Modulus-86 build thread

[henrylrjr]

I've only done work for myself over the last four years. But recently I started enjoying running the machines again, so at 75 years old, I'm ready to take on new jobs.

 

[tomchr]

Sounds like you should set up a website and make yourself available if you haven't already. Or start a thread in the Vendor's Bazaar. 🙂 Custom machining for the discerning DIYer...

If you need help or suggestions with the business end of things just toss me a line.

Tom

 

[henrylrjr]

This was my old site that I thought would be long gone. But then again "once on the web always on the web". Custom Machining and Billiard Cue Joint Protectors - home

 

[henrylrjr]

I still have some pieces of the exotic woods and preservatives if anyone is interested. The woods would make cool knobs.

 

[rsqt]

Hello Tom,


are both channels of the 49720 driven/needed if the THAT1200 is removed?


Thank you


Regards
Ralf

 

[rsqt]

hello Tom,
I want to test the Sparkos, the smaller SS3602 and the bigger SS2590PRO. Must I change the 49720 or the OPA227 or maybe both?



My modified Oppo BR-Player has the LM4562 inside and there it was a really big improvement with the Sparkos.



Thank you


Regards
Ralf

 

[tomchr]

are both channels of the 49720 driven/needed if the THAT1200 is removed?

No.

I want to test the Sparkos, the smaller SS3602 and the bigger SS2590PRO. Must I change the 49720 or the OPA227 or maybe both?

The OPA2277 has no impact on the sound unless you believe you can hear below 0.1 Hz.

Using just about any of the discrete opamps in place of the LME49720 is guaranteed to make the Modulus-86 unstable.

You would need to find a discrete opamp that has exactly the same open loop gain curve as the LME49720 to make it work. Best of luck with that... 😉

Tom

 

[rsqt]

Hello Tom,


it is only a trial, not more.


Thank you


Regards and a nice weekend
Ralf

 

[tomchr]

it is only a trial, not more.

Your time. Your money. Your risk.

By far the easiest is to just drop in a dual opamp instead of the LME49720. Then verify stability by running the amp to clipping with a sine wave and an 8 Ω load. See if the amp recovers as you turn the signal level down from clipping.

Then run a 10 kHz square wave into the amp. Aim for maybe 10-20 Vpp on the output. Load the amp with 8 Ω and add capacitance in parallel. I usually go in a logarithmic sequence: 1.0, 2.2, 4.7 from 1.0 nF to 1.0 or 2.2 uF.
Above 220 nF you'll start to see a bit of ringing. That's the resonance between the load cap and the output inductor. At the higher load caps, the ringing will persist for a period or two, but will always die down. If the amp is unstable it'll break into high-frequency oscillation.

If you don't have access to an oscilloscope I would not attempt to fiddle with the LME49720.

You're doing this entirely at your own risk. Don't come crying to me when the amp goes unstable and burns up your unubtanium beryllium tweeters. You've been warned.

Tom

 

[rsqt]

Thank you Tom.

I have several of these Sparkos. I do not need to buy them. I test it with a Fane wideband speaker with 38cm (linearised buy a DSP). This is my testspeaker, very difficult to kill.

Thanks
Ralf

 

[chrisb]

You’ll definitely need to report back on this experiment, with ‘scope photos if possible.

 

[rsqt]

Hello Tom,

what happens if the coupling cap is smaller than 2,2uF when THAT1200 is removed?
Is it a kind of highpass and the frequency rises?

Than you
Kind regards.
Ralf

 

[tomchr]

You should not use a coupling cap if you take the THAT1200 out. The LME49720 needs a DC path to ground. I suppose you could use something like 47 kΩ from pin 3 of the LME49720 to ground (which you can find on the decoupling cap just south of the IC). Then the LF pole will be set by the input cap and that 47 kΩ.

Tom

 

[mahtew]

hi there;




do you know a store where LM317lz / LM3886TF is obtainable? i was postponing my order with parts needed for assembly of modulus86... and now mousers shows availability of lm3886 not sooner than 09/2022.

i'm thinking about buying some cheap amps based on lm3886 and disassemble them for parts if nothing changes.

 

[tomchr]

There are three package versions of the LM3886:

LM3886TF/NOPB - Insulated (plastic) package. Lead-free solder on pins. - Mouser: 926-LM3886TF/NOPB.
LM3886TF - Insulated (plastic) package. Leaded solder on pins. Mouser: 926-LM3886TF.
LM3886T/NOPB - Non-insulated (metal-back) package. Lead-free solder on pins. Mouser: 926-LM3886T/NOPB.

They all have the same IC inside. The difference is only in the packaging. The LM3886T/NOPB will need a shoulder washer and a thermal washer (a TO-247 sized one usually fits) as the metal back is connected to the negative power supply. The others just need a thin coat of thermal grease.

Of these, the LM3886TF is most commonly in stock, so I usually point to this as the alternate. Mouser shows more arriving in a month.

Mouser just shipped 10 of the LM3886TF to me. I'll need a couple for my own use, but I'm happy to part with the rest.

Alternatively, B+D Enterprises has the LM3886TF/NOPB as well: B+D Enterprises - Home page

Tom

 

[tomchr]

BTW: Mouser shows 5878 in stock for the LM317: 595-LM317LCLPR.

Tom

 

[asuslover]

You should not use a coupling cap if you take the THAT1200 out. The LME49720 needs a DC path to ground. I suppose you could use something like 47 kΩ from pin 3 of the LME49720 to ground (which you can find on the decoupling cap just south of the IC). Then the LF pole will be set by the input cap and that 47 kΩ.

Tom

Post 5515 Tom....

 

[solderbynumbers]

Another MOD86 to be built!

Hello forum, hi Tom!


I'm happy to become a part of the mod86 community. As my user name suggests, I don't know much about electronics but enjoy soldering and I'm well equipped with leaded solder and collophonium, a decent solder station and a hot air station. I've successfully soldered some Atmega chips and mosfets onto smd motor controllers for rc multicopters. They all flew well and only crashed because of the piloting skills, not the soldering.


All mouser parts have arrived apart from the lm3886 which are scheduled for early September. Tom sent the boards last week so they should arrive soon. Just need a power-86, maybe a soft start unit and an enclosure. Oh and I guess some wires, connectors and "stuff", maybe a VU meter for some 70s look and feel.


Any suggestions on the internal cabling? Can I use piece of microphone cable for the signal input? Apart from the recommended wire gauges, would you recomment highly flexible cables with many fine wires or stiffer ones with fewer, thicker wires? Since there is not much mechanical stress/movement I'd go for stiffer cables but I have lots of the highly flexible around.


Thanks Tom for the quick replies to my emails! Just wondering, is everybody trusting this amp to the degree that nobody seems to bother with a speaker protection?Also, a soft start is recommended above 150-200VA. I'. using the Hammond 1182N22 Dual Secondary, Open 160VA, 44V CT@3.64A. Again, nobody seems to bother with a soft start (or only a few). What's the general view on this?


Regards
Stefan

 

[Daveh061]

hi there;

Hi Mahtew, I notice that RS has the LM3886T version in stock, they seem to stopped stocking the TF. RS Online Poland has 3886T in stock, Dave

 

[tomchr]

Post 5515 Tom....

And for everyone's benefit, here's a direct link to Post #5515.

How would I know that's what you had going?

Note point 3:

Add 47 kΩ from pin 2 to pin 1 on the input of the MOD86.

and point 5:

Connect pin 3 to pin 6 on the (now empty) THAT1200 socket.

That establishes a DC path to ground for the LME49720.

What was your issue again?

Tom