Many don't know it, but the rotation of toroidals can make a noticeable impact on the performance, since the direction of the flux leakage due to leakage inductance is changed. Apparently Bryston fine-tunes each of their amps individually in the factory. Of course the better the transformer quality the less this will have an effect, but it's audible even on a 20VA one in a valve preamp. Sadly the thickness of the secondary windings makes it a little impractical, but it's well worth toying around.
Here's a picture of Jozua's Mk2 with one of my boards placed on top for comparison. All the dimensions, mounting holes etc. are a perfect match! It was the original idea and carefully measured out but it's good to see it checks out.
Even though they won't be put in this particular amp, they will be used to upgrade his other Mk1 to a Mk2.
Here's a picture of Jozua's Mk2 with one of my boards placed on top for comparison. All the dimensions, mounting holes etc. are a perfect match! It was the original idea and carefully measured out but it's good to see it checks out.
Even though they won't be put in this particular amp, they will be used to upgrade his other Mk1 to a Mk2.
Attachments
PWatts said:
You are right about that PWatts,
Im aware of that stray fields round toroidals due to flux leakage. Thats the reason I always suggest potting in a mild steel can when asking for a price quotation.
Regarding thickness of secondary windings and toying around: its not an issue, due to it will be flexible leads on this transformer, if going as far as to a group buy (GB)
Nice boards, yes. Have you had the opportunity to listen to it yet?
Regards
Mark A. Gulbrandsen said:
If that was that easy.........................
I dont think it is possible for John to duplicate a Krell transformer due to all core and winding data is "classified" e.g. no data are available for people outside Avel or Krell. To do a duplicate, he must know exactly the core size, type, material used, supplier, windings, etc....
If he was, or are capable to duplicate an original Krell transformer, all other transformer companies would be able, but unfortunately, there is no data available for the originals.
Regards
Well it's a transformer not a nuclear reactor! None of the secondaries besides the main ones are important to the sound quality, and who cares what size the core is. Go for a proper VA rating, good quality core and all the refinements possible, pot the thing and there you go.
BTW, all the Krells I've measured up measured differently between the two channels with the same inconsistencies; this is likely because of each transformer's other secondaries (relays etc.) not operating the same. IMO use a transformer with a single set of secondaries and a small extra one for the rest.
I haven't fired up the boards for listening yet, but hope to do so in due course.
BTW, all the Krells I've measured up measured differently between the two channels with the same inconsistencies; this is likely because of each transformer's other secondaries (relays etc.) not operating the same. IMO use a transformer with a single set of secondaries and a small extra one for the rest.
I haven't fired up the boards for listening yet, but hope to do so in due course.
Hi Flod,
transformer designers are professionals. That is their bread and butter.
They will know what each other are producing and whether they have a market that wants cheap or medium cost or gold plated.
They can easily arrive at a set of parameters that will give a very good copy of the Krell original even though it might not be identical.
The problem is, we do not see the cost of each individual option nor do we see the cheaper way to get 95% of that performance.
If we sat down with the designer, he could take us through all the options with their pros/cons/costs and we/he could come up with either an expensive copy or a mid price copy but certainly not a cheap copy. Cheap = small core & little copper and no extra windings and no screen and no guass band and no vac impreg of either the core nor the windings, etc. We could even come up with better than the Krell original but none of us could afford it.
I bought a couple of Ebay 1kVA 74+74Vac transformers (Terralec UK). With a bit of work and careful re-taping I now have a four secondary 37Vac transformer. Now that I have tested it and found the voltage drops in a full ClassA design biassed to 2.6A, I have an option, re-build the next one identical or add 22turns to each winding to give +-52Vdc when fully biassed. It will cost one reel of 1.7mm wire (about £7 +£40 for the pair of 1kVA). There are ways that work.
edit.
just like Patrick, I went for a small 9Vac supplementary 50VA transformer to which I can add as many windings as the various ancilliaries need.
transformer designers are professionals. That is their bread and butter.
They will know what each other are producing and whether they have a market that wants cheap or medium cost or gold plated.
They can easily arrive at a set of parameters that will give a very good copy of the Krell original even though it might not be identical.
The problem is, we do not see the cost of each individual option nor do we see the cheaper way to get 95% of that performance.
If we sat down with the designer, he could take us through all the options with their pros/cons/costs and we/he could come up with either an expensive copy or a mid price copy but certainly not a cheap copy. Cheap = small core & little copper and no extra windings and no screen and no guass band and no vac impreg of either the core nor the windings, etc. We could even come up with better than the Krell original but none of us could afford it.
I bought a couple of Ebay 1kVA 74+74Vac transformers (Terralec UK). With a bit of work and careful re-taping I now have a four secondary 37Vac transformer. Now that I have tested it and found the voltage drops in a full ClassA design biassed to 2.6A, I have an option, re-build the next one identical or add 22turns to each winding to give +-52Vdc when fully biassed. It will cost one reel of 1.7mm wire (about £7 +£40 for the pair of 1kVA). There are ways that work.
edit.
just like Patrick, I went for a small 9Vac supplementary 50VA transformer to which I can add as many windings as the various ancilliaries need.
Yes... Bob Groger has volunteered to take over the shipping of the stuff and also the sale of the remaining boards. I will be shipping all the stuff to him on Monday. I think he will be making a post about it here very soon.
I'll be lucky if I have time to finish mine sometime in the next year....
Mark
Interesting to see whats the answer from John would be. Probably I will not receive an answer before Monday.
I think we should go for a model (e.g. the first quotation) rated like this:
If available: vacuum impregnated core
+ vacuum impregnated windings.
primaries,
2 of each: 0-110-115-120V (@ a totals of approx. 1000VA)
Secondaries
2 of each: 0-38V bifillar winded (total of 925VA= >12A each 38V winding)
2 of each 0-12V (@ 2-3A for a regulated power supply for the drivers and inputs)
Aux:
2 of each: 0-14V (could be ±12 dc regulated @ 1A for logic, relays, control, fans, lights etc)
+ electrostatic screen between prim. and sec.
+ gauss-band on outer rim of transformer.
Flexible leads, absolutely
Potted in a mild steel-can if possible (preferable, but not necessarily)
-----------------------------------------------------------------------------------
How about that? is this the transformer needed, or had it to be simpler?
What would be an acceptable price for a unit (or a pair)?
Regarding primary voltage of 38V, it is calculated at a current output of 12Amps and if calculated by a regulation of 95%, this voltage would rise to approx. 40V @ zero current.
It means the power rails voltage could vary by approx. (40V x √2)-2V ≈54.6V @ no bias, no audio outputs, to (38V x √2)-2.5 ≈51V @ 2.6A bias and full audio power output
Audio power outputs would then be calculated to be, in 8 ohm approx. 163W and in 4 ohm approx. 310W, and in 2 ohm approx. 560W (if transformer could handle 20 Amp and 40V in peak).
Regards
I think we should go for a model (e.g. the first quotation) rated like this:
If available: vacuum impregnated core
+ vacuum impregnated windings.
primaries,
2 of each: 0-110-115-120V (@ a totals of approx. 1000VA)
Secondaries
2 of each: 0-38V bifillar winded (total of 925VA= >12A each 38V winding)
2 of each 0-12V (@ 2-3A for a regulated power supply for the drivers and inputs)
Aux:
2 of each: 0-14V (could be ±12 dc regulated @ 1A for logic, relays, control, fans, lights etc)
+ electrostatic screen between prim. and sec.
+ gauss-band on outer rim of transformer.
Flexible leads, absolutely
Potted in a mild steel-can if possible (preferable, but not necessarily)
-----------------------------------------------------------------------------------
How about that? is this the transformer needed, or had it to be simpler?
What would be an acceptable price for a unit (or a pair)?
Regarding primary voltage of 38V, it is calculated at a current output of 12Amps and if calculated by a regulation of 95%, this voltage would rise to approx. 40V @ zero current.
It means the power rails voltage could vary by approx. (40V x √2)-2V ≈54.6V @ no bias, no audio outputs, to (38V x √2)-2.5 ≈51V @ 2.6A bias and full audio power output
Audio power outputs would then be calculated to be, in 8 ohm approx. 163W and in 4 ohm approx. 310W, and in 2 ohm approx. 560W (if transformer could handle 20 Amp and 40V in peak).
Regards
PWatts said:I would change that 0-38V to 0-39V to be purist.. the Mk2 I have has 55VDC unloaded with 230V mains, but the toroidals configured for 240V, so the correct voltage would be 57VDC.. not that it would make an audible difference though
Hehehe, I started at 39V, but when calculating I thought this was a to high value so I change it/lowered it to 38V. Guess we will be landing on 39V, is this ok?
Besides this any other comments PWatts? pricing etc ?
What would be a reasonable/ decent price for a custom made transformer like this one ? Anyone..........
Regards
Hi,
I have 240:37Vac and my loaded rails are 2V to 2.5Vdc low.
so 38V to 39V sounds about right.
BTW,
the total VA of all the windings can exceed the rated VA by quite a margin.
The predicted output is likely to be slightly higher.
I got 311W into 4r0 from 37Vac (~+-49.5Vdc).
Your assumption on regulation is well out.
A poor quality (high regulation) 1kVA will be about 4 to 4.5%.
A good one <=4%. My Terratec are 4% and the Noratel are 3.9%.
A very good one that is specified for low regulation (=extra copper and extra cost) <3.5%
5% target is far too high.
You could specify regulation as a way of ensuring we get adequate quality.
I think the spec we read about the very large Krell original indicates they specified a low regulation figure.
Ask if the guass band removes some or most of the requirement for a steel potting can.
I have 240:37Vac and my loaded rails are 2V to 2.5Vdc low.
so 38V to 39V sounds about right.
BTW,
the total VA of all the windings can exceed the rated VA by quite a margin.
The predicted output is likely to be slightly higher.
I got 311W into 4r0 from 37Vac (~+-49.5Vdc).
Your assumption on regulation is well out.
A poor quality (high regulation) 1kVA will be about 4 to 4.5%.
A good one <=4%. My Terratec are 4% and the Noratel are 3.9%.
A very good one that is specified for low regulation (=extra copper and extra cost) <3.5%
5% target is far too high.
You could specify regulation as a way of ensuring we get adequate quality.
I think the spec we read about the very large Krell original indicates they specified a low regulation figure.
Ask if the guass band removes some or most of the requirement for a steel potting can.
A simple 2*40Vac 1kVA retails for about £55 to £60 in the UK.
Adding on all the extras must raise this to the range £80 to £100 at retail cost.
But can we expect a wholesale discount if we bulk buy?
How much -10%, -20%, maybe -30%
If they offered to supply @ $140 including shipping we would be getting bargain compared to UK prices.
Adding on all the extras must raise this to the range £80 to £100 at retail cost.
But can we expect a wholesale discount if we bulk buy?
How much -10%, -20%, maybe -30%
If they offered to supply @ $140 including shipping we would be getting bargain compared to UK prices.
AndrewT said:
Andrew thanks for your comments.
Yes I know it wasnt right to calculate for 95% regulation, but I had no value at hands. But this does not change the output voltage at rated current (12A) only at zero current, right?
I have specified the transformer to be a "Low Flux Design" and it means all your considerations will be implemented in this model (larger core than what would be expected, thicker magnetic wires etc...).
Regarding pricing, ok, its a price level to start from. Lets see whats coming up next week.
Regards
Mark A. Gulbrandsen said:
As soon as I get the diodes and boards I will begin the task of packaging and shipping. It would really help if anyone expecting something would email me with a quantity paid for and your shipping address. Euro shipments can hopefully go to one member and be redistributed from there.
bobgroger at hotmail dot com
Bob G.
I don't think that John has tha ability to do transformers in cans. All the transformers that John winds for Leach and for Smart Devices are all standard toroids. I think that if you want them in cans you'll probably have to do it yourself. It will certainly add to the expense if he can (pun intended) do it. I was quoted about $130.00 to $150.00 for a simple dual secondary toroid. Extra windings will add to the cost.
Mark
Mark