CJ
Well-known member
you could wind a ten turn 5 VAC heater wind next to the 6.3, you have to isolate them with good insulation as the 5 volt heater sits at the B+ voltage,
Fender Blues DeVille Pwr XFMR + Amp Rebuild
- Thread starter CJ
- Start date
Help Support GroupDIY Audio Forum:
CJ
Well-known member
The Super Reverb runs a tube rect and sounds great, we are going to run an 82 Ohm 10 watt cathode resistor along with a negative fixed bias voltage to imitate that 5U4 compressed sound,
here is the print for this guy, they are running the core at 16 K Gauss which is too high for a pwr xfmr, we want about 12.6 KG, so we will up the primary turns to 300 and use a single strand of #21, which will make for more room which we need for the increase in secondary turns. DCR will go up to about 2 ohms, but that is still ballpark for a pwr xfmr, and primary inductance will go up also which means less excitation current. The lower flux will radiate less junk into the preamp circuit,
and we will put in a CT for the heater wind since we will use an even number of turns.
if we put this wind between the pri and sec, it will act as a shield with the grounded CT.
here is the print for this guy, they are running the core at 16 K Gauss which is too high for a pwr xfmr, we want about 12.6 KG, so we will up the primary turns to 300 and use a single strand of #21, which will make for more room which we need for the increase in secondary turns. DCR will go up to about 2 ohms, but that is still ballpark for a pwr xfmr, and primary inductance will go up also which means less excitation current. The lower flux will radiate less junk into the preamp circuit,
and we will put in a CT for the heater wind since we will use an even number of turns.
if we put this wind between the pri and sec, it will act as a shield with the grounded CT.
Attachments
CJ
Well-known member
inside the pressboard tube was this thermo switch in the pic,
that is why the primary was on the outside, so the deformity cause by placing the switch in the coil would not deform any other winds.
usually thermo switches are seen in offshore transformers, but this was done in a good shop, the wire did not look like it was too cheap, and the winding was pretty good, Schumacher must have a QC guy over there full time.
the switch says 141 Degrees C on it.
that is why the primary was on the outside, so the deformity cause by placing the switch in the coil would not deform any other winds.
usually thermo switches are seen in offshore transformers, but this was done in a good shop, the wire did not look like it was too cheap, and the winding was pretty good, Schumacher must have a QC guy over there full time.
the switch says 141 Degrees C on it.
Attachments
CJ
Well-known member
we got 289 turns of #21 primary down in 5 layers = 58 turns per layer,
this is an upgrade of 289-224 = 65 turns, we did not hit our target of 12.6 K Gauss but we did drop from 16 KG to 14 Kg which is fine for barn roof steel which has a B max of 21 KG unlike grain orient which sits at 18 KG.
DCR went from 1.4 ohms to 2 ohms, which is fine, we saved 3 layers of space by using single #21 instead of #22 bi-fi, #21 will due 810 ma at 1 ma per circular mil which is conservative. DCR was also helped by putting the primary wind on the bobbin first, which is standard for power transformers.
To add a few more turns to hit 12.5 KG would mean starting the coil with a partial layer and we do not want the uneven winding surface.
We need space for additional HV secondary turns, the bobbin allows more turns per layer than a layered paper coil allows due to the act that we do not need margins, just wind from flange to flange.
laying down two wraps of 7 mil Nomex so we are good for 14 KV break thru voltage. That is from the primary to the heater wind which now has a grounded CT. Will the Nylon bobbin do 14 KV from wind to core? Probably not but we should still pass the 1.5 KV hi-pot test.
Two 6L6GC tubes going all out might make 60 watts on a good day, so we have 60 VA plus the heaters which is another 12 to 14 VA so figure 74 VA = 0,61 amps, add some transformer loss and call it 1 amp.
Old Super amps used a 2 amp use, later on a 3 Amp fuse, Primary copper loss will be about 1 amp into 2 ohms = 2 watts which is fine.
this is an upgrade of 289-224 = 65 turns, we did not hit our target of 12.6 K Gauss but we did drop from 16 KG to 14 Kg which is fine for barn roof steel which has a B max of 21 KG unlike grain orient which sits at 18 KG.
DCR went from 1.4 ohms to 2 ohms, which is fine, we saved 3 layers of space by using single #21 instead of #22 bi-fi, #21 will due 810 ma at 1 ma per circular mil which is conservative. DCR was also helped by putting the primary wind on the bobbin first, which is standard for power transformers.
To add a few more turns to hit 12.5 KG would mean starting the coil with a partial layer and we do not want the uneven winding surface.
We need space for additional HV secondary turns, the bobbin allows more turns per layer than a layered paper coil allows due to the act that we do not need margins, just wind from flange to flange.
laying down two wraps of 7 mil Nomex so we are good for 14 KV break thru voltage. That is from the primary to the heater wind which now has a grounded CT. Will the Nylon bobbin do 14 KV from wind to core? Probably not but we should still pass the 1.5 KV hi-pot test.
Two 6L6GC tubes going all out might make 60 watts on a good day, so we have 60 VA plus the heaters which is another 12 to 14 VA so figure 74 VA = 0,61 amps, add some transformer loss and call it 1 amp.
Old Super amps used a 2 amp use, later on a 3 Amp fuse, Primary copper loss will be about 1 amp into 2 ohms = 2 watts which is fine.
Attachments
CJ
Well-known member
to build an anchor you lay the tape down adhesive side up, and wind turns over it, then fold the tape back over. Thin poly tape actually is stronger than thicker glass cloth tape as the cloth tape does not stick to the wire as good. The poly tape is like trying to rip a plastic shopping bag in half, the harder you pull, the stronger it gets.
Attachments
CJ
Well-known member
we like to lay down some cloth tape on the sides to keep wire from digging down into the margins as the nylon flange will flex outward a bit as you add layers. This also puts some additional insulation between the lead breakouts which we are bringing out of the side of the bobbin. Most people bring the leads up along the inside edge and put tape over the lead, but this does not provide much insulation strength and also makes it possible to rub the wire with additional turns, which could nick the enamel.
Attachments
CJ
Well-known member
we tried some hefty #17 for the heater wind but did not like the way it went down, took up too much valuable space that we need for the secondary winds, and we did not like the bulge caused by the CT splice wire so we will reduce that by using a piece of #26 and taping over and under it with cloth tape.
We used a bi-fi wind of #21 which will be good for 1.6 amps at 1 ma per mil, but this is such a short wind and DCR so low that we can double that easily. With the additional primary turns, we had to add heater turns, we came out at 15.5 turns to reduce heater voltage by 10 percent from 6.9 volts, since we can not have a half turn, we will wind 16 turns and hope that the slightly lower circular area of the wire will add some additional DCR over the stick wind. We can always kill a little heater voltage with a power resistor, but we can not make up for low voltage, do not want to get stuck with 5.8 or 5.9 volts, Would rather have 6.4 volts.
We used a bi-fi wind of #21 which will be good for 1.6 amps at 1 ma per mil, but this is such a short wind and DCR so low that we can double that easily. With the additional primary turns, we had to add heater turns, we came out at 15.5 turns to reduce heater voltage by 10 percent from 6.9 volts, since we can not have a half turn, we will wind 16 turns and hope that the slightly lower circular area of the wire will add some additional DCR over the stick wind. We can always kill a little heater voltage with a power resistor, but we can not make up for low voltage, do not want to get stuck with 5.8 or 5.9 volts, Would rather have 6.4 volts.
Attachments
CJ
Well-known member
hard to tell from the pics but we picked up maybe 1/32" by rewinding the heater, that is almost enuff for 2 layers of #26 which means an additional 200 turns of HV secondary wire.
after a few failures, you get an idea about how much room you need to finish off a coil without having the lamination's cut through your insulation and short out your coil.
the bottom pic is after the heater rewind.
we ran out of 3 mil Nomex layer insulation so more is on the way, 7 mil was too thick and would have built the coil up too quick.
after a few failures, you get an idea about how much room you need to finish off a coil without having the lamination's cut through your insulation and short out your coil.
the bottom pic is after the heater rewind.
we ran out of 3 mil Nomex layer insulation so more is on the way, 7 mil was too thick and would have built the coil up too quick.
Attachments
> we can not have a half turn
You can. Try it.
Pick a PT working at say 10 turns per Volt. Wind one 360 turn of any convenient small wire. It induces 0.1V.
Look at the 360 degree turn. It is really "two windows". Reduce it to 200 degrees but still two windows, it is still 0.1V. Reduce it to 160 degrees, i.e. one window, and suddenly it is 0.05V. You can reduce to a single pass through one window and still 0.05V.
I have not thought how this affects bringing the leads out. I guess it means coming out the other side of the bobbin, which is electrically OK but may be inconvenient in practice.
You can. Try it.
Pick a PT working at say 10 turns per Volt. Wind one 360 turn of any convenient small wire. It induces 0.1V.
Look at the 360 degree turn. It is really "two windows". Reduce it to 200 degrees but still two windows, it is still 0.1V. Reduce it to 160 degrees, i.e. one window, and suddenly it is 0.05V. You can reduce to a single pass through one window and still 0.05V.
I have not thought how this affects bringing the leads out. I guess it means coming out the other side of the bobbin, which is electrically OK but may be inconvenient in practice.
CJ
Well-known member
^^ I will have to give that a try,
got some 3 mil Nomex in, decided to go with some #33 wire to make sure we have enough room for a ull Wave CT secondary, this will introduce about 100 ohms into each leg, copper loss will be about 100 ma x 200 ohms = 2 watts, so we are spreading out the turns to keep them cool,
this resistance will introduce some sag into the supply which should mellow out the attack of the guitar, somewhat like a cathode resistor.
going with about 719 turns inside and 720 outside to make up for slightly decreased linkage,
insl. good for 3 KV per layer,
got some 3 mil Nomex in, decided to go with some #33 wire to make sure we have enough room for a ull Wave CT secondary, this will introduce about 100 ohms into each leg, copper loss will be about 100 ma x 200 ohms = 2 watts, so we are spreading out the turns to keep them cool,
this resistance will introduce some sag into the supply which should mellow out the attack of the guitar, somewhat like a cathode resistor.
going with about 719 turns inside and 720 outside to make up for slightly decreased linkage,
insl. good for 3 KV per layer,
Attachments
CJ
Well-known member
CJ
Well-known member
ready to lace up the coil and check voltages!
plenty of leads with the addition of the Half Wave CT system and bias tap,
so we did a two layer breakout on the busy side with a healthy chunk of polyester left over from the original, which is more puncture proof than Nomex.
primary and heaters on one side, HV and LV taps on the other,
plenty of leads with the addition of the Half Wave CT system and bias tap,
so we did a two layer breakout on the busy side with a healthy chunk of polyester left over from the original, which is more puncture proof than Nomex.
primary and heaters on one side, HV and LV taps on the other,
Attachments
CJ
Well-known member
core laced up 1 x 1, tape the leads down so they do not cross paths,
clamp the un-varnished core so it does not wake up the neighborhood,
we have about 302 - 0 - 302 for the HV secondary, unloaded, so with the DCR of 100 ohms each leg and at 80 ma 6L6GC plate current, we should drop about 8 volts AC when the amp is cranking,
add in the 1N4007 drop and we have 302 - 8.5 = 293.5 x 1.39 = 408 volts, OPT might chew a few volts so figure 406 VDC of B+ .
preamp tube current draw and other losses should get us that 400 VDC target voltage.
bias tap came out to 46 VAC unloaded, that should give about 60 volts into the first cap with the half wave rectifier, we add some filtering and a pot and it will drop right in there,
heater voltage came out to 6.74 volts unloaded so that should drop down to 6.2 to 6.3 and we now have a CT for hum reduction without the 100 ohm virtual CT .
clamp the un-varnished core so it does not wake up the neighborhood,
we have about 302 - 0 - 302 for the HV secondary, unloaded, so with the DCR of 100 ohms each leg and at 80 ma 6L6GC plate current, we should drop about 8 volts AC when the amp is cranking,
add in the 1N4007 drop and we have 302 - 8.5 = 293.5 x 1.39 = 408 volts, OPT might chew a few volts so figure 406 VDC of B+ .
preamp tube current draw and other losses should get us that 400 VDC target voltage.
bias tap came out to 46 VAC unloaded, that should give about 60 volts into the first cap with the half wave rectifier, we add some filtering and a pot and it will drop right in there,
heater voltage came out to 6.74 volts unloaded so that should drop down to 6.2 to 6.3 and we now have a CT for hum reduction without the 100 ohm virtual CT .
Attachments
CJ
Well-known member
happy with the voltages, we slip in a 7 mil Nomex wrapper and get ready to solder up the flux belt,
dry out at 200 F for a few hours, dip and bake overnight and we are ready to build our Custom DeVille power supply, low voltage for tube life and reasonable volume for the 4-10 cabinet so it does not rattle the chassis apart, and a little sag built in so we do not have to add cathode resistors,
dry out at 200 F for a few hours, dip and bake overnight and we are ready to build our Custom DeVille power supply, low voltage for tube life and reasonable volume for the 4-10 cabinet so it does not rattle the chassis apart, and a little sag built in so we do not have to add cathode resistors,
Attachments
CJ
Well-known member
got a rewind print,
6.3 V grounded CT wind replaces copper E shield to keep line garbage from getting into the B+
HV wind features a rev wind to keep the earthy CT side of the wind on the outside as to act as a shield to keep HV from radiating into circuit,
and the 32 V CT kind of does the same thing,
6.3 V grounded CT wind replaces copper E shield to keep line garbage from getting into the B+
HV wind features a rev wind to keep the earthy CT side of the wind on the outside as to act as a shield to keep HV from radiating into circuit,
and the 32 V CT kind of does the same thing,
Attachments
CJ
Well-known member
CJ
Well-known member
hooked it up and took some voltages,
B+ unloaded: 415 V-dc
Heater: 6.55 VAC across two 6L6 tubes,
this should drop down once we add the 3 preamp tubes which should draw an additional 0.3 x 3 = 0.9 amps,
Bias Voltage: minus 62 V-dc unloaded
Reverb Opamp Voltages: 16-0-16 VAC
sure is nice to get the parts count down with the full wave circuit and the dropped B+, two diodes instead of four, two plate/screen caps instead of 4 and no voltage divider to build, and the CT heater saves two 100 ohm resistors, and the bias tap gets rid of the voltage doubler that is used on the original Blues DeVille.
B+ unloaded: 415 V-dc
Heater: 6.55 VAC across two 6L6 tubes,
this should drop down once we add the 3 preamp tubes which should draw an additional 0.3 x 3 = 0.9 amps,
Bias Voltage: minus 62 V-dc unloaded
Reverb Opamp Voltages: 16-0-16 VAC
sure is nice to get the parts count down with the full wave circuit and the dropped B+, two diodes instead of four, two plate/screen caps instead of 4 and no voltage divider to build, and the CT heater saves two 100 ohm resistors, and the bias tap gets rid of the voltage doubler that is used on the original Blues DeVille.
Attachments
CJ
Well-known member
may as well mutate this thread into the amp build also,
built a plate for the tube sockets, was gonna shock mount it with grommets but thoughts of rotted grommets and a loose board with high voltage quickly changed that idea, better to shock mount the whole chassis,
got the plate bolted in and ready to wire,
built a plate for the tube sockets, was gonna shock mount it with grommets but thoughts of rotted grommets and a loose board with high voltage quickly changed that idea, better to shock mount the whole chassis,
got the plate bolted in and ready to wire,
Attachments
CJ
Well-known member
got the preamp tube heaters wired,
was expecting the transformer voltage to drop down from 6.55 to 6.3 with the addition of the 1 amp draw of the three 12AX7a tubes but we only got 6.5 but that is reasonable for late night mains voltage,
that dual #21 wind is pretty stiff, taking off another turn would mean 6.1 volts and we lose the balanced CT.
changing the wind to one strand of #21 instead of the bi-fi wind would do the trick if we were picky about the 0.2 volts.
plate volts are sitting at 395 with a bias of minus 42 volts on the JJ 6L6GC's. could probably use 16 more HV sec turns to hit 400 on the dot.
using the Bassman 5F6 bias scheme (no pot) with a solder tab for a trimming resistor. we have a 22 ohm cathode resistor for the two 6L6 tubes, this was chosen to give a reading of 1 V-dc when the tubes are biased right,
1 V / 22 ohms = 45.5 ma idle current.
60% of 30 watts (max 6L6 disp) = 18 watts
18 W / 395 V = 45.5 ma
will the 22 ohm cathode resistor add any mojo to the sound? probably not. this is a push pull circuit so degeneration currents will cancel unless we have some severe clipping,
got some .022 / 600 V orange drops ready to feed the grids, and we have 21 volts feeding 16 volt zeners for the reverb opamp supply, this cuts down on the hideous heat of the standard DeVille which had those trace lifting Dale pwr resistors in there. We used some 270 ohm 3 watt jobs in there, they have about 5 volts across them so 5/270=18.5 ma, so "eye squared are" loss is 0.092 watts so 3 watts is way over kill unless an opamp shorts out, we are not going to pull in any relays, been there, done that, don't want to go there,
FET switching for Clean/Lead switching is the way to go here, if we protect the gates.
we are going to go with a standard Bassman 5F6 for the clean channel and a Pig Nose G40V for the dirt, which is a 5F6 with an extra stage featuring a 10 K cathode resistor, so we keep the 3 tube format and will keep the common tone stack idea so it will match the chassis holes on the stock face plate.
this wire is awesome! pre-tinned and waxed, so no fluffy hairballs and no having to twist and tin before soldering, very close to original Blackface wire>
https://www.ebay.com/itm/12-FT-Vintage-Style-Cloth-Push-Back-Wire-22-awg-for-Guitars-22-ga-12-feet/290749267603?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2060353.m2749.l2649
was expecting the transformer voltage to drop down from 6.55 to 6.3 with the addition of the 1 amp draw of the three 12AX7a tubes but we only got 6.5 but that is reasonable for late night mains voltage,
that dual #21 wind is pretty stiff, taking off another turn would mean 6.1 volts and we lose the balanced CT.
changing the wind to one strand of #21 instead of the bi-fi wind would do the trick if we were picky about the 0.2 volts.
plate volts are sitting at 395 with a bias of minus 42 volts on the JJ 6L6GC's. could probably use 16 more HV sec turns to hit 400 on the dot.
using the Bassman 5F6 bias scheme (no pot) with a solder tab for a trimming resistor. we have a 22 ohm cathode resistor for the two 6L6 tubes, this was chosen to give a reading of 1 V-dc when the tubes are biased right,
1 V / 22 ohms = 45.5 ma idle current.
60% of 30 watts (max 6L6 disp) = 18 watts
18 W / 395 V = 45.5 ma
will the 22 ohm cathode resistor add any mojo to the sound? probably not. this is a push pull circuit so degeneration currents will cancel unless we have some severe clipping,
got some .022 / 600 V orange drops ready to feed the grids, and we have 21 volts feeding 16 volt zeners for the reverb opamp supply, this cuts down on the hideous heat of the standard DeVille which had those trace lifting Dale pwr resistors in there. We used some 270 ohm 3 watt jobs in there, they have about 5 volts across them so 5/270=18.5 ma, so "eye squared are" loss is 0.092 watts so 3 watts is way over kill unless an opamp shorts out, we are not going to pull in any relays, been there, done that, don't want to go there,
FET switching for Clean/Lead switching is the way to go here, if we protect the gates.
we are going to go with a standard Bassman 5F6 for the clean channel and a Pig Nose G40V for the dirt, which is a 5F6 with an extra stage featuring a 10 K cathode resistor, so we keep the 3 tube format and will keep the common tone stack idea so it will match the chassis holes on the stock face plate.
this wire is awesome! pre-tinned and waxed, so no fluffy hairballs and no having to twist and tin before soldering, very close to original Blackface wire>
https://www.ebay.com/itm/12-FT-Vintage-Style-Cloth-Push-Back-Wire-22-awg-for-Guitars-22-ga-12-feet/290749267603?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2060353.m2749.l2649
