metalwork
Hi Susan,
A couple of quick metalwork questions:
(1) Where did you get the metal boxes used for the 75W version? I've tried the Rittal boxes from Farnell, but the finish is rougher and the boxes aren't exactly the same as the ones you have shown.
(2) Your original black cases for the low power version are great looking. I realise it's quite a while ago but can you give me a pointer for a supplier?
Thank,
James
Hi Susan,
A couple of quick metalwork questions:
(1) Where did you get the metal boxes used for the 75W version? I've tried the Rittal boxes from Farnell, but the finish is rougher and the boxes aren't exactly the same as the ones you have shown.
(2) Your original black cases for the low power version are great looking. I realise it's quite a while ago but can you give me a pointer for a supplier?
Thank,
James
A question, mostly for Susan, any reason not to use a negative supply and p-channel MOSFETs?
Peace
Peace
Dear All,
Thanks for the input on PC sound cards.
I chose the E-Mu 1212M as it is 192 kHz sample rate so I can do FFTs up to 96 kHz.
Unfortunatly it is not so good at low levels:
but I suspect most sound cards will be like this.
(sorry for the big pic, but anything smaller doesn't show the detail.)
=========
Hi wrenchone,
Yes, I took a deep breath and spent on the 24 bit sig generator, plus I bought the dongle as well.
It really comes into it's own with looking at stuff like the intermodullation between the fundament and the harmonics.
BTW One can run multiple 24 bit generators to do fancy intermodulation tests.
Plus selecting the resolution from 16 to 24 bits is quite interesting.
Best wishes,
Susan.
Thanks for the input on PC sound cards.
I chose the E-Mu 1212M as it is 192 kHz sample rate so I can do FFTs up to 96 kHz.
Unfortunatly it is not so good at low levels:
but I suspect most sound cards will be like this.
(sorry for the big pic, but anything smaller doesn't show the detail.)
=========
Hi wrenchone,
wrenchone said:
Yes, I took a deep breath and spent on the 24 bit sig generator, plus I bought the dongle as well.
It really comes into it's own with looking at stuff like the intermodullation between the fundament and the harmonics.
BTW One can run multiple 24 bit generators to do fancy intermodulation tests.
Plus selecting the resolution from 16 to 24 bits is quite interesting.
Best wishes,
Susan.
Re: metalwork
Hi James,
1. Metal boxes for Zeus 75 are from RS - who are offline until Sunday afternoon 🙁 .
RS # 258-6714 - 150 x 300 x 80
RS # 258-6736 - 200 x 300 x 80
They are Schroff "Inline" terminal enclosures.
I use them "upside down" and remove the sealing gasket so that there is airflow from the base. The handles are fitted using the holes normally for wall mounting.
2. West Hyde metal enclosures from Farnell, Instrument "Image" in steel.
Farnell # 939-808 - 400 x 250 x 90 - UKP 47.75 + tax.
I used different feet to those supplied. But these are no longer made so I am going to have to turn the next lot out of aluminium bar.
Best wishes,
Susan.
Hi James,
nemestra said:
1. Metal boxes for Zeus 75 are from RS - who are offline until Sunday afternoon 🙁 .
RS # 258-6714 - 150 x 300 x 80
RS # 258-6736 - 200 x 300 x 80
They are Schroff "Inline" terminal enclosures.
I use them "upside down" and remove the sealing gasket so that there is airflow from the base. The handles are fitted using the holes normally for wall mounting.
2. West Hyde metal enclosures from Farnell, Instrument "Image" in steel.
Farnell # 939-808 - 400 x 250 x 90 - UKP 47.75 + tax.
I used different feet to those supplied. But these are no longer made so I am going to have to turn the next lot out of aluminium bar.
Best wishes,
Susan.
Roscoe Primrose said:
Actually, I was thinking (Boy that darn smoke detector is annoying) about that. What about suing a negative supply with n-channel MOSFETs? That'd allow you to do away with the mica insulators on the MOSFETs, increasing your heatsink effectiveness... Of course, your input and output trannies now have to have good enough insulation for your power supply voltage, and if you're using the output as a choke rather than a transformer you'll have power supply rail voltage at your speaker output terminals (probably a bad thing in EU terms)... Food for thought...
Anyway, I ordered 32 IRFZ48Vs from DigiKey (0.84USD/each) today. They're only 60V units, but the RDS is REAL low (0.012ohm). I don't need much power since my speakers are about 104dB/w/m, I'll probably run about a 24V supply, I've got a couple 24Vct/4A transformers just begging for something to do...
Peace
Hi Roscoe,
Thanks for your posts.
Sorry the matchingwasn't so good.
I have run with 100 mV diference between the mosfets. Not so good at low level but it does work.
The ST mosfets are relativly cheep and I have been fortunate in getting 3 matched pairs out of only 10 parts if you can fit the plastic package instead of the TO3s.
As far as I can see there is no reason at all. Should work just as well upside down.
N parts are more readily available with more types to choose from (and better for negative ground vehicles if one is going mobile).
But if you have some P-channel then why not?
V-bias is obviously negative as well.
Please let us know how you get on 🙂
Thanks.
Best wishes,
Susan.
Thanks for your posts.
Sorry the matchingwasn't so good.
I have run with 100 mV diference between the mosfets. Not so good at low level but it does work.
The ST mosfets are relativly cheep and I have been fortunate in getting 3 matched pairs out of only 10 parts if you can fit the plastic package instead of the TO3s.
Roscoe Primrose said:
As far as I can see there is no reason at all. Should work just as well upside down.
N parts are more readily available with more types to choose from (and better for negative ground vehicles if one is going mobile).
But if you have some P-channel then why not?
V-bias is obviously negative as well.
Please let us know how you get on 🙂
Thanks.
Best wishes,
Susan.
Roscoe Primrose said:
Well, I got to spend some more time listening to it today. The CD walkman I had initially used just wasn't up to the task, it worked a lot better with the headphone output on my iBook. It really does show some promise, although it seems a little rolled off at the high-end, now the trick is going to be figuring out whether the input toroid or the output toroid is the biggest offender. If I can dig out and older Cd player with a more hefty headphone output (i.e. one that isn't severely volume limited for liability reasons) I have some very nice 600:5Kct transformers I can substitute for the inputs, don't currently have any good in-house choices for outputs... I'll get more serious about transformers after I build the IRFZ48V-based monoblocks, I'll actually have enough room then, which is a real problem now...
Peace
Hi Roscoe,
Thanks for your further post and thoughts 🙂
Yes, this could be done or if one has an internal enclosed heatsink (as many domestic amps seem to have these days) one could have the heatsink at V+ volts. Running at 40 to 50 volts is not normally a lethal hazard.
I would agree that from a EU regulations point of view having a large DC offset floating around on speaker leads would probably get frowned upon.
I haven't gone in that direction as I have tried to keep everything as simple and straightforward as possible to make the design as accessible (and safe) as I can for everyone, even those with only basic skills.
The Joy of DIYaudio is that there are as many ways of making a project as there are people who reach for their soldering irons (and for my amp one could make the whole thing using terminal blocks so even soldering skills aren't required).
People with more experience can of course elaborate as they wish 🙂
Excellent.
Don't forget that the mosfets can swing twice the power rail because of the swinging inductor transformer action, so a 30 volt supply is actually a 60 volt supply as far as the mosfets are concerned.
I use 60 volt STP55NE06 mosfets for the preamp/line driver (I got a good matched pair from the first three I shook from the tube of 25) running with about +24 volts DC.
As you are thinking of using relatively low voltages you could also consider running your amplifier off a pair of 12 volt lead acid batteries (deep cycle are better if to hand) which should greatly improve the mains cycle rejection as you are using very sensitive speakers (which type may I ask?).
From my full range FFTs (43Hz to 96 kHz) with a noise floor down at -150 dB one can clearly see the mains cycle harmonics running up into the kilohertz (plus the computer/display switch mode RFI higher up). This is why I decided to build my Zeus 75 amp using separate steel boxes to shield the supply and the amp.
Best wishes,
Susan.
Thanks for your further post and thoughts 🙂
Roscoe Primrose said:
Yes, this could be done or if one has an internal enclosed heatsink (as many domestic amps seem to have these days) one could have the heatsink at V+ volts. Running at 40 to 50 volts is not normally a lethal hazard.
I would agree that from a EU regulations point of view having a large DC offset floating around on speaker leads would probably get frowned upon.
I haven't gone in that direction as I have tried to keep everything as simple and straightforward as possible to make the design as accessible (and safe) as I can for everyone, even those with only basic skills.
The Joy of DIYaudio is that there are as many ways of making a project as there are people who reach for their soldering irons (and for my amp one could make the whole thing using terminal blocks so even soldering skills aren't required).
People with more experience can of course elaborate as they wish 🙂
Excellent.
Don't forget that the mosfets can swing twice the power rail because of the swinging inductor transformer action, so a 30 volt supply is actually a 60 volt supply as far as the mosfets are concerned.
I use 60 volt STP55NE06 mosfets for the preamp/line driver (I got a good matched pair from the first three I shook from the tube of 25) running with about +24 volts DC.
As you are thinking of using relatively low voltages you could also consider running your amplifier off a pair of 12 volt lead acid batteries (deep cycle are better if to hand) which should greatly improve the mains cycle rejection as you are using very sensitive speakers (which type may I ask?).
From my full range FFTs (43Hz to 96 kHz) with a noise floor down at -150 dB one can clearly see the mains cycle harmonics running up into the kilohertz (plus the computer/display switch mode RFI higher up). This is why I decided to build my Zeus 75 amp using separate steel boxes to shield the supply and the amp.
Best wishes,
Susan.
Hi Roscoe,
I would suspect the input toroid placing a load on the headphone driver.
The output toroid should be good for at least 50 kHz. The input perhaps 25 kHz. But all this depends on the way they have been made and the core type, some transformers (including most mains EI types) just don't work.
There isn't much step up in these for a single stage amp but if you don't get enough gain they would definitely make good input transformers for a 2 stage amplifier, using the 15VA toroids as the interstage.
I have been considering 115+115 : 55+55 mains toroids as a possibility as the big ones seem to work okay to over 20 kHz even when used as proper transformers not just swinging inductors.
However they start to get expensive if paying full distro price and at this point making proper quad-filar wound output transformers would be better.
If you only want 1:1 or 2:1 step down then quad-filar winding with 18 AWG magnet wire by hand is not difficult and can be done "in the hand". The 1:4 with dual chamber bobbins is a bit more fiddly and really needs to be mounted on a spindle with a turns counter to match the two sides..
... bench space, the final frontier!
Best wishes,
Susan.
Roscoe Primrose said:
I would suspect the input toroid placing a load on the headphone driver.
The output toroid should be good for at least 50 kHz. The input perhaps 25 kHz. But all this depends on the way they have been made and the core type, some transformers (including most mains EI types) just don't work.
There isn't much step up in these for a single stage amp but if you don't get enough gain they would definitely make good input transformers for a 2 stage amplifier, using the 15VA toroids as the interstage.
I have been considering 115+115 : 55+55 mains toroids as a possibility as the big ones seem to work okay to over 20 kHz even when used as proper transformers not just swinging inductors.
However they start to get expensive if paying full distro price and at this point making proper quad-filar wound output transformers would be better.
If you only want 1:1 or 2:1 step down then quad-filar winding with 18 AWG magnet wire by hand is not difficult and can be done "in the hand". The 1:4 with dual chamber bobbins is a bit more fiddly and really needs to be mounted on a spindle with a turns counter to match the two sides..
... bench space, the final frontier!
Best wishes,
Susan.
Susan-Parker said:
Actually, I was thinking of using an LT1084CP (5A adjustable regulator) at about 24V. Speakers are Altec 604-8Gs on open baffles, with Linkwitz Phoenix-derived subs below 45hz...
Susan-Parker said:
Probably...
Well, it turns out I misremebered, the 600:5Ks are center-tapped on the 600 side, not the 5K side. I did pull some 600ct:20Kct out of a box, they're in there now, and a huge improvement over the 15VA toriods. Have them wired up with only half the 600 winding (150:20Kct) at this point. I seem to be missing a bit of the bass below 40Hz, but other than that the sound is quite good now... With these transformers the difference between the cheap CD walkman and the iBook is VERY noticeable.
Turns out the 225VA output toriods I'm using are 115+115:18+18 (I just found the data sheets this AM) and I'd been running the secondaries in parallel. With a stepdown of almost 13:1, it's no wonder I wasn't getting much output
I got the outputs surplus, so they were cheap (and, I dodn't buy them for THIS project, so the amp I have running now is still a no-cost object (as opposed to cost is no object 😉)).More like chassis space, the final frontier
Peace
…
I even came further… and got excellent sound without the power supply;
Got two 300W toroidal transformers “pri” 220V, “sec” 2X12V,
Pri I connected to CD phones output, sec into a wideband high efficiency speaker.
And you know what?
I HAVE NEVER BEFORE HEARD SO GOOD SOUNDING DEVICES.
The same situation was when I connected pri to line output of CD.
Of course the pressure of the sound was not too high, but for night-listening, when the city noise calmed, it was even too high, so then I had to connect the transformers via 470k Alps potentiometers.
I was astonished that such a simple act caused such a good result. Of course it is not a brave idea for designing, but it might be useful for some reasons/occasions. Maybe it was a primitive way, but the results were good.
Probably the output transformer from a tube amplifier would be a better choice, however it would be harder to get, and more expensive.
I am curious if the 2,3-way speakers with the filters, like are general used everywhere, will sound loud enough? Could anyone try and write?
Atached is this little stupid circuit
I even came further… and got excellent sound without the power supply;
Got two 300W toroidal transformers “pri” 220V, “sec” 2X12V,
Pri I connected to CD phones output, sec into a wideband high efficiency speaker.
And you know what?
I HAVE NEVER BEFORE HEARD SO GOOD SOUNDING DEVICES.
The same situation was when I connected pri to line output of CD.
Of course the pressure of the sound was not too high, but for night-listening, when the city noise calmed, it was even too high, so then I had to connect the transformers via 470k Alps potentiometers.
I was astonished that such a simple act caused such a good result. Of course it is not a brave idea for designing, but it might be useful for some reasons/occasions. Maybe it was a primitive way, but the results were good.
Probably the output transformer from a tube amplifier would be a better choice, however it would be harder to get, and more expensive.
I am curious if the 2,3-way speakers with the filters, like are general used everywhere, will sound loud enough? Could anyone try and write?
Atached is this little stupid circuit
Hi Roscoe,
Thanks for your further post 🙂
Sounds good to me.
Ah, nice.
Those subs look interesting.
The toroids are fine for trying the amp out, but properly designed wideband audio transformers do normally work better.
I find that the bass is still there, but the quality is different. It's much tighter.
However I don't have anything like the subs you use (our living room is too small for a start).
🙂
Ah!
This is very pleasing as one of the nice things about this design is that it can be made from "bits to hand".
Understood.
Since using high resolution FFTs I have become much more aware of the effects of mains power line frequencies and their harmonics being picked up by the audio. This is why I have gone over to separate steel cases for the power supplies.
Best wishes,
Susan.
Thanks for your further post 🙂
Roscoe Primrose said:
Sounds good to me.
Ah, nice.
Those subs look interesting.
The toroids are fine for trying the amp out, but properly designed wideband audio transformers do normally work better.
I find that the bass is still there, but the quality is different. It's much tighter.
However I don't have anything like the subs you use (our living room is too small for a start).
🙂
Ah!
This is very pleasing as one of the nice things about this design is that it can be made from "bits to hand".
Understood.
Since using high resolution FFTs I have become much more aware of the effects of mains power line frequencies and their harmonics being picked up by the audio. This is why I have gone over to separate steel cases for the power supplies.
Best wishes,
Susan.
Hi Padamiecki,
Thanks for your post and sharing your findings 🙂
The transformers are impedance matching and trading volts for current.
Many outputs give a substantially better performance with a higher impedance load (which reduces current) particularly when being used with a good voltage swing.
Transformers seem to have had a bad press but they do have their virtues.
Not stupid, neat 🙂
BW,
Susan.
Thanks for your post and sharing your findings 🙂
padamiecki said:
The transformers are impedance matching and trading volts for current.
Many outputs give a substantially better performance with a higher impedance load (which reduces current) particularly when being used with a good voltage swing.
Transformers seem to have had a bad press but they do have their virtues.
Not stupid, neat 🙂
BW,
Susan.
Hi Susan! Thank you for learning me to wind my own transformers/chokes.
Here is a picture of my latest centertapped bifilar winded choke. I think it would be good for about a 400 watt Zeus(-ish)-amp.Im thinking of building one of those to, but not with this one.
Here is a picture of my latest centertapped bifilar winded choke. I think it would be good for about a 400 watt Zeus(-ish)-amp.Im thinking of building one of those to, but not with this one.
An externally hosted image should be here but it was not working when we last tested it.
Well, I got the first one of my purpose-built Zeus-type monoblocks built, hopefully I'll have time to finish the other one tomorrow...
Peace
An externally hosted image should be here but it was not working when we last tested it.
Peace
Hi Circlomanen,
My pleasure.
Thanks for the picture. That is a fine looking bit of iron 🙂
Even allowing for perspective I imagine it must weigh at least 10 kilos.
Best wishes,
Susan.
Circlomanen said:
My pleasure.
Thanks for the picture. That is a fine looking bit of iron 🙂
Even allowing for perspective I imagine it must weigh at least 10 kilos.
Best wishes,
Susan.
Hi Roscoe,
Thanks for your post and picture 🙂
Looks good.
Did you fold up your chassis yourself?
I have a small bench brake, but it is only for 12" / 300 mm max, which includes the sides and any turnovers. Still it is useful to have despite being very heavy for it's size (I can't lift the 24" version).
Hope your progress has been fruitful.
Best wishes,
Susan.
Thanks for your post and picture 🙂
Roscoe Primrose said:
Looks good.
Did you fold up your chassis yourself?
I have a small bench brake, but it is only for 12" / 300 mm max, which includes the sides and any turnovers. Still it is useful to have despite being very heavy for it's size (I can't lift the 24" version).
Hope your progress has been fruitful.
Best wishes,
Susan.
Susan-Parker said:
No, it's a Hammond chassis, cheap, but they work...
I hate metalwork. If it wasn't for that, I'd have a lot more amps around here (the wife thinks I already have way too many, so maybe my aversion to metalwork is a good thing...).
I've made some progress. Got the metalwork for the second amp done today, and got most of the major components mounted. Got sidetracked on transformers, found a pair of Hammons 1628SEs (some 11lb/5kg transfomers designed for single-ended tube outpurs, 5k:4/8/16 ohms, sound like crap) that I'm going to re-wind with a quadfilar winding of 19AWG. Got one of those apart and unwound today...
Peace
Hi Roscoe
I am not desperately keen on sheet work myself. Mainly because of the problems with cutting it down to size. No matter how careful I am I always seem to manage to cut myself on a sharp edge, as well as break a nail or two.
I do like lathe work though.
Great 🙂
Best wishes,
Susan.
Roscoe Primrose said:
I am not desperately keen on sheet work myself. Mainly because of the problems with cutting it down to size. No matter how careful I am I always seem to manage to cut myself on a sharp edge, as well as break a nail or two.
I do like lathe work though.
Great 🙂
Best wishes,
Susan.
Dear Susan,
First of all, I'd like to thank you for sharing this interesting and promising topology with community. Currently my interest lies in more radical SE approach with o/p coupled with choke or transformer. I couldn't help but notice that your SE testing rig uses transformer without air gap. Perhaps core saturation is a cause of such high distortion and enefficiency (7.5 Vac with almost 2amp quiescent current, IIRC)? And if I may ask, what type of core do you use in PP version? I can't remember this from your web pages, although I may be wrong.
WBR, Andrew.
First of all, I'd like to thank you for sharing this interesting and promising topology with community. Currently my interest lies in more radical SE approach with o/p coupled with choke or transformer. I couldn't help but notice that your SE testing rig uses transformer without air gap. Perhaps core saturation is a cause of such high distortion and enefficiency (7.5 Vac with almost 2amp quiescent current, IIRC)? And if I may ask, what type of core do you use in PP version? I can't remember this from your web pages, although I may be wrong.
WBR, Andrew.