12B4 help needed

[PixelPlay]

Hi
I am desperate for some help with 12b4 and psu design, I am on a mac and therefore unable to use Duncan amps PSU calculator. So guys I am humbly asking for some help in order to have the optimum B+ before a C4S.
As I am building this from the ground up I have only a lundahl LL1651 (42Ω/500v, .43A, 4X 0.1Ω/6.6v, 3.1A) and some bottle head C4S's. I need to build a suitable PSU with the Lundahl's specifications.
I have poured over every available schematic here and on google, however without the availability of PSU calculator, I am unable with my experience to create a suitable PSU supply and subsequently the overall schematic. I have not yet found someone using this transformer for this preamp. I am having a hard time sleeping, with of the schematics constantly on my mind.
As a matter of interest, is it possible to create a 12b4 preamp that is fully balanced?

I have thought long and hard for not wanting to appear as some sort of freeloader before posting this. With all of the experience of the people here I am hoping that someone is willing to give me a hand in creating my first 12b4 line stage.

Thank you for your time

Greg
Attached is the Transformer details:

 

[mikvous]

Hi Greg,
I'm not an expert so I don't know what's a c4s but a single 500v winding without center tap it's way too high for this purpose.
Bare in mind that after rectification you'll get something around 500X1.4 volts.I see on the transformer's specs that some pins are shorted I can not find out if you can do something with them.
Anyway I'll tell you what I use in my 12b4 pre.
Transformer is 225 v primary with 250-0-250 secondary @ 200ma
and a separate one with 15v @ 2a for the heaters.
High tension rectified through a pair of UF4007 followed by a simple CRC (220uf-120ohm-220uf) followed by the attached trimmed to give 255v.
2x10 k/5w in parallel on the plate and 500 ohm on the cathode bypassed with a good quality 470uf/25v cap paralled with a 0.22 uf good quality cap.The sound is something else believe me.

Michael

 

[zigzagflux]

Do you have any ideas of what B+ voltage and current you want to produce, or are you looking for advice in that area also? Any idea as to regulation/ripple targets?

 

[PixelPlay]

Hey guys thanks for your prompt reply.
The optimal B+ voltages with the C4S (type of CCS) is 300VDC. The Ripple needed should be very low due to the microphonic nature of the 12B4. The Lundahl can be used as a center-tapped 250-0-250 however they recommend using the 500v due to the double C core windings. The 4 6.6v windings I think can be configured to 2 12v supplies.

Thank you for your help
Greg

 

[Brett]

Yes it can be fully balanced: I've used this before and posted about it, IIRC with designs. CCS loaded plates with or without OPTs is trivial.

As for the 12B4 being extremely microphonic, not that I've found.

 

[dsavitsk]

As I am building this from the ground up I have only a lundahl LL1651 (42©/500v, .43A, 4X 0.1©/6.6v, 3.1A) and some bottle head C4S's. I need to build a suitable PSU with the Lundahl's specifications.

You haven't said where you want to run the tubes, so I am going to estimate about 20mA per side with the CCS as a plate load.

I have poured over every available schematic here and on google, however without the availability of PSU calculator, I am unable with my experience to create a suitable PSU supply and subsequently the overall schematic.

In Morgan Jones, there is information on how to do PS calculations by hand.

As a matter of interest, is it possible to create a 12b4 preamp that is fully balanced?

Yes, but you'll want to use output and input transformers. Then, something like this will work: http://raleighaudio.com/Figure_16.jpg

I use this design as my preamp with Cinemag input transformers, Magnequest outputs, and 12A4 tubes -- which are kind of like a 12B4 with a little more gain (sort of like 1/2 of a 5687). The gain of the 12b4 is probably too low here without a step up on the input.

I am hoping that someone is willing to give me a hand in creating my first 12b4 line stage.

see below

Hey guys thanks for your prompt reply.
The optimal B+ voltages with the C4S (type of CCS) is 300VDC.

This is the max B+ above which the transistors will go up in smoke. I'd shoot for slightly less than 300 max without any load, and more like 250V or less with a load.

The Ripple needed should be very low due to the microphonic nature of the 12B4.

The CCS will effectively block all of the ripple, so getting it low is not that important. This is, after all, half the point of the CCS.

The Lundahl can be used as a center-tapped 250-0-250 however they recommend using the 500v due to the double C core windings.

Using it as 250-0-250 is much better here. I'd use a tube rectifier (probably an EZ80 or EZ81 since you have 6V windings) to drop a few volts. Then, use a basic CLCRC filter, or a CRCRC. For the first C, keep it low, like 5uF to keep voltage down. For the next two C's, anything from 100uF to 220uF is fine. The bigger, the more ripple you'll get rid of. The choke can be just about anything. I simmed with a 9H and 5H, and both seemed fine. The resistor, again, can be just about anything. I ran a sim with a 220R as that was what the software used and up to 500R. The point here is that with the CCS the PS is not that important.

So, with EZ80 -> 5uF -> 9H (100R) -> 220uF -> 500R -> 220uF @ 40mA, I got a nice ~275V with 0.0005V of ripple, which is way overkill.

 

[PixelPlay]

Thank you,
Yes I was thinking of 20-25ma with the CCS.
I will Have a look for the morgan Jones publication.
I will get this up and running and only then attempt to make a version fully balanced.

I will give the PSU design a go.

Thank you again for your help

Greg

 

[Brett]

Re: Re: 12B4 help needed

Yes, but you'll want to use output and input transformers. Then, something like this will work: http://raleighaudio.com/Figure_16.jpg

I use this design as my preamp with Cinemag input transformers, Magnequest outputs, and 12A4 tubes -- which are kind of like a 12B4 with a little more gain (sort of like 1/2 of a 5687). The gain of the 12b4 is probably too low here without a step up on the input.

This was similar to how I ran mine when I needed a preamp, both with CCS plate loads (Pimm and 2450) and with the Tx's as anode loads and a CCS in the tail. Gain will likely be more than sufficient with the 12B4's in this application. 25mA/tube, somewhere around 150Va-k is where they sound best and measure well. Try a LED string for bias on each tube a la SY's Red Light amp. Don't forget to raise the filaments 40V and for some reason DC, always sounded better with these.

 

[dsavitsk]

Re: Re: Re: 12B4 help needed

Try a LED string for bias on each tube a la SY's Red Light amp.

Generally I'd agree, but it is unnecessary here as the signal current does not go through the bias resistor. It is effectively fixed bias with a resistor.

 

[Brett]

Re: Re: Re: Re: 12B4 help needed

Generally I'd agree, but it is unnecessary here as the signal current does not go through the bias resistor. It is effectively fixed bias with a resistor.

Unbypassed, it's going through the cathode resistor. Bypassed it's not, but I'd rather not bypass a cathode with an electro if I can avoid it, hence the LED string is it'll still have an effective lower Z than an unbypassed resistor and is cheaper than a good cap.

 

[Miles Prower]

I have poured over every available schematic here and on google, however without the availability of PSU calculator, I am unable with my experience to create a suitable PSU supply and subsequently the overall schematic.

Rectifier Applications Handbook. Everything you need right here. Better than that PSU calculator thingy since it's always best to know why you're doing what you're doing.

Looks like the Lundahl power xfmr will do just fine here. A bit more complicated since it's a "universal" type xfmr.

 

[PixelPlay]

Thanks for the help.
I realize now that the transformer was probably not the right choice. The reason for the 500v was for me to be able to use it with other projects etc at a later date if I so desired. Lesson learned.
I would love to be able to use the 500v and not have to use the 250-0-250 centre tap therefore using the feature of the duel c core. I will give the book a good read.

Cheers
Greg

 

[Bandersnatch]

Re: Re: Re: Re: 12B4 help needed

Generally I'd agree, but it is unnecessary here as the signal current does not go through the bias resistor. It is effectively fixed bias with a resistor.

I'll have to disagree. Current drawn from the power supply will be constant. Current at the cathode will be CCS minus the load current. So we're back to needing a low impedance voltage like we'd get with a cap bypass or an LED.

A MOSFET CCS as poineered by Gary Pimm

http://www.pacifier.com/~gpimm/schematics.htm

Or done with depletion-mode MOSFET's and eliminate the need for battery bias( or at least do without if desired ).

The mu output puts the power supply back in the signal business, but eliminates the cathode bias resistor from it( as in no need to bypass it). Current in that part of the circuit is independant of loading.
cheers,
Douglas

 

[PixelPlay]

To those Audio Mac users, I found some software for macs that opens windows programs within OSX (CrossOver), so I am now able to run Duncan amps PSU calculator without installing windows. Finally I can start modeling the PSU needed for the 12b4 C4S.

Can anyone suggest a high quality manufacturer of power transformers so that in the future I don't just buy an off the shelf one. My Lundahl LL1651 I am sure will be fine however having one made for each application would end up being far less hassle.

Thanks for your input
Cheers
Greg

 

[Bandersnatch]

hey-Hey!!!,
I'd go with what ever Hammond 3-series is closest to your design. I've been adapting Tektronix 'scope Iron to my latest projects, and find close is quite easy to achieve. get the biggest one with the right voltage.

But then again, Heyboer TX in Michigan has always done an excellent job on my custom Iron orders...signal, output, or power. Shipping might be a bit ov an issue, so perhaps a local winder?
cheers,
Douglas

 

[Brett]

Can anyone suggest a high quality manufacturer of power transformers so that in the future I don't just buy an off the shelf one. My Lundahl LL1651 I am sure will be fine however having one made for each application would end up being far less hassle.

Greg,

If you only need a B+ and a filament winding, hit the yellow pages and thry the local guys. haurbuch have made me toroids very reasonably and I think they have some EI designs in their catalogue.
The closer it is to a standard spec, the cheaper it'll be.

 

[PixelPlay]

Re: Re: Re: 12B4 help needed

This was similar to how I ran mine when I needed a preamp, both with CCS plate loads (Pimm and 2450) and with the Tx's as anode loads and a CCS in the tail. Gain will likely be more than sufficient with the 12B4's in this application. 25mA/tube, somewhere around 150Va-k is where they sound best and measure well. Try a LED string for bias on each tube a la SY's Red Light amp. Don't forget to raise the filaments 40V and for some reason DC, always sounded better with these.

Hi Brett
You said to raise the filaments 40V and for some reason DC?
That is high?

Cheers
Greg

 

[Brett]

Filaments sounded better with good clean DC than AC. No idea why as I never investigated further. The 40V is a reference to floating the entire filament supply at 40V above ground. Std practice for me and explained better than I can ATM by MJ in Valve Amplifiers.

 

[PixelPlay]

Ah, I understand.
Thanks again for your help
Cheers

Greg

 

[planet10]

To those Audio Mac users, I found some software for macs that opens windows programs within OSX (CrossOver)

CrossOver (WINE for the newer Intel Macs) requires "well-behaved" Windows sw. Nice to know the Duncan Amps software does. I'd be interested to know if MathCAD (MJK worksheets), and TubeCAD run in it.

dave