REDDI DI

[abbey road d enfer]

a phase reverse is important for sure,

AS CJ said, it's very easy to implement with a DPDT switch. That may not be strictly necessary because most DAW's offer the possibility to reverse polarity ("phase"). Opting between a hard switch or a soft button is a matter of taste.

and ground lift...

The question of ground lift is nort as simple as it seems. There are potentially two loops that may generate hum. That happens when the guitar is connected to both an amp and the DI box. The two loops are:
one between the amp and DI "grounds"
one between the mixer and the DI box

The switch shown by CJ in his new version can break only the latter loop.
In fact, since the output is floating, a ground switch is not strictly necessary. I would install a low-impedance RC there (typically 100 ohms in parallels with 0.1uF, one side to pin 1, the other to chassis ground - not to audio ground).

In order to solve any problems related to amp-to-DI loop, the audio ground of the DI box should be made floating.
A switch should be installed between the audio ground and the chassis of the DI box.
CAVEAT: I'm not saying it is entirely safe and legal...

 

[raymancem]

What is Output transformer primary wire diameter ,
Secondary wire diameter
power transformers ,
Primary + Secondary ohmic resistance
Number of laps What is?

 

[CJ]

freq chart>

 

[Strawtles]

Is it possible someone will realize a PCB?
Cheers

 

[ruffrecords]

Is it possible someone will realize a PCB?
Cheers

Several people have used my poor man's tube gain make up PCB for this project.#

Cheers

Ian

 

[madriaanse]

these guys took a while, tried a lot of winding structures, finally got it,

100 EI horizontal mount,

ready for the varnish tank, DCR about 960 and 135,    8K:600,

Sweetness! Will you be selling these?

M.

 

[mutterd]

[

Sweetness! Will you be selling these?

M.

yeah - I'd be interested as well - Im making 1 of each of these circuits... got the PCB's from Ian (thanks!!!) and am waiting for the VTB2291 to come back in stock for the 6299 version... but I would love a hand wound transformer for the 6n1p circuit...

I had been thinking about trying one of these for the CJ circuit:
http://cinemag.biz/output/PDF/CM-2810.pdf

but i'd rather give the money to you...

 

[Strawtles]

Thank You Ian

 

[Fablab]

Hello, has anyone compared the sound of the 9600:600 Carnhill transformer with the original one?
I have used these transformers (the ungapped version) with my REDD 47 clone preamplifiers and I like them a lot

 

[CJ]

are you using it for bass?

you might be liking the saturated sound,

if you are running 2 sections of 6N1P for about 15 ma like the stock circuit,

then your DC flux is going to be 0.4 pi * 3800 Turns * 15 ma / (gap + (magnetic path length of 75 EI lams / effective perm for silicon steel))

or about 50,000 gauss, which is 2.8 times the max gauss for 29 ga M6  (Si Lams)

AC flux at 20 hz is only 1400 gauss for 15 volts into the primary of the xfmr, so no big deal there,

you have 480 Henries but not with that DC, so you really need to butt stack that core or  you will have mega distortion at the low end,
first did a  75 EI  OPT for a tube DI, did not have the 100 EI lams yet so now we have 2 versions, the 100 EI is not so rolled off on the  top end, and the bigger core means you have that low end rumble if you want it, but this 75 EI OPT sounds jiust fine,

if you figure that Carnhill uses the same 600 ohm winding structure as the LO 1166 for their 9600:600, then you have about 950 turns on the sec, so 9600/600 = 4:1. so figure 4 times 950 turns = 3800 turns primary, so i was not to far off from the Carnhill on this first attempt, same core, about the same pri turns, but it has a gap, simpler winding structure which is why it rolls of, just sec-pri-sec instead of sec-pri-sec-pri-sec like the Carnhill, the 100 EI version i did goes sec-pri-sec-pri sec but the sec's are paralleled like the Telefunken V76 OPT, so you are good to 60 K even with the 1 K pot, which tends to roll off the top of  transformers with simpler structure,

here is the 75 EI that is easy to wind,

 

[CJ]

just a note, if you use the ungapped Carnhill with the primary windings in parallel,
then DC flux will be cut in half to about 25 k gauss, still above the 18 k gauss limit of Si steel, but not as horrendous as 50 k gauss for the primaries in series,

note that DC flux goes up with increased turns if the current is kept the same,

and AC flux goes down with increased turns if the voltage is kept the same,

DC amp turns go up when DC current is kept constant and turns go up,

AC amp turns go down when turns go up,  why?

inductance is proportional to turns^2, so as turns go up, inductance goes way up, which means the AC current will go down quadratically as turns go up in a linear fashion, the net effect being a reduction in AC amp turns as turns go up,
because AC current will be going down quicker than the turns are going up,

a fixed core can only handle so many amp turns, increase amp turns and you need a bigger core, 

 

[alexc]

I too would break my vow not to do any more new builds if CJ was thinking of making a few of his 'special' traffos available ...  to the faithful .. for use in this design.

It would be a fitting last diy to have a CJ special in there ... for the discerning bassist/guitarist

Super simple but very, very classy.  'powered by CJ' you might say.

... just saying ... 


Regardz

 

[CJ]

working on a Neve type OPT for a tube DI,

same lam as the 1166/2567 = 75 EI,

sec wire size and turns kept the same,

same winding structure = 1/2 Sec - Pri Reverse - Sec - Pri - 1/2 Sec

10 ma DC in circuit test, different dual triode so a little under the 15-16 ma in the 6N1P circuit,

here is an inductance chart for different core arrangements, 1/2 by 1/2 means half the lams go in one way, the other half from the other side, 1/3 means core divided into 3 stacks, 7 x 7 means we alternate 7 lams one way, 7 the other way, etc,

looks like the 1/2 by 1/2 gives the best inductance for 10 ma, but for 15 ma DC i would guess that the no gap arrangement would give max Henrys,

No Gap = butt stack with no insl in gap,

Carnhill VTB 2291 i believe is the gapped version of a 9600: 600

http://www.audiomaintenance.com/acatalog/Output_Transformers.html

http://www.audiomaintenance.com/downloads/carnhill_design_guide.pdf

 

[CJ]

here is the low freq plot showing the effects of the stacking technique for a tube DI,

you could use the ungapped version for the DI box, but a 6 string bass might suffer with the 30 Hz B string, 40 Hz E string might be ok, saturated sound could be cool,

 

[CJ]

inductance and reactance curves for various gaps on the 100 EI lam, 10 ma DC,

bigger gaps increase hum,

different gaps shape the tonal response of the transformer,

high end stays mostly the same, but as bass is added and subtracted, the overall sound changes,

 

[alexc]

Fascinating graphs - thx CJ

 

[CJ]

no problem!

busted the 5963 tube i was using in this DI circuit,

was drilling a hole for a cathode cap bypass switch, which works pretty good,

stuck a 6BQ7a in there, sounds pretty good,

specs>

6N1P 
Amplification Factor  33
Transconductance  7500
Plate Resistance  4400 Ohms
Heater Current  600 ma
Cathode Resistor  220 Ohms

6BQ7a
Amplification Factor  35
Transconductance  6000
Plate Resistance  5800 Ohms
Heater Current  400 ma
Cathode Resistor  220 Ohms

old TV tubes like this are pretty cheap, this one is $1.79 at Antique Electronic Supply

 

[alexc]

Yeah - I too get a lot of satisfaction using some of the off-the-beaten-tracks tubes from time to time

My favs are the 6-8W dissipation single triode 9pin bottles - they work very well driving Edcors output traffos and such. There's loads of different ones to try! And lots of different amp topologies to check out, too.

That one is new to me - I'll have a read.

 

[CJ]

here is a DIY Carnhill  VTB 2290 if you feel  like rolling your own,

pri DCR came out higher because i did not stretch a bigger gauge to breaking point,

sec DCR is about the same,

stock pri  DCR is 350 each winding, i think i have about 430 each,

sec turns are adjusted to give equal voltage output on all windings,

you can either flip the bobbin to get the 3-4 rev wind, or reverse spindle direction and start at the right,

this is a gapped transformer

it sounds very nice, a bit more mellow than the EI 100 version,

a dual chamber bobbin was tried, this flattened out the response in the 50 kHz to 100 K hz region, so a single chamber bobbin will work just fine and give you more winding space and make taping and insl. easier,

 

[CJ]

here is an inductance vs low frequency graph for the DIY 2290 wired for a 2400 ohm primary,

let us say you are running the xfmr with the primaries strapped in parallel,

what would be the difference if you only used 1 primary?

just a slight drop in inductance at 5 hz,

however, i did notice a slight increase in hum while using only 1 primary,