• The Vendor's Bazaar forum is for commercial offers and transactions. Only unmoderated members can post here.

    diyAudio provides this forum for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members. Use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Modulus-86: Composite amplifier achieving <0.0004 % THD+N.

Jump to Latest
#2,361
AlexQS said:
@Johnny2Bad: Thanks much for that. I've learned a lot. I used to teach a Hotel Law course (among other hospitality related courses) at a College. You explain things very well Jon, and I appreciate it.

@Tom: I hope that others are like me, and will always buy from you simply because so many aspects of your business are "DoneRight". Everything from design, testing, educating customers, ordering, shipping, clear instruction, support, customer service,.. I feel like I could go on. My point is that your level of professionalism makes me "want to" do business with you, and support your efforts to make a living at helping us to enjoy music at a high level of refinement and fidelity.
Click to expand...

+1. Well said. Thanks Johnny, I certainly learned something there. And of course I'm in agreement that Tom's engineering is second to none!

Best,
Anand.
 
#2,365
The Parallel-86 (now discontinued) uses the two halves of the LM4780 (also discontinued) in parallel, thus, needs the caps to minimize the standing DC current between the two amp halves. The MOD86 and MOD286 do not, thus, don't need the caps. They're different designs. Same architecture. Different design. Different design -> different components.

Tom
 
#2,367
I found it odd that the images of Poseidon's Voice's build (link provided at the Neurochrome website for this amp board) appear to show him connecting XLR pin 1 to the amp board rather than to the chassis at the XLR jack. Surely only pins 2 and 3 should be connected to the amp board (and only a two-slot connector needed).

(Tom, note that your link to Imgur doesn't work but rather links to the Reddit content again.)
 
#2,370
I agree that Pin1 should go to enclosure.
But that is not the way Tomchr designed his amplifiers.
And the reason he gave for adopting this unconventional arrangement was to allow the RF attenuation filters to operate.

We have argued this in the past, but there seems to be no incentive to move the first stage RF attenuation to the enclosure entry point and thus keep to the Pin1 to enclosure as adpoted by everyone else.
 
#2,373
Note that RF attenuation does not get mentioned until page9 where he states
What happened is that a lot of people connected pin 1 to their internal zero volt reference (infelicitously called GND). Instead of shunting away circulating currents into the chassis,
Click to expand...
In his layout B.Putzeys does bring the filter caps onto the PCB and then connects them to the plane which in turn is connected to all six Pin1 of the IN/OUT XLRs. Shown on page12.

The interference MUST be taken to the enclosure.
 
#2,374
My assembly was an implementation of Self's low noise balanced to single-ended circuit. The XLR connectors were soldered directly to this input board. Pin 1 connected to chassis immediately at the XLR connector - the connection was integral to the selected XLR part. 100r and 100p were used to filter signals from pin 2 and 3 to chassis via pin 1 at the board (i.e. as quickly as possible).

Are you suggesting that in Tom's "assembly" the third connection we see from PCB to XLR is how PCB 0 volt / GND is connected to enclosure? I'm presuming only from looking at the photos that the connection is to XLR pin 1. It's not absolutely clear that this is the case. There's also not enough detail to show that XLR pin 1 is in turn connected to chassis.
 
Last edited:
#2,375
SGK said:
My assembly was an implementation of Self's low noise balanced to single-ended circuit. The XLR connectors were soldered directly to this input board. Pin 1 connected to chassis immediately at the XLR connector - the connection was integral to the XLR part. 100r and 100p were used to filter signals from pin 2 and 3 to chassis via pin 1 at the board (i.e. as quickly as possible).
Click to expand...
That describes how it should be done.
Pin1 to enclosure.
RF filtering to the enclosure.
Don't bring the interference in on cables or traces that will spill some emi internally.
Attenuate at the chassis if possible.
 
#2,378
AndrewT said:
No, I am confirming the conventional Pin1 to enclosure that all "switched on" professionals now use.

I am also confirming what Putzeys and Ott have said. Attenuate the interference by taking the interference current to the enclosure.
Click to expand...

Ok. I don't see any conflict between that and Tom's circuit. Unless you are saying the cable between the chassis XLR and attenuation circuit could be "too long". Or that any such cable is too long, and the circuit board must use a panel mount XLR. Since pin 1 is still connected to chassis right at the connector (in addition to the RF attenuation circuit).
 
#2,379
Per the design documentation:

Pin 1 connects directly from the XLR pin to the enclosure or chassis tab. It continues from there to the RFI filter on the board to provide a drain for the RF.

I agree that it would be better to have the RFI filter directly on the XLR connector or integrated into the connector, but I wanted the Modulus-86 to be a turnkey solution. Having to build a board and then add a bunch of "fly-by-wire" components here and there will be perceived as a half-baked idea even if it does provide better RF immunity.

Those interested in the details, please consult AES48-2005.

Thanks for the tip about the broken link. I'll fix that.

Tom
 
Last edited:
#2,380
bobakrep said:
Unless you are saying the cable between the chassis XLR and attenuation circuit could be "too long". Or that any such cable is too long, and the circuit board must use a panel mount XLR. Since pin 1 is still connected to chassis right at the connector (in addition to the RF attenuation circuit).
Click to expand...

My guess, based on Andrew's advice in other threads, is that he'd find any length of wire to be too long. Am I right so far Andrew?

There's some validity to that argument - at least in theory. Any lead length between the RFI filter and the chassis will add inductance, hence reduce the effectiveness of the filter. It also increases the risk of having that wire radiate some of the filtered RF back into the chassis.
That's basically the price of the convenience of having the filter on the board rather than having to mess with soldering individual components onto the back of an XLR connector. Nothing prevents the builder from doing exactly that - moving the parts from the board to the XLR connector - should they see fit. That's the beauty of DIY. Perspective, guys... Perspective. 🙂

In general I don't find RF to be that much of a problem in residential settings. The main interference sources are GSM cell phones and relays on inductive loads (furnace fans, refrigerators, etc.) turning into wide-band RF spark gap transmitters. The RFI filter, as implemented on the Modulus-86, completely eliminates these issues in my experience, even with a 20-30 cm shielded cable connection from the input connector on the MOD86 board to the XLR input connector within the chassis.

That said, should I choose to market the MOD86 commercially, I would make an input board with the XLR connector and RFI filter on it and attach that directly to the chassis and then run a shielded cable connection to the MOD86 board. You may also notice that my HP-1 headphone amp has the filter right by the XLR connector.

Tradeoffs, tradeoffs...

Tom
 
Last edited:
You must log in or register to reply here.
Top Bottom