Wind them yourself is the lowest price option. Buy cores on Taobao - I've used EP19 and EI25 with good results, you'll need some very fine magnet wire.
Use an input transformer - has the added benefit of allowing SE outputs to drive balanced ins. Plus a reduction in CM noise.
+1
Audio Transformers By Cinemag Inc.
Wind them yourself is the lowest price option. Buy cores on Taobao - I've used EP19 and EI25 with good results, you'll need some very fine magnet wire.
I've tried winding inductors myself but I have Trump hands it seems. Have you bought any on Taobao?
With 24K input impedance, 2.2uF (3Hz corner) silver/oil caps will sound good. However, I've never tried transformers so I'd like to give that a go.
Nobody that I've seen on Taobao is offering finished transformers using ferrite cores. There are one or two who offer small steel cored trafos, tried a sample of that but not really suitable. Drop me a PM if you're interested in getting some ready-made ferrite cored ones as I have an idea for doing business in transformers 😀
PM sent ! 🙂 I saw these - seems to be what I need except for the impedance, so not really: https://world.taobao.com/item/36037344558.htm?spm=a312a.7700714.0.0.biJvkC#detail
Last edited:
Thanks! There are times when I think Taobao is doing a Jedi mind trick on me.


Those Taobao trafos look fine, assuming that the stated 35H primary inductance corresponds to reality 🙂 They probably won't have the best CMRR though as they look to have the primary wound over the secondary. On second thoughts it looks like they've addressed this with an interwinding screen connected to the laminations. Nice job - be sure to ground that pin to chassis and not to the local 0V
Save
Last edited:
Magnetic noise - yes for sure. Keep them well away from EI mains transformers. Toroids are less of a problem except where the wires lead-out. Turn the lead-outs to point well away from the audio trafos.
Wouldn't these transformers pickup noise easily as they aren't shielded?
I think you can order with the casing for magnetic shielding.
Attachments
Nice, you have a direct link?
https://world.taobao.com/item/541836460157.htm?spm=a312a.7728556.2014080708.6.3IDltC
https://world.taobao.com/item/54186...spm=a312a.7700824.w4002-10656758547.51.e13TwA
The correct link for the 10K:10K input transformer.
https://world.taobao.com/item/54186...spm=a312a.7700824.w4002-10656758547.51.e13TwA
The correct link for the 10K:10K input transformer.
@ Skylab,
Great find! I was looking for something like that sometime ago but could not find it (at TB that is). Thanks!
Edcor does have transformers like these for purchase but have to be customer ordered. The price is higher too of course. However, the Edcor transformer do have very detailed spec though.
Regards,
Cheers and happy new year. 🙂
Have you tried injecting some 4G signal either by proximity coupling or with an audio cable connected between 4g-device and amplifier?
I have not. What should I expect if I do? When we last spoke on this I did not have a 4G device but now I do.
Always this rise in distortion. What is creating it. What does it take in terms of eulution to get rid of it?
An externally hosted image should be here but it was not working when we last tested it.
http://www.3e-audio.com/wp-content/uploads/2017/01/THDN_vs_frequency_20W.jpg
As far as I am aware it is also a result of diminished loop gain/feedback being available for linearisation of the overall amplifier performance.
The low power distortion performance is excellent, even better at sub 1 watt levels (which is where most of us do our listening). The 3rd order rise only really becomes significant at higher output power levels.
I have wondered if a composite amplifier could be arranged, such as 'The modulus', using the additional capabilities of a high performance op amp to help linearise what the amplifier itself has trouble with.
I have not. What should I expect if I do? When we last spoke on this I did not have a 4G device but now I do.
Burst noise at least, especially coupled into the opamps.
- Home
- Amplifiers
- Class D
- TPA3251d2