Larger transformers should be able to handle current swings better and have a higher capability but may not necessarly sound better in every case. It is all in the excution of the design. Avil-Lindburg [sp?] are the ones I used when I had some custom designs made up years ago; they are excellent and were used by Mark Levinson in some of their amps. Large and expensive then, I suppose they are more so if still avalable.
If your amp runs in pure class A, there is no benefit in upgrading the transformers besides slightly less internal heating of the transformers.
If your amps are the more popular class AB or AB1, depending on how good was the design to begin with, you may hear more dynamics (increased headroom) with larger transformers.
You will certainly need to upgrade the rectifiers, because a higher wattage transformer usually has less series impedance, therefore it will deliver more instantaneous current to the rectifiers during the power up cycle.
Also, you may need to reduce output stage bias, because an over-specified transformer may output up to 10% over-voltage. In summary, this is not as simple as it seems.
Here is a good read about choosing a transformer supplier;
I found your reply intriguing. Why would`nt a class A amp not improve with a larger transformer?
Because Class A runs at full output all the time while AB varies with demand. Since the transformers already handle full output bigger ones will not improve the sound. An insight into why well designed Class A amps are so large and expensive.
Thanks, Stanwal, excellent explanation.
Class A amps are fully on (conducting 100% of the rated power) most of the time, therefore the power transformer, rectifiers and power supply capacitors must be over-specified, in order to supply the high current demand, stay relatively cool, not buzz or fail.
On the contrary, Class AB amps pass a fraction of the maximum power most of the time, while playing music. On musical peaks, a larger transformer and stiffer power supply MAY help output more peak power.
Some amplifiers are current-limited by the choice of output devices or the size of the heatsinks. In that case a larger transformer would be of little help.
I suggest that the OP contact the amp's designer for guidance.
Larger transformers will have a much larger turn-on current inrush.
Not necessarily bad, but should be considered. Maximum from the wall, even a 20a circuit is what......1800va?
For larger transformer(s) to be effective, this amp will need a dedicated circuit.
Need some more infor. What class is the amp? Listening habbits and speakers will give some idea of actual power needed.
I thought North American power is 120 volts. That means a 15 amp circuit x 120 volts will equal 1800 watts and a 20 amp circuit will equal 2400 watts. If I was going to be using close to the limit for either scenario, I would opt for several dedicated circuits.
1. Unless you are upgrading and altering the circuit designs within your amp, the new transformers must have the same turns ratio for voltage in and out. In other words, whatever the original input and output voltages are in your current transformer, that is what it must be for your new transformers. With Higer VA rating, that means for the example I stated above, that the current output will be different.
2. Current in-rush must be considered. you would need a slow start circuit.
I have found more benefit by upgrading my power supplies, larger capacitors, separate power supplies, etc. and upgrading the output transistors and pre-drivers to newer, more linear powerful transistors. I have also added Threshold style heat sinks to my amps and altered the bias accordingly to more class A operation. Amps such as Bedini, Threshold, etc. really benefit from this.
By-the-way, you can bias for various levels of Class A. What most people feel is class A is bias whereby the output drivers are producing rated current with no signal. you can bias to lower amounts also. Say, for example the amp is a 100 WPC amp. Nelson Pass on his web site gives several examples of heat sinking, transformer characteristics, amp design, etc. so take a look at his web site, but, for full class A, this means that the output drivers are biased for current output that equates to 100 WPC for the amp. you can bias for 5 WPC, 10 WPC, 50 WPC, etc. This totally depends on the transformer capacity, other circuit considerations and mostly proper heatsinking. To have true full output class A, you need rediculous heat sinks and they cost a ton of money. This is why most amps are not true full output class A, but are biased to class A for lower wattage.
So, if you really like your amp and want to upgrade it. Try upgrading the power supplies first. Then look into new more linear output transistors and pre-drivers. for tube amps, that is a totally different story.
Firstly lets get this Class-A and AB thing sorted out.
Class-A: This is where the transistors that operate both the negative and positive going parts of the waveform are ALWAYS on. The reason why Class-A amplifiers get hot is because the transistors are always ON.
Class AB: is where the transistors are only on half of the time, they switch on and off, cycling between the two halves of the waveform. The good thing here is that the transistors are OFF half the time hence less heat.
The problem now is crossover distortion! This is where the 2 halves 'don't quite meet up properly'.
Stanwal - it has nothing to do with full power or AB varying with demand.
So the question to whether a larger VA rating can improve the audio is YES. This is partly because as the VA rating increases the transformer 'regulation' becomes more efficient. Also (and remember VA = Volt Amperes) as you increase VA your current 'tap' increases allowing your amplifier to draw more current when needed.
If you imagine - a standard desktop computer draws approx 25-35A for a few milli secs when asked to perform a BIG number crunch. - this is the equivalent of a large transient in music (kick drum, bass guitar pluck, timpany on drugs)
How much current does a preamp need????
Well, I'm not a Naim Audio lover BUT if you listen to any of there preamps with their smaller power supplies and go to a bigger power supply the improvement is BIG. This is the same for any amplifier design (A,AB,B etc).
Class-D (again not a fan) improves with larger VA transformer - just ask Hypex
Remember, to get the best from a larger VA rated transformer you must upgrade the rectifiers and filter caps accordingly or else it's bottleneck time.
Dcarol, why do you think it does not? A true chass A amp draws 4 watts ALL THE TIME for each watt of output. So a 100 watt a channel amp will draw a constant 800 watts from the wall. An AB amp idles at a few watts and rarely draws its maximum current. Therefore the transformers in an A amp have to be up to ALWAYS supplying full current as opposed to doing it once in a great while. It is obvious that class A amps require more robust parts to deal with the constant heat generated; they actually run cooler at full output than at idle. Class A amps never see a sharp increase in current draw because they are already drawing their maximum; how then would bigger transformers improve the performance? Yes you could increase the power rating but that would mean everything else would have to be bigger also. I thought the origional question was just changing the transformer and I have seen nothing in what you said to convince me that it would.
high-end audio reproduction doesn't work like that. If it did work the 'theoretical way' then a bigger power supply or larger VA rating on a transformer on a preamp would do absolutely nothing!
As we all pretty much know though, it does make a difference.
Your calculations do not take into account instantaneous/ transient current draw - the current draw during fast, large musical passages which happen in a few milliseconds. This is where a larger VA rating is needed. If you have ever built an amplifier whether Class-A or AB and calculated the exact power requirements and even been generous adding an extra 1.4 multiple, you will still hear an improvement doubling the VA rating.
If the theory and maths could explain everything then how does high-end audio work? Theory and maths does not explain differences in audio nor does it explain the reason why a larger supply helps the audio reproduction in just about every piece of audio equipment. Ask any audio designer...he'll/she'll be able to get so far but then after that it's all 'guess work'.
stanwal: I thought the origional question was just changing the transformer
Dirk: Changing the transformer to something larger is changing to a larger VA rating.
It is your original explanation to class-A and AB that is questionable. Class-AB 'varies with demand' and Class-A 'runs at full power' and also the idea that they are all large and expensive - they are not all large and expensive.
I'm not asking you to take anything I say onboard, just that it is not as simple as you make out - if it was then we would all have the same system.
Dcarol, it appears that you are not addressing Stanwal's point about a class A amplifier. I agree with him- if the transformer for such an amplifier is up to the task in the first place, a larger transformer is not going to make a difference, high end audio or no. If the transformer is not up to the task, then the amp is either not class A or its one with a reputation for eating transformers. In any case it may be moot as a class A amplifier is 'high end' anyway...
Unless you are upgrading and altering the circuit designs within your amp, the new transformers must have the same turns ratio for voltage in and out. In other words, whatever the original input and output voltages are in your current transformer, that is what it must be for your new transformers.
This comment is not entirely true. Power transformers can have voltage drops across their windings, a lot has to do with the internal temperature rise of the part during operation- transformers with more temperature rise have higher internal voltage drops. If you go with a transformer with more VA, you may well have to *reduce* the turns ratio to obtain the correct output voltages under load.
In any case, while on the surface it might appear to be a simple upgrade, the fact of the matter is that doing such a change is an R&D project and should be viewed in that light, especially when the initial results fail to satisfy :)
I might add that I was one of the origional Krell dealers and have also sold Class A amps from Musical Fidelity and Stax. I had a preamp with a custom power supply using 2 Avil Lindberg torrodial transformers and very large caps built in the early 80s; my ASR Basis phono stage uses a 50 lb power supply and my MF Kw line stage weighs 50 lb or so and will put out 50 volts. I also use an aftermarket PS for my MF V-DAC. I am well aware of the importance of power supplies; my point was that JUST the subistution of bigger transformers in a design that already had adquate ones was not a guarentee of better sound. As Atmasphere points out it is a lot more complicated than just making a simple subistution.
Atmasphere is correct. I was describing transformers whereby you want the same rail voltages and are just swapping transformers with one with higher VA ratings, but the voltage in/out (turns ratio) is the same. If you change the turns ratio from the original transformers without playing with the power supply circuitry, your power supply rails voltages will be different. Guys, and ladies, "always on" describes that the device is always turned on. it does not mean that it is biased towards full output capability. You can bias where the devices are "just" turned on (class A) or you can bias where the devices are cooking at full rated current output (full power class A). If I bias a 100 WPC amp on at 25 WPC, it is biased class A at 25 WPC, but it is still class A. Just not 100 WPC class A.
Thank you for the much-appreciated in-depth responses.
I should have also indicated, in my very limited knowledge of the subject, that I am aware that it is not a simple swap-out of one pair of transformers for largers ones. This naturally
Ultimately, though, as I've learned from all of you, is simply trust one's own ears. My custom main speakers - Van L Speakerworks/Quartet - are absolutely wonderful and about as far as I could have them designed (Deuland VSF Copper caps/resistors/cast inductors/SilTech gold internal wiring/WBT connectors/Eichmann Toppers/Scan Speak Revelator tweeters & custom dual-motor 5.25" bass drivers. My 'sub-bass system' is a pair of REL G-1's. My apologies for the boasting. I'm just extremely proud of this set-up. The intent with this Q & A is to [hopefully] make sure I'm not leaving anything on the table when it comes to getting the fullest reproduction available.
Thanks to all again.
The intent with this Q & A is to [hopefully] make sure I'm not leaving anything on the table when it comes to getting the fullest reproduction available.
Well heck, you could try out some tubes in that system :)
Response to Magfan - System is on a dedicated 20a circuit with 12ga solid core wire. Current amp is standard Class A/B design, V-caps, 14ga solid core silver internal p-t-p wiring. Speakers efficiency is 90db+ (conservative). Listening is largely on weekends. Music is jazz, big-band, rock, blues, piano, electronic (Eno, Bob Holryd, Harold Budd, John Hassell, etc.)
Thank you very much for your information sent 9.13.11. I agree with your response and will make sure my amp builder is on top of the rectifer & filter cap aspects (I'm reasonably sure he probably knows, but I'll talk to him nonetheless).
I've narrowed down the selection to a pair of Plitron 1500Va's. Their being potted and and being "low noise" are what sold me. After much research it appears they've got a reputation for quality. This is not to nay-say anything of the other manufacturers out there but I am comfortable with the decision. The challenge now is to try and avoid musical DT's for however long the build will take....
A household circuit is rated for something like 80% of capacity for full time use and only short term @100%.
I don't know rule-of-thumb, but a total of 1000va of transformer will put out a BUNCH of power continuously and still allow dynamic peaks....which are at least 10db and frequently higher than the average level.
Doesn't 10db amount to 10x the power?
Unless you have a HIGHLY reactive speaker, chances are you are playing electrical games with yourself.
Speaker 'efficiency' will be expressed as a percentage. Of power turned into sound, while Sensitivity is the 'db per watt', which also is related to impedance since a watt (voltage measure) is different at different impedances.
Somebody straighten this out if I've made a hash of it.