Building on the ZMA upgrades previously done by others on this forum (Vyokyong, Jeff of A and various musings from Steve Deckert), I ventured on a comprehensive ZMA upgrade program myself. It turned out to be a two year project. In previous posts I mentioned some of my earlier works on the ZMA and CSP3 and went silent for a while, while the upgrade process continued. Here is a comprehensive outline in one place starting from scratch for the ZMA. Observations on how the new parts affected the outcome and costs are included.

The objective of the upgrade process was essentially to leave Steve’s original design unchanged, it being proven and undoubtedly good but to upgrade each and every individual part. From a careful study of the internals in the original ZMA I concluded that Steve used parts of a reasonable quality every where, with key areas being treated with better/above average parts as an option (e.g. Type II Jupiter caps etc.). Since he has a business to run, this entirely makes sense to keep the price down thus allowing many more to enjoy these amps. I conclude from this experience that from the base original amp any upgrade is a case of diminishing returns, but for those willing to spend there are returns no doubt  and all these small returns add up for a impressive gain in the end at a price. In my case the total expenditure was about $4000 ( I did it all myself). It would perhaps work out to about $6000 if done by a technician as the number of hours spent was easily about 100 hours over two years. It takes considerably more time to upgrade than to build a new amp from scratch (once a prototype or design is completed).

Coming back to the upgrade for the most part I did a simple one by one replacement of (almost) all components in my ZMA with parts of the same specs as the original ones but of substantially higher quality. i.e. same capacitance or resistance but made of better quality material. Some new components costs were perhaps 2 to 25 times the original components. It might surprise you that in spite of the accomplished sound produced by the factory ZMA some of the resistors within can be bought for as little as 50c and up. Each of my resistors are now $6 to $50 and since there are very many of them it adds up both in price and final quality. Because price alone does not guarantee quality a lot of research and experience from others on this forum and other forums was considered before buying the parts. But good quality also cost more as there is no such thing as free lunch. The trick is to separate the hype from actual good quality, so in buying the various parts I was not so much concerned about the make and manufacturer of parts and the subtle differences between manufactures but leaned more towards upgrades of material and technology say from electrolytic or metalized film caps to copper foil and from copper foil to silver foil capacitors and from ceramic to non-inductive wire wound resistors and so on and then ventured to find a reputed manufacturer who sold these upgraded material parts at the best price. This was to weed out paying exorbitant sums for brand name and hype alone.

As I said before the whole enhancement process took me about two years as it was done one or two parts at a time. Most of the time went towards the research associated with each new part purchase and to allow for distribution of the costs over time. Time was needed as well to judge if the just replaced component(s) had indeed contributed to improvement - after adequate burn in. Only in one instance I had to reverse my upgrade and seek another component. Encouraged by each success (most times only incremental improvements were noted after each part change for quite a substantial cumulative gain described towards the end of this write-up), I have now come at a point where about 95% of the parts have been upgraded, the only constant being the original design.

Overall outline of what was done with details of each following:
•      Internal power supply path upgrade
•      Internal Power noise filtering upgrade
•      Bias current path upgrade
•      Input signal path upgrade
•      Amplification stage upgrade
•      Output signal path upgrade
•      Start-up/Shut-down enhancements (Bleed resistor modifications)
•      B - bus upgrade
 
Internal power supply path upgrade:
-      The original non-branded IEC female connector was replaced with Furutech IEC inlet FI-03 (R) Rhodium plated connector ($27). This resulted in the grip and fit with the male connector from the power cord to be significantly better. Additionally there is now better electrical conduction due to its pure copper prongs and more robust internal conductors for the fuse clamps.
-      Synergistic Research Blue Fuse ($150)
-      FURUTECH AC Cable FP-TCS31 Alpha ($157/meter) 12 gauge wire installed between the above new IEC connector to the power toggle switch replacing the original 14 gauge wire.  
-      Much as I would have liked to have changed the power switch, the original one was left in as I could not find a better one that exactly fitted in the slot.
-      The original generic full wave rectifier diodes were replaced with Schottky Ultra Fast 10A 1200V diodes ($110). These diodes cost more than 20 time the original ones but are worth it (see below for how they contributed to improvement).
-      Mundorf Mlytic 4700 uF HP+ 4 pole low ESL/ESR capacitors ($550) replaced the large red Mallory 2000 uF power capacitors. Note here that just because there are large capacitors in the power supply it (the capacity) is of less consequence if that stored power cannot be readily and fluently delivered when needed by the downstream circuitry i.e. delivered without transient dips in supply voltage and current due to lack of charge/discharge speed. High ESR/ESL caps than act as a choke point (in spite of the high capacitance value) due to their slower charge and discharge time constants.
-      Additional 47 uF and 1uF silver foil and copper foil capacitors ($300) added to the 4 poles of the 4700 uF Mlytic caps to more readily absorb the 120 Hz diode switching noise and to additionally speed up the power supply delivery. (The idea being that these faster smaller metal foil caps provide for the first micro seconds of power delivery when the music instantaneously demands it and then the slower larger electrolytic 4700 uF Mundorf Mlytic caps take over supplying bulk power minimizing the dips (minimize current ripple). BTW though I said “slower”, the Mundorf Mlytics by themselves are quite a bit faster than the original large Mallory red caps. The only thing going for the Mallory red caps is I believe that they will last 20 to 30 years where as the Mundorf Mlytics HP+ will last about 10-15 years. But these are very easy to change – no soldering required while the benefits in the meantime are enormous.
These 3 upgrades together (Schottky Ultra Fast diodes, Mundorf Mlytic 4 pole caps paralleled with fast copper/silver film caps) resulted in a step discernible overall top to bottom, side to side improvement: significantly reducing background noise and was just the tonic my speakers needed; especially the woofers – it was an ah! moment. A jump in overall performance was noticed right away on power up without the additional benefits that comes with adequate burn in – bass improvements (cleaner, tighter and deeper – brought the drums in the room) being the most significant, but also in the midrange and the high frequencies areas.
-      Wires connecting the Mlytics caps from the rectifier and also those to the regulator tube (RCA OA3) anodes were upgraded to Furutech POCC 14 gauge wire.
What I like about this is that the results were not only felt but also measurable. There is 3 to 4 dB gain from the woofers, 2 to 3 dB gain from the mid-range drivers and 2 - 3 db from the tweeters drivers with the dB meter placed 2 inches from each driver (all other settings and music played being the same), background also got way blacker and everything is more relaxed, less labored, fluent and natural.
Power supply is the key to improving everything else downstream – the sine qua non, the without which not.

Internal Power noise filtering upgrade:
-      The 4 x 47uF F&T capacitors on the ground rail (- B rail replaced) with 4 x Mundorf 47uF 500Vdc MLytic HV ($52).
-      These were additionally by-passed with 10 uF and 0.1 uF Obligatto Ultra Premium Gold film caps ($100) to remove a larger spectrum of power supply noise.

In theory these upgrades will reduce background noise further. In my case the improvement was not much, in fact hard to notice. Perhaps because by this time my power supply - both internal within the amp as described earlier and externally - was already clean and fast. Perhaps the laws of diminishing returns were kicking in. I experimented with disconnecting and reconnecting the by-pass caps (on one side) there was a sense that noise floor got lowered when these caps were connected but so imperceptibly that it could be psychological. I would imagine that for those in areas where the grid electricity supply has noise or without a dedicated power circuit in the house for audio and with no power conditioning equipment, it would make more of a difference.

Next … Bias current path upgrade and more

Input signal path upgrades:
-      The original solid brass RCA connectors were replaced with WBT 0210-AgMs pure silver RCA jacks ($650 – ouch!)
-      Master Volume stepped attenuator bypassed since I control the volume at the source (DAC). The less the attenuators in the signal path from the source equipment to and within the amp the better. These attenuators are the single largest drag on the sensitive and delicate input signals. If parts of your input signal coming from the source gets stripped those parts cannot be amplified further down the line in the circuitry. Once lost its gone, the downstream ZMA amplification stages cannot amplifier lost signal fractions. Besides due to the law of conservation of energy I think that loss is converted to heat and noise in the circuitry. No point in having high grade interconnects and speaker cables if there are multiple choke points (attenuators, low wattage resistors, slow caps) within the signal path internally. The weakest point in your whole chain from source to speakers including internal circuitry in the amp, pre-amp and source equipment will manifest itself in the music you hear i.e. the weakest point dictates the sound. That means that one weak point in the chain will restrain the rest of your high end expensive system.
Of all the upgrades done by me, bypassing the attenuator gave the best bang for the buck. Almost a free one, only labor cost. If your overall system allows it get rid of unnecessary attenuators or simply bypass them so that you can go back, in the future if your system changes and you need one (at least one attenuator is required). Note that you will need a 80 to 100 ohm resistor connected to ground to maintain input impedance so that your source can more easily drive the amp without the source overheating.
Signals coming from the source through the interconnects and RCAs connectors now go directly to the input tubes (A12/B12) bypassing the master volume control. I used multiple ultra pure silver (6N) lintz construction tone arm silver wires ($400). In short the signal path now is DAC >>> Silver WBT RCAs >>> to 6N silver lintz construction wires internally >>> input tube (no attenuator in between)

This combination of upgrades brought out qualities in music that I listened to for years that I did not know existed. Familiar voices are far more clear and have more character e.g. the drawing in and out of the singers breath and subtle vibratos in the singer voice that were not apparent before appeared, pitch has more range, tone more authentic – life like. String instrument leading and especially trailing reverberations continue longer as in reality. Finger rubs on strings, hollowness of drums and reverberations of the drum diaphragm showed up in one case when they were not so apparent before. These accompanying natural phenomena make you believe that live music was being played within the room just in front of you. Sound stage improved significantly. Even whiskey and beer is more enjoyable while listening to music now.

Next ….. Amplification stage upgrade

This picture shows the Mundorf Mlytic 4700 uF HP+ 4 pole low ESL/ESR capacitors on the left and the large red Mallory 2000 uF power capacitors on the right.

The Mundorfs are shorter in height though have more than 2 times the capacitance compared to that of the Mallory caps.

I did a comparison of the sound on the left channel (Mundorf) to that of the right (Mallory). The gains in db on the left side was about 3 dB for woofers and slightly lower gains for the midrange and tweeters. This test was with only the one left Mundorf  Mlytic 4700 uF HP+ 4 pole capacitor replacing the left red Mallory's all else remained the same.

Wow, thanks for sharing all the details of your extensive modifications. Truly an exceptional level of changes, thanks for sharing with us. As Jeff said, definitely the best sounding ZMA bar none. Hope you are enjoying it immensely.

So in addition to Steve's ZMA 25th Anniversary mods, what would you say are the three "relatively simple" additional changes you could recommend to other ZMA owners to improve their amps sound quality? Or in other words, what were the changes you made that resulted in the biggest improvement in sound quality?

This might provide insight for DIY'ers and/or future modification packages Decware might consider.  Thanks.

HK  

Bonny,

Thank you again for sharing and your thoughtful suggestions. Since I will be sending in my ZMA for 25th Anniversary mods next month, I'll request some additional modifications per your suggestions.

I have a CSP3 preamp, ZROCK2 and ZDAC, so three sources to control volume and ride the gain.

Removing the attenuator makes sense, I almost had this done to on my original build, so I'll add this.

I'll also get the RCA input jacks replaced. Having heard the benefits of WBT RCA terminations on some of my ZenWave IC's this also makes sense. I believe Parts Connexion and/or VH Audio have some of the WBT Nextgen models on sale.

Regarding power caps, I believe Steve includes the option of replacing standard caps with new Nichicon NX 4700's. Are these the type of caps you were referring to? I know you selected Mundorf caps, but these are pretty expensive and I prefer the 20 year lifespan, unless there is a significant SQ difference between the two.

Feel free to share any additional thoughts or suggestions. Also, does replacing the power IEC input make much of a change? My power cord is pretty heavy.

HK

Thanks Bonny

Can you provide a source for purchasing the Mundorf Myltic 4700 uF caps and the Mundorf catalog item number if possible. It appears Sonic Craft only carries up to 3300 uF caps and HiFi Collective is currently out of stock.

Are these difficult to install myself? Is aditional wiring/soldering required?  I believe they have screw terminals.

Apologies for all the questions but I want to get it right.

HK

Here is a picture of my ZMA taken before this process began 2 years ago for comparison.

HK,
The seller I bought from in Germany has no 4700 uF Mlytic caps  currently.

But he has 3300 uF Mlytic caps.

https://www.ebay.com/itm/MUNDORF-MLYTIC-MLSC-HP-3300-f-500VDC-Kondensator-4-pol-...

I found another listing from France. The price is quite good as it is a unused resale. Seems like it is 60% after customs duty and shipping of what I paid back then.

https://www.ebay.fr/itm/223462562200?rmvSB=true

Continuing on ….

Amplification stage upgrade:
-      All resistors to and from the input and inverter tubes replaced with either low noise 2W Audio Note resistors ($150) or 12W Mills resistors ($150) e.g. 330ohm output tube grid resistors with 2W versions of the Audio Note resistors replacing 0.5W original resistors.
-      Signal capacitors replaced with 0.1uF 600Vdc Jupiter Copper Foil ($126) and Jupiter 0.022uF 600Vdc Copper Foil ($76) capacitors
-      Additional parallel 0.01uF Duelund copper foil capacitors ($150) added across the above signal capacitors.


Output signal path upgrade:
-      All resistors to and from the output tubes replaced with 12W Mills Resistor MRA-12 Series ($100). Originals were 2W or less.
-      A 12W Mills resistor of 330 ohms added between the screen grid pin and Ultra linear tap of the output transformer based on Steve’s 25th anniversary mod. Here I had to do a few reversals, experimented with 220 ohms, 480 ohms finally settling with 330 ohm.
-      Original solid brass speaker binding post replaced with WBT 703 pure silver ones ($700 – again ouch!).
-      Furutech POOC 12 gauge wires added to the negative speaker binding post to – B bus.

Of all of the above, the speaker binding post change to WBT silver ones was the distinguishable one in that it further enhanced the experiences outlined in the last paragraph of the input signal path upgrade.
Where as for the other upgrades listed here they enabled lower level listening without loss of micro detail. This was a nice experience (kind of a ah! moment on experiencing this fact) as it suited my listening behavior better since I prefer low level listening most of the time with the occasional high volume play when in the mood. Having said that to entirely attribute this phase of upgrades to the ability to listen at a lower volume may not be totally correct as I am sure all of the other stages helped as well in that if the other stages were not done this low level listening experience would not be possible.

Next .... Power on/ off enhancements - bleed resistor upgrade

Power on/ off enhancements - bleed resistor upgrade:
-      Replaced the original single 220 Kilo ohm resistors per side with 330 Kilo ohm (3 parallel 1 mega ohm 5 W resistors in parallel to create more flow paths) resistors per side.

These resistors are there so that when the power is switched off the large power capacitors and all other caps in the circuitry can smoothly discharge their residual stored power to ground. The down side is that during operation they lower the overall circuit voltage and will introduce noise into the –B rail. This noise backwashes into the signal path as the returns of both the power supply and signal circuit are connected to this same return rail (more on minimizing this later).

As you can see I increased the bleed resistor resistance slightly from 220 kilo ohm to 330 kilo ohm to minimizes parasitic power losses during operation. In turn this will reduce overall circuit voltage drop. So more power is now being used instead of wasted. Because voltage drop is reduced it in turn will reduce the current drawn by the various in the internal circuits for the same power output (P=V x I) which in turn reduces heating of internal components (Q=I² x R) for the same power output. This is waste heat i.e. power that is not being utilized.

When power is now turned on the OA3 regulator tubes do not make any sound at all and achieved steady conduction state status within 5 to 6 seconds vs about 9 to 10 seconds earlier. The ammeter needles now start ramping up a few seconds earlier and achieve their steady state value a few seconds faster. On powering off the OA3 tubes stay lit for about 3 seconds before flickering off with less flickers than before. The ammeter needles oscillate and deflect a lot less. These needles used to oscillate and deflect almost full scale earlier when power was turned off. With PSvane KT88 tubes in and biased at 70 mA one of the ammeters would get pinned to the extreme end on turning power off. A tap on the ammeter glass would bring it back (this was the reason I thought of slightly upping the resistance of the bleed resistors which in turn helped in other ways as listed in the paragraph above). Now the needles deflect  about 90 percent of full scale on powering off and come back on their own. Important to note that this pinning of the needle was not happening on the original un-modded amp it started after I did some of the upgrades. I attribute it to more energy being stored in the additional bypass/parallel caps and higher capacity power caps installation besides steps were taken to lessen the wastage of power as mentioned earlier. BTW I have long since moved on to Tung-Sol 7581A's and Black Boxed anode triple mica 6N5P's and bias at 58 mA.


Next ….. -B bus upgrade ….. the last one ...….. after which I'll summarize the overall cumulative effect.

Overall conclusion:
The overall result is simply sublime. A fantastic original amp in its own right got exponentially better. Listening at all volume levels - high and low volume - got better. The question is, is it worth the $4000 or so spent? that would depend on each person, for me it is given that it’s now going to be enjoyed for the next 2 decades or thereabouts. When I first bought the ZMA (with all original upgrade options) I thought this was it. After the initial euphoria settled, I realized at lower and at higher volumes I did not feel as satisfied as when listening at mid volume levels on my system. My PSB T3 speakers are 89 dB efficient (91 dB anechoic) but capable of going down to 24 Hz and up to 22 kHz. The alternative was to buy very high efficiency speakers but I have listened to some and though impressed the lack of deep bass at the one end and sparkle at the other when playing music that has those elements did not satisfy as much. High efficiency speakers I found were excellent for a vast range of music but not all. Admittedly I did not try out very many of them so I could be wrong.

Here is another fact that demonstrates the usefulness of these upgrades done by me. Initially I was compelled to bi-amp the PSB T3 speakers with the unmodified ZMA driving the tweeters and mid-range drivers and a Solid state Cary Audio amp driving the woofers and TakeT Batmaster super tweeters (reason being the Cary audio went down to 18 Hz and up to 90 kHz (-3 dB) as tested on my audio analyzer while the ZMAs range was 28 Hz to 52 kHz). The unmodified ZMA would start clipping at mid-volume levels if it was used to drive the T3s on its own (not bi-amped). After the upgrades with the ZMA alone I now drive the T3 at beyond mid-levels into the lower high volume levels before clipping becomes apparent.

As I mentioned earlier the returns are diminishing from the base factory amp. As can be seen one has to spend a bit to go to the next level. My experience is that changing a part here and there will improve things for you but not to the level that takes your music a quantum step up in level. The rest of the circuit still acts as a chock point not allowing that upgraded part to bloom to its full potential. From my experience I can say that each upgrade makes the previous upgrades better and the previous upgrades made the most recent one better as more and more weak links are removed. Until all is done it is a wandering weak link (another old part become the weak link) and then that weak link will define the quality of the music that is output.

My music listening experience now has that feel good factor one gets when listening to a live show in a quiet room (best of both worlds). Besides 3D imaging and spatial improvements, there is also what I now call the magical 4th dimension which to me is a complex concoction of timing, speed, dynamics, rhythm, density, separation of instruments, voice layering (foreground, background) all within that increased density, each distinct and more life like. Additional nuances in long familiar music have appeared. More over everything is more naturally reproduced with less strain and more perfectly placed in 3D space and time.

Future upgrade considerations:
I am suffering from upgrade fatigue and am going to take a long break and just enjoy the system now. In preparation for the future the only thing, I think that will make a further step jump in performance after all of the above will be an upgrade of the input and output transformers, especially the input transformer as it is a significant part of the source power. I am have trouble finding high quality ones which will seamlessly fit in the original space. If anyone has done this or has ideas it would be awesome if you could share it.

Well folks that’s it from me, hope some of you get inspired by this experience of mine. It is not so daunting money wise and effort wise if done over a period of time.

The end reward is worth it.

Cheers
BonnyA

I have been wrestling with solving the mystery of why a wholesale change to larger wattage resistors improved the dynamic response of my system.

Here is my half baked theory: larger wattage resistors (but of same value) have a lower RC/RL time constant (𝜏) so it helps in facilitating transient rise and ebb of current within the audio circuitry. The resistance value being the same the end voltage drop across the resistor is the same but it reaches this end value faster. Music being dynamic if the previous signal does not reach its peak before the next instance of signal it gets superimposed and so mis-represents the original signal affecting the outcome.

Feel free to chime in on this by agreeing or disagreeing with it or stating your own. It will help in understanding this phenomenon.

Because manufacturers do not test the intrinsic RC or RL time constant for resistors or at least state it in their specifications this is only a guess at this point. I have an idea of how to crudely test this theory. Will get back if I decide to go ahead with the test until then it is just a theory of why it works.

Glass fuse voltage drop reading. The voltage was fluctuating up and down between the 0.45 to 0.95 mV range due to I presume the dynamic nature of music and the fuses limitations. This is a snap shot somewhere in between in that range when the camera shutter opened.

The humble copper wire reading. 0.0000 V drop! no matter how high I went even briefly tried going in the clipping range. See picture.

Archie, thank you for bringing this up, looking back my explanation was not clear because I took things for granted.

It seems like those readings in the millivolt region are insignificant but it depends on what it is about. Millivolt when it comes to power say a battery is indeed insignificant in fact very insignificant. I am not interested in voltage drop per say. A voltage drop of 0.95 millivolts on a 120 volt supply gives us  119.999905V which is absolutely OK and a waste of time writing about because our grid voltage fluctuates much more than that. But here we are trying to indirectly establish resistivity and here millivolt is significant. Any sort of voltage drop across a conductor especially across only an inch of a conductor, indicates it is not a pure conductor there is resistance in it compared to a pure conductor. That is why absolute audiophiles that have loads of money will buy power cables with 5N even 7N pure copper in it and even beyond going for 5N silver cables. If I did the same test between say a 5N copper cable versus a 7N copper cable I will need I am guessing a multimeter that is sensitive to 25 digits after the decimal. But audiophiles want every last bit of performance so they go for the 7N copper cable. It depends on what one is satisfied with and what one is willing to pay. Even with different material conductors the difference is not much in terms of voltage drop: say with a hypothetical mild steel cable versus a pure copper cable the voltage drop will be far less than a glass fuse versus SR Blue fuse but I would not want a mild steel cable powering my equipment. I have not tried it nor would I want to.
 
Having said that in terms of difference of readings between a glass fuse versus say a SR Blue fuse here I was interested in by how much of a factor is a glass fuse is worse than a SR Blue fuse and the factor is between 2 to 3 times worse. It is obvious both these fuse type still work and largely do their job, but I am talking about trying to extract performance way beyond the normal.

You said "I generally don't question when someone on this Forum says they hear something but I am always curious/suspect of the why"

I must say I am the exact opposite, I sometimes wonder when well meaning folks say that so and so aspect has improved without backing it with measurements or science, I wonder if its rationalization (our brains are finicky) because that person has spent money on it, so wants to hear that improvement and the brain obliges. Where as measurements are a fact that does not depend on the brain’s subjectivity. Going by your statement above I figure you are skeptical of everything I said in all the of above posts because of my liberal dosage of science backing (the why) and measurments. Reluctantly I am forced to tell you why I take measurements where ever it is easily possible and back things by theory. I am an engineer (I am sure there a many more on this forum), when we talk among ourselves at trouble shoot sessions in a professional setting we always say "that which we cannot measure we cannot improve" so it is ingrained in me from my professional life being a daily habit for a few decades now.

You also said “I have a hard time understanding how a great sounding amp can sound "twice as good," for example”.

I do not deny that the ZMA is a good sounding amp. I mentioned it all along but have an issue with saying that it is the ultimate and cannot be improved any further. Not sure if you have costed most of the original parts (resistors, capacitors, solid state rectifiers, brass connectors etc.) in the ZMA, the parts within are of average (let me stress – not poor but average) quality. Only some capacitors are slightly above average included as an option. I have no issue with it as built, if I was running a business, I would perhaps do the same.These parts more than adequately do the job. It is the direct and straightforward circuitry that does the magic. I have largely not touched the circuitry just improved the parts and removed impediments like attenuators that I do not need in my audio system as a whole.

BTW I think you are right to question everything because I get the feeling that it is genuine and based on your experience and I respect it and have no issue with it, just giving you my side of it hopefully without causing you to feel I am challenging what you said. And off course it is obvious you do not have to spend money on something you do not believe in. I put the information out there for folks to benefit from my experience and together have a better experience as I have from some of them.  

So my question is WHY?
Why not buy something different? Did you buy the amp for it's casework?
Is there something about the design that you like but don't think that the execution was good enough?
Maybe I don't understand because I'm not an engineer. I'm just paid to be one.

Donnie,
From my personal experience with "modding" the ZMA, I can honestly say it is STILL A ZMA.  If you recall, I (and Vyonkyong before me)  actually installed a power supply bypass modification (amongst some other key component upgrades) before Steve created the Anniversary mods on the ZMA that are now offered from Decware.  

I think the answer to your question is simple...  Just like Steve believed that the ZMA topology had greater potential.  This motivated the R&D which later resulted in the current ZMA Anniversary additions. Bonny (and myself) embarked on a journey that took us down a slightly different path but ended up at the same destination - a much improved version of the ZMA that we loved then, and love even more now!!

Bonny has undoubtedly gone to the extreme but at the end of the day, Steve himself determined that the genius in his design could be bettered, just like all the amps he currently offers with the Decware Anniversary "mojo" sprinkled about.  

Lest not forget, a ZMA it was..  and a ZMA it still is.   It is Steve's creation and regardless of what "tweaks" are applied, the design credit was birthed by Steve (and I'm sure he has the stretch marks to prove it) and for that I am thankful.

I live in Canada.

You can still come.  I don't think there are plans for a Northern Border wall ... yet.  

I have installed the adapted 6 gauge wire "fuse" in the fuse slot again and have been running it this way for the last 10 days and intend to continue it. Hoping Murphy's law does not come back to bite me as has in the past with other stuff. Having drunk from the poisoned chalice I am finding it difficult to go back.

Folks seriously do not try this one, depending on your level of OCD - mine is obviously high you too might not want to go back. I do not leave sight of the amp as a mitigating factor.

My modus operandi for now is when trying out new tubes is to have a fuse in and after a period of say 100 hours with those tubes go back to the wire "fuse" but always being in the vicinity of the amp when it is turned on. I am not so much worried of the amp getting destroyed, do not want to cause a electrical fire and burn the house down.

Quote:


Is this really true? Confirmed through independent testing?

I do not own an ZMA, but I am concerned that statements like these are an issue if not really true. I have seen similar statements about other tube amps in the 25-60 watt range on the internet. I don’t want to be all about measurement or specifications. More reason to stay with amps I already have. I do not care if they do not measure 20-20K Hz, but I hope they would.

Cool. I see a response of 20-80 kHz for my amps on the product page, but I do not see something similar on the ZMA page.

Do you use a preamp?

I thought about bi-amping, but I haven’t gotten serious about it. I see a lot of folks using Class D amps on the bass.

Folks I took another step in this quest to further improve the ZMA.

I have now replaced the large Mundorf Mlytic 4700 uF HP+ 4 pole caps with 6800uF ones of the same make and type. As can be seen in the attached picture they tower over the transformers in the background.

There is a lot written elsewhere and on these forums as to why the power supply should be many times the actual power draw from the audio equipment it servers. e.g.electrical panel (dedicated circuit 20 A or 30 Amp circuit), wiring (10 or 8 gauge), high quality wall receptacles, high quality and capacity power conditioners etc. etc. Many of us have actually experienced the benefit of having a large power supply chain.

In keeping with this philosophy I wanted to increase the internal power supply within the ZMA for the longest time as this could be a choke point. Since Steve was limiting his upgrade to 3300 uF in the 25th anniversary mods I was hesitant, in spite of the feeling that that limit was more a business decision rather than a sonic one. I was hesitant also because I was wondering if the gains - if any - would be discernible enough given that my ZMA has improved so much already. I wondered if there was possibly no more room for improvement available. Time and time again it always got better with thought full upgrades so much so that I have now concluded that there is no such thing as a limit. Time and time again I thought the end was near and the next upgrade was going to be a waste but time and time again there were noticeable improvements as I mentioned in this thread. I now believe there is no hard ceiling, to use a cliché, the sky is the limit.

I believe the power supply within the ZMA does not have much excess available margin and that is why many on this forum have debated on limiting the bias current to 50, 60, 70, 80 mA or whatever and at what point the input transformers would be stressed and fail. I think the internal power supply is too close to the requirement of the amp circuitry, so suffers when the dynamics of music being played calls for instantaneous power boost.

Back to the point, this upgrade of installing higher capacity capacitors (3 times stock or 2 times 25th anniversary upgrade) made all aspects of music even more clear and precise. Extremely faint leading and trailing edges of every micro morsel of sound across all frequencies came into existance, bringing about a completeness and therefore realism.

I am impressed enough that I am now considering increasing the power capacitors capacity manifold !!! say 50,000 uF or may be even 100,000 uF or more, about like in very high end systems costing above $50,000 were the ranges are in the Farad range (not micro Farad range). I now have a lot of spare caps lying around – 2 each of the original red 2000 uF, the Mundorf 4700 uF and additional 10,000 uF that I bought sometime ago for another project which I shelved (DC offset eliminator) that I can utilize for starters. For space reasons, I am going to have to put it all on a separate “Power” box/base with additional higher capacity power transformers on it in as the ZMA transformers will most certainly not be able to handle the initial surge current as the capacitors charge. My preliminary thinking is leave the ZMA input transformers to only supply the various tube filaments and the new external high capacity transformers on the separate Power unit do the heavy lifting. The intention is also to include a CLC or better noise filter. Note currently the original ZMA has no standard filtering techniques being employed. Instead the ZMA utilizes large somewhat slowish electrolytic caps to “dampen” the noise rather than a full pronged attempt to “filter out” the noise so there is another reason to do this. Although the 47 uF F&T caps and the new 25th anniversary caps on the B – rail are an attempt to filter out some noise but is nothing like say a good old CLCRC filter combo.

Picture of the additional capacitors in place.

Looking at it perhaps the title of this thread should be changed to ZMA - Extreme upgrade. ha!  At the outset of this adventure I did not think it would go this far!

This a temporary set up. I'll have to think of something to make it look more pleasant until I do a more dramatic 50,000 uF change which will require higher capacity power transformers and ancillary "Power Supply" box.