Power amps & R core transformers
- Thread starter chord
- Start date
337alant
Negatively Biased
There was a group buy on PFM by TimH for R Core transformers for the Hackernap Power amps several years ago, so there was quite a few used with good results
I also bought some and used them in my NCC200 voyagers which sounded excellent.
When I built my NCC300 mono power amps I used the same R Cores in them and again the amps sound fantastic
These are the transformers R-CORE-TRANSFORMERS.pdf (shilchar.com)
voyager 13 by Alan Towell, on Flickr
IMG_3316 by Alan Towell, on Flickr
Alan
I also bought some and used them in my NCC200 voyagers which sounded excellent.
When I built my NCC300 mono power amps I used the same R Cores in them and again the amps sound fantastic
These are the transformers R-CORE-TRANSFORMERS.pdf (shilchar.com)
voyager 13 by Alan Towell, on FlickrIMG_3316 by Alan Towell, on FlickrAlan
PigletsDad
My intelligence test came back negative.
R cores give better isolation between windings, at the expense of increased leakage inductance. This means a supply using R-cores will tend to have a high impedance, but less noise. For front-end devices this is a clear win. If using separate supplies for the front-end of a power amp, like in a HackerNap, still a win. For the output stage, the PSRR is usually high, so noise is not such a big deal, but higher impedance will hurt power delivery into difficult loads.
All design is trade-offs, and I would pick a toroid for a high power application myself.
Finally, the external field of a toroid is low to very low, while you don't get such good geometrical cancellation for an R-core. This could be an issue if you put the transformer close to high gain circuits like a phono stage.
All design is trade-offs, and I would pick a toroid for a high power application myself.
Finally, the external field of a toroid is low to very low, while you don't get such good geometrical cancellation for an R-core. This could be an issue if you put the transformer close to high gain circuits like a phono stage.
Arkless Electronics
Trade: Amp design and repairs.
Transformers have no effect on sound quality. That simple.
I suppose, you are joking.
I just tried to change the toroidal trafos in my monoblocks.
Both ones have the same electric / physical parameters; but made by different brands.
One pair is Talema/Nuvotem; other pair is Noratel.
300VA each.
Some years ago I changed them because of the hum; now I inserted back the originals to check the sound/hum.
Hums are almost the same... a light difference, yes, really. Not very suprising.
Sound?
Enormous difference.
I can not produce this overtones-deviation via my 2k $ speakercable vs. 10 $ pure wire.
So, that's why I feel a small free area here to get a better sound.
Anyway, I see a little conspiration about toroid-glorification. The size and efficiency and field-insultation is OK. Brands know it very well; it's a serious economical question.
But what about the sound? I don't know yet.
If you show the toroidal winding system to a 6 years child, she/he will say 'Sunny'. Yes, there will be radial gaps always. Always. No solution; it's a geometrical issue. The old EI and the newer R has independent and no-gap wiring.
Arkless Electronics
Trade: Amp design and repairs.
ROFL
sq225917
Bit of this, bit of that
VA, impedance, physical hum and radiated electrical noise. Stuff em in a small box next to sensitive high gain circuits and it might just be audible, might.
Put em in a power amp 20cm from a circuit with 90db+ of psrr and it makes naff all difference unless they're run so hard the rails clip, or you saturate the core, in which case spec the right parts.
Put em in a power amp 20cm from a circuit with 90db+ of psrr and it makes naff all difference unless they're run so hard the rails clip, or you saturate the core, in which case spec the right parts.
S-Man
StrivingON
Badly designed toroids can emit considerable magnetic interference.
Magnetic saturation is caused by primary voltage (integrated over time) being too high. Too few primary turns, dc on the mains or operation at lower frequencies than intended can result in this problem.
Toroids tend to be designed to operate near to magnetic saturation in order to keep the dimensions small. They also saturate more sharply than EI types, which cannot operate as close to saturation because the magnetic path is constricted at the corners. When they saturate they tend to spray out a magnetic field with 100Hz + harmonic content related to the rectification.
I suspect a lot of the above relates to R-cores as well.
Quality and suitable design is more important than the type IMO (maybe
).
Magnetic saturation is caused by primary voltage (integrated over time) being too high. Too few primary turns, dc on the mains or operation at lower frequencies than intended can result in this problem.
Toroids tend to be designed to operate near to magnetic saturation in order to keep the dimensions small. They also saturate more sharply than EI types, which cannot operate as close to saturation because the magnetic path is constricted at the corners. When they saturate they tend to spray out a magnetic field with 100Hz + harmonic content related to the rectification.
I suspect a lot of the above relates to R-cores as well.
Quality and suitable design is more important than the type IMO (maybe
).
a.palfreyman
pfm Member
Some thoughts:
We all tend to over-spec transformers for DIY builds. My NCC200 based amp uses a pair of 250VA 25/0/25 transformers for essentially a 60W/ch rms output at full chat. If the amp is idling along at say 0.5-1W nominal output with transients of say 10-20W, then with bias currents etc the supply is only delivering say 5-6W typical for most of the time. Is it possible that what some hear with different transformers is the effect of partial saturation in the core at such low supply (transformer) loading? Do say Canterbury Windings transformers have a larger core than 'XYZ' transformer for the same VA rating? If so, that will probably 'get out of saturation' at a lower supply loading than 'XYZ' transformer?
We all tend to over-spec transformers for DIY builds. My NCC200 based amp uses a pair of 250VA 25/0/25 transformers for essentially a 60W/ch rms output at full chat. If the amp is idling along at say 0.5-1W nominal output with transients of say 10-20W, then with bias currents etc the supply is only delivering say 5-6W typical for most of the time. Is it possible that what some hear with different transformers is the effect of partial saturation in the core at such low supply (transformer) loading? Do say Canterbury Windings transformers have a larger core than 'XYZ' transformer for the same VA rating? If so, that will probably 'get out of saturation' at a lower supply loading than 'XYZ' transformer?
Tube OPT cores are extremly sensitive area; why should we think that PSU transformer cores have less importance?
What I tried to avoid the hum: 'O core' toroid; I ordered 1×600VA with two 35V outputs; (basically was 2×300VA with 4×35V) and asked 265V / 40V winding to get lower excitation.
Arrived well; output is exactly 35V; nice work and no warming and absolutely noiseless. No any hum. Never. Almost impossible to find out if 'on' or 'off'.
Sound?, well, finally I disliked. Not bad; excellent bass power, on rock music no question; but classical-mid section has no that miracle what I like.
So, now, the noisy, over-saturated toroids are at work.
Complicated job we have...
S-Man
StrivingON
Naim claims that a low impedance transformer will result in better sound because the smoothing caps charge faster and therefore the conduction angle is smaller. They reputedly have fairly noisy transformers, perhaps because they are running at high magnetic flux density to get the lowest possible impedance. This seems to be the opposite to your thoughts above, but tallies with @chord's observations.
I wish I knew what really matters in transformer design!
For low power applications I find a clear benefit in using EI cores.
For power amps, I have two favourite transformers: a 1KVA toroid which tends to be a bit of a hummer (quite possibly wound for lowest possible impedance) and Les' monster EI.
Just another thought...
Me and a pal tried some toroidal output transformers in his KT88 PP amp (I have the same amp). The measured bandwidth of the toroids was quite a bit better, which resulted in less overshoot on square waves. The toroids could also manage slightly higher power output due to their lower losses. The turns ratios were identical.
BUT... the "valve magic" had gone when we auditioned the amp fitted with the toroids
.
JensenHealey
pfm Member
Well, for one thing..
...tube OPT have to carry the full audio signal - a very wide frequency range - 1000:1 to cover 20Hz - 20KHz for instance.
The PSU transformer? - 50Hz only.
a.palfreyman
pfm Member
Lower distortion seemingly with your torroids...
We are listening the modified PSU voltage/current on our speakers.
Or not?
At tube amps it's not true.
S-Man
StrivingON
The current flow is very peaky, due to the non-linear load of the rectification and smoothing caps. There are plenty of other frequencies involved.
It's quite a complex system if all the parasitics are included.
DoctorRad
pfm Member
AIUI, one thing that will happen in this case is that the current draw from the mains will be higher for a shorter period of time, which would throw greater interference / voltage sag back onto the mains supply, possibly affecting other components.
There is a lot of analysis of the whys and wherefores of audio mains supplies - albeit in this case for a preamp - in the first part of Ben Duncan's AMP-01 text. The stuff about power supplies starts on page 53 of the PDF (page 55 of the magazine).