Tír Na HiFi

abraxalito wrote:John sent me a link over to this thread, said I had an admirer here so as an incorrigibly attention-seeking guy I couldn't resist signing up.

Glad to see you here.

We will all learn something from your participation.

jkeny wrote: Does this mean that is the impedance that the source will see driving the trafo & the load being driven will also see?

That's pretty much the same question I'm asking myself about them. Transformers look like an inductor to the driving circuit at lowish frequencies. The value of this inductance is sometimes called the 'shunt inductance' or 'magnetizing inductance' or 'primary inductance'. Given its an inductor the impedance of it depends on frequency. So the question is 'What frequency is the impedance at?'.

Being a 1:1 winding ratio means that it doesn't present a lighter load for the opamp output stage of a DAC to drive?

Right, there's no impedance transformation effect when the ratio is unity. So in practice the load will be somewhat heavier than with no transformer in circuit, due to the shunt inductance.

BTW, does the small sizes of these trafos (28X23X16mm) (18X14.5X13mm) suggest steel cores rather than ferrite?

The small size could be obtained by ferrite too. What suggests steel to me is tradition - I've not so far seen any vendor offering off the shelf ferrite cored transformers for audio use. Even where ferrite is traditionally used (switching power supplies) off the shelf solutions are extremely rare because in power supplies, each application is different. One technical reason ferrite isn't used is lowish inductance compared to steel.

jkeny wrote: My question is - I know what low frequency I want (20Hz) but how do I know what Volt-seconds I want?

Volt-seconds means a product of volts with seconds, multiplying them together. The lowest frequency the transformer needs to handle gives us the 'seconds' part of this product. The voltage that the transformer needs to handle at this lowest frequency gives us the 'volt' part.

Seeing as the thread's about input transformers to amps, the voltage an amplifier normally handles (for full output) is around 1VRMS. So if you want the amplifier to produce its full output voltage when 1VRMS is applied to the transformer the 'Volts' part of this equation becomes 1.41 (its the peak voltage we need). If you want to connect this transformer direct to a CD player (say in an application where you're using a digital volume control within the CD player) then you might want a higher voltage capability because usually CD players output 2VRMS maximum. But then again you might stick with 1VRMS and accept that at maximum you'll experience distortion - the transformer will saturate at low frequencies. The distortion the listener encounters will be a strong incentive to 'turn it down'.

A transformer that handles higher voltages needs more turns so as a DIYer it pays not to overspecify this voltage requirement.

How does V-s translate into what we hear?

At this point I get lost with your questions as in this context its about electrical signal levels, not sound. So I have no idea how to formulate an answer.

abraxalito wrote:

nige2000 wrote: gnd loop through monitor earth is a common one

Yes and also the monitor contains a switching power supply so its an 'aggressor' in EMC terms, throwing noise into the system.

as the one we tried on the soekris didn't work that great nor sounded good, think we concluded it was the low output of the soekris?
could be a cool solution to funky grounds in hifi

Right, didn't you mention the Soekris wasn't buffered so had an output impedance in the region of a kilo-ohm or something? Trafos really do need to be driven by lowish impedances, particularly those examples which were designed for headphones. An optimized line-level trafo would do somewhat better.

im not sure if i changed the soekris output from the stated stock "Single ended 1.4V RMS, Zout 640R" when i used 3.3 v for the r2r instead of stock 4v
i dont recall any loss of volume levels

i suppose i should be looking at a 1.5v rms capable transformer for this app?

is most dac chips circa 2vrms? meaning they will require at least that capability?

i think ive got a mild grasp of the concept although im completely lost in regards the method of physically winding the transformers ill have to youtube it :)

maybe if you did a an example solutions based on 2v rms

abraxalito wrote:

jkeny wrote: My question is - I know what low frequency I want (20Hz) but how do I know what Volt-seconds I want?

Volt-seconds means a product of volts with seconds, multiplying them together. The lowest frequency the transformer needs to handle gives us the 'seconds' part of this product. The voltage that the transformer needs to handle at this lowest frequency gives us the 'volt' part.

Seeing as the thread's about input transformers to amps, the voltage an amplifier normally handles (for full output) is around 1VRMS. So if you want the amplifier to produce its full output voltage when 1VRMS is applied to the transformer the 'Volts' part of this equation becomes 1.41 (its the peak voltage we need). If you want to connect this transformer direct to a CD player (say in an application where you're using a digital volume control within the CD player) then you might want a higher voltage capability because usually CD players output 2VRMS maximum. But then again you might stick with 1VRMS and accept that at maximum you'll experience distortion - the transformer will saturate at low frequencies. The distortion the listener encounters will be a strong incentive to 'turn it down'.

A transformer that handles higher voltages needs more turns so as a DIYer it pays not to overspecify this voltage requirement.

How does V-s translate into what we hear?

At this point I get lost with your questions as in this context its about electrical signal levels, not sound. So I have no idea how to formulate an answer.

OK, thanks
The reason why the rest of the question makes no sense is because I was interpreting V-s as V/s - in other words volts per second or the risetime of the trafo - it's speed in handling transients

jkeny wrote:

abraxalito wrote:

jkeny wrote: My question is - I know what low frequency I want (20Hz) but how do I know what Volt-seconds I want?

Volt-seconds means a product of volts with seconds, multiplying them together. The lowest frequency the transformer needs to handle gives us the 'seconds' part of this product. The voltage that the transformer needs to handle at this lowest frequency gives us the 'volt' part.

Seeing as the thread's about input transformers to amps, the voltage an amplifier normally handles (for full output) is around 1VRMS. So if you want the amplifier to produce its full output voltage when 1VRMS is applied to the transformer the 'Volts' part of this equation becomes 1.41 (its the peak voltage we need). If you want to connect this transformer direct to a CD player (say in an application where you're using a digital volume control within the CD player) then you might want a higher voltage capability because usually CD players output 2VRMS maximum. But then again you might stick with 1VRMS and accept that at maximum you'll experience distortion - the transformer will saturate at low frequencies. The distortion the listener encounters will be a strong incentive to 'turn it down'.

A transformer that handles higher voltages needs more turns so as a DIYer it pays not to overspecify this voltage requirement.

How does V-s translate into what we hear?

At this point I get lost with your questions as in this context its about electrical signal levels, not sound. So I have no idea how to formulate an answer.

OK, thanks
The reason why the rest of the question makes no sense is because I was interpreting V-s as V/s - in other words volts per second or the risetime of the trafo - it's speed in handling transients

i was thinking of it like a refresh rate

what v-s will you need?

not lookin forward to doin hundreds of turns with wire like thin fragile hair

nige2000 wrote: i was thinking of it like a refresh rate

what v-s will you need?

not lookin forward to doin hundreds of turns with wire like thin fragile hair

The V-s I would need is 2V at the DAC output but I would be happy that this dropped to 1V on the output of the trafo
Yea, I'm not too patient these days - as Richard will attest to :) - & the thought of winding 100s of turns of fine wire only to have it break near the end is my worst nightmare

I found these trafos on ebay which I checked with Richard - they are a steal @ €6 for 10 (including posting from HK) & seem too good to be true which he also thinks http://www.ebay.de/itm/10X-Audio-Transf ... 2151385804
They say 600:600 ohm impedance, nickel core, 800 turns of 0.06mm wire, etc for 0.50c each - hard to believe
Richard says that the primary inductance is too low to be driven by a normal line stage & are equivalent to his headphone trafo that we heard between his smartphone & headphones
Model: EI14
Material: Nickel alloy
AC impedance : EI14 600 : 600 Ohm
Item : Audio transformer
inductance : 290mH (±20%)
DC resistance : 135(±%)
Windability : Double-wire winding
Quality &wire diameter : QA-1 0.06MM
Primary coils : 800 turns
secondary coils : 800 turns
Alternating-current impedance value : 600
Size: Approx. 14x12x13mm(L*W*H)
Length of pin: 12mm

Does "Double-wire winding" mean bifilar winding i.e two wires wound together as one?
Although, depending on how they are wound (no varnish used, etc) it might be possible to unwind some of the 800 turns of wiring? Prefer unwinding to winding fine wire :)

They also have a 1300:8ohm version for the same price but I don't know how suitable they are - little info given about them: http://www.ebay.de/itm/10x-Audio-Transf ... 2216204859
Name: Audio transformer
Model: EE14
AC impedance : 1300:8
Size: about 13mm x 11mm x 6mm
Quantity:10pcs
Frequency Type: High Frequency

Zooming in on the pics seems to show they are wound on laminated sheets of ? - looks like there might be some rust in the pics (ferrous) - does the wiring that can be seen in some pics look like it's 0.06mm

Maybe he can say what use they might be suitable for & if unwinding some turns might be an easy way of modifying them?

primary is on the inside aint it?
if it is cant do anything with it

confused now..... on a step down is the secondary have more windings or less?

out of curiosity we could order them ebay ones

Did a quick search on Taobao for trafos like your eBay ones, found these -

https://item.taobao.com/item.htm?spm=a2 ... =13#detail

Do you think they look similar enough? Same EI14 size and they say they're 600ohms. If so I'll order some up as they're stupid cheap here (like 0.15euro) and reverse engineer one.

My worry is that the laminations will be held together by varnish so almost impossible to disassemble.... but it'll be a cheap experiment nevertheless.

nige2000 wrote: im not sure if i changed the soekris output from the stated stock "Single ended 1.4V RMS, Zout 640R" when i used 3.3 v for the r2r instead of stock 4v

Is the Soekris internally balanced? If so then a trafo which converted from balanced to unbalanced and adjusted the output level to suit whatever amp you're using would be ideal. I'd expect reducing the Vref to 3.3V from 4V would reduce the output level by under 2dB, not particularly noticeable.

i suppose i should be looking at a 1.5v rms capable transformer for this app?

Would be right if the Soekris is internally just one DAC element per channel.

is most dac chips circa 2vrms? meaning they will require at least that capability?

Most of the modern ones are like that, from ES9023 and onward. Other older ones have balanced outs they'd be better off having a transformer which uses the balanced capability which might be more than 2V in total.

maybe if you did a an example solutions based on 2v rms

Good idea, I'll see if I can work up a worked example. Tell me what the narrowest wire you're comfortable using is and I can choose a big enough core to accommodate that.