the Poor Man 660 support thread

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
The 5687 heaters run of 0.9A each (IIRC) - they will get very hot.
distortion with 6.4V and no distortion with 6.3V is strange ....
 
Sounds like tubes that still need to be burned in, or some other problem.  Slight filament voltage variation and distortion are not related IMO.    We are in tolerance here either way. 
 
Siegfried Meier said:
Doug, can you get me the actual part # of what you used?

Thanks!
Sig

Sorry, I can't as I don't need and haven't ordered one.  I glanced in the catalog months ago when this came up, and posted about it then.  There's a wide variety of tapped resistors available.  Ohmite, I think.  25 or 50W 1 ohm. 
 
Gachet said:
Did some people noticed that the 5687 are becoming very hot after 15 minutes.
The plastic socket are litteraly  cooked by the heat of the valve.Is that normal?
I think I will change those sockets to ceramics ones.
we mentioned before somewhere that you have to use ceramic sockets because of the 5687 heat ...
hey, send us some pics of your finished PM ;D
 
hi

first : thanks for sharing this. regulated heater supply is a must.  ;D ;D ;D
i experienced lot of "mystic" / weird behaviours with gear from the 50's & 60's because of unregulated heater supply.

moby :
Don't be confused with parallel parts Its because I wanted few footprints for part on my PCB...
i am confused ! ;)
is R1 = R2 = R3 ? just use one of them 3 ?
is there 3 because you want 3 footprints on the pcb ? (for component placement, matter of size... ?)
I don't understand...

by the way, looking at the PSU PCB, i'm wondering, where do you plug you heater supply regulator ?
beetween the rectifier and the PSU PCB ?

and if you have time to post values of component, you'll make a bunch of dyiers happy ;)

thanks ;)
 
OK, I found some time and updated the schemo. Sorry, I don't have a PCB file, I accidentally deleted that from my disk, but I'm sure if you ask Volker he will do something. Can you Volker?  ;)
SLOWTURNONLarge.jpg
 
Hehe, I started to work from the datasheet on my own this morning :)

4700µF is a little bit overkill, I would like to keep the PCB as small as possible.
how many watt do you suggest for R1 ? (R8 isn't needed in my opinion)
looks something like this at the moment:

heaterreg-prev01.png

 
Great Volker. Well, R8 is highly recommended because it protect LM338 from oscillating. It's more important than R1  ;) Please don't exclude that. With 12V and 2.6A consumption you will have something around 1.5W dissipation. So , 3W will work for both R1,R8 but I went with 5W.
 
ok, let me source some parts and packaging options.
will include R8.
those 5W square resistors ?
single sided for home edging, or double sided for manufacturing ?
what I have learned is you need to use parts easily available all over the world,
but this makes it somehow slow ... ???
but ok, working on it
 
Volker, I think that most of guys will like to have finished PCB's . I did it for myself already but it's up to you to decide. You can include that to all PCB's you offer since it's universal filament PSU and can be used in various projects  ;D
Parts, well , I bought all of parts locally so mine source won't work for this. I went with 5W ceramic power resistors. I's widely available. If you want me to "trace" the parts I can do that but nothing fancy is included here. BTW , if you decided to go with finnished PCB's maybe you can include the heatsink on it. Same as PM main PSU
 
bassculture said:
regulated heater supply is a must. 

I disagree with this.  I have owned around 10 types of vintage tube limiters with unregulated filament supplies, and it's not been an issue.  I turn on around 120 tubes at my studio every day.  Not a regulated filament supply among the pieces.  Haven't replaced a tube due to any sonic problem in close to 10 years.   Never personally seen a filament failure in an operating tube amp in 20 years of operating tube gear. 

Lots of old tested tubes in my gear.   New tubes have a burn in period, and can seem problematic for the first 40-50 hours of operation.  Leave the thing on for a week and then start listening more critically. 

PRR has already written a good amount about these filament issues in other threads. 

But you will likely build a regulated filament supply, and it will work.   You will have increased the complexity of the piece and increased the odds of other failure due to increased parts count. 

I have seen regulated supplies fail more than unregulated types, and regulated portions of tube limiters fail where the unregulated sections didn't.  I've seen the failure of DC portions of filament supplies that used both AC and DC for different parts of the amp.  I've never seen a filament contribute to the problem.   

Buy the $9 multi-tap resistor, fine tune it's adjustment for an hour and forget about it.    Im 100% sure that will work perfectly; it has for many decades in many amps.   It's not a 'paint by numbers' part of the equation, unfortunately.   1/10th of an ohm can make a large difference, so it can't be predicted. 

A final point:  I challenge anyone to find another tube limiter thread here in which it's been decided that regulated filament supply is a must.   LA-2A, Federal, BA-6A, 436C, etc etc.  As far as I can tell this is simple fear of a variable part value, a part that works fine in the prototype builds.     
 
Hey emrr,
I presume this is the adjustable tap resistor you mean:
http://www.mouser.com/Search/ProductDetail.aspx?qs=J6Z1PhGvUbzIlcveIcozLA%3d%3d

If not, I presume that it'd be suitable anyway.I don't think I'll bother with a regulated heater supply either.

Rob
 
I have owned around 10 types of vintage tube limiters with unregulated filament supplies, and it's not been an issue
must be that your area has a stable mains voltage . mine not  ;) The current-regulated filament supply does have one other advantage – it protects the filament from over-current at startup, which can prolong the life of the filament. That's not my opinion that's a fact.
Hey emrr,
I presume this is the adjustable tap resistor you mean:
http://www.mouser.com/Search/ProductDetail.aspx?qs=J6Z1PhGvUbzIlcveIcozLA%3d%3d
No, it's a 1ohm resistor and you need something higher, say 1.5 or 2ohm so you can adjust that since you will need around 1ohm for stereo unit. this one can do a job for both, stereo or mono units http://www.mouser.com/Search/ProductDetail.aspx?qs=ecf2eIk0DojJWcSH%252bYlZdg%3d%3d
 
Moby said:
I have owned around 10 types of vintage tube limiters with unregulated filament supplies, and it's not been an issue
must be that your area has a stable mains voltage . mine not  ;) The current-regulated filament supply does have one other advantage – it protects the filament from over-current at startup, which can prolong the life of the filament. That's not my opinion that's a fact.

I think it borders on a pointless impractical fact; how long are we talking?  Again, reference PRR's commentary on the subject.  Tube filaments have a design center life that is quite long.  Like I said, I haven't seen a filament failure in 20+ years.  Be sensible; how much will you pay (time and $) to put together a regulated filament supply versus buying a decent stock of replacement tubes (that you may never need to use)?  Again, where are the other tube limiter builds here that have been modified for regulated filament supply?  Or the examples of failures because of the lack?  I can't find one example.  My mains voltage fluctuates about 112 to 128; I think not so stable. 

I presume this is the adjustable tap resistor you mean:
http://www.mouser.com/Search/ProductDetail.aspx?qs=J6Z1PhGvUbzIlcveIcozLA%3d%3d
No, it's a 1ohm resistor and you need something higher, say 1.5 or 2ohm so you can adjust that since you will need around 1ohm for stereo unit. this one can do a job for both, stereo or mono units http://www.mouser.com/Search/ProductDetail.aspx?qs=ecf2eIk0DojJWcSH%252bYlZdg%3d%3d
[/quote]

That's the type of resistor, yes. 
 
My mains voltage fluctuates about 112 to 128; I think not so stable.
Cool, than you will have a filament voltage swing of +/- 0.8V... If it's OK with you to supply from 5.5-7.1V than you go with that but don't bother with rheostat. Just drop 1ohm power resistor and it will work. I'm not sarcastic, just suggesting. ;D
 
Moby said:
OK, I found some time and updated the schemo. Sorry, I don't have a PCB file, I accidentally deleted that from my disk, but I'm sure if you ask Volker he will do something. Can you Volker?  ;)
SLOWTURNONLarge.jpg
Is there a real need of Q1 to stabilize the tension on R4 ?
Wouldn't a single cap, let's say 10uF, paralleled with R4 do the job ?  I think the PSU is over regulated (it's always better than under regulated  ;)) But I think it would work without Q1 and R6... And certainly without C1,R1 and R8...

If it does, Volker, the job is done... Copy the DAOC PSU and here you are, LM338 and LM350 are pin compatible !!! ;)
 
emrr said:
Be sensible; how much will you pay (time and $) to put together a regulated filament supply versus buying a decent stock of replacement tubes (that you may never need to use)?  Again, where are the other tube limiter builds here that have been modified for regulated filament supply?  Or the examples of failures because of the lack?  I can't find one example.  My mains voltage fluctuates about 112 to 128; I think not so stable.   
I agree I've never seen a tube filament fail...
I disagree, 7 tubes getting old fast for one filament fail is a real risk...
Can you find brand new 6BC8 or 5687 ? No, the best you'll find will be NOS.
Are NOS tubes secure : no. There's minor risks of glass porosity, real risks of bad treatments (by shipping or ex-owner).
If your tube fails, the problem is 100% yours. I prefer to repair sometimes my PSU with brand new IC's than to change my "not-nowadays-built" tubes. But that's a matter of taste...
 
Is there a real need of Q1 to stabilize the tension on R4 ?
Wouldn't a single cap, let's say 10uF, paralleled with R4 do the job ?  I think the PSU is over regulated (it's always better than under regulated  ;)) But I think it would work without Q1 and R6... And certainly without C1,R1 and R8...
Q1 works in the slow turn on circuit. I didn't tried without it  :-\ If you try please tell me if that works like slow turn on. Yes, capacitors are bit larger than needed but front C-R-C of useful. 2200uf will work great for 2.6A filament.
 
Moby said:
Q1 works in the slow turn on circuit.
I was planing to do a "DAOC like" filaments PSU but the slow turn on is a great idea, maybe totaly useless (DAOC works without it) but a great idea !!!  ;)
I assume that nearly any PNP with a 100+ hfe will work in this circuit...
 

Latest posts

Back
Top