Which supply are you using?
The Modulus-86 presents a balanced load on the supply. It may be a little skewed (say 10 mA difference) until everything settles completely after startup (takes 15-20 seconds). After that, it's a solid 55 mA (give or take) per side. With a music signal or sine wave, I've always seen a symmetric current draw, so I can't think of anything in the MOD86 circuit that would cause the excessive droop you're seeing.
Are you sure the supply caps are good? If one is dried out, you'd have extreme ripple on that side.
Excessive DC offset (as in 10s of volt of offset) can cause this condition as well. That can happen if pin 1 shorts to pin 2 on the THAT1200 due to a solder bridge. In that condition, you get V- on the output.
Tom
The Modulus-86 presents a balanced load on the supply. It may be a little skewed (say 10 mA difference) until everything settles completely after startup (takes 15-20 seconds). After that, it's a solid 55 mA (give or take) per side. With a music signal or sine wave, I've always seen a symmetric current draw, so I can't think of anything in the MOD86 circuit that would cause the excessive droop you're seeing.
Are you sure the supply caps are good? If one is dried out, you'd have extreme ripple on that side.
Excessive DC offset (as in 10s of volt of offset) can cause this condition as well. That can happen if pin 1 shorts to pin 2 on the THAT1200 due to a solder bridge. In that condition, you get V- on the output.
Tom
Last edited:
You can use 120 Ω for the two input resistors. They're part of the RF input filter. In theory they'll give slightly higher noise, though, other noise sources dominate so you won't notice. Also they'll shift the RF filter pole down slightly in frequency. Again, not enough to worry about.
The only concern I have is that an imbalance between these resistors will slightly degrade the common-mode rejection ratio. I'm assuming the Z-foils are pretty tight tolerance (probably way better than ±1 %) so it should not be an issue. Mismatch between these two resistors would be my only concern for values higher than 100 Ω.
Do note that I'm getting the specified performance without resorting to fancy components. There's no need for Z-foil resistors but if you have them I don't see much harm in using them.
Tom
The only concern I have is that an imbalance between these resistors will slightly degrade the common-mode rejection ratio. I'm assuming the Z-foils are pretty tight tolerance (probably way better than ±1 %) so it should not be an issue. Mismatch between these two resistors would be my only concern for values higher than 100 Ω.
Do note that I'm getting the specified performance without resorting to fancy components. There's no need for Z-foil resistors but if you have them I don't see much harm in using them.
Tom
Hi Tom,
Thank you very much for response. VARs are 0,1% and lowest noise. From my experience I found that they do the job (less noise => better sound) when placed on directed signal path and nfb positions. I actually doubt that such low value would generate audible noise anyway, but still ... this is DIY in the end ...
I got two more questions:
1. Can I use 22k instead of 20k on R13? I have 22k VAR
2. Do I need to use dip-8 sockets? Or could solder opamps directly? Are the sockets are needed to prevents opamp overheating during soldering or for opamp rolling?
In the end I am just "stupid" user following whatever have been posted on your website. One of links there refer to Poseidons Amp. On its pictures I can find VARs almost in all places (signal path, components critical for stability)
.So Tom, your message is a bit confusing to me (maybe not only to me) from one side you say fancy components are not necessary from the other on your website I can find pictures of Modulus 86 with such as reference. Maybe you can think of alignment on that to make it consistent. I just got cognitive dissonance
Anyways Tom I value your work a lot. It is fantastic design. Thanks. Cannot wait to build it and make it run ...
Actually, I think Tom should remove my build link from his website (it won't alarm me in the least!). I can only imagine how many of the aforementioned questions he is having to answer for. As I have said before, I made a large order for several other projects through TCC and included the Modulus 86 in that order. I am a fella with a large discretionary income and spending habit! My implementation was purely based on conjecture and not rooted in science whatsoever. Remember the "datasheet" with the naked VAR's is full of marketing jargon. As an aside, the folks at TCC are very pleasant to deal with. Their resistors look pretty and they're very exacting with their matching (0.1%) which in most builds including the Modulus, is completely unnecessary.
Best,
Anand.
Best,
Anand.
Last edited:
Fair enough Anand. Now it is clear for me. Thanks. I think your link could be left on Toms website. Have very nice pictures. What you wrote could be used as quote on Toms website as the disclaimer for your build in order not to create discrepancy between what Tom states and uses as build reference.
Anyways, my implementation is also purely based on conjecture and not rooted in science whatsoever
Last edited:
1. Can I use 22k instead of 20k on R13? I have 22k VAR
Yes. You'll get slightly higher gain by doing so.
You don't need to. If you'd rather not use sockets, just delete them from your cart when you order.
Opamp "rolling" (aka random opamp substitutions) is **NOT** recommended. Best case you degrade the performance of the amp. Worst case, the amp will oscillate. Composite amplifiers are not suitable for random component substitutions.
In the end I am just "stupid" user following whatever have been posted on your website. One of links there refer to Poseidons Amp. On its pictures I can find VARs almost in all places (signal path, components critical for stability)
I've changed the text to read:
The build process was documented in great detail by one of my builders. Do note that this builder went overboard on component selection with “TCC Naked Z-foil” resistors and such. This is completely unnecessary from a technical performance perspective as the Modulus-86 will provide world class performance with the specified parts as shown in the image above. The fancy components do make for a good looking amp and impressive marketing copy, however. His slide show is available here: Poseidon’s Voice – Neurochrome Modulus-86 Rev. 2.1 Build.
Please let me know if that clears up the confusion or if you'd rather see the text removed altogether. I rather like the slide show. Thanks to Anand for making it available. If the slide show causes more confusion than explanation, it should go, though. Just let me know what you think.
You're welcome.
I'm surprised that people want to modify Tom's Modulus86 or Parallel86. They are the best best amps I've heard and I've listened to top of the line Krell, Levinson, Meridian and Linn. Tom's amps with good, source components, speakers and room consideration make for outstanding music reproduction for me.
Finally got around to ordering my boards earlier this week. Given how fast I progress, I should finish them before the election....in 2020
My next decision is on a case. I have an old Alesis RA150 which I was hoping to re-use to simplify things, but I'm not sure I can fit everything in. I do have a couple other metal boxes I could use for the xformer and/or PS. Has anyone done a 2-box Mod-86 build? If I did go that way, I assume the best idea would be to simply move the xformer external and keep the PS board close to the amps to minimize supply lead impedance - right?
The alternate enclosure would be the Dissipante 2U. For folks that have used this, did you get the additional baseplate, or just mount stuff to the bottom of the case?
Final decision would be whether to run with an existing 25V xformer that I have or back off to the recommended 22V or 20V units. For the extra ~$50 shipped it's probably not worth pushing things, particularly if I can recoup some of the cost by selling the existing one (I think it's a 3225 or equivalent).
My next decision is on a case. I have an old Alesis RA150 which I was hoping to re-use to simplify things, but I'm not sure I can fit everything in. I do have a couple other metal boxes I could use for the xformer and/or PS. Has anyone done a 2-box Mod-86 build? If I did go that way, I assume the best idea would be to simply move the xformer external and keep the PS board close to the amps to minimize supply lead impedance - right?
The alternate enclosure would be the Dissipante 2U. For folks that have used this, did you get the additional baseplate, or just mount stuff to the bottom of the case?
Final decision would be whether to run with an existing 25V xformer that I have or back off to the recommended 22V or 20V units. For the extra ~$50 shipped it's probably not worth pushing things, particularly if I can recoup some of the cost by selling the existing one (I think it's a 3225 or equivalent).
dwk123,
I used a Dissipante 2U case, sadly the 400mm depth model instead of 300mm. Everything in my dual mono build would've fit more snugly in the 300mm model, and I should've fit 100VA transformers instead of 200VA. You can see my build from the link on Tom's webpage or here. I did not use an additional baseplate but drew everything out on FPE. If possible I greatly prefer the simplicity of keeping both the transformer and PS in the same enclosure as the main amplifier to minimize supply impedance and avoid grounding issues. Besides, the PSRR of the Modulus 86 is über high.
Best,
Anand.
I used a Dissipante 2U case, sadly the 400mm depth model instead of 300mm. Everything in my dual mono build would've fit more snugly in the 300mm model, and I should've fit 100VA transformers instead of 200VA. You can see my build from the link on Tom's webpage or here. I did not use an additional baseplate but drew everything out on FPE. If possible I greatly prefer the simplicity of keeping both the transformer and PS in the same enclosure as the main amplifier to minimize supply impedance and avoid grounding issues. Besides, the PSRR of the Modulus 86 is über high.
Best,
Anand.
Last edited:
Thank you for your kinds words. I really appreciate it.
Mind if I put that quote on my website, Henry?
Tom
If you're going through the trouble of two enclosures, I suggest building mono blocks.
Putting the supply separate is about the worst you can do from a supply and ground impedance stand point. Granted the Modulus-86 has stellar PSRR, thus, is not all that sensitive to supply perturbations, but still... There's no reason to shoot yourself in the foot.
There are pictures of both approaches in this thread. Either approach is valid.
2x25 VAC will result in about ±35 V rectified. That perfect for an LM3886-based amp, such as the Modulus-86, driving an 8 Ω load. For 4 Ω, I recommend using a 2x22 VAC transformer (-> ±28-30 V rectified). The reason for this is the thermals. You won't get the full 65-68 W into 4 Ω with a ±35 V supply. The LM3886 will start to engage the SPiKe protection once the output power gets into the 55-60 W range. In all reality, the difference between running the Modulus-86 on ±35 V and on ±28-30 V is marginal, however, to meet the published specs, you need to run ±28-30 V. That's what I recommend from an engineering standpoint.
Tom