OK, another novice question. I've resumed stuffing my boards and have reached the diode stage. Now I know that the diodes have to be put in the right way round and I can see the black ring that marks the cathode on the diodes. On the PCB there is one square hole and one round hole for each diode. Can someone tell me which is which?
Thanks, Ian
OK, found the answer, the markings on the PCB where the diodes go have an extra line for the cathode end.
Thanks, Ian
OK, found the answer, the markings on the PCB where the diodes go have an extra line for the cathode end.
Last edited:
The square holes are the cathode end for the diodes. The short rectangles on the silk-screened part representation on the PCB also indicate the direction of the cathode.
Edit: Be careful when you install the electrolytic capacitors as the squares holes there are used to indicate the positive lead for the part.
Edit: Be careful when you install the electrolytic capacitors as the squares holes there are used to indicate the positive lead for the part.
Last edited:
Well, thanks to Tom's excellent instructions and support here I've more or less finished stuffing the boards. Mouser missed one ceramic cap off and I wrecked a resistor removing it from the wrong hole 
. Replacements should be here in the next day or two. I've not fitted the LM3886 chips yet, as advised, that will wait until I have the boards positioned in the case.
I don't suppose there are any scruffier builds around but on the plus side the solder joints are bright and there doesn't appear to be any short circuits. Of course that needs to be tested. Given my previous soldering experience was mainly with speaker cables and the odd tonearm rewire I'm quite pleased with this and have certainly learned a lot. If it works then that will be a bonus
I've managed to borrow a Variac so I'm off out to get a fuse holder and some fuses - I can at least test the power supply.
Wish me luck!
. Replacements should be here in the next day or two. I've not fitted the LM3886 chips yet, as advised, that will wait until I have the boards positioned in the case.
I don't suppose there are any scruffier builds around but on the plus side the solder joints are bright and there doesn't appear to be any short circuits. Of course that needs to be tested. Given my previous soldering experience was mainly with speaker cables and the odd tonearm rewire I'm quite pleased with this and have certainly learned a lot. If it works then that will be a bonus
I've managed to borrow a Variac so I'm off out to get a fuse holder and some fuses - I can at least test the power supply.
Wish me luck!
Echo, echo, echo.....
Ian just said he was going to use a fused variac. That's actually better than the mains bulb tester. Using the variac, dial up the voltage while monitoring the current to the amp. The amp should draw very little current, so aside from a blip on the ammeter when you change the mains voltage, you should hardly see any power consumption. If there is a short circuit, you will see the current rise proportionally to the mains voltage as the mains voltage is increased. You'll see significant power draw even at low (say 10-20 VAC) mains voltage.
Tom
That's an impressive jump in project complexity and exactitude for what you've done previously. Next thing you know you'll be doing SMD work with Tom's Maida and Filament regulators.
Yep. Nice progress. Good work.
Tom
I have been listening to my quad Mod86 with my LXminis for a few days now. I have built both projects together, each needing the other to complete the system. I have no other components to switch in for reference, so I find myself in a situation where I cannot distinguish between amp and speaker in a subjective listening test.
To complicate things, I became concerned at one point about the D-A-D-A conversions involved with my digital source and the MiniDSP's analog input, so I bought an external sound card and built the low gain version of the amp. For now, I am am doing PEQ with JRiver and settings shared in a thread at OPLUG. I have a preloaded MiniDSP 2x4 which I bought with my LXmini parts purchase and have built pseudo differential cables.
I am not always happy with what I am hearing - sometimes yes, other times no, depending on the recording. There is some debate regarding the accuracy of the JRiver DSP settings over at OPLUG, so the obvious first step is to fall back to the Linkwitz recommended build and switch in the MiniDSP. The problem is that 20 dB of gain is not enough to get the volume I need to subjectively compare the two systems. De-soldering R13 and replacing it is not a big job but I would prefer to easily switch back and forth.
What is involved with building a switch into the amp? Can I somehow build both resisrors and a jumper into each board?
To complicate things, I became concerned at one point about the D-A-D-A conversions involved with my digital source and the MiniDSP's analog input, so I bought an external sound card and built the low gain version of the amp. For now, I am am doing PEQ with JRiver and settings shared in a thread at OPLUG. I have a preloaded MiniDSP 2x4 which I bought with my LXmini parts purchase and have built pseudo differential cables.
I am not always happy with what I am hearing - sometimes yes, other times no, depending on the recording. There is some debate regarding the accuracy of the JRiver DSP settings over at OPLUG, so the obvious first step is to fall back to the Linkwitz recommended build and switch in the MiniDSP. The problem is that 20 dB of gain is not enough to get the volume I need to subjectively compare the two systems. De-soldering R13 and replacing it is not a big job but I would prefer to easily switch back and forth.
What is involved with building a switch into the amp? Can I somehow build both resisrors and a jumper into each board?
Last edited:
Assuming we're talking about Modulus-86 Rev. 2.0, it's R12 that you need to play with.
R12 = 1.05 kΩ --> 26 dB gain
R12 = 732 Ω --> 29 dB gain
R12 = 511 Ω --> 32 dB gain
I wouldn't fit a switch. That's a pretty sensitive part of the circuit.
The MiniDSP 2x4 can only deliver 900 mV RMS out, so you'll need 26 dB of gain to drive the MOD86 to clipping with ±28 V supply rails. I'd probably go with 29 dB to get a little more oomph for those high dynamic range recordings.
Lower gain is better for noise, though.
My Focusrite Saffire PRO24 outputs 2.8 V RMS and I have no trouble driving the LXmini to the excursion limits with the default Modulus-86 gain of 20 dB.
Tom
R12 = 1.05 kΩ --> 26 dB gain
R12 = 732 Ω --> 29 dB gain
R12 = 511 Ω --> 32 dB gain
I wouldn't fit a switch. That's a pretty sensitive part of the circuit.
The MiniDSP 2x4 can only deliver 900 mV RMS out, so you'll need 26 dB of gain to drive the MOD86 to clipping with ±28 V supply rails. I'd probably go with 29 dB to get a little more oomph for those high dynamic range recordings.
Lower gain is better for noise, though.
My Focusrite Saffire PRO24 outputs 2.8 V RMS and I have no trouble driving the LXmini to the excursion limits with the default Modulus-86 gain of 20 dB.
Tom
I've been playing with the possible layouts for the components inside the case.
Arranged like this I can route the signal wires along the left and power cable along the right of the chassis which seems like a good idea to me. Any reason not to do this? There's a hole for a switch in the front of the cabinet so there will be some quite long runs of power cables.
Re the switch, it has four connectors:
The centre pin of each pair has a number on it, 1 and 2. The upper pins are unmarked. There is a little hole in the chassis above the switch, presumably for an LED. For power to the PSU I presume I use one pair left or right and maybe if I use install an LED, that goes on the other pair. Have I got this right?
Cheers, Ian
P.S. I've checked continuity with the switch in on position and each pair is connected to each other but not the other side. In off position there is no continuity either side so that is presumably as expected.
Arranged like this I can route the signal wires along the left and power cable along the right of the chassis which seems like a good idea to me. Any reason not to do this? There's a hole for a switch in the front of the cabinet so there will be some quite long runs of power cables.
Re the switch, it has four connectors:
The centre pin of each pair has a number on it, 1 and 2. The upper pins are unmarked. There is a little hole in the chassis above the switch, presumably for an LED. For power to the PSU I presume I use one pair left or right and maybe if I use install an LED, that goes on the other pair. Have I got this right?
Cheers, Ian
P.S. I've checked continuity with the switch in on position and each pair is connected to each other but not the other side. In off position there is no continuity either side so that is presumably as expected.
Last edited:
Yep, rating is fine, thanks.
More new skills. Had to drill out the holes made for phono plugs to accept XLR sockets. Whoever recommended step drill bits, thanks, a surprisingly easy job as was cutting a thread for the M3 machine screws - another first!
Oh, and despite my concerns about enlarging the hole for the power socket, on proper examination I found it is already equipped with a fuse holder
More new skills. Had to drill out the holes made for phono plugs to accept XLR sockets. Whoever recommended step drill bits, thanks, a surprisingly easy job as was cutting a thread for the M3 machine screws - another first!
Oh, and despite my concerns about enlarging the hole for the power socket, on proper examination I found it is already equipped with a fuse holder
I'm confused. From the Rev. 2.0 design doc:
Higher Gain Options
The Modulus-86 is intended as a turnkey analog subsystem and will provide world class performance when assembled according to the bill-of-materials provided. A few builders using preamplifiers with lower output voltages may consider increasing the Modulus-86 gain to the THX standard 26 dB. The gain of the Modulus-86 can be increased to 26 dB by changing R13 to 40.2 k" ±1 % (Mouser P/N: 660-MF1/4DC4022F).
I have built the default 20 dB version of the amp and get all the volume I need when driven by my Saffire PR026. I want to increase the gain to 26 dB to be to replace the Saffire with the MiniDSP 2x4. I anticipate needing to go back and forth between the two configurations while comparing the output of the two configurations.
You do not have this problem having built both systems simultaneously. From what I read, you used measurements from one system to configure the parametric EQ of the other.
Just out of interest, what are the options for implementing a switch, and what are their potential problems?
As Andrew says, and to be a little more explicit, no, the other pair is not for an optional LED. A double pole switch is intended to switch both live and neutral.
As long as you keep the mains wiring tightly coupled (twisted pair is an option), you should be fine.
Re the switch, it has four connectors:
Grab your ohmmeter and measure the resistance between the terminals. As Andrew points to, you have a DPST switch. With the switch OFF, you should have near infinite resistance between all terminals. With the switch in the ON position, you should be able to find the two switch pairs. You'll have continuity (low resistance) between two pairs of terminals.
If you don't have an ohmmeter (or digital multimeter), I suggest getting one. It's a handy tool to have. However, it is also possible to make a continuity tester using an LED in series with a 1 kΩ resistor, powered by a 9 V battery.
Connect the incoming live wire to one switch pair and the incoming neutral wire to the other switch pair. The other end of the pairs go to the power transformer.
If you want to fit an LED for power-on indication, I'd use a 5.6 kΩ, 1 W power resistor in series with the LED and power it from the -28 V rail (yes, the negative rail).
The reason I'd use the negative supply rail is that the negative rail also controls the MUTE function in the LM3886. Hence, the negative rail should collapse before the positive for proper operation of the MUTE circuit.
I guess I should have read your post until the end.
Yes. Your switch works as designed.You'll get slightly better stability as well as lower noise floor if you change R12 rather than R13. Hence, I now recommend changing R12 to control the gain.
The feedback network is probably the most sensitive part of an amplifier, so introducing switches there would not be my first choice. You could build the amp with 26 dB of gain and just dial the volume down by 6 dB in the Saffire MixControl software. You lose 6 dB of dynamic range by doing so, but that is an option.
Yeah, I tend to be a "set-and-forget" kinda guy rather than a tweaker. After 10+ years as an IC designer, I've rather had my fill of tweaking...
To configure Reaper and ReaEQ to make the computer+Saffire PRO24 work as the XO for the LXmini, I started with the MiniDSP configuration and ported the values into ReaEQ. The only snag was that the shelving filters were defined in terms of centre frequency and Q in MiniDSP, whereas, ReaEQ uses centre frequency and bandwidth (or maybe I have the two reversed).
Once I had the coefficients dialled in, I verified the frequency response of the MiniDSP 2x4 versus the Saffire + ReaEQ and made sure they were line-on-line on the plot.
But as you say, I did have the luxury of building the two systems at the same time.
Tom
You have so much room in there that the distance between signal and power cables should not be a real issue (Tom pointed this out a while back). I tried to locate my transformer closer to center so that the weight is more or less balanced when the amp is picked up. You might also consider leaving open the option of adding two more channels at some later date (and using the full potential of all that heatsink).
And, although I sheepishly admit to having bought my chassis up front, I really like Andrew's approach which he has expressed a couple of times on this thread, that is, to wire the components of the amp in a line, test it, then fold it into a box, and test it again to make sure you have not degraded its performance. A kind of linear origami, or entorology
.