The Parallel-86 boards do occasionally pop up for sale in the Swap Meet. Perhaps a Want To Buy post would do the trick. I don't have any Parallel-86 boards left.
The MOD286 that I expect to have available shortly before Christmas will provide performance similar to that of the Parallel-86. It uses the LM3886. It's all SMD, though. For some that's a feature - for others a drawback.
Tom
The MOD286 that I expect to have available shortly before Christmas will provide performance similar to that of the Parallel-86. It uses the LM3886. It's all SMD, though. For some that's a feature - for others a drawback.
Tom
hello,
just one question: why don't you use a shunt regulator as supply for the opamps? wouldn't it be preferable?
best regards
just one question: why don't you use a shunt regulator as supply for the opamps? wouldn't it be preferable?
best regards
One customer has found shunts preferable subjectively I understand. Doubt they'd change any measurements.
How relevant is this breathtaking amplifier performance, accomplishing distortion figures five or six orders of magnitude below that of the loudspeaker actually making the sound?
I can't speak for anyone else, but knowing that I have an amplifier that is pretty much as good as it can get within its power envelope gives me a certain satisfaction. That is worth something both in pride of ownership and in the easily confused wet squishy thing between my ears. I personally don't care if I can hear the difference as I KNOW my reference amplifiers are clean.
But that's just me.
But that's just me.
Well, in my case, I'm driving a set of Klipschorns with it. 105dB/1W/1M. If I put my ear right up to the tweeter I can hear very very slight hiss with open inputs on the Modulus-86. More than about a foot away, nothing. I haven't tried shorting the inputs.
I wanted something that was quiet yet had punch. With those efficiency levels, the Modulus-86 more than fits the bill.
I'm super happy with it.
I wanted something that was quiet yet had punch. With those efficiency levels, the Modulus-86 more than fits the bill.
I'm super happy with it.
I do agree that some marketeers consider shunt regulators to be the holy grail. However, I base my design decisions on engineering and science rather than marketing. A shunt regulator would not offer any technical advantages over the regulators I'm currently using.
Tom
If they have, then they haven't told me about it. I haven't read about it here either.
The MOD86 is not well suited to random chip/component swapping. I strongly suggest building it according to the instructions. That said, the regulators are easy to experiment with, so as long as you know what you're doing, have at it!
I'm able to get world class performance with the components specified. I see no technical reason to make any changes.
Tom
It'll get quieter with the inputs shorted as that eliminates the noise from the input impedance.
Cool. I'm glad you're enjoying it. Thanks for sharing.
Nah. Not just you. You're in good company here. 🙂
Tom
I agree that is probable the best starting point. You always want one reference before any changes are made. Even if changes are later made just to see what happens. I quite often only make changes in a single channel first, if the playback starts to have some audible effects, it is much easier to then go in and see if the measured data differs in any way. If you hear a difference that generally relates with hearing sound image shifting, but cannot see a difference in normal measurements, that is when it becomes necessary to look into different measurements.
Sent from my iPhone using Tapatalk
Its also my understanding they weren't intending to pass on that kind of feedback to yourself.
The measured (recall you made this measurement yourself) single-digit mV RMS of noise on the opamp rails looks to me to be a technical reason to lower the supply impedance. Particularly as you were claiming (haven't looked recently) 'low noise regulators' for those op-amp supplies. Normally 'low noise' in the context of regulators means noise in the uV region, not mV.
I didn't measure mV of noise. I measured mV of load-induced ripple on the supply. That's Ohm's law acting on the output current and output impedance of the regulator.
This mV level of ripple then sees more than 120 dB of power supply rejection from the driving opamp before coming near anything that can affect the output of the amp.
The output noise of the voltage regulators is in the 10s of µV. This also is attenuated by the PSRR of the driving opamp.
I understand the desire to geek out, but there are bigger fish to fry, honestly. Maybe getting better speakers...
Tom
This mV level of ripple then sees more than 120 dB of power supply rejection from the driving opamp before coming near anything that can affect the output of the amp.
The output noise of the voltage regulators is in the 10s of µV. This also is attenuated by the PSRR of the driving opamp.
I understand the desire to geek out, but there are bigger fish to fry, honestly. Maybe getting better speakers...
Tom
Last edited:
I wonder how many of us who've built toms amps would be interested in fiddling with them on the basis of a subjective opinion of a third party, that we've never heard of, when the expectation is that the numbers remain the same?
I don't exactly see an orderly queue forming...
I don't exactly see an orderly queue forming...
I know... It was supposed to be a SECRET too! Until Abrax blew it and spilled the beans. 😀 
Tom
Tom
Tom
When I built my amps (a while ago), they were driving an 8 ohm speaker load. I'm looking at changing the speakers, and the likely candidates will be a 6 ohm load. What (if any), changes would I need to make to the current setup? I'm thinking it's just the transformer so that I get a lower rail voltage (I can't remember exactly what I went for, but the rail voltage would have been somewhere in the middle of what you'd recommend for an 8 ohm load)
If it is just the transformer, what secondary power ratings would I need? Thanks
When I built my amps (a while ago), they were driving an 8 ohm speaker load. I'm looking at changing the speakers, and the likely candidates will be a 6 ohm load. What (if any), changes would I need to make to the current setup? I'm thinking it's just the transformer so that I get a lower rail voltage (I can't remember exactly what I went for, but the rail voltage would have been somewhere in the middle of what you'd recommend for an 8 ohm load)
If it is just the transformer, what secondary power ratings would I need? Thanks
+100.
I'm going to start work on my MOD 286 this weekend, tricked out with Sussumu resistors Tom recommends in the BOM. Good enough for me!
I don't think some people here understand the crazy amount of PSRR Tom's Modulus design has, it makes all these "theoretical" PS mods a very moot point. Gilding the Lilly that is already gilded.
Best,
Anand.
I'm with Simon. My bridged Mod 86s are so good I wouldn't dream of messing with them.
I'm currently making a pair of Nelson's VFET amps ( a once in a lifetime opportunity that I couldn't pass (ha ha) up). People say that they are the best of his amps, so it will be very interesting to compare them to the Mod 86s. It's actually hard to imagine anything significantly better than the Mod 86s, but maybe those triode-like VFETs will do something magical. I suspect though that they will just be different rather than better . . . .
Further to this post, I went for 35V rails for my 8 ohm speakers. Having looked back at the Power Supply documents on your site, I reckon 30V rails are about the maximum I should go for = 90W Ps (if I got the calculations correct)
I generally recommend ±28 to ±30 V for 4 Ω operation. ±35 to ±36 V for 8 Ω. A happy medium is probably around ±32 to ±33 V for 6 Ω.
Note, however, that the amp will survive fine if you power it with ±36 V and drive a 4 Ω load. You just won't get the full rated 65 W out of the amp before the SPiKe circuit engages. Expect around 50 W under those conditions before the chip starts to show signs of overheating. That's why I recommend ±28 to ±30 V. That's the best overall compromise between output power into a variety of speakers and power dissipation in the LM3886.
Considering that you already have the amp built, I wouldn't change a thing. OK, so you won't get full output power. Big deal... If you were building from scratch, I'd follow my recommendations above.
Tom
Note, however, that the amp will survive fine if you power it with ±36 V and drive a 4 Ω load. You just won't get the full rated 65 W out of the amp before the SPiKe circuit engages. Expect around 50 W under those conditions before the chip starts to show signs of overheating. That's why I recommend ±28 to ±30 V. That's the best overall compromise between output power into a variety of speakers and power dissipation in the LM3886.
Considering that you already have the amp built, I wouldn't change a thing. OK, so you won't get full output power. Big deal... If you were building from scratch, I'd follow my recommendations above.
Tom