Warren Jones Posted February 21, 2022 Posted February 21, 2022 33 minutes ago, davewantsmoore said: These two advantages are often over stated IMHO. For #1, a reduction in non linear distortion is only relevant if the distortion was already too high.... and if it was, then it's likely something else was already quite wrong. #2 is also not a given either. The electrical impedance phase may be better with reactive components inbetween the driver and the amplifier. In fairly broad terms, we should expect the filters with the same transfer functions will sound the same, no matter whether they are made from individual amplifiers per driver, and/or if there are reactive components between the amplifier(s) and the drivers. That is, assuming everything else is OK. But.... that points to the main difference. Some filter shapes are impractical to build with "passive"... or even "analogue active" filters. It really comes down to what filter shapes does your speaker need/want .... and how are you going to build them. .... or would just using the filter shapes you could make from a "typical" speaker level passive filter ..... be a big difference / detriment. ... and obviously, there's neither a universal answer to that, or a completely objective answer to it. My posts are based on education and 45years experience designing and repairing electronics. #1 An amplifier that is fed 2 signals 100Hz and 1kHz regardless of how linear will produce IMD products. The 100Hz will modulate the 1kHz and produce AM sidebands at F1+F2 and F1-F2 this is a fact. Narrow band amplifiers will still produce IMD products but they are so close to the fundamental that they are lost in the noise floor. #2 One example - Any filter is stable as long as source and load Z are stable. If a speaker is driven hard and the voice coil heats up the driver impedance will change. If the driver has a passive XO between it and the amplifier the XO frequency will also change. With the amp driving the speaker directly there will be no change in XO. I don't see how a passive component (capacitor or inductor) can improve the transfer function over having the components before the power amplifier, any impedance variation in the speaker driver due to either frequency or temperature will have no effect on the XO transfer function. Filter shapes can be an issue for passive XO as the driver Z changes a few octaves either side of the XO frequency so does the transfer function and phase. Zobel networks can mitigate this at the expense of complexity and sensitivity requiring more power to drive them. Not all filter shapes will sound the same it will depend on too many variables, component selection XO topology and how difficult the speaker is to drive. This is especially so for the low frequency XO where the components required for a speaker level XO are large and expensive. TI released a whitepaper many years ago testing capacitor distortion and the best performers with the lowest distortion were film caps, in a line level XO these would be a few $ but for speaker lever they get expensive very quickly.
davewantsmoore Posted February 21, 2022 Posted February 21, 2022 42 minutes ago, Warren Jones said: My posts are based on education and 45years experience designing and repairing electronics. #1 An amplifier that is fed 2 signals 100Hz and 1kHz regardless of how linear will produce IMD products. Yes, but this is only relevant if the IMD is too high. So, if we have a "passive crossover", with IMD which is not too high...... then any IMD reduction from using seperate amplifiers is irrelevant. 42 minutes ago, Warren Jones said: I don't see how a passive component (capacitor or inductor) can improve the transfer function over having the components before the power amplifier By having a flat impedance, it could make for less change. If the change in impedance vs temperature is significant .... then sure, a passive XO may not perform properly, but this is unusual in typical hifi. 42 minutes ago, Warren Jones said: Not all filter shapes will sound the same it will depend on.... I suspect you may have misunderstood my point. If the filter shapes are the same .... then they will sound the same (irregardless of how they are achieved). You're providing examples of where they may not actually be the same..... and yes, that is the core benefit. Getting the filter transfer function that you want to achieve. The oft cited "easier to drive as no passive components in the way, and less IMD" are both in practise not very significant. (NB - I'm about as big of a proponent of "active crossovers" as I've ever met .... so I don't mean to "defend passive filters"..... just being very realistic about what the "this will always be an improvement" type of benefits are to them). 42 minutes ago, Warren Jones said: too many variables, component selection XO topology and how difficult the speaker is to drive. This is especially so for the low frequency XO where the components required for a speaker level XO are large and expensive. TI released a whitepaper many years ago testing capacitor distortion and the best performers with the lowest distortion were film caps, in a line level XO these would be a few $ but for speaker lever they get expensive very quickly.
ENIGMA Posted February 22, 2022 Posted February 22, 2022 (edited) Well ! As a MiniDSP user i looked at this thread thinking i would find some information on potentially trying out a different power supply, haven't seen any real advice on that topic ............. Curious as to how many of you actually heard or use the MiniDSP - i see Warren has one. I would be getting on 10 years use with mine. Yes, definitely a difference if you leave it on 24/7 - i normally do this. And, Warren has correctly indicated the benefits you gain from going down the path of an active DSP set up. The most understated point is that you can hear any adjustments you make on the fly . I've had my setup measured and calibrated - it was also surprising to see that even after spending a day on this process that final adjustments were made from listening as well. That aside, whilst i also agree with davewantsmore on the passive side of things, i have to say Dave the DSP angle is far more sophisticated level and the mere fact there are individual amps driving individual drivers on discrete frequency ranges certainly impacts on dynamics. No restrictions especially in my case where i have 10 watt tube amps on mids and 380w solid state amps on bass assigned accordingly. From my perspective with an active 4 way sensitive horn system that resembles an avatgarde trio there is an improvement from using this type of DSP setup. I also experienced this in the car audio scene. Lastly, Vinyl (analog) with a digital crossover works just fine. Cheers, E. Edited February 22, 2022 by ENIGMA 1
andyr Posted February 22, 2022 Posted February 22, 2022 33 minutes ago, ENIGMA said: Well ! As a MiniDSP user i looked at this thread thinking i would find some information on potentially trying out a different power supply, haven't seen any real advice on that topic ............. Curious as to how many of you actually heard or use the MiniDSP - i see Warren has one. Perhaps you missed this in my post on the previous page? "And re. your original question: "Ideal Power Supply Option": * yes, I noticed an increase in SQ when I used a 12v Sbooster to power my miniDSP 10x10 HD, instead of the supplied Meanwell SMPS. * but having changed my miniDSP setup from the 10x10 HD to: an all-digital nanoDIGI 8x8 an A2D converter on my phono stage (input) a 4-way digital source selector and 3x E30 DACs for the output ... I now use 5 and 12v Meanwell SMPSs on all these devices - however, between each SMPS and the wall socket is an intermediary device, consisting of an Isolation transformers plus a hash filter." 35 minutes ago, ENIGMA said: Lastly, Vinyl (analogue) with a digital crossover works just fine. Cheers, E. +10! Andy
ENIGMA Posted February 23, 2022 Posted February 23, 2022 Thanks Andy, i did miss it with all the other posts.
Warren Jones Posted June 5, 2022 Posted June 5, 2022 On 23/02/2022 at 9:24 AM, ENIGMA said: Well ! As a MiniDSP user i looked at this thread thinking i would find some information on potentially trying out a different power supply, haven't seen any real advice on that topic ............. I have 4 MiniDSP's and all of them use onboard SMPS so powering the MiniDSP with an LPS is IMO not worth the effort as the LPS will be switched inside the MiniDSP. All of the I/O cap are electrolytic and unbiased so will be producing higher distortion than film caps. I only use the digital inputs on my MiniDSP 4x10HD so I only removed the output caps on the RCA outputs and added wire links to the PCB. This mod is audible and produces NO turn on/off thumps.
Recommended Posts