+1
It looks like Potterma has a 1/4" headphone jack connected. Are you using the MOD86 as a headphone amp by any chance? Just curious.
I've had a few questions about using the MOD86 as a headphone amp. I find it to be rather overkill for that application, but hey. Each to his own. Some headphone manufacturers recommend 1-5 W for their headphones and should anybody need to use their MOD86 for that, I'll be glad to help.
Just beware that the MOD86 has enough power to catch your headphones on fire. Just saying...
Tom
I know this has been asked/posted before, but it's hard to find the exact post...
What is the acceptable power supply voltage range for the Mod-86?
IIRC, somewhere along the way Tom reported success with the ConnexElectronic SMPS300RE at +/-30V on his Mod-86. I was actually thinking of using the +/-24V version. Are there any downsides to using +/-24V, except for the obvious lowered power ceiling?
This would be for a primarily nearfield application. I sit about 1m away from my 85ish dB efficient speakers, and generally prefer moderate levels (typically less than 75 dB according to a free smartphone SPL meter).
So quick back-of-the-envelope math says I'm generally asking for less than 1 watt per channel. 20 watts/channel is probably more than enough for the rare occasion when I turn it up and still have headroom for dynamic range.
I'm leaning towards the Connex SMPS for the practical reasons of smaller physical size and low weight (as this will sit on my desk). I also have good experience with these power supplies in other builds. Lower power also should reduce thermal management requirements as well, which would also be a pro in my application.
Am I overlooking anything? Thanks!
What is the acceptable power supply voltage range for the Mod-86?
IIRC, somewhere along the way Tom reported success with the ConnexElectronic SMPS300RE at +/-30V on his Mod-86. I was actually thinking of using the +/-24V version. Are there any downsides to using +/-24V, except for the obvious lowered power ceiling?
This would be for a primarily nearfield application. I sit about 1m away from my 85ish dB efficient speakers, and generally prefer moderate levels (typically less than 75 dB according to a free smartphone SPL meter).
So quick back-of-the-envelope math says I'm generally asking for less than 1 watt per channel. 20 watts/channel is probably more than enough for the rare occasion when I turn it up and still have headroom for dynamic range.
I'm leaning towards the Connex SMPS for the practical reasons of smaller physical size and low weight (as this will sit on my desk). I also have good experience with these power supplies in other builds. Lower power also should reduce thermal management requirements as well, which would also be a pro in my application.
Am I overlooking anything? Thanks!
I recommend ±28 V for operation with 4 Ω loads. ±30 V would be OK as well, assuming a few volt drop across the LM3886. This limit is determined by the min guaranteed output current of the LM3886 (7 A; 7 A into 4 Ω -> 28 V).
Correct.
No harm at all. You'll get lower output power, but you knew that. You'll also get better reliability as everything will operate at lower temperatures.
Yay. Someone did the math.
I fully appreciate that and agree with your conclusion.The only thing to be leery of with the Connex supplies is that they're specified in "music power", i.e. their 300 W supply will supply enough power to power a 300 W class D amp, assuming a crest factor of 2.5. This means their "300 W" supply is really only a 100 W supply if you use a resistive load. I found this out the hard way. Thankfully, the fried diodes that resulted weren't expensive and pretty easy to replace.
The 100 W the SMPS300RE is able to provide is more than enough to power a stereo MOD86 amp, though.
Tom
+1
It looks like Potterma has a 1/4" headphone jack connected. Are you using the MOD86 as a headphone amp by any chance? Just curious.
I've had a few questions about using the MOD86 as a headphone amp. I find it to be rather overkill for that application, but hey. Each to his own. Some headphone manufacturers recommend 1-5 W for their headphones and should anybody need to use their MOD86 for that, I'll be glad to help.
Just beware that the MOD86 has enough power to catch your headphones on fire. Just saying...
Tom
Sharp eyes, Tom. That is indeed a 1/4" TRS connected to the output. When the chassis makes its appearance, it will also sport a 4 pin XLR output for balanced headphone out.
Sounds silly running headphones off a speaker amp, doesn't it? Truth be told, some headphones deserve a speaker amp to reach their full potential. Notably, the HifFiMan HE-6, and other orthodynamic headphones.
To ensure I would never run into clipping on the HE-6, I targeted 25 V peak as the minimum. In previous measurements using the HE-6 driven by a Krell KAV-500, I discovered that peak voltages could exceed 20 volts! No, I don't crank 'em that loud all the time
but when you really want the Taiko Drum experience, you gotta crank it! So, a 5 W RMS power recommendation from certain manufacturers is not mere marketing hype. Its a solid recommendation.With headphones, especially the orthos, finding an amplifier that will deliver clean undistorted peak voltage becomes the dominant issue. HE-6 nominal impedance is 50 ohms, so current is much lower than when driving speakers.
I'm running ± 35 V rails into the MOD86 and have measured 31.5 V peak out into 50 ohm resistive load (pre-amp ran out of steam in SE mode). No worries about clipping here! And, with plenty of margin, I recon THD and IMD will stay at insanely low levels, even when pushing the music loud.
But what about "real" headphones, you ask? One of the other headphones I have driven with the MOD86 is a pair of AKG K7XX (Massdrop exclusive) phones. Dynamic drivers and considerably lower power requirement, but its the best I've ever heard them sound, and I had plenty of range on my passive pre-amp knob, so I'm confident that when I jump to balanced input I'll still have plenty of range for them.
One note on pseudo-differential cables: Love 'em! My DAC has balanced out, but my pre-amp is SE only. I wasn't sure that running pseudo-differential from the DAC to the pre would be of much help, but it really is. I have a small refrigerator in my office, on the same circuit as the power line conditioner all the music equipment is plugged into. Whenever the compressor kicks on, I'd get this "brrrrrrrrrrrrrtttttttttt" until things settled back down. So far, running the pseudo-differential cable from DAC to pre and not a peep. Very cool! Almost had to move the beer fridge! Tragedy!!!
Haven't spent a lot of time with it yet, obviously, as I just got it up and running over the weekend, but I look forward to more detailed listening.
Michael
Speaking of power supplies. I was poking around for power transformer alternatives. I normally recommend the Antek Inc. AS-2222 (2x22 VAC, 200 VA), which is easy to get in North America, but prohibitively expensive to ship to Europe. The Hammond 1182N22 (2x22 VAC, 160 VA) is a suitable candidate for a stereo MOD86 amp as well. Mouser carries Hammond, so it should be obtainable across the world.
Tom
Tom
I will be getting mine from airlink Chassis mounting toroidal transformers
They also have options for electrostatic shield (nice to have) and goss band for the paranoid.
They also have options for electrostatic shield (nice to have) and goss band for the paranoid.
Just FYI: I'm down to seven R2.0 boards, so if you want your boards to be identical to what you have already and are considering buying more, please do so now.
R2.1 will switch the LME49710 to use the TO-99 metal can footprint as TI has discontinued the DIP version. As a result, R2.1 will be a few bucks more to build. It will provide the same stellar performance as R2.0.
Tom
R2.1 will switch the LME49710 to use the TO-99 metal can footprint as TI has discontinued the DIP version. As a result, R2.1 will be a few bucks more to build. It will provide the same stellar performance as R2.0.
Tom
Comparable toroids available in US?
Checked Antek, Avel, Maryland Toroid and Plitron, and they do not have such options for shields and gauss band available (Antek has, for some smaller values), except at order minimums which are not for the hobbyist. Does anyone know of one-off toroid winders in the US with such capability to add shield and Gauss band?
Checked Antek, Avel, Maryland Toroid and Plitron, and they do not have such options for shields and gauss band available (Antek has, for some smaller values), except at order minimums which are not for the hobbyist. Does anyone know of one-off toroid winders in the US with such capability to add shield and Gauss band?
I was under the impression that they should be used only from SE to XLR, not the other way around. Reason being that if you have balanced outputs and tie - to GND there will be possibly a lot of current that flows from - to GND. Given that it is voltage out you will create a kind of short circuit.
Anyone care to verify/refute this?
Last edited:
Cool!
So you're seeing 20+ V peaks into 53 Ω headphones? That's 7.5 W peak (3.8 W RMS). InnerFidelity measured them to require just shy of 20 mW to reach 90 dB SPL. Something tells me you like loud music!
Very cool!
I knew there was a reason I designed in a differential input on the MOD86.
My experience is similar, actually. I run four channels of MOD86 driving my LXmini speakers. Fridge/furnace/laser printer/whatever turns on and not a peep.
Tom
The short answer is: It depends.
Some differential outputs are actually two single-ended circuits driving the (+) and (-) outputs of the differential link in opposite phase. If you short the (+) to ground, you may end up grounding the input signal for the inverter that generates the (-) signal. In this situation, you end up with no output.
The other scenario is that you ground the (-) output, which causes the inverter driving the (-) output to load the (+) output with a non-linear load. This causes a degradation in THD on the (+) output, so you get crappy performance.
THAT Corp. solved this with their OutSmarts™ output topology used in the THAT1646. That's one of the reasons I use this chip in my DIFF PRE 8x2 differential preamp. The THAT1646 and THAT1606 are the only differential line driver circuits, that I'm aware of, which will handle grounding of either (+) or (-) of the differential pair without causing a degradation in performance.
To summarize: Whether you can go from XLR (differential) to RCA (single-ended) depends on the line driver circuit. You can always go from RCA to XLR. In some cases you get performance degradation. In others, like the THAT1200, you don't. That's one of the reasons I use the THAT1200 in the MOD86.
Tom
Really? Antek's AS-xxxx series has the electrostatic shield. The purple wire connects to that. They're also surrounded by some rigid metal band, which I assumed was a Gauss band (aka belly band).
This is from the write-up on the AS-2222:
The AN-xxxx series does not have shields or bands.
Tom
As per Bill's previous recommendation I am using Airlink transformer with electrostatic shield. I have no way of knowing if this is better or worse than other transformers but amp is totally silent - when not playing music of course.
Service is excellent.
Pardon my flip-flopping here. I received an offer I could not refuse. There will be another last run of the Parallel-86 boards. This will be the ultimate run. The last run. Really. I mean it this time... As you know TI has discontinued the LM4780 and the LME49710NA, so the Parallel-86 is sunsetting as well.
I will get about a dozen Parallel-86 boards made. Nine of them will be sold in a combo with the LM4780 and a piece of SilPad 1500ST. The rest will be sold as bare boards. Once they're gone, they're gone.
Yeah, yeah. I said that last time too and look what happened! Don't count on another board order. You want boards, please order boards.
I have opened up for backorders on my website. I don't have inventory management on backorders, so I hope I can manage the stampede manually. If you'd like to reserve your boards now, please place a backorder through my website.
Thanks,
Tom
As you know TI has discontinued the LM4780 and the LME49710NA, so the Parallel-86 is sunsetting as well.I will get about a dozen Parallel-86 boards made. Nine of them will be sold in a combo with the LM4780 and a piece of SilPad 1500ST. The rest will be sold as bare boards. Once they're gone, they're gone.
Yeah, yeah. I said that last time too and look what happened! Don't count on another board order. You want boards, please order boards.
I have opened up for backorders on my website. I don't have inventory management on backorders, so I hope I can manage the stampede manually. If you'd like to reserve your boards now, please place a backorder through my website.
Thanks,
Tom
Years ago I was listening to Sony MDR-7506 headphones plugged into a 70s receiver (where the standard headphone jack is connected to the power amplifier's output through 220 ohm resistors. Listening to a scratchy LP, the clicks I heard through speakers became cowbells in the phones! These obviously have a really bad resonance when driven with high impedance, and I could still hear it when driven with as little as 8 ohms. So some phones really need to be driven with high-damping-factor low-output-amplifier impedance levels to have their "flat" frequency response....
I've had a few questions about using the MOD86 as a headphone amp. I find it to be rather overkill for that application, but hey. Each to his own. Some headphone manufacturers recommend 1-5 W for their headphones and should anybody need to use their MOD86 for that, I'll be glad to help.
Just beware that the MOD86 has enough power to catch your headphones on fire. Just saying...
Tom
As far as directly driving headphones from a power amplifier, I find that a scary idea, not just because of the phones catching fire (there's a story about that allegedly happening to a drummer in "The Mixerman Chronicles,"). At that point, the volume level surely extrapolates to way beyond 120dBSPL - many phones can actually go that high, and hearing damage is certainly possible. I've had the idea of driving headphones through a power amplifier and a low-value resistive divider (say, 8 ohms and 0.5 ohms) to limit the power so that worst case the max SPL is something "reasonable."
Cool!
So you're seeing 20+ V peaks into 53 Ω headphones? That's 7.5 W peak (3.8 W RMS). InnerFidelity measured them to require just shy of 20 mW to reach 90 dB SPL. Something tells me you like loud music!
Tom
Well, not all that loud, I'd say. A "fun" listening level for Dire Straits "Money for Nothing" measures 86 or 87 dB (C weighted, 1 kHz sine wave). I get just under 1 Vrms, so with my 60 ohm set of drivers, that's about 16 mW. So, the majority of the time I'm nowhere near the 20 V peak kind of numbers.
Peak power on "Money for Nothing" stay below 50 mW.
The > 20 V peak requirement for building the amp was really for those rare occasions when Taiko drums or Chinese drums just ache to be turned up (Yim Hok-Man "Master of Chinese Percussion" for example).
Sure, things can, and do, go wrong driving headphones directly off of speaker amps, but there is a cadre of folks who swear by it, especially for ortho's. Headphone forums (or at least one in particular) has a thread dedicated to such lunacy
I've been driving HE-6, Yamaha HP-1 with speaker amps of 125 WPC into 8 ohms for a few years. Careful attention to the volume control is crucial, however! Don't want to let the magic smoke out! One also has to be mindful of the connections involved. Some speaker amps do not handle driving a single ended configuration at all. Others, the MOD86 among them, are perfectly fine driving a single ended load.