dc filaments
Moderators: pompeiisneaks, Colossal
-
funkmeblue
- Posts: 485
- Joined: Mon Mar 13, 2006 4:17 am
- Location: akron, ohio
dc filaments
few questions about dc filaments......should I twist the wires as they are in ac filaments......should I run dc to the power tubes or just the preamp, can one use the 5 volt tap for the dc preamp heaters and the 6.3 volt for the ac power tube heaters (it looks like george allesandro did this in the blue tick gut shots I've seen).....I am building an "express" with 6sl7 preamp
- David Root
- Posts: 3540
- Joined: Fri Aug 04, 2006 3:00 pm
- Location: Chilliwack BC
1 others liked this
Re: dc filaments
I always twist the wires even if it's DC. It looks better and will help with residual AC ripple.
You don't need to run DC on the power tube heaters. I've done it but it's really not necessary. If your PT has the reserve, OK, fine.
Yes you can use the 5V AC tap for 6.3V DC heaters. You may need a small resistor or two to bring the rectified DC down to 6.3V. You could even use a doubler circuit, or an auxiliary 12VAC transformer and use 12VDC on the heaters. Good idea if you have a multistage hi-gain preamp. A SS regulator, usually a TO-220 size, makes life easier.
6SL7 in an Express circuit! Take it from me, I've done it, don't even think about trying this with AC heaters. The humbucking 12A*7 option ain't there! You MUST use 6.3VDC, preferably with a 1.5mH choke and two 10,000 uF caps and even then you may have hum problems if you don't use shielded signal wires between the input jack and the phase inverter. (This is a big PITA!) 6.3 VDC regulator is probably a better option too. Use a modern design, I like the Linear Technology line myself. Naturally you should also use a star grounding system with a single chassis ground, as per drz440. These old octals do sound great, but they were not designed for high gain at all, and are fussy about grounding in that situation too! You probably didn't see that in Allesandro's schematics, he's no dummy!
You don't need to run DC on the power tube heaters. I've done it but it's really not necessary. If your PT has the reserve, OK, fine.
Yes you can use the 5V AC tap for 6.3V DC heaters. You may need a small resistor or two to bring the rectified DC down to 6.3V. You could even use a doubler circuit, or an auxiliary 12VAC transformer and use 12VDC on the heaters. Good idea if you have a multistage hi-gain preamp. A SS regulator, usually a TO-220 size, makes life easier.
6SL7 in an Express circuit! Take it from me, I've done it, don't even think about trying this with AC heaters. The humbucking 12A*7 option ain't there! You MUST use 6.3VDC, preferably with a 1.5mH choke and two 10,000 uF caps and even then you may have hum problems if you don't use shielded signal wires between the input jack and the phase inverter. (This is a big PITA!) 6.3 VDC regulator is probably a better option too. Use a modern design, I like the Linear Technology line myself. Naturally you should also use a star grounding system with a single chassis ground, as per drz440. These old octals do sound great, but they were not designed for high gain at all, and are fussy about grounding in that situation too! You probably didn't see that in Allesandro's schematics, he's no dummy!
-
funkmeblue
- Posts: 485
- Joined: Mon Mar 13, 2006 4:17 am
- Location: akron, ohio
Re: dc filaments
can I combine the 6.3 and the 5 volt together and then use a regulator?
I really don't want to add a separate transformer. I used the hammond 272jx.....the amp sounds sublime, but I would like to tame the hum some. would I rectify both taps and them them combine them? Also my 5 volt has a center tap, does this get connected? Sorry so many questions, thanks for all the help
I really don't want to add a separate transformer. I used the hammond 272jx.....the amp sounds sublime, but I would like to tame the hum some. would I rectify both taps and them them combine them? Also my 5 volt has a center tap, does this get connected? Sorry so many questions, thanks for all the help
- David Root
- Posts: 3540
- Joined: Fri Aug 04, 2006 3:00 pm
- Location: Chilliwack BC
Re: dc filaments
With the reserve available in your 6.3V tap (8A) you have plenty left after the power tubes (2xEL34?) at 3A total, i.e. 5A AC. If you use a 6V fixed regulator eg 7806, and derating 5A AC *0.7071=~3.5A DC available. You don't need to use the 5V tap at all. Also, doing it this way will give you power tube heater voltage closer to 6.3V as it is loading up that tap, not the 5V tap.
Run the 6.3VAC tap leads to the power tubes/pilot lite and take a parallel feed off the pilot lite or first power tube. Use this to make your 6.3VDC.
For a rectifier I would use an ordinary Radio Shack or Mouser 25A 50V full wave bridge rectifier, as this size is big enough to handle rectification spikes and also has a metal case that attaches thru the chassis with a bolt, so your chassis is the heatsink. This is overkill, but bulletproof.
There is a very good schematic for this in TUT 1, page 2-14, Fig. 2-16 A. Note that the 6.3VAC center tap is NOT used. This circuit uses the 7806 so you'll get 6.0V not 6.3V but that's fine. Apex Jr. has these for $0.40 ea. The diodes D1 and D2 are 1N4001 type, nothing special. Nichicon electrolytics are good here. The 7806 will give you 1.0A DC which should cover the PI tube, the third preamp stage and the 6SL7GT nicely. You'll need about 6-8 sq. inches of chassis area for the components, depending on your layout. I stick the end of each cap to the chassis with double sided carpet tape, or you can lay them horizontally with a cable tag and adhesive square. You don't need a separate board, but you could do that too if you have space.
Finally, buy the best i.e. least microphonic 6SL7 GT you can find. If you have access to several and a tester that will test microphonics, so much the better. In my experience (I have built an amp with five 6SL7GTs in it) the most common NOS, the JAN Phillips ECG from the '80s is generally OK, but some older NOS can be better, and some worse, in this respect. I have not used any current production 6SL7GTs, so I have no knowledge on them.
BTW, there is a thread I saw y'day on one of these three forums re 272JX as being particularly "hummy" and not recommended. You might want to check that out too, before you start.
Run the 6.3VAC tap leads to the power tubes/pilot lite and take a parallel feed off the pilot lite or first power tube. Use this to make your 6.3VDC.
For a rectifier I would use an ordinary Radio Shack or Mouser 25A 50V full wave bridge rectifier, as this size is big enough to handle rectification spikes and also has a metal case that attaches thru the chassis with a bolt, so your chassis is the heatsink. This is overkill, but bulletproof.
There is a very good schematic for this in TUT 1, page 2-14, Fig. 2-16 A. Note that the 6.3VAC center tap is NOT used. This circuit uses the 7806 so you'll get 6.0V not 6.3V but that's fine. Apex Jr. has these for $0.40 ea. The diodes D1 and D2 are 1N4001 type, nothing special. Nichicon electrolytics are good here. The 7806 will give you 1.0A DC which should cover the PI tube, the third preamp stage and the 6SL7GT nicely. You'll need about 6-8 sq. inches of chassis area for the components, depending on your layout. I stick the end of each cap to the chassis with double sided carpet tape, or you can lay them horizontally with a cable tag and adhesive square. You don't need a separate board, but you could do that too if you have space.
Finally, buy the best i.e. least microphonic 6SL7 GT you can find. If you have access to several and a tester that will test microphonics, so much the better. In my experience (I have built an amp with five 6SL7GTs in it) the most common NOS, the JAN Phillips ECG from the '80s is generally OK, but some older NOS can be better, and some worse, in this respect. I have not used any current production 6SL7GTs, so I have no knowledge on them.
BTW, there is a thread I saw y'day on one of these three forums re 272JX as being particularly "hummy" and not recommended. You might want to check that out too, before you start.
-
funkmeblue
- Posts: 485
- Joined: Mon Mar 13, 2006 4:17 am
- Location: akron, ohio
Re: dc filaments
I picked up some new production tung sol 6sl7's from tube depot tested for microphonics and I am quite impressed so far.....this amp also sounds good with nos 6sn7....nice and clean
are there any links to the schematics you speak of?
6.3 volt ac into the 7806 then into 25A 50v full wave bridge rectifier to caps then to tubes....is that correct? Can I still leave the heater center tap connected?
thanks for all the help, I have built a few amps, but have never messed with dc heaters....this amp sounds great, but I think I can get it a little quieter
are there any links to the schematics you speak of?
6.3 volt ac into the 7806 then into 25A 50v full wave bridge rectifier to caps then to tubes....is that correct? Can I still leave the heater center tap connected?
thanks for all the help, I have built a few amps, but have never messed with dc heaters....this amp sounds great, but I think I can get it a little quieter
Re: dc filaments
Regulator before rectifier? I think you want that the other way around...
- David Root
- Posts: 3540
- Joined: Fri Aug 04, 2006 3:00 pm
- Location: Chilliwack BC
Re: dc filaments
You do indeed.
6.3 VAC taps into ~ inputs on the rectifier. (Centertap NOT CONNECTED to anything. Cap it off). Take +ve off rectifier to +ve of 2200 uF 10V filter cap. Run the -ve off the rectifier to the -ve of the 2200 uF cap. Connect the +ve of the 2200 uF cap to pin 1 of the 7806 regulator. This is the left pin of three viewed from the top of the regulator. (The bottom is metal and is the cathode. It must be insulated from the chassis with a mica sheet and attached to the chassis with a 4-40 machine screw and a nylon collar washer, thru the hole in the TO-220 size case. Sounds complicated but welcome to the wonderful world of modern electronics. In this case it's worth it!) Some silicone thermal grease between the mica and the chassis should be applied too. Google TO-220 mounting instructions and you'll see how it works.
Connect -ve of the 2200 cap to pin 2, the middle or reference pin. Connect pin 3, the right hand pin of the regulator, to the top of a 470 uF 10V cap. Connect its -ve back to pin 2 of the regulator. Connect a 1N4001 type diode across the 470 uF cap with the stripe end to the +ve end of the cap. connect a second diode from the +ve end of the 470 uF cap back to pin 1 of the regulator, with the stripe end on pin 1.
You now have 6.0VDC at 1.0 A across the 470 uF cap. Connect this to your PI heaters 4/5 and 9, disconnect the AC heater supply thereto, and if you haven't already balanced the preamp heater string at V1A, the first stage tube, put a 100 ohm 1% metal film on each side, twist the other ends together and connect them to the -ve of the decoupled power supply cap for that stage. This ensures no hum (yes, DC heaters can hum too in certain circumstances) and of course provide the required ground reference. The -ve of the PS cap should be the final star ground collection node. It should also have the input jack 1M grid resistor grounded to it. The final and only circuit chassis ground goes from this node to chassis. I use a #8 machine screw for this, but my middle name is "overkill".
Sorry to have to write all this out but I don't have a scanner.
6.3 VAC taps into ~ inputs on the rectifier. (Centertap NOT CONNECTED to anything. Cap it off). Take +ve off rectifier to +ve of 2200 uF 10V filter cap. Run the -ve off the rectifier to the -ve of the 2200 uF cap. Connect the +ve of the 2200 uF cap to pin 1 of the 7806 regulator. This is the left pin of three viewed from the top of the regulator. (The bottom is metal and is the cathode. It must be insulated from the chassis with a mica sheet and attached to the chassis with a 4-40 machine screw and a nylon collar washer, thru the hole in the TO-220 size case. Sounds complicated but welcome to the wonderful world of modern electronics. In this case it's worth it!) Some silicone thermal grease between the mica and the chassis should be applied too. Google TO-220 mounting instructions and you'll see how it works.
Connect -ve of the 2200 cap to pin 2, the middle or reference pin. Connect pin 3, the right hand pin of the regulator, to the top of a 470 uF 10V cap. Connect its -ve back to pin 2 of the regulator. Connect a 1N4001 type diode across the 470 uF cap with the stripe end to the +ve end of the cap. connect a second diode from the +ve end of the 470 uF cap back to pin 1 of the regulator, with the stripe end on pin 1.
You now have 6.0VDC at 1.0 A across the 470 uF cap. Connect this to your PI heaters 4/5 and 9, disconnect the AC heater supply thereto, and if you haven't already balanced the preamp heater string at V1A, the first stage tube, put a 100 ohm 1% metal film on each side, twist the other ends together and connect them to the -ve of the decoupled power supply cap for that stage. This ensures no hum (yes, DC heaters can hum too in certain circumstances) and of course provide the required ground reference. The -ve of the PS cap should be the final star ground collection node. It should also have the input jack 1M grid resistor grounded to it. The final and only circuit chassis ground goes from this node to chassis. I use a #8 machine screw for this, but my middle name is "overkill".
Sorry to have to write all this out but I don't have a scanner.
Re: dc filaments
David,
How critical is it that the TO-220 be isolated from the chassis? If the actual ground is also the chassis will it work? I've tried to look this up on the web, but have not found anything that explains this stuff.
Also - if I use the mica spacer and want to use thermal grease (actually Arctic Silver CPU thermal paste) do I put it on both sides of the mica? (reg side and chassis side)
BTW - I'm asking re: relay supply, not filaments. I have a 12.6V xfmr to bridge-rectify and regulate to 12V w/ a 7812.
Michael
How critical is it that the TO-220 be isolated from the chassis? If the actual ground is also the chassis will it work? I've tried to look this up on the web, but have not found anything that explains this stuff.
Also - if I use the mica spacer and want to use thermal grease (actually Arctic Silver CPU thermal paste) do I put it on both sides of the mica? (reg side and chassis side)
BTW - I'm asking re: relay supply, not filaments. I have a 12.6V xfmr to bridge-rectify and regulate to 12V w/ a 7812.
Michael
Wife: How many amps do you need?
Me: Just one more...
Me: Just one more...
- David Root
- Posts: 3540
- Joined: Fri Aug 04, 2006 3:00 pm
- Location: Chilliwack BC
Relay supply
Well, that changes a lot! Why not use a 12 VDC relay and run it (them) off the existing 12V heater supply?
On the TO-220, the metal base is the also connected to the cathode, so it MUST be electrically isolated from ground in order to function. For relay use, it doesn't need a heatsink, so you could just stick it to the chassis with doublesided tape. I use 3M VHB foam tape for this kind of application. It's used in the auto industry for permanently attaching trim parts. I originally used it for sticking pickup bobbins to my coil winder pad.
In a normal TO-220 chassis mount, the thermal grease is applied only between the lower side of the mica and the chassis. You want the upper side of the mica in direct contact with the cathode base for maximum electrical insulation.
On the TO-220, the metal base is the also connected to the cathode, so it MUST be electrically isolated from ground in order to function. For relay use, it doesn't need a heatsink, so you could just stick it to the chassis with doublesided tape. I use 3M VHB foam tape for this kind of application. It's used in the auto industry for permanently attaching trim parts. I originally used it for sticking pickup bobbins to my coil winder pad.
In a normal TO-220 chassis mount, the thermal grease is applied only between the lower side of the mica and the chassis. You want the upper side of the mica in direct contact with the cathode base for maximum electrical insulation.
You do not have the required permissions to view the files attached to this post.
Re: dc filaments
David,
Thanks!
The heater supply is 6.3V. This 12.6V xfmr is currently dedicated to relay and LED supply. (though I may consider driving the preamp tubes from it...) Please forgive my next question - why are the 2 grounds not the same thing? Can they separately connect to the chassis? Or are they supposed to be completely isolated from each other? In the pictures of ODS #124 it seems to show that the center pin of the regulator is grounded to the chassis via a terminal strip ground lug on the relay supply xfmr. Does this imply that the back tab of the regulator is really not connected to anything electrically (or would that be electronically?)?
Thanks for sharing your expertise!
Michael
Thanks!
The heater supply is 6.3V. This 12.6V xfmr is currently dedicated to relay and LED supply. (though I may consider driving the preamp tubes from it...) Please forgive my next question - why are the 2 grounds not the same thing? Can they separately connect to the chassis? Or are they supposed to be completely isolated from each other? In the pictures of ODS #124 it seems to show that the center pin of the regulator is grounded to the chassis via a terminal strip ground lug on the relay supply xfmr. Does this imply that the back tab of the regulator is really not connected to anything electrically (or would that be electronically?)?
Thanks for sharing your expertise!
Michael
Wife: How many amps do you need?
Me: Just one more...
Me: Just one more...
- David Root
- Posts: 3540
- Joined: Fri Aug 04, 2006 3:00 pm
- Location: Chilliwack BC
TO-220
The center pin on a TO-220 regulator is internally connected to the metal back. Both supply the chip's regulated OUTPUT voltage, so to use it "old world style" on a metal chassis, the back must be electrically isolated, or you're just dumping the output to ground. Having said that, the output is referenced to ground, in the typical regulator application thru an electrolytic cap.
Perhaps ODS #124 uses a different case design. There are older cases, or metal "cans', as they're known, but AFAIK they also have the output on the case, ie "live case".
I'm no expert on regulator chips, in fact the only two I do have any familiarity with are both TO-220 cases, the LT317A and the LT1086. The Fairchild "Stealth" diodes I use for HT and bias supplies are also TO-220 cases and all four of these devices have the "live back".
Perhaps ODS #124 uses a different case design. There are older cases, or metal "cans', as they're known, but AFAIK they also have the output on the case, ie "live case".
I'm no expert on regulator chips, in fact the only two I do have any familiarity with are both TO-220 cases, the LT317A and the LT1086. The Fairchild "Stealth" diodes I use for HT and bias supplies are also TO-220 cases and all four of these devices have the "live back".
Re: dc filaments
I thought that for the 7812 (actually, the whole 7800 series of positive voltage regulators) the center pin was ground. I believe on the 7900 series (negative voltage regulators) the center and right pins are switched (output and ground). Am I confused about something? This will be the first IC of any kind I've ever put in a tube amp of mine, so I'm just trying to get educated! Again, thanks for sharing your experience and expertise.
Michael
Michael
Wife: How many amps do you need?
Me: Just one more...
Me: Just one more...