Bottlehead Forum
Bottlehead Kits => Legacy Kit Products => Seduction => Topic started by: saildoctor on July 16, 2011, 12:58:02 PM
-
I've gotten to thinking about the possibilities of using gas tube regulation for my Seduction after reading some of the posts here. Paul J. thought one might be able to get ~30ma from the PT1 transformer by using a fullwave bridge and choke/cap filter. My Seduction came with a circuit diagram for the basic model, but the goings on of the c4s upgrade are still unknown to me. Does my upgraded Seduction have 4 x c4s loads? Most gas tubes have a 5ma minimum - so would the total power draw be 21ma? I traded in my computer for a mac a few months ago - so I'll have to go seek out a PC to run PSUD from. Unless anyone knows of any similar program that will run on OS X? Thanks for any help!
-
Here is a picture with the quad C4S board front and center:
http://www.bottlehead.com/store.php?crn=225&rn=431&action=show_detail
You'll want another single C4S feeding around 25mA to the reg tube and preamp circuit for best performance. Using a resistor is OK but the value may depend on your power line voltage, so some trial and error will be involved. I'd use the 105v reg tubes; they have the best performance and still have plenty of compliance above the plate voltage of the 6922 tubes.
-
hows about using something fun like a 6R7 coke bottle, or a larger 5U5G? or is the volage too off?
-
I've gotten my act together planning this out recently. I'm going to try out using a (25ma) C4S feeding a 0C3 (105V) reg tube and the 4x4ma C4S's of my upgraded seduction. One last question before I build - I'm wondering what happens with the circuit if the 0C3 fails to fire (not enough voltage/becomes damaged/etc.). That would leave a 25ma C4S feeding into 4x4ma of C4S downstream. Is that ok? Would it be a good/bad idea to add a high value resistor somewhere after the 25ma?
-
No problem - the 25mA C4S will just saturate and only drop a couple volts. I don't think you'll damage anything. No promises of course, I haven't tried it and I'm not planning to! I'll bet ti sounds great though - keep us posted!
-
It's up and working! I finished up two days ago and I've listened to a few records - sounds good! A big thanks to all the folks at bottlehead and on the forum for advise. The only previous mods to my seduction were installing new Obbligato gold film caps - 0.1uf coupling and 1.0uf output. I ordered a C4S manual from Eileen so that I could study up on what is going on with the circuit since I needed to make changes to the boards I had. Despite the manual being written for the old-style boards it was easy enough to look at the newer ones and figure out components and input/output/ground. In reading forum posts during the build I learned that the method of selecting R1 has changed since the C4S manual was written. The table within was based on different LED's than the hlmp-6000's currently used and has been replaced with a formula of R1 = 0.855/current(in amps). The board I used as well as the nylon risers holding itself and the filter choke came out of my Paramounts after I added the soft-start upgrade. I decided to use the side of the board that had a MJE5731A in the Q2 position because it had a higher power rating. The bridge rectifier (mounted to a terminal strip and is squeezed under the filter choke) is made of 4x SiC Shottky diodes, the negative side of the bridge output is grounded. The power supply is LCRC (with bleeder resistor!) and utilizes half of the original seduction power supply with the resistor replaced with 2k. With different R values one could change the output of the PS and use a different value gas reg tube. My value of 2k gives 152v (measured). The 0C3 i have plugged in right now regulates at 108v. I read in a forum post somewhere (I think?) that a C4S is happiest with at least ~30v voltage difference between input and output so this seemed appropriate. It's getting late - I'll post a little more info tomorrow for anyone who might want to do something similar to what I've done. I'm definately a happy camper right now! :)
-
Sweet! Update us after you have some hours on it!
-
To myself it sounds gooood! My monitors with Fostex fe207e's are not the last word in bass I'm sure, but what I am hearing sounds super. Seems like there is much more definition in certain ranges than before. I would definitely recommend trying this mod to anyone willing to drill an extra socket in their Seduction, and the extra holes for nylon standoffs. Check out the pictures I posted previously in this thread to see where I placed components. I wanted to keep the new socket centered on the top plate which required drilling out where the stock chassis ground was. So if anyone does decide to try this out please do not overlook drilling a new spot and installing chassis ground. Also, please keep in mind I have the anticipation upgrade installed which has a C4S load to going to each plate. If you do not have anticipation installed you may have to design a bit different - I'm not sure.
The following is more information (than you require probably) for anyone interested in proceeding. It allows an Anticipation upgraded Seduction to be modified to use a choke input filter feeding into a single C4S which in turn feeds a gaseous voltage regulating tube plus Anticipation C4S board, while keeping the same PT1 transformer that comes as stock. While the PT1 specs say 20ma max - that is most likely based upon a capacitor input filter. A choke input filter passes current at a more steady rate than the rapid swings of a cap input filter and produces less strain/heat on a transformer.
To recap the choke input power supply:
PT1> Bridge rectifier> L > C1 > R > C2 > C4S > 0C3 (VR Tube) > 4xC4S
Bridge Rectifier - I used 1A Schottky SiC (silicon carbide) that come in a TO-220 package. I liked the fact that they have low/no switching losses due to their operation. I soldered them into a bridge using a 5 terminal strip placed in the hole where the previous 1st section of power supply was located. Remember the center lug of the strip is grounded to the chassis! Also, the metal back/tab on these are NOT isolated from the internals, so they cannot touch anything including chassis and are at dangerous voltage. The negative side of the bridge output should be grounded. Both sides of the rectifier input attatch to both sides of the PT1 250V secondary. The center tap is not used and must not be grounded. I repurposed the stock center tap wire to connect the 0C3 to the input of the anticipation C4S board. http://en.wikipedia.org/wiki/Bridge_rectifier (http://en.wikipedia.org/wiki/Bridge_rectifier)
L - For filter choke I used a Hammond 157G (30H, 595ohm, 40ma). You could get away with a lower value choke so long as you stay above the critical inductance value. L = (supply voltage)/(load voltage in milliamps). That is the minimum below which voltage will spike for a given current going through. Bigger is better for ripple reduction I'm sure, so this seemed like the one to get for me. http://ken-gilbert.com/power-supply-design-henry-pasternack (http://ken-gilbert.com/power-supply-design-henry-pasternack)
C1 & C2 - I kept the 220uf Panasonics mounted to the stock terminal strip. Be sure to cut the old high voltage output STP wires leading out (but keep the ground wire inside connecting to the ground plane.) You could go much smaller if you wanted and still get a good noise reduction.
R - Replace the 1k resistor mounted between C1 & C2 with a higher watt value. This R can act as both filter component and voltage dropper. You can select a value which gives a filter output voltage 30V more than the VR Tube you wish to use or greater. C4S needs a certain amount of voltage drop or greater, plus a VR tube needs this extra headroom to fire up and function. I went with 2kohm. If you use a different choke than this build remember that its resistance will be different and the filter's output value will be different. Using Duncan amps PSU designer I came up with these resistance vs. output values:
500 ohm - 194V
1 kohm - 182V
1.5 kohm - 170V
2 kohm - 157V
2.5 kohm - 145V
3 kohm - 132V
C4S - I repurposed a dual C4S board that came out of paramounts during soft-start upgrade. I highly recommend getting a C4S manual from the fine folks at Bottlehead to understand how to properly use and obtain component values. Be aware that the new boards look different but the circuit layout is the same. If you have a newer board it will use HLMP-6000 LED's which are different than in the old manual. Because of this the table for determining R1 should be disregarded and instead use the formula R1 = 0.855/current(in amps). If you screw this up like I did and order the manual table1 value you can solder a second resistor of appropriate value across R1 to get in line with the formula. It can be lower wattage because it won't be passing much current. With the amount of current that will be going through I chose to use the MJE5731A for Q2 because it can handle a higher wattage and easily accepts a clip on heatsink. From forum posts of PJ's it appears the appropriate amount of bias current should be ~10% of the current being passed. Mine is set at 2ma assuming input voltage is 157V. The current passed in my build is set to 25ma. Note that VR tubes require 5ma minimum to pass through to stay alight when setting C4S current.
VR Tube - I used an 0C3 but looks like an 0D3 would work well too. Just make sure the filter output voltage is high enough according to whatever tube's specs. Bypass across cathode/anode with a cap < or = to 0.1uf. Above this value there be dragons. I used one of the 0.1uf orange drops that came as coupling caps in the stock kit. Read up on VR tubes! Fun! http://tubedata.milbert.com/sheets/141/g/GL874.pdf (http://tubedata.milbert.com/sheets/141/g/GL874.pdf)
The top of the VR tube should have both an input coming from the filter output and output wire going to the 4xC4S board input. If you decide to mount a choke on nylon mounts be sure to attach a wire going to ground to its chassis like done on the PT1. Also, I triple checked that the TO-220 rectifier diodes were not touching chassis - or anything else, and that no legs were soldered to the center grounded terminal lug. This build could be done without adding a C4S to feed, but more careful selection of R will be required. Refer to the VR tube manual to get that info.
I hope some find this information useful! I learned a lot during the design and execution of this build and it took awhile of searching through disparate sources to fully grok what I had got myself into. It sounds great though, I haven't been able to stop listening to vinyl for days!
-
Listening to The Thermals' - 'Now we can see' right now. Only one problem with my bottlehead system: I keep singing along to the music.
-
Singing along is a danger, especially for those of us with a Froggy voice. You keep wondering if there is something wrong with the bass.
-
I'm getting ready to do this to my Seduction. I'd been running it with 6N1P for some time and liked them and have been using a separate transformer for the filament. To do the 0C3 regulation, I'm repackaging the power supply in a cast aluminum enclosure. I just saw the comments about the the possibility of using the PT-5 for the Seduction in this thread: http://www.bottlehead.com/smf/index.php/topic,1822.15.html .
It looks to me that using it instead of the PT-1, sticking with the full-wave bridge rectifier using the Cree SiC diodes into a choke-input filter and otherwise keeping the same setup that SailDoctor used (while adjusting the R between the two C's to set the voltage going into the pre-0C3 C4S) should work well. AND the PT-5 would allow me to power the filments of the 6N1P without a 2nd transformer AND also allow me to go to regulated filaments.
Sounds like a win all the way around. Paul J, any others, have I missed something here? I have the PT-5 in my cart, but want to get some feedback before I pull the trigger.
Greg in Mississippi
P.S. I know the Eros would be better choice rather than re-working my Seduction this way, but part of why I'm doing this is the repackaging... the main Seduction circuit will also go into a cast ally box (similar to the recommended Hammond one for additional shielding) because I need to make the unit a bit less wide and longer to fit in my now-available area. I can do this with my several-years-old Seduction and not feel too bad. I'd feel REALLY bad chopping up an Eros this way.
-
Another question occurred to me... is there much or even any benefit going to the Bottlehead FC-1 choke instead of the Hammond 157G SailDoctor used for the PSU choke?
TIA again!
Greg in Mississippi
-
The 157G is 30 henries, the FC-1 is 0.030 henries.
The PT-5 should work well; an experiment is planned sometime in the next few months. A choke-input filter with a fullwave bridge may not produce enough voltage to fire the 0C3. I may experiment with a simple Zener regulator for the high voltage as well.
-
Hi Greg, the main reason I used choke-input was to get more current from the PT-1 without melting it down. The 0C3 will need a minimum of 5ma on top of the stock current being pushed through the new C4S and then there is it's own bias current. That adds up to more than the 20ma cap input rating.
If you use the PT-5 you could use a cap input filter because it has a much higher current rating - I think you would wind up with enough voltage to make it work. You might want to model it with the Duncan amps PSU modeler first. I did a search of the forum to try and find the PT-5 primary/secondary DCR values but didn't notice them? You could still add a choke or two later in the filter too.
It's totally worth it btw. Mine is sounding awesome with this and a RIAA cap upgrade. I'm using some early 60's Tesla 6922's with gold grids. I'm already scheming what to do next. Too bad there's not enough room under the hood to squeeze another PT-1! :)
-
Paul & Kerry, thanks for the replies.
PT-5 is ordered.
I was confused at the voltage thing, but looking at the PT-1 again, it makes more sense... it looks to have 125v-0vct-125v (or 0v-250v where the CT is not used), where the PT-5 is 0v-135v, right? I should have cracked open my Seduction manual first, all would have been clear!
So now I see how the PT-5 may not provide enough voltage... I can try it choke-input, cap-input and voltage-doubler to see what gives me sufficient voltage. The PS caps I have are 150uf/350v Black Gates, so I have a little headroom there.
I do have the Hammond 157G choke already... DUH about the difference in inductance rating between it and the BHead one, I always have had issues with decimal points (sorta like dyslexia!). If it doesn't get used as the input element of the PS, it'll likely go between the two caps.
This rebuild will have a few cap upgrades too, both couplers & RIAA. Very curious to see how it sounds.
Again, thanks. I'll post pix & impressions when done.
Greg in Mississippi
-
The PT-5 has plenty of voltage and current. For the high voltage, use a full-wave bridge rectifier into one of the stock 220uF caps, then the 157G, then at least one 220uF cap (maybe 2 in parallel). This will leave you with ~165V of raw B+, which is plenty to light up an 0C3. I'd set your dropping resistor for 50mA, so 165V-105V=60V, 60V/0.050mA=1200 Ohms.
For power, 1200 Ohms * (0.050mA^2)=3 Watts, so use a 10 watt resistor.
If you used a voltage doubler, this resistor would need to be quite large.
For the heater supply, I'd use a bridge of the 1N5818 diodes, then a 10 Ohm/5W Rheostat wired as an adjustable resistor, then a 10,000uF cap. Start with the rheostat at half its rotation, plug your two tubes in, then use your meter to dial the voltage to 5.9V.
At some point, it becomes easier and less expensive to build an Eros ;)
-
Paul,
Thanks for the info... it will be invaluable.
One question on the filament supply... I was intending to reuse the stock 1N5818's & 10kuf cap and then go into a Linear Tech regulator (likely one LT3080 per tube, but I could go with a single LT1083 instead for simplicity). I figured if it is good enough for the Eros, it is good enough for my Seduction. Bad move?
Then on doing this rebuild instead of just getting an Eros...
1. I have a 5.5" x 12"-14" space available. My Seduction stock won't fit, but I can restructure it to do so. Looking at the Eros, I don't see any easy path to fitting it in that space.
2. I already have the Seduction and this won't cost much in parts (About $200 for the iron and the cast ally boxes versus almost $740 for an Eros... and I need to cast boxes to repackage it for fitting in the available space anyway).
3. I'm already in build-mode, doing a couple of solid-state phono stages to try too... but both are current input for LOMC's only, this turntable has an arm known for being a challenge to setup correctly (Older Well-Tempered Classic), and I don't want to risk a higher-buck LOMC while I"m learning the arm. This build flows naturally with the other two.
Again, thanks for the info and the advice.
Greg in Mississippi
-
A linear regulator is good. Set it up for 6.0 volts, not 6.3 - this gives a longer life and potentially less noise as long as the plate current is low, as it is in Seduction.
We tried the LT3083 (3-amp version of the 3080) and couldn't get the noise (hum/buzz) down; I suspect the higher impedance voltage setting node is sensitive. Just a word of caution; we're still experimenting but no answers yet. The 1083/4/5 series have behaved themselves. Just read the notes in the spec sheet, especially about using the right kind of capacitors.
I had forgotten the earlier experiment with the PT-1. PT-5 into a FW bridge should give slightly more than PT-1 in FWCT, so a choke-input supply is possible if you use a current source to feed the shunt reg gas tube. (With a resistor, the current becomes a strong function of the power line voltage, so I don't recommend it.)
-
Paul,
Thanks. Sorry, I didn't say before, but I will be using a current source to supply the 0C3... I have a couple of the BH CCCCS kits from back in the day when I thought I could convert my Seduction to dual-mono.... until I realized the clever design ya'll put into your Seduction current source upgrade.
Thanks for the tip on the LT3083. I didn't even know about that variant... will have to get some. I use the LT3080 to replace LT1083's in some audio-ish applications and found it a nice upgrade there. Pity! I'll stick with the LT1083/1085 line then.
As I said, the PT-5 is on order. BTW, roughly when might that order ship, holidays & such?
Finally, got a bit of a disappointment last night... I had purchased some caps for upgrading the RIAA a few years back, but never put them into my Seduction as I didn't have my vinyl setup running then. I got the bag out & checked them on my trusty B&K & all were too far away from the target values to use. So I think I'll go the route of the AMTRANS polyprops from Parts Connexion... parallel 2 5600pf for the .0112uf value and 4 8200pf for the .033uf value. I'll get a few extra of each value so I can make 1% combos. Of course, their holiday shutdown means I won't see these any faster than the end of next week! More delays!
Again, Thanks!
Greg in Mississippi
-
Part of my goal in my own build was to add the 0C3 but keep it all within my existing Seduction chassis with the same tranny. Well... actually I think if I had to explain why I needed another box to power the first box my housemate may have tried to smother me as I sleep. Ha! That's great that you're going for it - bet it will be super quiet!
-
Part of my goal in my own build was to add the 0C3 but keep it all within my existing Seduction chassis with the same tranny. Well... actually I think if I had to explain why I needed another box to power the first box my housemate may have tried to smother me as I sleep. Ha! That's great that you're going for it - bet it will be super quiet!
LOL!
My housemate is a saint about my hobbies... and you'd agree if you saw how much my different projects and sprawling stereo have taken up space in our home!
We'll see about the quiet. This is for my smaller, downstairs system where it shares a shelf with the turntable. Whenever I get the main, upstairs system going again, the phono stage has it's own stand and there's plenty of room for the full-sized Seduction or larger.
Part of the reason for the repackaging is to fit it into a space. only 6" wide or so... but also, it sits next to my computer music server. While I've done a lot of mods to that to reduce conducted & radiated EMI/RFI, it will be only a few inches away. I know I may have to power down the server to play records... I can live with that. It may be I only have to power down the highest-polluting item, the small screen, which has its own supply and can be powered off by itself. But part of the reason for the re-boxing is to better shield the whole thing.
One issue I'm grappling with is how to deal with mechanical vibrations. The cast ally boxes ring... they'll get damped. Then I'm rebuilding the Seduction active circuit on a piece of Corian... that will get damped too and it'll have a suspension using Herbie's Audio Product material. It's somewhat of an experiment, but a fun one... and having your gas-tube regulator upgrade to go in at the same time is just another excuse for doing it! Thanks for publishing that!
Greg in Mississippi
-
...So I think I'll go the route of the AMTRANS polyprops from Parts Connexion... parallel 2 5600pf for the .0112uf value and 4 8200pf for the .033uf value. I'll get a few extra of each value so I can make 1% combos. ...
Another way to go is to use 0.030uF and 0.010uF caps, changing the resistors 66.5K to 75K and 9.64K to 10.7K. Precision caps are often available in only a few values, but 0.010uF is usually one of them, and you can parallel three of them to get 0.030. Sometimes you get a price break for 10 caps, so you throw out the worst two and sort the rest. The target is actually 0.0297 and 0.01011, and 73.7K, as long as you're sorting...
-
Paul,
Thanks, that's very helpful. It'll give me some flexibility in cap selection.
I'll be using 6N1Ps... I assume that these values don't change, but is the change of the RIAA termination resistor from 1M to 500K affected?
Also, I have .033uf and .011317uf as the nominals for the stock Seduction RIAA values... those right?
Again, and as always, thanks!
Later!
Greg in Mississippi
-
Paul,
Thanks, that's very helpful. It'll give me some flexibility in cap selection.
I'll be using 6N1Ps... I assume that these values don't change, but is the change of the RIAA termination resistor from 1M to 500K affected?
Also, I have .033uf and .011317uf as the nominals for the stock Seduction RIAA values... those right?
Again, and as always, thanks!
Later!
Greg in Mississippi
Yes, I'd change the load resistor to compensate for the 6N1P's higher plate resistance.
Take 60pF off of C2 (0.011257uF) - 50pF for Mr. Miller, and 10pF for wiring, socket, etc. capacitance. If you need to be REALLY picky! I get 66pF total for the 6N1P, according to the US Svetlana data sheet - net 0.011251.