All You Tube Guru's

[craigmillard]

Hi Guys

Im still designing and building this 4 tube preamp in my spare time and have hit a bit of a blocker and not sure how to get around it.

The preamp has a Cathode Follower in it and will have a higher cathode-grid voltage than they should have Vh-k(max)=100V as they are 6n2p-ev tubes. Normally an elevated heater center tap would work and i could boost the heaters up to 75V or so, but this preamp i want to build from cheap parts and no transformers. Therefore i built the Max1771 SMPS which works great and allows 40mA and 300+ V for the plates from a 15V power supply. For the heaters i was going to use the same power supply but run 2 * 2 tubes in series with a dropping resistor to get 6.3 volts on each.

How can i elevate the heaters in this configuration? Is it even possible?

Could i put 2 100ohm resistors across the +15v and 0 rails of the power supply and create a virtual center tap and then inject the main plate voltage through a voltage divider?

Any ideas:) Would love to get this up and running this weekend 
Im hoping to create a 4 tube preamp with fairly standard parts that can be built without mains AC involved apart from a shop bought power supply (laptop etc) like the little big amp!

Cheers Guys!

[merlinb]

Could i put 2 100ohm resistors across the +15v and 0 rails of the power supply and create a virtual center tap and then inject the main plate voltage through a voltage divider?

I think the LV and HV share ground don't they, in which case you can't elevate, unless your SMPS is transformer isolated?

Easiest thing is probably just to make sure your CF heater is at the top of the series string, so it rests at 15V. Then you'll have about 85V Vhk, which isn't too bad. The tube won't die instantly or anything, it may just go crackly after a year or two, or it may work fine for years.

[tubegeek]

It would depend on whether the SMPS has a reference to the 0V of the 15V supply, which it most likely does.

A way to do this that is a slightly different approach:
Assuming you are in a 120VAC wall voltage country:
Take a 120:12VAC transformer from the wall. Take a 240:12VAC transformer and turn it around (12:240.)

Rectify and filter the 240V winding for your high voltage supply. Use the 12V winding for your heater supply and elevate it to 75V with a divider from the high voltage supply. You could even rectify and filter the 12VAC if you want a DC heater supply, and just connect its ground reference to the elevated connection as well.

Here is a similar arrangement, this elevates the 12VAC off of a connection to a voltage-doubler supply.

[craigmillard]

cheers guys:)

Merlin: yup they do share a ground and i did think that would cause an issue! So you think i can get away with running them a bit over?

Is my working out correct? my plate voltage will be around 300v so cathode voltage about 150, if i have the CF first in the chain then it will run at  about 12.6v so vhk = 137.4v? The tubes recommended is vhk=100v?
If i used 12ax7 there recommended vhk=180, these would have an easier life wouldn't they?

tubegeek: I am in the UK so 240v but had looked at this before but want to try and avoid any transformers in this build. Tis a good way to do it though, i wonder if i could use just a filament transformer with dual secondaries 12.6 and 6.3 and use the 12.6 to feed the smps and 6.3 for the heaters?!

[tubegeek]

i wonder if i could use just a filament transformer with dual secondaries 12.6 and 6.3 and use the 12.6 to feed the smps and 6.3 for the heaters?!

Excellent solution. Then you can put the 6.3 V at any DC level (including 75V) and pull your SMPS from the 12.6 VAC winding without trouble. My interpretation of your description is that you would first want to rectify the 12.6 VAC before using the SMPS to boost it to high voltage - the SMPS will want a DC feed. Two diodes (or four if there is no center tap on the 12.6V winding) and a cap will get you there: you can expect something like this:

[merlinb]

Is my working out correct?

Yes

The tubes recommended is vhk=100v?
If i used 12ax7 there recommended vhk=180, these would have an easier life wouldn't they?

There is no difference between the heater construction of the 12AX7 and any other small signal valve. The data sheet ratings just crept up to sell more valves. You can expect your valves to manage just as well as any 12AX7. Vhk is not a hard limit, its more of a recommendation; I wouldn't sweat it.

[MetalGuy]

It's well known fact now that newer production 12AX7 die when used as cathode follower. This problem applies to different brands of tubes like, EH, JJ, Tungsol, some Sovteks  and others so my advice to you is don't use them as CF and I'm speaking from experience.
So which ones are OK then. Believe it or not the chinese tubes will take the abuse but even then if possible elevate heaters especially if you're building for someone else.

[merlinb]

It's well known fact now that newer production 12AX7 die when used as cathode follower.

I think that is more usually due to grid conduction than heater-cathode failure, isn't it? And they generally die only in very-high HT amps like Mesa.

[iccaros]

It's well known fact now that newer production 12AX7 die when used as cathode follower. This problem applies to different brands of tubes like, EH, JJ, Tungsol, some Sovteks  and others so my advice to you is don't use them as CF and I'm speaking from experience.
So which ones are OK then. Believe it or not the chinese tubes will take the abuse but even then if possible elevate heaters especially if you're building for someone else.

I have not seen this issue, nor have I read about it, a cathode follower does  not stress the tube any differently as far as I know, I have read that a 12AU7 is better for that role due to current sourcing, If I understand my NEETS and Merlin's page correctly.  

discussions on this topic
http://www.ax84.com/bbs/dm.php?thread=394196

and
http://www.freewebs.com/valvewizard1/dccf.htm


Edit.. Sorry Merlin, I submitted right on top of you..

[MetalGuy]

I have not seen this issue, nor have I read about it, a cathode follower does  not stress the tube any differently as far as I know, I have read that a 12AU7 is better for that role due to current sourcing, If I understand my NEETS and Merlin's page correctly.   

The CF issue now is not a new one and it was discussed at several forums. Manufacturers have confirmed the issue as well.
The CF does stress the tube by exceeding the heater to cathode voltage.

I think that is more usually due to grid conduction than heater-cathode failure, isn't it? And they generally die only in very-high HT amps like Mesa.

No it's exactly a heater-cathode failure. Even with lower voltages like 250-300V depending on how it's biased you can exceed the h/k voltage depending on the tube brand.
AFAIK few years ago the russians did some changes in their technology and since then tubes started to fail. EH recommends now +/-100V h/k voltage.
I had different  EH tubes from several years in a row. The earlier ones are still working as CF, some batches from the same year are OK, some not. Later (I was still unaware of the issue) they started to fail immediately after power on. Same applies to other brands as well so if you design from scratch and have a CF elevating heaters is highly recommended.

[defaced]

The preamp has a Cathode Follower in it and will have a higher cathode-grid voltage than they should have Vh-k(max)=100V as they are 6n2p-ev tubes.

I'd consider re-allocating the CF triode to do something else and using a MOSFET as a source follower in place of the CF.  It gets rid of the Vh-k issue entirely and gives you another triode to play with.  IME (with high gain amps, SLO, Recto, VHT, et al), you will not notice a difference in tone. 

[iccaros]

The CF issue now is not a new one and it was discussed at several forums. Manufacturers have confirmed the issue as well.
The CF does stress the tube by exceeding the heater to cathode voltage.

I am not saying your wrong, but can you point out some information?  I would like to read up on it so I could avoid issues. Currently I have 12au7 in my CF, except home builds, but I have a some  6N6P (got cheep ) I was going to start to use..

Defaces info is always good, just feels so wrong to mix....

[defaced]

Whimp

[iccaros]

IF I could just keep people from looking.. They already frown on a SMPS...

[Jdansti]

"All You Tube Gurus..." Wasn't that a Mott the Hoople song?

[merlinb]

Later (I was still unaware of the issue) they started to fail immediately after power on.

That's most likely a grid conduction issue, as I suspected, which can occur in any tube,not just new production tubes. I too would interested to see the manufacturers' comments, otherwise we might be venturing into internet legend territory... The trouble with 'well known facts' is that they frequently turn out not to be facts!

[MetalGuy]

One discussion is here:

http://music-electronics-forum.com/t9514/

but it's not the only one.

I was buying directly from Electro Harmonix in bulk and when I started having issues with their tubes I wrote them an email about that. They confirmed that their tubes have +/-100V h/k voltage that's why they will fail as CF. Then I contacted several internet retailers. Some of them were unaware of this issue others knew the problem.
Because I don't feel like playing lottery since then I started elevating heaters and even those EHs that failed immediately before (from the same batches) are still working without any problems (several years now). Later it was confirmed by others that Tungsol, JJ and all tubes with spiral filaments like Sovtek LPS for example also fail as CF.
I also noticed that some tube amp manufacturers use chinese tubes in CF positions (Bogner Ecstasy comes to mind). It turns out that currently only chinese tubes can handle the abuse even with those crasy voltages as found in Mesa DR.
So I'm not making this up or rely on rumors. This an advice coming from long practice and directly from a manufacturer. Whether you're going to consider it is entirely your choice.

[merlinb]

One discussion is here:
http://music-electronics-forum.com/t9514/
but it's not the only one.

Hmm, a bit 'forum-y'. Lots of people saying "I asked a well-known vendor/manufactuers/tech and he confirmed what I already said in a previous post". But no one saying WHICH vendor/manufacturer/tech and actually posting their email word-for-word.

I'm not saying these tubes don't fail- I'm sure they do by the reports. What I want to get to the bottom of is if it really is h-k failure or grid failure, or just a bad batch at the time, or something else. Just saying "my CF died" doesn't tell you anything. Forums are generally not the places to determine precise technical details; everyone's an armchair expert.

[MetalGuy]

I don't keep messages several years old but you can always email a manufacturer and ask them. Maybe this will answer some of your questions:

http://www.newsensor.com/pdf/electro-harmonix/12ax7eh.pdf

I'll be elevating heaters when I have CF. Period.

[merlinb]

I'll be elevating heaters when I have CF. Period.

Dude, chilax, no one's calling you a liar!