[PSP]

I am going to convert an LF55 to LF25 Platinum to drive my Orion+ tweeters (two tweeters in parallel per side, Zmin=2.5R).  I plan to rebuild the amp power supply and would like your ideas on what diodes are best suited for this application.  My line voltage is ~117VAC and I will be using 18-0-18 Plitron toroidal xfmrs.

Stealth?  Schottky???  Others? 

Many thanks,
Peter

[AKSA]

Peter,

I would go for Schottkys - they are now widely available.  You could also use the Philips Ultra Fast Soft Recovery diodes, the BYx series.

You need a minimum of 75PIV rating, at 10A per diode.  Many of these diodes are actually two diodes encapsulated in one TO220 package with three leads, with a common cathode, just like the BYV and BYQ Philips series of UFSR I use in all my amplifiers.  With three such packages, you can make a bridge, but be sure to have at least a 10A rating on EACH diode as the peaks for this supply could be high (determined pretty much for the storage cap following the rectifier).

As a general rule, 5A copes well with 4,700uF per rail, but you will need a 10A rating for 2 x 4,700uf per rail.

[andyr]

I am going to convert an LF55 to LF25 Platinum to drive my Orion+ tweeters (two tweeters in parallel per side, Zmin=2.5R).  I plan to rebuild the amp power supply and would like your ideas on what diodes are best suited for this application.  My line voltage is ~117VAC and I will be using 18-0-18 Plitron toroidal xfmrs.

Stealth?  Schottky???  Others? 

Many thanks,
Peter

Peter,

Why don't you simply use the ones you have on your LF55 PS?  In my new active system, I have an LF55 for the mids and an LF25 for the ribbons - both have exactly the same PS PCB components.  The only difference is that the power tranny has lower-voltage secondaries and there is a resistor change on the LF25 PCB.

(And with my original AKSAs, it was the same.)

Regards,

Andy

[PSP]

Andy,
Yes... that's what I had planned to do, but before I start soldering I wondered if anyone here had taken a serious listen to some of the newer diodes suitable for the LF power supply.  This is a fairly innovative group, yes?

Peter

[MikeC]

Hi Peter

I used International Rectifier Schottky's in all my amps. These are 3-pin drop-in replacements for the diodes Hugh supplied. The theoretical advantage is very fast switching and essentially zero recovery time, and thus power supply pollution. They never used to be available in voltages much over 45V, but that has rapidly changed due to the proliferation of switching power supplies where their very low forward voltage drop and thus higher efficiency is as much of an advantage as reduced EMI.

If you want, I can look up the designations of the ones I used. They are/were readily available from Digikey.

Regards

Mike

[PSP]

Mike,
If you have part numbers, that would be dandy.

Digikey is just up the road from me, so it's very convenient.

Many thanks!
Peter

[AKSA]

Hi Peter,

Try this web page:

https://ec.irf.com/v6/en/US/adirect/ir?cmd=eneNavigation&N=0+4294841628+4294871718

It has the appropriate selection, some nine as I recall.  You'll have to do some research on the packaging, but the minimum current is 15A, which is beefy.  I've chosen 100PIV.

What is interesting is that nowhere in the datasheet is switching time discussed.  I'd be looking at tns around 25nS, and this figure is available on all the Philips BYQ/V series, but NOT on the Schottkys, which surprised me.

Hope this helps,

Cheers,

Hugh

[MikeC]

Hi Peter

I checked through my old files. I used the IR 16CTQ100 in TO220 package. They were $1.54 each when I bought them (I bought 16). I did a check on the IR web site, as I didn't recognise the part numbers on the link Hugh posted, and these are now divested to Vishay. However, Digikey has the 16CTQ100PBF (lead free) from IR Vishay and they are now $2.78 each, discounted to $1.79 if you buy 10 or more. I used these in both my midrange and tweeter amps. I needed a higher current device for my bass amps (the supplied rectifiers were possibly not man enough) and used the ST Micro STPS40150CT (40A device) here. These are also still available from Digikey, but feel free to investigate the current IR devices as well.

From Hugh

What is interesting is that nowhere in the datasheet is switching time discussed.  I'd be looking at tns around 25nS, and this figure is available on all the Philips BYQ/V series, but NOT on the Schottkys, which surprised me.

See http://en.wikipedia.org/wiki/Schottky_diode

The most important difference between P-N and Schottky diode is reverse recovery time, when the diode switches from non-conducting to conducting state and vice versa. Where in a P-N diode the reverse recovery time can be in the order of hundreds of nanoseconds and less than 100ns for fast diodes, Schottky diodes do not have a recovery time, as there is nothing to recover from. The switching time is ~100 ps for the small signal diodes, and up to tens of nanoseconds for special high-capacity power diodes. With P-N junction switching, there is also a reverse recovery current, which in high-power semiconductors brings increased EMI noise. With Schottky diodes switching instantly with only slight capacitive loading, this is much less of a concern.

I can't say that I heard a difference in the AKSA's from using Schottky's as there was a period of over a year between my house being cleaned out  and my new, completely diferent system with Orions being completed. I previously had a completely different turntable, phono stage, CD transport & DAC, pre-amp and speakers, so even leaving out the time difference, too much had changed to ascertain differences in rectifiers. However, I can remember hearing the difference previously when a friend had changed from fast/soft recovery diodes to Schottkys'; there was a reduced etch to the high frequencies and a more "organic" feel to the music than previously. Seeing as I was building a largely DIY "state of the art"    system from scratch, it seemed like a good idea to include these. I still plan to upgrade the diodes in the GK-1 as well one day.

Hope this helps.

Regards

Mike

[AKSA]

Thanks Mike!!

0.1nS seems pretty good to me - that's quite an improvement on 25nS on the BYQ28E - hmmmmm,  thinking.......

I often think about your appalling theft of most of your possessions.  Certainly housebreaks happen in Oz, quite a lot in some areas, but in twenty years of living in the NE suburbs of Melbourne we've lost one small air compressor, to a building contractor I believe.

I remember being outraged at the time, but really, this is a typical record here.  Perhaps you might like to consider crossing the pond, Mike?

Cheers,

Hugh

[whubbard]

Hi Peter.

Is there any reason you aren't going to keep it a LF55, as thats what I'm going to do to drive my orion+ tweeters.

-West

[PSP]

Hi West,
Since the Orion+ specifies two tweeters per channel wired in parallel, they present a low impedance load to the LF55.  To cope with the low impedance Hugh recommended that I reduce the LF55 rail voltage, and since I already had a couple of 18-0-18 torroids on hand, we decided to use those to make an LF25.  While I had the case open I also replaced the input caps with 0.47uf Sonicap Platinums.

I biased the LF25 while driving both tweeters at high volume and things went extremely well... no audible distortion, no perceptible temperature rise on the heatsinks, and the musical result (most likely due to the teflon input caps) has been extremely positive.

Since you are building Orions and will use the GK-1 (if I remember right), please take a look at my earlier post regarding increasing the input impedance of the Orion ASP (from 10k2 to 36k) so that the GK-1 has an easier time driving the load (  http://www.audiocircle.com/index.php?topic=49938.msg454211;topicseen#msg454211  ).

PM me (or MikeC) when you begin to build your ASP; one of us can describe the very simple modifications required.

Peter

[whubbard]

hi peter,

thank you very much. I had done the calculations earlier for the drivers in parallel, but I figured the LF55 would be fine. However it seems like I should swith it to the 25. Luckily I haven't bought the toros yet. What would you recommend for the best performance? (still the 18-0-18)

Did you leave you LF100 unchanged for the peerless bass drivers?

I will most likely not begin on the building until june, but I will make sure to contact you about the modifications.

Thanks

[PSP]

Hi West,
I have used Plitron transformers for most of my amps (no special reason) although for the last amp (100N+) I used Avel-Lindberg (with the dollar tanking, buying good ol' USA stuff looks better and better).

I am currently using two channels of 100N+ to drive the Orion's four Peerless woofers (there are several old posts on this, but briefly, I reduced the 100N+ rail voltage from 50v to 43v and beefed up the power supply so every part was good for at least 10 amps, preferably 20 amps.  The LifeForce amps are a different beast, so you should get Hugh's opinion on running LF100's into 2R5).  In the next month or so I will evaluate the benefits of adding another 100N+ on Orion bass extension and punch in my system.   

Peter

[whubbard]

Peter,

Thank you very much.
Have you noticed any difference in the Plitron or Avel-Lindburg toros...I was going to be buying Plitron, but you make a good point about the dollar. 
Hopefully things will get better soon.

Thanks.

Hugh,

Would you mind lending some of your very useful knowledge on the subject of 'beefing' up the Lifeforce 100?

Thanks

[PSP]

West,
I've not noticed a difference between Plitron and Avel-Lindberg xfrmrs, but I have not done any kind of a fair listening comparison.  Plitrons have a potted center, filled with "looks like epoxy" with a 1/4" hole drilled through the center.  There is a nice rubber pad underneath, so mounting is a snap... you just drill 1/4" holes through your chassis and get a long enough bolt (I use non-magnetic stainless steel bolts with a locking washer).  Plitrons look cool (for a transformer, at least).  Plitrons are made to order, so it usually takes me 4-5 weeks to get Plitrons.  Avels look a bit more industrial.  Also, mounting the Avel is slightly less elegant (it uses a mounting washer and pad on top, nothing on the bottom if I remember correctly).  Avels are usually in stock, so you can get them quick.

FWIW, Dave Dlugos (Planet10) who long ago guided me through my first tweaks (on my old NAD 7020, which still makes lovely music in my wife's little system) advocated strongly for Plitron.  He was also the guy who first suggested that I check out AKSA amps, so he must know something, yes?

Peter

[aurelius]

Hi Peter,

Hope you are well. 

FWIW, I've been driving my Orion+ tweeters with a 55N+ for > 6 months to no ill effect.  The amount of power the tweeters get under normal program material is so small that I'm not too concerned about overloading the AKSA.  Not withstanding the previous comment, I am interested to get your impressions of any sonic advantage gained by converting to LF25.

Are you doing anything with the LF55s other than lowering suply voltage and changing input cap?  I recall that with the AKSA55 N+, there were a few other tweaks required to ensure correct bias with the lower supply voltages.

I would also note that you could easily get away with 0.1uF Sonicap Platinums for a tweeter amp.  This would raise the pole from ~8Hz (as supplied by Hugh) to around 36Hz.  The influence on the signal at point of crossover (1440Hz) would be minimal, and would cost less than 1/2 the price.  I know this is not too usefull now that you already have them in, but you may be able to swap them out in the future?

Very interested if you notice any diference between 1 x 100W at the bottom end, versus 2 x 100W.  I am still using 2 x 55W and do not feel lacking in the bass department.  Please update with any observations.

Just out of interest, are you still using AD826?  I have just recently switched out OPA2107s for LM4562s and think that I am happier.  Yet to try the ADs.

Regards,

Mark

[andyr]

FWIW, I've been driving my Orion+ tweeters with a 55N+ for > 6 months to no ill effect.  The amount of power the tweeters get under normal program material is so small that I'm not too concerned about overloading the AKSA.  Not withstanding the previous comment, I am interested to get your impressions of any sonic advantage gained by converting to LF25.

Regards,

Mark

Hi marcus,

If you've connected your O+ tweeters in parallel (which I understand is the recommended way?) then you are getting down to about 2.5ohms, right? 

This is quite a bit less than the impedance Hugh recommended for the AKSA 55s (and I think also the LF55!  ) and hence when I set up my 3-channel amps for my Maggies, I used AKSA25 modules for the ribbons (2.9ohms).  As I understand it, if you don't need the "power" to drive the load then going down in power by reducing the DC rails, enables the amp module to produce much more current at the lower impedances without getting stressed.

Also Hugh made a comment at one stage that an AKSA 25 was "like a SET with balls!!" - ie. there certainly was a sonic advantage!!    So I chose to do this!

Are you doing anything with the LF55s other than lowering supply voltage and changing input cap?  I recall that with the AKSA55 N+, there were a few other tweaks required to ensure correct bias with the lower supply voltages.

Yes, there are a couple of minor changes associated with the lower DC rail voltage.

I would also note that you could easily get away with 0.1uF Sonicap Platinums for a tweeter amp.  This would raise the pole from ~8Hz (as supplied by Hugh) to around 36Hz.  The influence on the signal at point of crossover (1440Hz) would be minimal, and would cost less than 1/2 the price.

Although, yes, one can save money by having different values for C1 ... does having these different C1 values in an active setup introduce undesirable phase changes between low, mid & high amps? 

Regards,

Andy

[aurelius]

Hi Andy,

I must get around to visiting you one day and listening to your system. 

As I understand it, if you don't need the "power" to drive the load then going down in power by reducing the DC rails, enables the amp module to produce much more current at the lower impedances without getting stressed.

Understood.  By presenting a 2.5ohm load, I run the gauntlet that I may drive too much current through the output devices and destroy them.  The reality is that it is rare that there is more than a few watts going through the tweeters.  I would, hovever consider the change if there are sonic advantages.

Although, yes, one can save money by having different values for C1 ... does having these different C1 values in an active setup introduce undesirable phase changes between low, mid & high amps?

Potentially yes.  I have done the calculations and figured out that using a 470nF cap leaves a residual 0.3deg phase shift at 1440Hz.  If all amps use the same cap, then they all experience the same phase shift relative to one another.  Using a 100nF cap leaves a residual 1.43deg phase shift, so at the crossover point, the tweeter amp would be out by approx 1.13deg. relative to the mid range amp.  Not sure that this would be too detrimental, but your point is definitely worthy of consideration.

[Tliner]

Hi Guys,

Very interesting on how to tweak an AKSA amp. I know very little about amps but if the mods alter the phasing of the drivers be carefull. I design speakers and xovers. Basically a 1.3deg phase shift of a tweeter away from what is I assume to be satisfactory can cause the slam of percussion and lower piano and even detail of bass guitar to become smudgy. Not as crisp as when the phasing between the tweeter and the next lower driver is better than 0.5deg. Ideally the phasing of all drivers should be 0deg phase shift. But that is very hard to achieve due to the manufacturing tolerances of components. Also the group delay between drivers can be affected as phasing (nearly one in the same criteria) particularly if you are using separate amps for each driver. The xover values may have to be recalculated if pahsing is altered too much. Only your ears will tell you if the mods are an overall improvement. I have found that tweakers modify everything to the final output stage but rarely check the xovers for system synergy.

Have fun,

Cheers,


Laurie

[aurelius]

Thank you Tliner for the benefit of your experience... I guess I'll stick to keeping all input caps the same value.