RCA 96-A from scratch

[DaveP]

The front panel is now designed and drilled for components and brackets.

All the bracket screws/holes had to be countersunk and back-filled.



I used a special metal filler for the first time (seems to work!) but I'll know more if it sands down flat OK.

This is the back view of the front panel showing the brackets which attach it to the rest of the amp.



I sealed the threads with red nail varnish as the special stuff from RS has prohibitive transport.  I felt a little bit weird examining the  nail varnish section in our local super marche, but in our gender fluid age who cares? :

Next job is to prepare the front panel for spraying and lettering.

DaveP

 

[DaveP]

The area under the masking tape has been polished to a mirror finish to take the lettering.

The rest has been sanded with 240 paper to give the primer something to grip.

The metal filler sanded down very nicely as hoped.



I gave it two coats of grey metal primer an hour apart, this will now cure overnight, before I paint it black.

DaveP

 

[RuudNL]

I also use nailpolish to secure screws.
But my wife always makes jokes when we are in a shop.
Then she says (more than loud enough for everybody in the shop to hear):
"Are you sure that this is the right color for you?"
"Is it the same nailpolish you are always using?"  ;D  ;D  ;D

 

[DaveP]

I also use nailpolish to secure screws.
But my wife always makes jokes when we are in a shop.
Then she says (more than loud enough for everybody in the shop to hear):
"Are you sure that this is the right color for you?"
"Is it the same nailpolish you are always using?"

Next time it happens say "No, this is my mistress's favourite"  Touche'! ;D

DaveP

 

[RuudNL]

;D

By the way: I really enjoy the pictures of your builds. It is always nice to see real craftsmanship!

 

[rackmonkey]

Looking good, Dave!

What brand of metal filler did you use? I've used a type of Loctite before that I wasn't happy with. It seemed to leave a haze even after curing that showed through two layers of primer. Doesn't appear you had that problem.

 

[DaveP]

I used Evo-Stik  metal epoxy putty.

It comes as a kind of grey swissroll and you cut a slice and knead it (with plastic gloves on)  until it's a uniform colour, then you have a minute or so to use it before it sets.  The point of a thin knife is handy to apply it.

I found it stuck well as I filled the countersunk holes, then I lightly used a damp cloth to remove some of the excess.

I don't know how long the remainder will last now, but the core seems insulated from the rest somehow.

DaveP

 

[scott2000]

I've used a type of Loctite before that I wasn't happy with. It seemed to leave a haze even after curing that showed through two layers of primer. Doesn't appear you had that problem.

Maybe a pass with some glazing putty would help??

 

[rackmonkey]

Thanks, guys. I'll try the glazing putty next time, Scott.

 

[DaveP]

Thanks for the tips, but I'm not sure that product is available in Europe.

The front has now had two coats of matt black.



I used matt because it will be varnished later.

The mirror finish panels will have lettering to be protected as well as the finish itself, hence the varnish.

I can now proceed with the rest of the chassis.

DaveP

 

[matriachamplification]

Amazing! Learning a lot from this thread.

 

[DaveP]

The chassis metalwork is on the way to being finished, I'll post on that in a day or two.

Time to turn to the Power Supply.



This is a schematic I have used several times (660's) because it's simple and it works.  Because the current is around 70mA it's asking a lot of an EL34, but I have several 6080's and one of those paired up will work fine.  The original 96A used a separate power supply, but this chassis is big enough to take it onboard.

I use several 240 to 2x55V 50VA toroids for this design as the under used regulation  gives a voltage boost.
These are the provisional values of the components, some may need tweaking on test.

DaveP

 

[DaveP]

I have made the sides and power supply wall and the removable base plate so far and I'm preparing the back panel to take the various in/out sockets.  This is the view of the base.



The accuracy is about 0.25/0.5mm which is the best I can do with hand tools in a reasonable time.  This is the power supply housing where the toroids are mounted vertically.  I will assemble the transformers on the side plate as a complete unit before attaching to the rest of the chassis.  Connections will be accessable via a removable back panel.



This is the front view showing the large size of the 96-A, and my version is several inches smaller!  Ian  wants to rest another unit on the top so I will provide a top to go over it.



The dimensions are all settled now, although not all the panels have been cut to size yet.  I can now work on the physical layout of the components on the main chassis.  This can't be rushed as you have to imagine all the connections and components in place.



I hope to cut and fit the remaining panels over the next few days.

DaveP

 

[DaveP]

All the panels are now cut and fitted so the chassis is now finished.........



The large back panel gains access to the meter and switches, and the ventilated panel on the left is for access to the power supply.

The power supply now has to be assembled and fitted.

DaveP

 

[core13]

Designers keep the g2 voltage reasonably constant to get a sharper knee on the compression curve.  The Altec  436, for example, has no means to hold any voltages constant so the knee is very soft, this is fine for recording guitars, but it would not be any good for radio where there are penaties for exceeding your transmission power.  I don't want to use a tube regularor because it will affect the knee and change the character of the compression.  I hope that's helpful.

DaveP

do you think replacin the gz34 on a 176 with diode will change the knee an give an other flavor?
to use it as bus comp for exemple

 

[DaveP]

The 176 has an OB2 voltage regulator, so I would not expect changing the GZ34 for SS diodes to make any difference unless the OB2 was not working properly.  But using diodes would increase the supply voltage and some components would not like that!

DaveP

 

[emrr]

do you think replacin the gz34 on a 176 with diode will change the knee an give an other flavor?
to use it as bus comp for exemple

On the scent but not there:  older threads here discuss changing the side chain rectifier to SS diodes changes the knee relative to tube diodes.  PSU rectification apples to apples at same voltage does nothing, there’s no momentary sag under load in the vast majority of designs. 

 

[core13]

Thank you for your precision,

I am currently building a pair of 176 and it could be a nice mod to make the 6AL5 switchable with SS diodes

Dave your RCA 96 is awesome, the amount of work is huge
I also remember your fairchild.

thank you for sharing this.

 

[DaveP]

That might not have much effect, but switching out the OB2 would be likely to soften the knee.

A two way switch could disconnect the supply from the OB2 and connect another supply from a resistor instead.  This would make the knee similar to a 436 for example.

DaveP

 

[DaveP]

Well Covid-19 or not, the beat goes on!

Self-isolation is ideal for getting on with projects!

I have finished the Power Supply today with the exception of two dropper resistors that will have to be determined on test.



There is a 50VA TX for the DC heater supply at the top,  then two 110V TX's in series for the HT/B+ and finally a smaller TX for the elevated supply.  My chances of finding one TX to do all those jobs are zilch.

They all fit in nicely on the right hand side panel, this mounts them vertically and avoids any hum pickup.

The next job is to fit the major components on the main amp panel and settle the layout.

DaveP