Wanted to highlight a cool internal design detail of a pair of stacked PCBs that hold the circuitry for the speaker output speaker 5 way banana jacks combined with the switchable Thiele network (inductor/cap/resistor) combos that are switchable for 4 or 8 ohm output. These boards attach to the inside of the back panel.
The stacked PCBs allow the speaker out binding posts to be attached directly to large pads on the PCBs for the most direct connection possible. You can see the custom wound air-core inductors (wrapped in black shrink tube) installed on the top deck.
Here is the outside of the same PCB area:
The stacked PCBs allow the speaker out binding posts to be attached directly to large pads on the PCBs for the most direct connection possible. You can see the custom wound air-core inductors (wrapped in black shrink tube) installed on the top deck.
Here is the outside of the same PCB area:
It’s 1.5kHz XO frequency and for many of my speaker designs, I like the low DCR of iron core since it gives more punchy bass - plus compactness is important and I think it sounds great. I always use air core on midrange and tweeters.
Is there any situation in which you'd put an air inductor in the woofer filter? Or do you pretty much universally use cored inductors?
If you are building a high end speaker with a very low distortion (-70dB) driver like a Purifi, it may be something to consider. But most drivers have % levels of distortion so getting an ultra low distortion inductor won’t be noticeable. The other case is for use in high power pro audio applications where iron core inductors have distortion that is a function of the non linear memory of the signal that preceded the current signal when signal is large.
Sunrise in Carolina Beach from yesterday. I’m back home now but it was a nice break.
I have an idea for a product called the Line Noise Black Hole (LNBH). The idea is to have a line power filter to block noise sources from switch mode power supplies above 100kHz. This type of noise is ubiquitous in modern homes because we have so many things powered by SMPS (LED lighting, computers, TVs, etc) and even if we don’t have these noise sources, our neighbors do, and their line noise leaks into our house. I have developed a passive high order filter that is of order -130dB/octave above 100kHz. -130dB is about as close to a black hole as practical. I have ordered the components and will be assembling a P2P prototype to test. Here are 4 of the 10A capable inductors that will be used.
This is the predicted Bode plot:
This is the predicted Bode plot:
Hi Maty,
That’s a nice steep response and indeed works above 10kHz. How big is the inductor on the simulation that you provided?
That’s a nice steep response and indeed works above 10kHz. How big is the inductor on the simulation that you provided?
By Aleksandar (ATL Audio - before: ATL HiFi)
https://www.atlhifi.com/shop/assembled-pcbs/emi-rfi-common-mode-filter-assembled-and-tested-pcb/
I have a lot of DC and RF/EMI at home. DC Blockers (from ATL, I have the prototype -with EPCOS changed with expensive Mundorf- and others DCB from ATL) + Schaffner RF/EMI FN9244B (B: medical) inlet filters + Würth 150 kHz ferrites are my friends.
https://www.atlhifi.com/shop/assembled-pcbs/emi-rfi-common-mode-filter-assembled-and-tested-pcb/
I have a lot of DC and RF/EMI at home. DC Blockers (from ATL, I have the prototype -with EPCOS changed with expensive Mundorf- and others DCB from ATL) + Schaffner RF/EMI FN9244B (B: medical) inlet filters + Würth 150 kHz ferrites are my friends.
Last edited:
Actually, the above circuit is only a 2nd order and takes a decade to get maximum drop from 10khz to 100kHz. But a nice result if used for power lines although at 400kHz there is only -50dB attenuation.
Thanks.
Thanks.
Last edited:
It looks like a magnificent graph to me. If I had to start from scratch, this is the one I would buy along with other ATL gadgets.
In case anyone is interested, in my second audio system (computer, near field mainly) I had to add a new Schaffner filter for the Magnat MA 900 hybrid integrated that I bought in December. The first filter, 8 A, is before the power strip. Then I added a 3A one for the old amplifier. After adding a second of 3A the sound improved appreciably with the Magnat. It is assumed that a modern amplifier would have better filtering but in this case that is not the case.
8A, 3A, 3A. Remember: Spain 230 Vac / 50 Hz.
Looking for a matched pair of 12AT7 tubes to replace JJ's ECC81s.
In case anyone is interested, in my second audio system (computer, near field mainly) I had to add a new Schaffner filter for the Magnat MA 900 hybrid integrated that I bought in December. The first filter, 8 A, is before the power strip. Then I added a 3A one for the old amplifier. After adding a second of 3A the sound improved appreciably with the Magnat. It is assumed that a modern amplifier would have better filtering but in this case that is not the case.
8A, 3A, 3A. Remember: Spain 230 Vac / 50 Hz.
Looking for a matched pair of 12AT7 tubes to replace JJ's ECC81s.
Last edited:
I managed to repair a rare vintage ESS Eclipse 500A power amplifier this week. It showed up with the main power switch TRIAC shorted on and the hot power line conducting to the chassis with 150 ohms. Oh boy. The ground wire from the power cord was physically connected to the chassis with a spade lug under the power switch, but it didn't make any electrical connection due to the paint / coating. So from the factory I suspect most of these amplifiers are the same. The blown TRIAC was leaking to the chassis making this thing a potential killer. The 5k offset trim pot in the L channel went open when I turned it. The DC protection circuit tripped on every power up, blowing the protection latch fuse, 1/8 amp. Powered up with the series 150W light bulb it would run fine until the 1000 Hz test signal got to around 40 volts peak. This thing has +/- 100V rails. I eventually figured out the speaker DC protection circuit lowpass filter cap was measuring 0.5 nF and was supposed to be 8 uF. So it was tripping on audio frequency signals not a DC output. All good now. Only burned through four of my five 1/8 amp fuses to get there.
I found this youTube video to be very helpful. Skip to the 1 hour mark to get to the repair suggestions.
This amp is a beast. 250W per channel all day long. The left channel PCB is on top, and the right channel is below it out of view.
The collection of dead parts. That's the stud mounted power TRIAC.
I found this youTube video to be very helpful. Skip to the 1 hour mark to get to the repair suggestions.
This amp is a beast. 250W per channel all day long. The left channel PCB is on top, and the right channel is below it out of view.
The collection of dead parts. That's the stud mounted power TRIAC.
Last edited:
Hi Maty,
The response curve shows the amplitude and phase. The phase is the one that dips very rapidly but the amplitude is shallow because it is a second order filter (-12dB/octave). Hence, at 400kHz it is about -50dB and at 100kHz, it is about -30dB. What I was trying to achieve was -120dB at 200kHz (8th order filter with 4 inductors).
The response curve shows the amplitude and phase. The phase is the one that dips very rapidly but the amplitude is shallow because it is a second order filter (-12dB/octave). Hence, at 400kHz it is about -50dB and at 100kHz, it is about -30dB. What I was trying to achieve was -120dB at 200kHz (8th order filter with 4 inductors).
For a second order the graph is great. You have to be careful when attenuating too much as we can lose dynamic range. The previous and old amplifier did not tolerate a third filter, however the new one does.
I tuned my two audio systems by combining Schaffner filters and Würth ferrites. More or less, a Würth ferrite is equivalent to a 1/2 Schaffner filter in terms of chain addition.
On some devices I have: 1, 1+1/2, 2, 2+1/2. Now also a 3 (Magnat). There are also several DC Blockers with BIG caps in the chain.
I tuned my two audio systems by combining Schaffner filters and Würth ferrites. More or less, a Würth ferrite is equivalent to a 1/2 Schaffner filter in terms of chain addition.
On some devices I have: 1, 1+1/2, 2, 2+1/2. Now also a 3 (Magnat). There are also several DC Blockers with BIG caps in the chain.
Production crossovers for Vanguard Scout speakers arrived. Very nicely made - solid and heavy. One version has a bipolar electrolytic just so I can see if there is a difference. But I have already made up my mid to use the full size film cap.
