Wire resistance
Hi Tyimo,
Practically, one works this out by calculation. Length of wire to resistance per kilometer.
I.e. X turns of average L length = Y meters.
COPPER ENAMELLED WIRE
Gauge Dia Area Max-Current Ohms/Km
30SWG 0.32 mm 0.08 mm² 120 mA 223.0
27SWG 0.40 mm 0.14 mm² 194 mA 141.4
25SWG 0.50 mm 0.20 mm² 304 mA 90.48
24SWG 0.56 mm 0.25 mm² 381 mA 72.15
22SWG 0.71 mm 0.40 mm² 613 mA 44.39
21SWG 0.80 mm 0.52 mm² 779 mA 35.35
19SWG 1.00 mm 0.66 mm² 1.21 A 22.63
18SWG 1.25 mm 0.82 mm² 1.91 A 14.48
16SWG 1.50 mm 2.00 mm² 2.74 A 09.95
Note for transformers one should think in terms of about half the average current since the wire is wound in a bobbin and so has greater self heating (than a straight wire).
Best wishes,
Susan.
P.S. SWG is NOT AWG. There are plenty of conversion tables etc on the web.
Hi Tyimo,
Tyimo said:
Practically, one works this out by calculation. Length of wire to resistance per kilometer.
I.e. X turns of average L length = Y meters.
COPPER ENAMELLED WIRE
Gauge Dia Area Max-Current Ohms/Km
30SWG 0.32 mm 0.08 mm² 120 mA 223.0
27SWG 0.40 mm 0.14 mm² 194 mA 141.4
25SWG 0.50 mm 0.20 mm² 304 mA 90.48
24SWG 0.56 mm 0.25 mm² 381 mA 72.15
22SWG 0.71 mm 0.40 mm² 613 mA 44.39
21SWG 0.80 mm 0.52 mm² 779 mA 35.35
19SWG 1.00 mm 0.66 mm² 1.21 A 22.63
18SWG 1.25 mm 0.82 mm² 1.91 A 14.48
16SWG 1.50 mm 2.00 mm² 2.74 A 09.95
Note for transformers one should think in terms of about half the average current since the wire is wound in a bobbin and so has greater self heating (than a straight wire).
Best wishes,
Susan.
P.S. SWG is NOT AWG. There are plenty of conversion tables etc on the web.
Winding resistance
Hi Tyimo,
It's how I calculate my winding resistances.
I don't have anything (at the moment) to be able to measure accurately below about 10 ohms as lower than this the resistance of the test leads and clips starts to become significant.
Trying to measure sub ohm resistances is a losing proposition using a standard ohm meter.
For those who might be wondering about this if one wishes to actually get a real measured value would be to:
1. Use a good 1 ohm power resistor (i.e. 1% if possible) in series with the winding and put 1 volt across the whole. Measure the total and the remainder across the winding. The ratio of voltages across the resistor and the winding will give the winding resistance.
2. Otherwise if you have a calibrated current source, you can use that and measure the derived voltage across the winding. That by calculation will give you the resistance.
As far as I recall, by my calculation one primary winding of my EI120 transformer (i.e. two half sections in series) is about 0.8 ohms.
Best wishes,
Susan.
Hi Tyimo,
Tyimo said:
It's how I calculate my winding resistances.
I don't have anything (at the moment) to be able to measure accurately below about 10 ohms as lower than this the resistance of the test leads and clips starts to become significant.
Trying to measure sub ohm resistances is a losing proposition using a standard ohm meter.
For those who might be wondering about this if one wishes to actually get a real measured value would be to:
1. Use a good 1 ohm power resistor (i.e. 1% if possible) in series with the winding and put 1 volt across the whole. Measure the total and the remainder across the winding. The ratio of voltages across the resistor and the winding will give the winding resistance.
2. Otherwise if you have a calibrated current source, you can use that and measure the derived voltage across the winding. That by calculation will give you the resistance.
As far as I recall, by my calculation one primary winding of my EI120 transformer (i.e. two half sections in series) is about 0.8 ohms.
Best wishes,
Susan.
Member
Joined 2009
Paid Member
Multi-Channel ?
Hi,
Forgive my barging in....I am looking to build a 5-channel set up for fairly sensitive single-driver DIY speakers; it would be fed from the analogue out signals of a video source (DVD, Blu-Ray etc.).. The Zeus amp appeals to me for many of the same reasons it has generated interest for others.
But at 5 channels we are talking more money than I want to invest (custom transformers - unless I can wind them or substitute something off-the-shelf) and it would be a large and heavy amplifier not to say the least.
But I wonder if some of these concerns can be relaxed if I don't need to go below 80Hz (sub woofer XO) and if I need relatively low power (20W) ?
Hi,
Forgive my barging in....I am looking to build a 5-channel set up for fairly sensitive single-driver DIY speakers; it would be fed from the analogue out signals of a video source (DVD, Blu-Ray etc.).. The Zeus amp appeals to me for many of the same reasons it has generated interest for others.
But at 5 channels we are talking more money than I want to invest (custom transformers - unless I can wind them or substitute something off-the-shelf) and it would be a large and heavy amplifier not to say the least.
But I wonder if some of these concerns can be relaxed if I don't need to go below 80Hz (sub woofer XO) and if I need relatively low power (20W) ?
Hi,
I use a 10K linear pot.
No, can't really define a universal preset position as it depends on mosfet and diode types. Easy enough to do, set to minimum, then increase whilst measuring current.
I usually include a 0.1 ohm power resistor in the power supply rail, so 100mV = 1 Amp.
Best wishes,
Susan.
hoktuna said:
I use a 10K linear pot.
No, can't really define a universal preset position as it depends on mosfet and diode types. Easy enough to do, set to minimum, then increase whilst measuring current.
I usually include a 0.1 ohm power resistor in the power supply rail, so 100mV = 1 Amp.
Best wishes,
Susan.
Susan,
Long time no hear. Hope life is treating you well.
> Pretty much all mosfets have transconductance specs ... First though I am writing some test software to drive my HP8903B audio analyzer via GPIB-LAN so I can do some full range testing.
If you have a functions generator and a 2-channel storage scope, you can test any MOSFET or JFET full range without anything else other than a 0.1-ohm power film resistor on a heatsink (to measure drain current). I have been doing this for the last 5 years with over 3000 devices (FETs) of all types.
And to measure below 10 ohms in DC, I use a current source (LM317 based) which I calibrate using a reference resistor. Then you can just measure voltage across the resistor (Kevin's connections) using a DMM, and calculate resistance. Works like a treat.
Patrick
Long time no hear. Hope life is treating you well.
> Pretty much all mosfets have transconductance specs ... First though I am writing some test software to drive my HP8903B audio analyzer via GPIB-LAN so I can do some full range testing.
If you have a functions generator and a 2-channel storage scope, you can test any MOSFET or JFET full range without anything else other than a 0.1-ohm power film resistor on a heatsink (to measure drain current). I have been doing this for the last 5 years with over 3000 devices (FETs) of all types.
And to measure below 10 ohms in DC, I use a current source (LM317 based) which I calibrate using a reference resistor. Then you can just measure voltage across the resistor (Kevin's connections) using a DMM, and calculate resistance. Works like a treat.
Patrick
Re: Multi-Channel ?
Hi.
No problem. Interesting project.
Yes, very much aware of this. Which is why I go for monoblocks.
If you don't need the lower bass then the Zeus35 type transformers should be fine. These amps were originally designed with my sphere speakers in mind. The 200VA transformers are a lot smaller and lighter, so getting five in cabinet would be okay.
However you will need enough heatsinking for all five channels.
If you use 4:1 mode you can try using lower biasing (e.g. 250 mA per mosfet) which will reduce the quiescent power. Distortion will be higher though, so a bit of a balancing act needed here.
If your optical player has volume control inbuilt, then you can do a straight PP line driver. This can use a power toroid for the line driver transformer (as it doesn't need a gap) and in swinging inductors configuration, so that would save you a bit of money.
Lundahl (or Sowter) input transformers are really recommended for the line driver. Lundahl mixing desk line transformers quite often come up on that auction site people use from time to time
Best wishes,
Susan.
Hi.
Bigun said:
No problem. Interesting project.
Yes, very much aware of this. Which is why I go for monoblocks.
If you don't need the lower bass then the Zeus35 type transformers should be fine. These amps were originally designed with my sphere speakers in mind. The 200VA transformers are a lot smaller and lighter, so getting five in cabinet would be okay.
However you will need enough heatsinking for all five channels.
If you use 4:1 mode you can try using lower biasing (e.g. 250 mA per mosfet) which will reduce the quiescent power. Distortion will be higher though, so a bit of a balancing act needed here.
If your optical player has volume control inbuilt, then you can do a straight PP line driver. This can use a power toroid for the line driver transformer (as it doesn't need a gap) and in swinging inductors configuration, so that would save you a bit of money.
Lundahl (or Sowter) input transformers are really recommended for the line driver. Lundahl mixing desk line transformers quite often come up on that auction site people use from time to time
Best wishes,
Susan.
Member
Joined 2009
Paid Member
Susan,
Thanks for the quick response.
The Class-A heat sinking doesn't worry me (perhaps a case of too little knowledge being dangerous) but the cost/weight of the transformers do. The possibility of the simpler toroids looks promising.
I don't yet have the video source. I'm looking at Blu-ray with built in decoders and analogue out. But it has to be good analogue out
p.s. I'm just realizing that DIY amplifiers are tricky. Because you really have to build them and listen which means committing time and money. Technical specifications alone aren't sufficient and personal descriptions are very subjective. I think your amplifier is interesting because it's different, leaving open enough mystery that we hope for some magic in the sound that is hard to find off-the-shelf.
Thanks for the quick response.
The Class-A heat sinking doesn't worry me (perhaps a case of too little knowledge being dangerous) but the cost/weight of the transformers do. The possibility of the simpler toroids looks promising.
I don't yet have the video source. I'm looking at Blu-ray with built in decoders and analogue out. But it has to be good analogue out
p.s. I'm just realizing that DIY amplifiers are tricky. Because you really have to build them and listen which means committing time and money. Technical specifications alone aren't sufficient and personal descriptions are very subjective. I think your amplifier is interesting because it's different, leaving open enough mystery that we hope for some magic in the sound that is hard to find off-the-shelf.
Hi Patrick,
Very busy at the moment, not much personal time to spend in front of a computer (although I seem to do a lot of that at work).
Yes, thank you. Have been working full time since last May (2008), and slowly catching up with the ravages of having been self employed for so long.
I am looking at setting up something using my sound card since the resolution needed to precisely match parts is going to be difficult to achieve with my old Telequipment storage scope (still works, but is a bit twitchy).
Yes, that also works. Trick is getting a good reference resistor to start the measurement chain with. HP make some nice ones that can be picked up secondhand.
How are things with you?
Best wishes,
Susan.
Very busy at the moment, not much personal time to spend in front of a computer (although I seem to do a lot of that at work).
EUVL said:Susan, Long time no hear. Hope life is treating you well.
Yes, thank you. Have been working full time since last May (2008), and slowly catching up with the ravages of having been self employed for so long.
I am looking at setting up something using my sound card since the resolution needed to precisely match parts is going to be difficult to achieve with my old Telequipment storage scope (still works, but is a bit twitchy).
Yes, that also works. Trick is getting a good reference resistor to start the measurement chain with. HP make some nice ones that can be picked up secondhand.
How are things with you?
Best wishes,
Susan.
Hi Bigun,
Not so quick this time - so much to do, tick tock etc.
For the line driver in PP its a possibility to use off the shelf parts, for the output transformer a custom wound part is required to get the required ratios.
It is just possible that you might be able to find a pair a 300VA or so toroids that have four instead of two secondaries.
If you start looking at bespoke items like Plitron...
... the Sowter offerings suddenly seem a lot cheaper.
Note I haven't personally used any Plitron parts, and for their nominal solid state PP OP TX transformer the impedance ratio is too high.
Lundahl have a matching transformer, but then the pricing is such that there isn't really a big difference to the Sowter parts either.
BTW, if you have a supply of M6 EI120 laminations and the other bits of hardware for the big transformers you can always ask if you could special order just the wound bobbins from Sowter. You will save a bit on the price plus more on the carriage.
I would look at the new Oppo BDP-83 Blu-ray Disc player.
It's just about the only one that will play anything including SACD. It should be a good starting point, and you can upgrade the DACs later if needs be.
I modded my DV980H with a pair of PCM1794As, and the result is very good.
Yes
Thank you.
Best wishes,
Susan.
Bigun said:
Not so quick this time - so much to do, tick tock etc.
For the line driver in PP its a possibility to use off the shelf parts, for the output transformer a custom wound part is required to get the required ratios.
It is just possible that you might be able to find a pair a 300VA or so toroids that have four instead of two secondaries.
If you start looking at bespoke items like Plitron...
... the Sowter offerings suddenly seem a lot cheaper.
Note I haven't personally used any Plitron parts, and for their nominal solid state PP OP TX transformer the impedance ratio is too high.
Lundahl have a matching transformer, but then the pricing is such that there isn't really a big difference to the Sowter parts either.
BTW, if you have a supply of M6 EI120 laminations and the other bits of hardware for the big transformers you can always ask if you could special order just the wound bobbins from Sowter. You will save a bit on the price plus more on the carriage.
I don't yet have the video source. I'm looking at Blu-ray with built in decoders and analogue out. But it has to be good analogue out
I would look at the new Oppo BDP-83 Blu-ray Disc player.
It's just about the only one that will play anything including SACD. It should be a good starting point, and you can upgrade the DACs later if needs be.
I modded my DV980H with a pair of PCM1794As, and the result is very good.
Yes
Thank you.
Best wishes,
Susan.
Member
Joined 2009
Paid Member
Hi Susan,
I am going to pursue a SS amp solution in parallel. This means a steep learning curve for me as my background is not EE by physics (and that was decades ago). The advantage is that it's a fairly popular and well known approach so I can get lots of support during this learning.
I have not given up on the Zeus amp however...
I have a colleague, a big tube fan who was talking eagerly about a SE 'static induction transistor - transformer' amplifier today. I showed him your web site and he recognized your amplifier as kindred spirit. So I may have a partner in crime. I also learned that Hammond are a few miles down the road from where I live so perhaps something can be arranged.
I am going to pursue a SS amp solution in parallel. This means a steep learning curve for me as my background is not EE by physics (and that was decades ago). The advantage is that it's a fairly popular and well known approach so I can get lots of support during this learning.
I have not given up on the Zeus amp however...
I have a colleague, a big tube fan who was talking eagerly about a SE 'static induction transistor - transformer' amplifier today. I showed him your web site and he recognized your amplifier as kindred spirit. So I may have a partner in crime. I also learned that Hammond are a few miles down the road from where I live so perhaps something can be arranged.
Quadfilar or no?
Firstly Susan, I greatly admire your dedication and support for diy audio folk. Your amplifier design looks excellent, I appreciate just how many hours of work you're sharing. Please let people know what we can do in return.
I'm an op amp designer (bsee 1979). We got little or no education in transformers, particularly practical aspects.
I'm headed toward building a couple of Zeus amps to drive some A166 folded horn speakers (also yet to be built). I expect this project will run into next winter due to the items queued up first. I'm sure I will have more questions, but one specific one here (and enough wordy introduction).
I'm talking to a transformer company here in the USA that will sell me laminations and such (I will share their name if they allow--I don't know if they see me as profit or just good-will). They were recommending I buy one of their standard units until I got to the quad-filar part. If I follow correctly, the filar winding lowers leakage inductance? My speakers will be 8 Ohms (Fostex FE166E). They're also fairly efficient, I probably won't be using more than a fraction of the amp's power. What happens if I go with a more conventional layered winding?
Thank you,
Steve Smith
Brownfield, Maine
Firstly Susan, I greatly admire your dedication and support for diy audio folk. Your amplifier design looks excellent, I appreciate just how many hours of work you're sharing. Please let people know what we can do in return.
I'm an op amp designer (bsee 1979). We got little or no education in transformers, particularly practical aspects.
I'm headed toward building a couple of Zeus amps to drive some A166 folded horn speakers (also yet to be built). I expect this project will run into next winter due to the items queued up first. I'm sure I will have more questions, but one specific one here (and enough wordy introduction).
I'm talking to a transformer company here in the USA that will sell me laminations and such (I will share their name if they allow--I don't know if they see me as profit or just good-will). They were recommending I buy one of their standard units until I got to the quad-filar part. If I follow correctly, the filar winding lowers leakage inductance? My speakers will be 8 Ohms (Fostex FE166E). They're also fairly efficient, I probably won't be using more than a fraction of the amp's power. What happens if I go with a more conventional layered winding?
Thank you,
Steve Smith
Brownfield, Maine
Hi Gareth,
Thank you for your posts.
Anything that involves building something for oneself is to be encouraged.
My design is a bit simpler than most, but still as is often the case the devil is in the details and for my amps one of the devils is the use of good quality iron.
Ah! Having someone to share the soldering iron with is often helpful.
Let me know how you get on and be assured that I always do my best to answer any questions.
Best wishes,
Susan.
Thank you for your posts.
Bigun said:
Anything that involves building something for oneself is to be encouraged.
My design is a bit simpler than most, but still as is often the case the devil is in the details and for my amps one of the devils is the use of good quality iron.
Ah! Having someone to share the soldering iron with is often helpful.
Let me know how you get on and be assured that I always do my best to answer any questions.
Best wishes,
Susan.
Re: Quadfilar or no?
Hi Steve,
Thank you.
Transformers are often seen as a bit of dark art. For power engineering, electrical grid distribution and HV pulse work they are covered quite well in the literature. However getting info on wide band audio transformers can be another matter.
The use of quad filar winding is to get the best possible coupling between each of the primaries and between the primaries and the secondary.
This type of winding is only suitable because of the way I split up the windings and so each section is coupled 1:1. Normal tube transformers have different ratios and lots more turns, so have to be constructed differently.
The end result is an output stage that has a bandwidth of about 2 MHz (with standard MOSFETs). Now this might perhaps seem excessive, however it gives a good follower response to the reactive load of the speaker driver(s) and also helps maintain phase information in the harmonics.
Also using multi filar windings makes it a lot easier to get perfect match sections.
If you don't need lots of power i.e. 12W or so will do you, then 4:1 mode gives best performance and is the configuration I would suggest for horns.
Best wishes,
Susan.
Hi Steve,
sos said:
Thank you.
Transformers are often seen as a bit of dark art. For power engineering, electrical grid distribution and HV pulse work they are covered quite well in the literature. However getting info on wide band audio transformers can be another matter.
The use of quad filar winding is to get the best possible coupling between each of the primaries and between the primaries and the secondary.
This type of winding is only suitable because of the way I split up the windings and so each section is coupled 1:1. Normal tube transformers have different ratios and lots more turns, so have to be constructed differently.
The end result is an output stage that has a bandwidth of about 2 MHz (with standard MOSFETs). Now this might perhaps seem excessive, however it gives a good follower response to the reactive load of the speaker driver(s) and also helps maintain phase information in the harmonics.
Also using multi filar windings makes it a lot easier to get perfect match sections.
If you don't need lots of power i.e. 12W or so will do you, then 4:1 mode gives best performance and is the configuration I would suggest for horns.
Best wishes,
Susan.
Susan,
Good to hear that life is treating you well.
I am, probably like many others, busy trying to survive the economic crisis which hits the semicon industry particularly hard. That is the down side.
The up side is, I have more time for more crazy ideas. Here are a couple, a lot more to come in the next months :
http://www.diyaudio.com/forums/showthread.php?s=&threadid=135154&perpage=25&pagenumber=1
http://www.diyaudio.com/forums/showthread.php?s=&threadid=140488
Remember we talked about a transportless player years ago. Someone is making my dream come true :
http://www.diyaudio.com/forums/showthread.php?s=&threadid=140538
I cannot think of a better solution.
Patrick
Good to hear that life is treating you well.
I am, probably like many others, busy trying to survive the economic crisis which hits the semicon industry particularly hard. That is the down side.
The up side is, I have more time for more crazy ideas. Here are a couple, a lot more to come in the next months :
http://www.diyaudio.com/forums/showthread.php?s=&threadid=135154&perpage=25&pagenumber=1
http://www.diyaudio.com/forums/showthread.php?s=&threadid=140488
Remember we talked about a transportless player years ago. Someone is making my dream come true :
http://www.diyaudio.com/forums/showthread.php?s=&threadid=140538
I cannot think of a better solution.
Patrick
Transformer construction
I'm having a hard time convincing myself that I follow the winding details. Quadfilar is fine, that's not my question.
The 75 W output transformer page says:
"Both sides to have identical number of turns and length of wire i.e. mirror image twins."
Identical number of turns and length makes sense. Mirror image twins to me means that the left side of the bobbin should be wound in one direction and the right side the opposite direction.
I think I'm over interpreting--maybe this isn't what is meant by 'mirror image'. I'm not seeing a big advantage to having two winding directions; I'd appreciate either the details of why it is important or confirmation that it isn't crucial. The single chamber bobbin version of the transformer certainly doesn't do mirror image winding for half the turns!
Another question on the schematic. The transformer has three fat black bars, top, middle and bottom. This is the bobbin, right?
Simple questions,
Thanks,
Steve
I'm having a hard time convincing myself that I follow the winding details. Quadfilar is fine, that's not my question.
The 75 W output transformer page says:
"Both sides to have identical number of turns and length of wire i.e. mirror image twins."
Identical number of turns and length makes sense. Mirror image twins to me means that the left side of the bobbin should be wound in one direction and the right side the opposite direction.
I think I'm over interpreting--maybe this isn't what is meant by 'mirror image'. I'm not seeing a big advantage to having two winding directions; I'd appreciate either the details of why it is important or confirmation that it isn't crucial. The single chamber bobbin version of the transformer certainly doesn't do mirror image winding for half the turns!
Another question on the schematic. The transformer has three fat black bars, top, middle and bottom. This is the bobbin, right?
Simple questions,
Thanks,
Steve