John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
jneutron,
If that is your son you are teaching I think that is a great start to understanding how mechanical things work. My first experience as a child of about 5 or 6 was having my grandfather or uncles hand me what they thought were broken pocket watches to play with. I would proceed to dismantle them and put them back together and hand them back ticking away. Most often they just over-wound the dammed things and it was simple to fix that. Other times a gear was out of the jeweled socket and I just had to put the gear back in place. I still think to this day that is how I started down the road to mechanical design and understanding.
 
Member
Joined 2002
Paid Member
John,

I am using my worst transformer and diodes while exploring the filter network!

Even so the 60 hz is 108 db down from the DC output.

ES


:confused: How do you define this?
60Hz component is -64dbV (0,63mV) to –82dbV (0.08mV)
http://www.diyaudio.com/forums/attachments/analog-line-level/337613d1363896759-john-curls-blowtorch-preamplifier-part-ii-4-stages-stage.jpg

120Hz component is -30dbV (31.6mV) for cyan to –56dbV (1.6mV) for blue
180Hz component is -76dbV (0.16mV) >> to –104dbV (0.006mV) >>
240Hz component is -42dbV (7.9mV) >> to –80dbV (0.1mV) >>
480Hz component is -64dbV (0.6mV) >> to –102dbV (0.008mV) >>
Above this freq, peaks drop at 40db/decade (12db/oct).

For cyan, LF rectification components will increase 10-12 db more than HF components when psu is loaded dynamically.

George
 
Last edited:
:confused: How do you define this?
60Hz component is -64dbV (0,63mV) to –82dbV (0.08mV)
http://www.diyaudio.com/forums/attachments/analog-line-level/337613d1363896759-john-curls-blowtorch-preamplifier-part-ii-4-stages-stage.jpg

120Hz component is -30dbV (31.6mV) for cyan to –56dbV (1.6mV) for blue
180Hz component is -76dbV (0.16mV) >> to –104dbV (0.006mV) >>
240Hz component is -42dbV (7.9mV) >> to –80dbV (0.1mV) >>
480Hz component is -64dbV (0.6mV) >> to –102dbV (0.008mV) >>
Above this freq, peaks drop at 40db/decade (12db/oct).

For cyan, LF rectification components will increase 10-12 db more than HF components when psu is loaded dynamically.

George

60 Hz best case is .08 mV ripple, agreed, the DC voltage output is 16.25 Vdc so my eyeball of -108 db was wrong it is only -106.15 db!

However it is still a bit noisy for a preamp circuit. -56 db ripple @ 120 into a PSSR of -95 db with a gain of 60 db would give .03 mV of ripple or more than -90 dbm of 120 hz. since my noise floor should run around -134, just using CRC filters are a hard way to get there unless someone goes for CRCRCRCRCRCRCRCRCRC!

Now that filter would have around 20% voltage loss. Of course a 7815 needs 2 volts across it for a 13% loss. I will go for three CR stages into a regulator. The regulator works well below 2,000 hz and the filters do in most of that before it gets to the regulator. As real regulators can do 45 db, all that remains of concern is ringing under load. Of course this is all from a clean AC line, so the transformer EMI pass through and diode noise will come into play.
 
Last edited:
Afair, Mr Harrison used Guaiacum wood for the bearings of his clocks.
(aka lignum vitae, lignum sanctum, legno sancto, pockholz, pokhout, guajak, etc.)

Guaiacum is the hardest wood, also the hardest to work with.
To this day, Guajak bearings are still used and if kept wet constantly, superior in longevity to any modern variety.

I was pricing lignum vitae just yesterday. The local guys with good prices had some pieces starting around $600. Of course that was a bit more than 1/2 a board foot. (A board foot is 144 cubic inches!)(.00236 cubic meters for the rest of the world.)
 
Last edited:
My father described a pickling tank made of Lignum Vitae long boards bolted together with Hastelloy bolts and filled with aqua regia for pickling steel wire by the multi tonne load...
The wood made the whole tank, no other support apart from the bolts holding it together.

Must have used quite a bit of wood.
 
did someone say lignum vitae...

Depends on the ebony.
My current favorite is Macassar ebony, but much easier to machine & smooth than the 7' x 3.5' dining table I made in solid Bubinga.
Downside of Macassar is that it does ~$260/Bdft, $1300 for 5Bdft.

(Hardest as in handwork and lifting, weighs up to 95lb/cuft. Try white meranti by hand with a chisel, instead of a lathe)
 
Try working with Coco Bollo and let me know how you like the nice rash you can get from the oils in the wood. Related to poison oak I am told. And try and glue any of these oily woods plus teak and a few others, takes a strong epoxy and even then it is questionable. Most expensive woods I have worked with were true wood veneers such as Carpathian Elm Burl, mucho bucks for a single 4' x 8' sheet with paper backing.

Now back to electronics and John's saga about the phono stage or something of that ilk.
 
60 Hz best case is .08 mV ripple, agreed, the DC voltage output is 16.25 Vdc so my eyeball of -108 db was wrong it is only -106.15 db!

However it is still a bit noisy for a preamp circuit. -56 db ripple @ 120 into a PSSR of -95 db with a gain of 60 db would give .03 mV of ripple or more than -90 dbm of 120 hz. since my noise floor should run around -134, just using CRC filters are a hard way to get there unless someone goes for CRCRCRCRCRCRCRCRCRC!

Now that filter would have around 20% voltage loss. Of course a 7815 needs 2 volts across it for a 13% loss. I will go for three CR stages into a regulator. The regulator works well below 2,000 hz and the filters do in most of that before it gets to the regulator. As real regulators can do 45 db, all that remains of concern is ringing under load. Of course this is all from a clean AC line, so the transformer EMI pass through and diode noise will come into play.

Hello,

Have you tried 2 half wave rectifiers rather than a bridge?
It goes against convention but it maintains a balance for +/-
Referenced to ground.
A CT ground depends on the wiring more.
 
Depends on the ebony.
My current favorite is Macassar ebony, but much easier to machine & smooth than the 7' x 3.5' dining table I made in solid Bubinga.
Downside of Macassar is that it does ~$260/Bdft, $1300 for 5Bdft.

(Hardest as in handwork and lifting, weighs up to 95lb/cuft. Try white meranti by hand with a chisel, instead of a lathe)

Ebony Macassar 4/4 RGH KD - Two Pcs - Ebony Macassar by Hardwood Emporium

My local guys only get $70 a board ft! ( I do keep some around for trim, handles and knobs.)
 
Status
Not open for further replies.