I can't quite see what he's supplementing given the length and cross-section of the wires to the white caps. If those caps are intended to be HF rail decouplers then their effect is largely nullified by connecting them over long (i.e. high inductance) wires. Note these caps' values are 22uF and 6.8uF, not '22mF' and 6m8F as has been quoted - the abbreviation 'm' means 'milli'.
ack's system - end of round 1
- Thread starter ack
- Start date
I can't quite see what he's supplementing given the length and cross-section of the wires to the white caps. If those caps are intended to be HF rail decouplers then their effect is largely nullified by connecting them over long (i.e. high inductance) wires. Note these caps' values are 22uF and 6.8uF, not '22mF' and 6m8F as has been quoted - the abbreviation 'm' means 'milli'.
As an ex-designer myself I also consider those 'arguments' (they're assertions really) claptrap. The designer most certainly does not know everything and even if he did he's constrained by the marketing department which most likely sets the price-point at the very least, also quite likely the time scale available for development. Engineers like to continue to refine their designs but there does come a point where something has to reach market.. There are of course smaller companies where the engineer in charge is also the CEO but even then the CEO will be subject to the board....
View attachment 30531
I used these custom-made ones and they have exactly the same dimentions as the stock ones:
http://hificollective.co.uk/catalog/10000uf-100vdc-mlytic-electrolytic-caps.html
Check out the pics of my modded Stentors(The pics are large so will have to upload in several posts.)
Based on your comments and conviction, I just had to try these Mundorf electrolytics in my REL; but I also increased the capacitance from 2x10,000uF to 4x. They didn't fit directly into my board, but I made them fit, without much effort. I really wasn't going to modify the REL because I want to sell it and get a Magico S Sub, but then... then, I was itching. There was one issue that was bugging me about the BHC Aerovox capacitors, and that was their screw-on nature, and more than that, the washer that REL used to connect them to the board - the whole thing looked sloppy. So that's when I said the proverbial "what the hell, what have I got to lose, the sub is not worth more than $800 on the used market", and went ahead and ordered 4 of your custom Mundorfs. First of all, THANKS for pushing to get them built, nothing else would fit in there, and nothing that would enable me to double the capacity.
So, about those results. Well, either the new caps make better contact with the board, or the added capacity was truly needed, or the lower ESR/ESL helps, or all of the above... but the audible improvement is unmistakable, quite substantive and discernible. The first thing I noticed is that, at the usual volume setting I use on the REL, the sub was less audible than before... Hmm, the caps must be breaking in... Wait a day, two days; three days go by, no difference... I crank up the REL volume a notch, then two notches. With light material, I continue to hear of no sub presence and frequency interference outside its operating range... hmmm, how can this be... Then I play some heavy organ, just for the heck of it... Oh, such tight ultra-low frequencies, and overall, yes less interference by the sub in upper bass and midrange. I am scratching my head... why... Tighter control of the woofer? I still don't know. But I can easily attest to your impressions of more articulate sub-bass notes, and again, equally important, the much lower interference at higher frequencies (I cross it at 25Hz).
Then the measurements come in, and this is a tough one. My RadioShack SPL meter rolls off below 30Hz and above 10kHz, and I rely on published correction curves. Surely, at 20Hz, there is 2dB higher output, as a result of me cranking up the volume; the 50Hz measurement is unchanged; now that is interesting! Then I play some really challenging material, like the following (BTW, I oftentimes hold Dorian in higher regard than Reference Recordings):
Then the Poulenc Concerto for Organ, Timpani and Strings on Chandos; then analog - the results are the same. BTW, on the Jongen Finale, I got the most metallic sounding cymbal crashes coming out of this modified Alpha DAC yet - really impressive stuff.
So the journey cannot end here, can it... I see three electrolytics on the bass-summing board (which you don't have), that I don't like. They are 100uF/25V, and I am looking to replace them with film... except that film capacitors of that nature will be huge. Still searching the various manufacturers for various options, but any suggestions are more than welcome.
Thanks for turning me onto the REL mods!
Last edited:
Not by me, and I don't think they are. As I think some more about the much lower interference in the upper bass frequencies, I am beginning to suspect that the higher and higher-quality filtration is lowering line noise, and perhaps noise artifacts are what was really causing "leakage" into higher frequencies. But I am not sure... This is really perplexing me, because this was a welcome but unforeseen benefit.
You are very welcome! The thing to do now is to try and cross your REL at a higher frequency - say, 36 or even 41Hz - and be surprized again. That's exactly what happened in my case. Before mods - my Stentors were also crossed at 25 Hz. Now - the sweet crossing point is 41Hz.Based on your comments and conviction, I just had to try these Mundorf electrolytics in my REL; but I also increased the capacitance from 2x10,000uF to 4x. They didn't fit directly into my board, but I made them fit, without much effort. I really wasn't going to modify the REL because I want to sell it and get a Magico S Sub, but then... then, I was itching. There was one issue that was bugging me about the BHC Aerovox capacitors, and that was their screw-on nature, and more than that, the washer that REL used to connect them to the board - the whole thing looked sloppy. So that's when I said the proverbial "what the hell, what have I got to lose, the sub is not worth more than $800 on the used market", and went ahead and ordered 4 of your custom Mundorfs. First of all, THANKS for pushing to get them built, nothing else would fit in there, and nothing that would enable me to double the capacity.
So, about those results. Well, either the new caps make better contact with the board, or the added capacity was truly needed, or the lower ESR/ESL helps, or all of the above... but the audible improvement is unmistakable, quite substantive and discernible. The first thing I noticed is that, at the usual volume setting I use on the REL, the sub was less audible than before... Hmm, the caps must be breaking in... Wait a day, two days; three days go by, no difference... I crank up the REL volume a notch, then two notches. With light material, I continue to hear of no sub presence and frequency interference outside its operating range... hmmm, how can this be... Then I play some heavy organ, just for the heck of it... Oh, such tight ultra-low frequencies, and overall, yes less interference by the sub in upper bass and midrange. I am scratching my head... why... Tighter control of the woofer? I still don't know. But I can easily attest to your impressions of more articulate sub-bass notes, and again, equally important, the much lower interference at higher frequencies (I cross it at 25Hz).
Then the measurements come in, and this is a tough one. My RadioShack SPL meter rolls off below 30Hz and above 10kHz, and I rely on published correction curves. Surely, at 20Hz, there is 2dB higher output, as a result of me cranking up the volume; the 50Hz measurement is unchanged; now that is interesting! Then I play some really challenging material, like the following (BTW, I oftentimes hold Dorian in higher regard than Reference Recordings):
Then the Poulenc Concerto for Organ, Timpani and Strings on Chandos; then analog - the results are the same. BTW, on the Jongen Finale, I got the most metallic sounding cymbal crashes coming out of this modified Alpha DAC yet - really impressive stuff.
So the journey cannot end here, can it... I see three electrolytics on the bass-summing board (which you don't have), that I don't like. They are 100uF/25V, and I am looking to replace them with film... except that film capacitors of that nature will be huge. Still searching the various manufacturers for various options, but any suggestions are more than welcome.
Thanks for turning me onto the REL mods!
Interesting, so you are confirming that you also heard less interference, and the lack of it enabled you to cross at a higher point. So I am not imagining all of this... this is perplexing, how can it be... I guess the smoother the power supply, the more controlled the sub's behavior is.
Exactly my experience when I retired my 8 year old IcePower sub amps and replaced them with the current Hypex units. There seems to be less bass but actually it's just because a lot of overshoot is eliminated. Underdamped cone resonance really does go up into the midrange. Congrats Tasos
Exactly my experience when I retired my 8 year old IcePower sub amps and replaced them with the current Hypex units. There seems to be less bass but actually it's just because a lot of overshoot is eliminated. Underdamped cone resonance really does go up into the midrange. Congrats Tasos
Thanks Jack and ayreman, good to see additional confirmation of what I am hearing. So I just had to prove to myself that distortion is much lower now, and the simplest test that I could think of was to run 20Hz warble tones at full preamp gain, and sure enough, I did not get anywhere near the same amount of distortion and port-chaffing I was getting before; we are talking orders of magnitude less distortion and port noise.
Regarding underdamped resonances going up to the midrange, it is so true, and years ago when I was researching the woofers I would use, a typical example even from high end manufacturers such as SEAS was this popular one https://www.madisoundspeakerstore.c...xcel-w22ex-001-e0022-8-magnesium-cone-woofer/ which you will find in many speakers, and notice its frequency response going all the way to the midrange with a nasty peak at ~5kHz, with sizeable output even at 20kHz for a woofer:
It is wild driver behavior like this that forces complex crossovers into the designs, and is pretty ridiculous if you ask me.
steve williams
Site Founder, Site Owner, Administrator
Fun stuff Steve; and here's some more...
On the Alpha mods: I finally removed the four original Panasonic FM-series electrolytics from the power supply - which I was supplementing - and the sound is still slightly better; two of these have always been running way too hot for years, so they are probably not performing as well now. The Mundorfs are now the only power supply caps.
On the REL mods: I mentioned there are three electrolytics in the bass summation/amplification circuit that I didn't like, and these Daewoo have now been replaced with much higher quality Nichicon (10% tolerance). I have never heard such tight bass from the sub. I don't know if I am done yet, but I am probably done with REL subs going forward.
On the Alpha mods: I finally removed the four original Panasonic FM-series electrolytics from the power supply - which I was supplementing - and the sound is still slightly better; two of these have always been running way too hot for years, so they are probably not performing as well now. The Mundorfs are now the only power supply caps.
On the REL mods: I mentioned there are three electrolytics in the bass summation/amplification circuit that I didn't like, and these Daewoo have now been replaced with much higher quality Nichicon (10% tolerance). I have never heard such tight bass from the sub. I don't know if I am done yet, but I am probably done with REL subs going forward.
You have hot electrolytics in your Alpha PSU? I"m curious - how do they heat them up? Are there some nearby components dissipating heat? Coz in a normal mains linear supply such as the Alpha has, its really hard to make the ripple currents through the capacitors high enough to make them warm. Do you have pictures ack?
They are inherently hot (C6 and C7 on the left, below), probably because they are rated for 16V and 12V is applied through them, and I am sure there is quite a bit of current going through them. For years I was running an external fan over the Alpha, now it was time to act. I also decided to "surface-mount" the wires w/o a pin or other means, because the holes are only 1mm in diameter.
Here are the pictures before cleaning and applying electrical tape
Capacitors don't get hot from DC voltage across them unless that voltage is beyond their rating.
The transformer to the left shows two windings, one 10V and one 20V. The 10V winding will give a peak voltage of 14V on load so the 3900uF/16V is within its rating unless the loading on the trafo is very light (which makes the voltage rise). One possibility is that C7 heats up due to the close proximity of the regulator to its right in the pic. Seems its butted right up against its heatsink with minimal clearance. C6 is respectably distant from its associated regulator though so this couldn't explain why C6 heats up.
Interesting to see the details, thanks!
Incidentally I see no snubbers fitted across the many rectifier diodes, adding them (simple RC networks to suppress switch-off noise from those diodes) could be a fairly simple and straightforward mod which pays dividends in noise reduction.
C7 heats up quite quickly, so it's unlikely the adjacent transistor, which does not heat up as much - the top plate of both C6 and C7 have always been too hot. Also, one of the transformers is for the analog circuit (which has two sets of diodes as you can see on the left, and the two capacitors), the other for the digital (whose capacitors never heated up), and those capacitors are on the right. So I would say there are effectively three power supplies.
I'll look into that, thanks.
That's very interesting - I've never come across hot capacitors before in all my long experience with mains frequency rectified supplies. Which inclines me to think maybe the capacitors themselves aren't pukka ones. Could be that although the 3900uF is marked 16V it isn't really, its something lower. But if it was over-voltage then you'd see the metal top of the can start to bow outwards as internal pressure builds up. No evidence of that right? If the capacitor really is heating up purely due to ripple current that current is going to have to be very large. Let's take a look at a spec sheet for similar capacitors to yours - I'll get back to this post when I've done that.
That would be normal in such a product yes. But strangely the supplies are not symmetrical - one is double the other. Usually analog supplies are symmetric - equal in positive and negative. So the plot thickens....
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Co-Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |