short across the 230VAC line at all times due to the reversed polarity cancelling out the primary inductance of the transformer leaving only the 188 Ohm resistance at all times, not just at initial turn on.  If the line voltage selector switch were correctly wired, then the 115 volt selection would also blow fuses if the primaries were out of polarity with each other.  The DCR measurement won't catch that miswire.

 

[gyraf]

Remember to use a "T" type (slow blow) fuse. A "F" will blow instantly.

Jakob E.

 

[okgb]

Hey Jakob ,  that first picture that appeared of the 81 ,
It was actually taken by you at a trade show ?
Do you remember when ?

 

[Mylithra]

There was a spare original fuse in the fuse holder.. Did you go through that one as well?

 

[Steve Hogan]

2) Transformer dimensions (to 1/16 inch): diameter 2 1/2 inches, height 1 inch

3) The box's inside depth is a tad over 1 5/8 inches

4) The transformer is a doughnut, mounted with rubber washers above and below, and a steel disc above.

5) There is a dimple into the bottom of the box for the mounting hole, so this might make remounting the transformer in a different location slightly problematic. There is 1/4 inch of space between the transformer and the wires coming off the mains connector (if you bend them a little,) so the transformer could be 3 inches diameter in its current location.

There's not enough room in the 81 for that size, without remounting it using a different hole location - then you have to deal with the raised dimple from the old location.

Presuming that the 81,73 and 84 chassis are all identical, I have just made up a 3.5" OD "paper doll" to scale of my new transformer with max OD of 3.5".  It appears to me that it will fit over the original dimple just fine.  The new transformer's center hole is large enough to move it forward in the box to clear the line voltage switch and the AC inlet faston connectors.  It also clears the 3 unused circuit board standoffs in that area -- 1 at about 5 o'clock, 1 at 8 o'clock and 1 at about 10 o'clock.  Perhaps in your effort to eliminate the hum induced by the toroidal power transformer you loosened the securing bolt and moved it as far back in the chassis (away from affected EQ boards) as possible.  There is no question that one will have to carefully dress the new soldered connections to the Voltage selector switch.  The Old switch is wired with wires and heat shrink sticking straight out from the switch.  The switch terminals themselves, however, as far as I can tell from photos, are short enough to clear a 3.5" OD transformer.  One needs to wire the switch from the top and bottom instead of straight out towards the transformer.

I have now read an additional 10 pages of posts in this thread and I am up to speed on the Q4, Q5 transistor problem, the fact that the gain switch omitted the "OFF" position that Neve implemented to prevent positive feedback in the first amp and the discussions about remoting the toroidal power tranformer to get it away from the sensitive circuitry.  I also read about the power tranformer passing high frequency garbage from the line, so I will comment on that.

Because of the way a toroidal transformer is wound with layers of primary wire wrapped around the core and then the secondary wires wrapped around the core on top of the primary, there is excellent magnetic coupling between primary and secondary due to the windings being long and flat. This gives toroidal transformers low leakage inductance compared to conventional EI core transformers.  This construction also results in high capacitive coupling between primary and secondary. Thus high frequency noise on the power line (from switching power supplies, computers, etc.) can more easily get into the secondary.  This capacitive coupling can be minimized by careful arrangement of the primaries making the "groundy" neutral end of the primary winding end up on top, closest to the secondaries.  For best eliminination of this problem, however, a grounded copper foil shield can be installed between the primary and secondary. This essentially eliminates the capacitive coupling of high frequency garbage from the line into the secondaries. This type of shielded power transformer is used on a much bigger scale for "balanced power" installations.
The original Chinese transformer does not have a faraday shield, but I have designed one into my proposed replacement.  The shield is a separate issue from the low-flux density, no-hum magnetic design, but together they make a killer solution for these boxes IMHO.

Someone asked if anybody has a gaussmeter in one of the posts.  I do.  I have a 73 unit on the way to me.  I have a tranformer on the way to me.  I guess I really need to get someone who wants the new transformer in his/her 81 to send me that box so I can have it as a test bed and can measure everything in place for fit and electronic function.

A big part of the problem with just remoting the tranformer is that it radiates enough magnetic field to induce hum into other gear.  IMHO a much better solution is to attack the problem at its source by replacing the transformer inside the box with one that doesn't cause problems. That eliminates the offending radiator of stray magnetic field from the picture entirely, and it can't infect its own circuitry or the circuitry of the other gear racked on top and bottom of the preamp.

 

[crazydoc]

There was a spare original fuse in the fuse holder.. Did you go through that one as well?

And they are marked "T 250ma"

 

[crazydoc]

Presuming that the 81,73 and 84 chassis are all identical, I have just made up a 3.5" OD "paper doll" to scale of my new transformer with max OD of 3.5".  It appears to me that it will fit over the original dimple just fine.  The new transformer's center hole is large enough to move it forward in the box to clear the line voltage switch and the AC inlet faston connectors.  It also clears the 3 unused circuit board standoffs in that area -- 1 at about 5 o'clock, 1 at 8 o'clock and 1 at about 10 o'clock.  Perhaps in your effort to eliminate the hum induced by the toroidal power transformer you loosened the securing bolt and moved it as far back in the chassis (away from affected EQ boards) as possible.  There is no question that one will have to carefully dress the new soldered connections to the Voltage selector switch.  The Old switch is wired with wires and heat shrink sticking straight out from the switch.  The switch terminals themselves, however, as far as I can tell from photos, are short enough to clear a 3.5" OD transformer.  One needs to wire the switch from the top and bottom instead of straight out towards the transformer.

Thanks Steve - you are right. I just measured the hole location in the 73 and the 81 - identical, so I presume the 84 uses the same chassis also. And I hadn't taken in to account re-connecteing the voltage selector switch wires with a different orientation - it will certainly fit with millimeters to spare, especially if the new transformer can be moved forward somewhat.

Were you ever able to get ahold of Chance to get some units to work on? I'd be glad to send you one of my 81's (from Northern CA.)

Here's a pic set at a radius of 1 3/4".

 

[Steve Hogan]

Thanks Steve - you are right. I just measured the hole location in the 73 and the 81 - identical, so I presume the 84 uses the same chassis also. And I hadn't taken in to account re-connecting the voltage selector switch wires with a different orientation - it will certainly fit with millimeters to spare, especially if the new transformer can be moved forward somewhat.

Were you ever able to get ahold of Chance to get some units to work on? I'd be glad to send you one of my 81's (from Northern CA.)

I did reach Chance by phone.  He said that he was concentrating all his efforts on getting the last units shipped out, and that he would then be able to focus more attention on fixing the various issues as soon as that was done.

At this point in time, however, I believe it is best to just get examples of the three different units from owners who want the transformer upgrade and use those units to prove out the design, etc. Therefore, I would be delighted if you can send me one of your 81 units since I already have a 73 and an 84 in transit to me. It makes the most sense to send me a stock unit, without the shielding that you have added to your experimental unit.  That way I can correctly judge the impact made by simply changing the power transformer.

With the clearance issues resolved, the final specs for the replacement transformer(s) will be:
ID = 1.25" approx.
OD = 3.5" Max typically 3.4"
Height = 1.5" Max typically 1.4"
Faraday shield with separate lead
Ultra-low flux density design
Nominal Primary Voltage = 120/240 Vac
Min. Voltage to maintain regulation = 105/210 Vac
Max Voltage = 132/264 Vac

  I have designed the new transformer for 11 kGauss at 50 Hz and 9.2 kGauss at 60 Hz.  My regular low-noise audio power tranformers are typically 12 kGauss at 50 Hz, so these are exceptionally low stray field.  The winding resistances of my transformer design are roughly half of the original, so the regulation will be twice as good.  I am trying to set the secondary output voltage so that all the supplies will stay in regulation until 105VAC line and will not overvoltage the filter caps until 132 VAC line.  I also want to put enough volts on the phantom winding to be able to raise the regulator to 52 volts instead of 48 which will better power modern mics.

Because the transformer has less voltage droop than the old one, it will be necessary to up the voltage of the main filter cap for the Phantom supply to 100 V instead of 63.  The pre-regulator Filters in the 24V supplies should be 50 Volts instead of 35V to prevent over-voltaging the filter caps at high line voltages.  With the actual units in house, I can make adjustments to the secondaries to get it right.

 

[redddog]

So you seem to be pretty far along with this. Could you give an estimation of what we can expect in terms of cost for this mofo?

Thanks for all the work. Pretty excitin'.

 

[peteb]

There was a spare original fuse in the fuse holder.. Did you go through that one as well?

And they are marked "T 250ma"

Thanks,
Crazydoc, Mylithera, Gyraf, Steve Hogan and anyone else I missed,
It was as you guys said. I put a couple of 100mA Slow Blow fuses in and they're running well.
Hopefully I'm a better engineer that a tech.
Well as you might have guessed I'm not a tech at all.
I didn't use the original 250mA ones because I was on 240v.
Was following instructions to the letter. Except for the Slow Blow bit that is.
Interestingly one 73 has the loud input position noise and the other is very much quieter through the same position.
They both have the eq buzz but it's not in an area that I use very much so I can live with that.
I like the sound of these pres.
Thanks again for a great response to my query.
Cheers,
peteb

 

[gyraf]

Hey Jakob ,  that first picture that appeared of the 81 ,
It was actually taken by you at a trade show ?
Do you remember when ?

It was taken at the "797 Audio" booth at Frankfurt Music Show - 13. March 2008 at 13:51 CET

Jakob E.

 

[pohaku]

Steve,

I have a stock unmodded 81 I can send you.  Where should it go?

Great work BTW.

Thanks!

Henri

 

[okgb]

Sounds like the power xfmr will be a good general use
stocking kind of item .
As you've done a few of these you must have an  approx idea on cost ?
what would it be for this unit ?

And now that i try to use mine , the transister fixed the overall
eq hum / buzz so now you can tell the hum that is contributed
from those bands [ unusable ]

thanks & regards Greg

 

[Steve Hogan]

Sounds like the power xfmr will be a good general use
stocking kind of item .
As you've done a few of these you must have an  approx idea on cost ?
what would it be for this unit ?

thanks & regards Greg

At the present time I think that the transformer cost (as a separate part) will be in the $75 ballpark.

Once the preamps arrive, and I get a chance to get my scope probes on them, and get a chance to look at the PCB and cable layouts, I will have a lot better sense of both the "must fixes" and the optional improvements that can be made to really polish these preamps.

 

[Steve Hogan]

I personally appreciate the time your giving to this preamp fiasco. I hoping for the best.
Although something tells me this is not a typical DIY project. Its more of a engineers
puzzle than a simple circuit fix. But educating, nonetheless.

Thanks for the kind words. Since I upgrade and modify pro audio gear for a living, and end up fixing many problems that were originally designed into the gear, my approach is hopefully very scientific.  The transistor swap problem, for example is one that I want to see for myself.  It is true that substituting different transistors may have stopped the apparent oscillation, but I want to see the PCB layout, and figure out what is REALLY going on.  The power dissipation in these Q4, Q5 transistors isn't very high (unless it is oscillating), so a 2 Amp rated transistor isn't really necessary here. Q5 is just a current sink load for Q4 and it is set to 8.5 mA.  Q4 should only dissipate about 125 mW when it's working right.  All that to say that the large, metal can BC441 and BC461 are perhaps overkill to resolve this oscillation problem.  I have spent considerable time researching beefy transistors that have similar capacitance, gain-bandwith product and Hfe as the BC441/BC461, but will fit easily into the existing PCB.  I am not inclined to go for the TO-126 package as a solution (mechanical nightmare) when there are many appropriate TO-92 and TO-92L (tall package for audio driver) transistors that may work perfectly, and be a very close mimic to the Neve BC441/BC461 parts, which are hard to get and unnecessarily expensive.  I am seeking the very best, if you will, engineered solution.   I will let you know what I figure out as soon as I get my hands on the preamps that are in route now.

I received my first power transformer from a 73 yesterday and today I put a 79 turn test winding on it which enabled me to calculate the exact number of turns on the transformer.  Based on the number of primary turns, (1975T at 115V), and estimating the core to be 1" ID, 2" OD and 0.5" High, the 50 Hz flux density can be as high as 20 kGauss at high line, and about 19 at 115V. This is right at the limit of the core material, which is what you do when you are trying to squeeze maximum output from a tiny transformer.  Fine for an appliance, but for audio?  Compare that to 11 kGauss at 50 Hz on my new transformer.

The maximum field on the transformer is directly below and above it and it measures about 100 milliGauss at the bottom of the transformer (opposite the lead break out).  This is pretty much in line with an off the shelf design, which is very cost sensitive due to competition.

Since the highest concentration of stray flux is on the top and bottom of this transformer, I was thinking that the bottom of the steel chassis will conduct magnetic lines of force.  I am wondering if perhaps part of the problem of the inductors and/or potential loop areas in the PCB picking up hum may be due to the steel box acting as a giant pole piece to conduct the stray flux from the transformer to other areas where it causes grief.

Reducing the overall stray flux from the power transformer is a great place to start, but there may be some conduction that we may want to address as well.

More Later as the project progresses.

 

[okgb]

I swapped Q4 , Q5 for 2N2222A & 2N2907A
which were in metal cases and indeed the hum / buzz
noise went away now revealing the hum as the freq bands
on the inductor boards get close to mains freq , there also
sounds like harmonics in there as well .

It may be worth noting that there seems to be varying
degrees of these same problems with some people
I've got a 73 that mostly seems fine and another
that has a slight broadband buzz to it