10K:600 Vs. 15K:600 Ohm Carnhill Transformers

[ruffrecords]

Thank you very much for the detailed informations, but this issue is getting too complicated for my knowledge

I am building my preamps using TESLA E88CC tubes

I have read that other members have built their preamplifier with the EDCOR transformers (10K:600 or 15K:600) but I am considering to buy the Carnhill because I think thay are better than EDCOR
(but cheaper than the Sowter that are too expensive for me)

I agree. The Carnhill output transformers are very good. They are good value for money and their performance is very good too. I think they are ideal for you. They are definitely better than Edcor and probably just as good as the Sowters but much cheaper.

Edit: I use the VTB2291 in the EZTube Mixer but wired for 2K4:600 so I know these transformers work well.

Tesla E88CC tubes will be fine. I have tried these myself and I found them to be excellent performers.

Cheers

Ian

 

[Strawtles]

So I will buy the 9600:600 one
Just another things ... I have seen that they have the same transformer gapped or ungapped
What is the best for me?
Thank you

 

[CJ]

ungapped will be much better,

you have a cap blocking the DC,

if DC current ran thru the transformer, it would polarize the core and this would lead to a drop in the amount of signal the transformer could handle before saturating, a gapped transformer would handle a large signal before saturating if DC were used, but with a gap, the transformer's inductance goes way down which screws up bass response,

since we have the blocking cap, we retain full inductance with the ungapped core which means much better bass,


don't worry about all the transformer rhetoric, that was for me so i have a ref for building the output, or anybody else who might want to roll their own, if you hang around here long enuff, you will be spewing tech garbage like nobody's bidness, 

and remember that one good lab experiment is worth a thousand theories, so you really don't need to over analyze this stuff, you could get real info from a 1 hr experiment,

what is the part number on the Carnhill you are looking at?  i had problems getting any info off their website,

here is the cut sheet for the Tesla:

http://www.mif.pg.gda.pl/homepages/frank/sheets/183/e/E88CC.pdf

 

[ruffrecords]

So I will buy the 9600:600 one
Just another things ... I have seen that they have the same transformer gapped or ungapped
What is the best for me?
Thank you

As CJ says you definitely want the un-gapped version for this application. There is no dc flowing through the inductor so you don't need a gap and a bonus is that the inductance is much higher and bass response much better.

Cheers

Ian

 

[ruffrecords]

what is the part number on the Carnhill you are looking at?  i had problems getting any info off their website,

VTB2291. Best info is the Carnhill design guide which you can download from http://www.audiomaintenance.com/ although it contains precious little technical info.

Cheers

Ian

 

[CJ]

here is the schemo just for ref,

kind of a V72 ripoff but every thing is derivative nowadays anyway, so wtf, over?
they went cheap on the choke, don't blame then, 10 million turns in three sections, you need a special bobbin, ,

yes, there is the plate current right there, i need some bi-focals,

going to use a DJ8, got to wind an input, Mr. Hinson was tweaking the input ratio, i do not know what he had cinemag build for him, somewhere in between 1:6 and 1:7,

 

[CJ]

here is Winston's version  in case your interested>

 

[shabtek]

thanks cj! u da man

the only real diff I spot in winstons version are the cap values in the feedback-gain switch

 

[CJ]

Zobel network was tweaked for the cinemag stuff , that's about it,
oh, and some 0.1 caps across the filters,

FYI from Mr boogie:

"I can answer that question.  Easy.  The original transformer (an A92 of which a friend of mine just sold a pair I owned to a forum member) does NOT have a centre tap on the secondary.
What you see on the original REDD47 schematic is a cap and a resistor (100pF and 100K damping network) soldered to an un-used pin on the top of the can.  The pin that is used as this tie point is pin number 6 and it does not go to the winding.  Of course, if you use the CMMI-7C, which I did myself, then you would use a different and appropriate damping network as the resonance is higher on the CMMI-7C.
The CMMI-7C was a transformer I had a small hand in designing many years ago. I had had Marco who used to work for Reichenbach Engineering wind it for me.  It was based on an older Reichenbach design but I asked for larger input level handling and for the peak to be higher so that I could decide if I wanted to let it ring at some ungodly high place or damp it with a network.  The transformer was initially stamped as an R47-IT.  When Tom Reichenbach bought the assets of RE and turned the company around (along with David) into the great company that is today Cinemag, I had him make the same transformer for use in a few pre amps for UA and another company who shall remain nameless.  At that time, it got assigned the Cinemag number CMMI-7C."



here is a layout from a previos thread

http://twin-x.com/groupdiy/displayimage.php?pid=3026&fullsize=1

 

[CJ]

more info:

"These amps were NEVER loaded by 200 ohms.  They were built out to 200 ohms which is a big difference.

Mostly, the load on them was bridging.  Worst case would be if a fader was off, then they'd see the build out resistor and the 200 ohm of the fader.  So, a lowish load in that case but they're OFF...so who cares 

The whole idea was that the valve saw as little a load as practical."

"Upon reflection: If you can swing it, spend the extra $30 for the Hi-N. Cinemag.  In the scheme of your total parts cost, it's a fraction.
I'd say my new transformer sounds closer to that one than it does to the 50-50

I know you're using a PCB so it's a bit more complicated but there are other things too that affect the tone of pre amps.  Power Supply topology, component choices, layout, tubes, slight tweaks in H.T. or bias, different forms of bias, Feedback (don't use too much if you can help it)... etc...


The Cinemags handle a ton on signal and, in typical level use, are very linear.  Same for the CMMI-7C input if you have a stepped attenuator before it.  You won't be hearing much iron saturation.  The differences will be tonal shifts to various degress.  I've heard as much of a difference in a pre amp by changing the brand and voltage of a cathode bypass cap." 

"Dear Mr. Normal    (John Lennon called their first engineer Norman Smith "Normal"  so it's kinda, sorta apt.  )
None of the circuits in the desk operated into a 200 ohm load.  Each device 'presented' 200 ohms to the following circuit.  Except the oscillator which WAS loaded with 200.

Internal output impedance of the REDD47 (including the DCR of the original transformers) was about 50 ohms.  A 150 ohm resistor was added as 'build out' to bring it to 200 ohms.  Depended on what was following the unit as to what the actual load was.  Worst case: A following 200 ohm "T" attenuator or fader turned all the way off would load the amp with 200.  Add this 200 to the 150 build out so, worst case load was 350 ohms.  Again, this was when the amp was turned off.  Most times the amp is bridged by a higher load.


Not sure which spec sheet you're referring to, I have them all and there are quite a few but a 200 load on an original "built out" REDD47 would result in 6dB loss in gain.


Edit: which brings me back to a point earlier: measure the internal output impedance of your REDD47 and calculate backwards and you'll see that the anode source impedance to the transformer is WAY lower than 20+ K."
WB


RDH4 pg 519>
"loading results in higher wave distortion and secondary loading also results in higher transformer distortion, loading is generally undesirable, although unavoidable in some applications"

the effective load on the valve is given by R-L = R1 + R2/T*2, where

R1 =  resistance shunted across secondary
R2 = resistance shunted across secondary
T = N2/N1 = turns ratio

notice that the turns ratio is backwards from normal, ie, a 7:1 would be 1:7 in the formula

 

[Strawtles]

Thank you CJ

 

[CJ]

no problem, i learned a lot too!

just remember, the Beatles where famous in spite of their equipment, not because of it, they were always complaining about their sound, Lennon even told George Martin that he hated the sound of their records and that he wishes they could redoo everything,

same with Hendrix,  the Stratocaster is a real POS, hard to keep in tune, thin tone, noisy as heck, but he got the most out of it, .

 

[Strawtles]

You're absolutely right!
The proof is that all the recording made by the beatles (even those done in other studios) sound absolutely awesome, but what can I say ...  I fell in love with all the Abbey Road Machines and I'd like to realize all these machines 


An upcoming project that I'd like to realize is the EMI TG preamplifier ... but at the moment I think it is difficult to find the schematic

 

[ruffrecords]

@CJ, be careful using a 6DJ8, it doesn't really have enough heater cathode breakdown voltage spec and I think the max plate voltage spec is only 125V.

Cheers

Ian

 

[CJ]

thanks for the heads up Ian!

290 volts for a preamp stage is kind of high anyway, i guess they did that for gain,

i bet noise could be brought down by running maybe 240-250, 

a voltage divider to elevate the heater might be the ticket, less noise that way also,,

 

[ruffrecords]

thanks for the heads up Ian!

290 volts for a preamp stage is kind of high anyway, i guess they did that for gain,

i bet noise could be brought down by running maybe 240-250, 

a voltage divider to elevate the heater might be the ticket, less noise that way also,,

Heater elevation is the way to go. As for the HT voltage, if you want up to +22dBu (10V rms) at the secondary of a 7:1 transformer then you have 70V rms at the primary which is nearly 200V peak to peak.

Cheers

Ian

 

[CJ]

holy cow your right,  so when Sowter says you can get 26 db, he means with the sec's wired in series which cuts the turns ratio in half, or doubles the voltage at the sec, i get it now, thanks! you did mention this  gain boost up above,

this makes using a 4:1 a better deal than the 5:1 if you want to really hit the tape deck hard, as that Carnhill probably does not have split secondaries?

plus the high end will be better with less turns, and you can adjust for any small loss of bass by tweaking the cap feeding the parallel fed transformer,

looks like an open and shut case for going 4:1,


now if the Carnhill does have split secondaries, you could wire up a switch to get 2 differnt output levels and maybe a variation in sound,

 

[ruffrecords]

now if the Carnhill does have split secondaries, you could wire up a switch to get 2 differnt output levels and maybe a variation in sound,

It has split primaries AND secondaries.

Cheers

Ian

 

[alexc]

The Carnhill 10K:600 sounds good and works great in this circuit.
Take the extra gain and be happy! You need it for a red47 style circuit.


Worth the extra dollars over the Edcor imho.

And also thanks guys for the most interesting discussion of the morning 

Good question!

 

[CJ]

that Sowter is a 10K:200 with sec's in parallel and a 10K:600 with the sec's in series,

so it is a 10K pri anyway you slice it,

where do they get the 30K:600 ?

maybe they figure that if you load the transformer with a 600 ohm load when wired for 10K:200, that the reflected load becomes 600 * 49  (7:1 turns ratio squared) = 30K,

but if the sec inductance = 200 ohms at 20 hz, then loading it with 600 ohms will not make it a 600 ohm winding,

when you parallel 2 windings, the inductance stays the same as the turns are on the same core , it becomes kind of a bi fi winding, with the 2 sections in different places on the coil, you get better coupling which means less flux leakage, but it is still a 200 ohm winding, so i think the 30K to 600 spec is bogus,

you can load a transformer Down, but you can not load it Up,

load it down and it will reflect a lower pri z, but load up a 200 ohm winding and it is still a 200 ohm winding,  does that make sense?  ???

there is no such thing as an infinite ohms load, if you have air between the terminals of a 200 ohm winding, the reflected Z to the pri can not be infinite, the pri Z that the tube sees will be limited by it's inductance,

this means the Carnhill 10K will be a perfect match for the Redd 47,