Fairchild 670 DIY In Progress

[CJ]

thanks for the kind words you guys, keeps me going on this,

was gonna order a Fender cap can and put over these atoms, but too small, so get some 22 ga and some tin snips plus the energy saver soldering iron and we are set to fab our own cap box that also doubles as a Sola constant voltage transformer decoy, so we deviate from the norm with the axial caps instead of the two cans, which worries me because of failures witnessed in the TV biz where can insulators get compromised some how and you have a direct short to ground, so missing two cap cans, and the fancy gold insulator cardboard, but we keep the retro look with the fake Sola,

 

[CJ]

here is the can installed over the caps, which come after the choke in the regulated pwr supply.
i lined the can with fish paper, just in case the heat shrink on the caps wears out and shorts to the can, as there are two caps in series to ground, so the top cap case is sitting at half the B+, which is only about 200 dc, but im jus sayin,

 

[CJ]

on the flip side, we soldered a few things after the dreaded fab work on the can,
saved a lot of space by locating the 140uf/350 caps on the other side of the chassis, rubber grommets on the thru holes for the wires, got the pwr supply almost wired, and a few components on the signal amp,

using ero caps from West Germany, remember the Wall? 

nice epoxy end caps, which was top notch back in the age of wax and paper,

got some funky resistors too, don't know what kind, like a 3 watt carbon comp only metal on ceramic, 1950's or around there, values still match the color codes,


more tube sockets will arrive today so we can clean up even more sphaghetti, spell check on spheghetti, 

 

[okgb]

Clean up on isle 5 ,    wtf over ?  Go Cj Go Cj 

Are you still off the grid ?  this will warm your home , though you may want to
centrally locate the control room after this , nice progress !

 

[CJ]

still on the grid, yes, and the cathode, 

soldered a few wires today but the room mate drank some vodka and took some tylenols after his girl ditched him, thats why the lady is a tramp i guess,

so afer we called the looney wagon i soldered the control output primaries to the B+, not really worth takin a pic of,

went to the Sowter site, noticed a resistor that is of questionable value on the redrawn Drip schemo,

there are some voltage dividers that sit on the 240 regulated line, one is a 27L/100k for the screens of the 12Bh7a tubes, the other is a 100K/100K divider that is used to elevate the heaters of the  EL34 and the pentode that are us3ed in the reg circuit, there is no heater on the gas discharge tube,

the top resistor on the Drip heater divider circuit reads 27 K instead of 100K, so watch out for that if you use that nice redrawn schemo,

also, Sowter says the bias supply choke is 5 henries, Drip says 1.5 Henries, so either a mod here or a misprint, and also Sowter mis labled the pwr supply choke as a bias choke but no biggy,

re located the B+ control pot, 250K, also changes it to 100K because 250 K gives too wide a span,
orig position requires about 2 feet of wire that sits at 240 volts, to allow trimming from the lower front panel, this seems like tooo much pwr supply wire running around the chassis so i moved the trimmer right next to the el34 pass tube which dropped the wiring and stray C down about 2 feet or 200 pico farads, which ever comes first,

note that you can use an ef86 in place of the expensive mil tube that drives the EL34.
same pinout and everything.

also, Sowter sells the UTC Y 6431 replacement but has it at 1:17 ratio,

the ratio is actually 1:10, the transformer is identical to the UTC A-10, only no taps, same lams, just has a different can,

cj

 

[bockaudio]

went to the Sowter site, noticed a resistor that is of questionable value on the redrawn Drip schemo,

there are some voltage dividers that sit on the 240 regulated line, one is a 27L/100k for the screens of the 12Bh7a tubes, the other is a 100K/100K divider that is used to elevate the heaters of the  EL34 and the pentode that are us3ed in the reg circuit, there is no heater on the gas discharge tube,
the top resistor on the Drip heater divider circuit reads 27 K instead of 100K, so watch out for that if you use that nice redrawn schemo,

Both my original 660 & 670 schemos say 27k(top)/100k(bottom)/ divider for 6973/6V6 screens yielding "154"v really 171v.
True the gas tube "has no heater connection" as drawn on either 660/70 original drwgs.

also, Sowter says the bias supply choke is 5 henries, Drip says 1.5 Henries, so either a mod here or a misprint, and also Sowter mis labled the pwr supply choke as a bias choke but no biggy,

I also have 1.5Hy for the choke on my originals but how could more hurt? FWIW The 660 has no choke at that location.

re located the B+ control pot, 250K, also changes it to 100K because 250 K gives too wide a span,
orig position requires about 2 feet of wire that sits at 240 volts, to allow trimming from the lower front panel, this seems like tooo much pwr supply wire running around the chassis so i moved the trimmer right next to the el34 pass tube which dropped the wiring and stray C down about 2 feet or 200 pico farads, which ever comes first,

I've done about a foot unshielded more than once with no probs. but yes the 100k better than 250k.

note that you can use an ef86 in place of the expensive mil tube that drives the EL34.

finding a quiet one can be hard these days.
A dual triode works as well, with rewiring.

also, Sowter sells the UTC Y 6431 replacement but has it at 1:17 ratio,

the ratio is actually 1:10, the transformer is identical to the UTC A-10, only no taps, same lams, just has a different can,

cj

I have Triad HS-14 or UTC A-19 for this location.

 

[kepeb]

addictive thread so far!
seems like every time i check, this is bounding forward.
you dont hang about do you CJ
good on ya

 

[CJ]

there are two voltage dividers, the screen supply has the 27 k,

the heater divider has the 100k,

i have heard rumors of a triad being in there instead of the UTc, i bet they used whatever they had, these things were too expensive to wire, i bet that is why they quit making them, i have never seen a pic of a 670 with anything but a y 6431, so if you have some documentation or pictures showing a triad, that would be interesting to see,

ef86 with noise would be bad as a preamp, but as a regulator tube i would think that any noise will be filtered out by te caps,

 

[bockaudio]

there are two voltage dividers, the screen supply has the 27 k,

the heater divider has the 100k,

right the screen is 27k/100k and the heater bias is 100k/100k

i have heard rumors of a triad being in there instead of the UTc, i bet they used whatever they had, these things were too expensive to wire, i bet that is why they quit making them,

I have seen both kinds i n factory wired jobs.

ef86 with noise would be bad as a preamp, but as a regulator tube i would think that any noise will be filtered out by te caps,

guess I'm paranoid about noise after working on mics all day.

 

[CJ]

i hope the drip guys are not using the 17k as shown in the schemo, it will put the heater of the pentode at 170 volts which is above the 84 volts on the cathode held by the discharge tube,
yes i think i went by the mic booth at AES, correct?

 

[CJ]

just a few more wires,hooked up the outputs to  the plates and B+, rebuilt the pwr supply turret,

you can spy the B+ adj pot by the rear turret,

back to full time so quicker progress ahead,

 

[hop.sing]

That looks soooo much better and is definitly much more fun than the drip bulid. Great thread.

 

[bockaudio]

That looks soooo much better and is definitly much more fun than the drip bulid. Great thread.

P to P =awesome. Fixable. Experimentable.

 

[alexc]

Got to love those vertical turrets - never saw those before.

It's clear they can really make things more tidy in a complex build like this.

Nice one! Another tip for my own pp take on this classic.

This is going to be one really nice unit.

Cheers CJ

 

[radiance]

Yo, yo CJ 'sup?
Thought you is going full time on this one....suspense is killing....

 

[CJ]

had to relocate a turret, but i bought a new pair of old stock USA Xcelite 57CG 140 mm needle nose which is making life good again, this job requires a good set of needle nose,

i am adding 2 additional turrets because i have them and the original wiring jiob on the 670, i am finding out, leaves much to be desired.

this was obviously Rev 1, and they built a wire loom for hooking this thing up,

i did not make a wiring diagram for this as i should have, would have saved a lot of rework, but i think i got a good layout on channel 1 control amp, so we can copy that on channel 2.

they had to with the hassle of the matrix circuitry, too many wires as it is,

ran out of 1 meg resistors but just got back from the store so we cool,

just ordered the 12bh7a's, out of RCA so i go for the GE.
orig circuit has 100 k plate resistors for 12bh7a, this is way off the standard 10 k you see like on the LA2a, so i was not worried about subbing in a well matched pair of high wattage 92K resistors, there is a reason they are starving the plates, probably in a PRR thread somewhere,

home prices are up for the 6th straight month in a row,

it looks like the DCR of the bias choke is being used as a resistor, never a good idea as the dcr of a cheap choke is not exactly a precision resistor, if no choke, you have about 30 volts dc worst case on  a 100 ohm dual pot to ground, which should read 100 ohms to ground no matter what the setting is, so 30 volts across 100 ohms gives 1/3amp into 100 ohms is 5 watts, so we add the 800 ohms from the choke, 30/900 = 33 ma, which is about 1 watt of power, most of this will be seen by the choke, so the choke will get hot after it warms up,

with solid 1n4004 instead of selenium, we need to drop the xfmr sec voltage down,
so we will dial that part of the supply in with a variac and see what transformer we need, there are other voltages for biasing cathodes, so it all needs to add up.

i had the same sprague black beauty 0.05 and 0.1 at 600 coupling caps, but after i hooked a 9 volt battery in series with a micro amp vu meter, i noticed they were all leaky, so don't spend a fortune on evil bay for those things, most of the ones i have seen are pretty bad.

ok pictures at 11, or as soon as i find the usb cord,  :-X

 

[buildafriend]

<---- waiting for 11 to come

 

[CJ]

ok we have the buss bar thing going in, wires are going away, gonna be ready for a time constant circuit pretty soon and of course mo iron, going for utc a 26 and triad hs 52, gonna wind the y-6431, need another 6.3 aux heater trans and a choke and pwr trans for the bias supply,

added an extra turret which you can see on the far right, the 670 with the most tube turrets wins, right? 

changed the 0.1 caps to a smaller model but same value, figure the smaller the better for response time, a big roll of foil will act as an inductor, so we need a quick cap, silver foil gives more juice per area i think, have to look that up,

the dual 180 k control amp input panel mount location requires that the 12ax7a grid wires get routed thru the big bundle and up to the pot, lots of wire there that we can shorten, not good to have bare grids attached to a wire antenna,

the buss bar thing is a pleasure to work with, anything that needs grounding, like cathode resistors, grid resistors, well you just wrap that wire around the buss bar way up there out of the way of everything and melt away,

since both the audio and control signals are going thru that wire only, the whole chassis is eliminated as a noise source, a chassis will have transformer noise and external radiated noise running thru it, and this is a big chassis, so good idea on the buss bar, plus, this being a balanced circuit, you have even more noise rejection.

also, the buss bar adds a lot of strength to the turret system, once that wire is hooked up, the whole chassis gets stiff, plus you can hook your volt meter ground alligator up to it and it will stay there,

using ero foil caps and big ceramic resistors, the big guys are 4 watts with color bands, which is pretty weird,

gonna fire up the main pwr supply pretty soon, gz34 by JJ, any good?

mullard gz34 going for 200 bucks on evilbay so scratch that, JJ is 19 bucks so lets see if we can blow one up, 

 

[CJ]

if you do a production run on something lie this, you can pre wire the turrets and this will speed up production and make it easy to wire,

this is the cathode/tert resistor ladder for the 6973 output tubes,

points deducted for the color bands going opposite ways, 420 wiring is a blast!  ;D

 

[volker]

The JJ GZ34 works perfectly fine. Its behaviour is a tad differently though from vintage specimen. The internal resistance seems to be lower and hence also the voltage drop. But since you have a voltage regulator in there that shouldn't matter too much.