Gates 28 CO New Build

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

DaveP

Well-known member
Joined
Nov 8, 2005
Messages
3,019
Location
France
In two weeks time I will be starting the recreation of a 1947 Gates 28CO.

This is an interesting compressor as it is based on the changing resistance of light bulbs, which makes a change from the many vari-mu designs that came after it.  It started life in 1937 as the 17B



But that had no provision for release so after WW2 in 1947, Gates came up with the 28CO which incorporated an adjustable release control.  This was unusual in that it was derived in a similar way to Vari-Mu types but they used low noise 6J7's instead of Vari-Mu tubes, equally unusually they applied the negative voltage to grid three of the pentodes instead of Grid 1.  I have tested this and it does work ok; it is a very subtle effect.



The original is a large unit about 7u high but I will be making it 3u size with modern components.  It still uses 5 audio transformers though!!!

This is the original 28CO:-







Thanks to EMRR for the pix.

I expect the 28CO to have an interesting and unique sound.
best
DaveP
 
Here we go again. I love these threads from you!
What happened to the SA-39B, haven't heard anything about it since it crossed the Atlantic?

Best
Magnus
 
Thanks Magnus,

Doug has it at EMRR and it has been used for some recording.  He has a lot on his plate at present so this is a long term venture.

best
Dave
 
Man, these threads are always such an inspiration. Just the other day I discovered a small metal shop down the road that has a sheet metal folder, and I thought, "I should make an enclosure DaveP style." I especially like how esoteric this compressor is. I'll be watching this project closely!
 
I have been working on the layout so I thought I would share the general principles I use when designing a point to point layout (probably a dying art).  Some are obvious, others not so.

1.  Power TX's as far away as possible from low level signals.
2.  Use Toroidal power TX's wherever possible.
3.  Keep audio input socket well away from audio output socket.
4.  Keep tubes far enough away from each other and components to avoid over heating.
5.  Keep tubes near enough to each other so components can reach sockets (~2"/ 50mm)
6.  Keep push-pull layouts symmetrical to avoid inductive imbalance in wiring.
7.  If possible, use output TX's to screen power TX's from low level signals.
8.  Put power TX's in screened box if possible.
9.  Have plenty of spare tag strips available for the prototype.
10. Only attach chassis to earth ground at one point.
11. Do not use chassis for earth returns.
12. Use chassis for connecting all screened cable earths and in/out sockets.
13. Either use a star one point earthing point or a heavy busbar.
14. If using a busbar, connect the low current connections upstream and heavy returns nearest earth connection.
15. Insulate the busbar from the chassis (grommet holes in screens if passing through them).
16. Do not run signal wires along heater wires or earth returns.
17. Cross signal wires at right angles.
18. Use DC heating for tubes or a centre-tapped AC supply with twisted wires.
19. If using several busbars, join them nearest the earth power socket.
20. These principles make a symmetrical layout similar to the schematic the most promising.
21. Screen as much of the audio switching and controls as possible.
22. Metal tube cases grounded to chassis.
23. Adjacent transformers at 90 degrees to prevent field pick-up.

If you follow these guidelines the layout will almost design itself, but be prepared to compromise.

I expect I've forgotten something, but that's basically it.
best
DaveP
 
Nice list.  I'll add that can caps were often laid out for their screening properties, and metal tubes are useful in areas with a propensity for AC field pickup. 
 
I like to use the metal can caps for the mains filtering in tube builds with sufficient space.

A wide variety of GE motor run caps, big and small are available at 'surplus centre' in the US for very reasonable prices. They are almost .. architectural .. in the visual statement.

And I do find they help in shielding.
 
I've started on the chassis now.
Had to do lots of trial layouts as there is a lot to fit in, like 5 audio transformers and 3 PS toroidals.



The power supply will go in a screened box in the space on the left, two on top and one below.  A piece of studding holds both toroids on the same axis.

Not much to show at the moment, but you can see where most of the tubes are going, OPT's are central.  I keep the protective film on for as long as possible for marking out and avoiding scratches.



This screened section at the front will house the meter, control lamps and usual switches and pads.  The thick section front panel will be screwed onto this.  There will be special angled sides to support the weight.

best
DaveP
 
Been working hard metal bashing.



This is the front screened section finished.  Should be enough grommets for all the wiring that needs to come through.

The reinforcing angles all in place and the housing for the power supply.



This shows how the Toroids fit in, half of B+ TX at front and heater TX in middle, heatsink for 6.3V supply schottky diodes at rear.



All those octals to cut out.... Aargh.  Still have to decide where to put the 6H6....  No suggestions required ;)



Back panel cut out for audio in/out power and fuses.
Probably more pics tomorrow.

best
DaveP
 
Oh yeah - looking like another DaveP classic build in the making    8)

So very cool!
 
One hours filing later and we have 9 neat octal holes....I have resolved to buy a hole punch!



I've fitted the rear sockets but just placed the transformers for the pic.  The tx's were specially made for me by Livinginthepast in the UK.  I could have bought them from the USA or Canada but the shipping costs more than the transformers.  In the first amp with 6F6's, the OPT has a 10k primary with separate 8 ohm and 600 ohm secondaries.  The 2nd amp has an 8k primary with 600 secondary.

I shall do some more assembly next week before starting on the front panel.

best
DaveP
 
Thanks Alan,

Assembly has begun....  ::)



Here you can clearly see there are two amps, the nearer one also has a transformer input but it will be behind the front panel near the output control.  The HT and LT transformers are now fitted in their screened box.




I promised myself a hole puch for future octals, but they are an odd size ~26mm and there is nothing like that available.  Nearest is 27.5mm but they would slop about in that size.  I have decided to get a conical hole cutter instead, they go up to 30mm.

I decided to fit a screen for the PS section, not sure if its going to be needed, but its much easier to put in now than later.  Put in plenty of grommets just in case.




I pre-wired the sockets before fitting, makes life easier that way.  There are 2 massive HT caps for the B+ and 2 equally massive LT caps for the heaters, they all fit in neatly down the side so that the busbar can connect all the negative terminals in a short run.  The original has 2 chokes and 3 small caps, I did the maths and 2 big 470uF will just better the ripple level over the stock PS. Not sure exactly where the heatsink for the schottky diodes is going yet, but at least I have a choice!

By next week I will have the tag strips in place and the heater wiring and other busbars, after that I should be clear to go for components.  Even though a lot of thought goes into the layout, its still possible to realise you've made a C**k up until you get to a certain stage in the assembly, I wouldn't be rash enough to say I'm home and dry yet, but its starting to look as though I might be ???

best
DaveP
 
I love the *three* power supply toroids, neatly canned in a row :)

It will be interesting to hear your asessment of the interstage traffo from LivingInthePast - those are not easily found items.

Looking good - all shiny like.

Cheers
Alec
 
Thanks Alex,

Yes, I have to use 3 toroids (one for LT) as no-one makes a high voltage toroid over here, so by using 2 x 2x55V in series I can get over 240V before the bridge rectifier.  They are oversize so the regulation allows a higher voltage than the spec.

They are OPT's actually and the spec is 20-20kHz, one is 8K:600 and the other is 10K:600:8, I'm very pleased with them and the guys at Livinginthepast have been very helpful to me.

best
DaveP
 
Dave - can you clarify a bit on the role of the light bulbs ?

I can see reasonably conventional sidechain :

- audio interstage transformer secondary with ground referenced drive to a sidechain transformer primary.

- sidechain transformer secondary with balanced winding drives HiZ the recitifier tube, and the centre tap driving a time constant network to provide a CV, which would be a positive  negative dc varying voltage.

- CV modulate the screens - not sure if it's providing the entire screen voltage or just modulating a seperate screen dc supply.

and the audio path  is :

- t-pad attenuator to transformer primary, feeding a balanced gain reduction amp of 2 stages
- first is the GR amp, second is a fixed buffer feeding first interstage transformer
- audio continues thru the interstage traffo, to a seperate ground-referenced winding, and drives the lamp network and then to the next interstage trafffo - in a sort of asymetrical kind of way

- followed by a straight foward balanced line amp with balanced rotary attenuator at the input

--

So what is the lamps' varying resistance (as they light up/down) modulating? And why does it seem to be organised into 2 different resistive arms of the network?

Does it further impress some non-linear thing onto the audio? Is it a 2 stage waveform thing?

Looks very interesting indeed.


 
Alex,

The gain reduction is done mainly by the Lamp Bridge, there is only a small amount done by the negative voltage on G3.  The main purpose of the G3 voltage is to provide some release control, gain reduction is minimal because 6J7's are not remote types.

Although the transformers look like they are in a typical "Interstage" role, they are in fact ordinary output transformers, the first one could drive speakers with its 8 ohm output.  This is used to drive the lamp bridge which has a similar loading.

The bridge works like this, one lamp has a 7.5 ohm res in series and the other has 15 ohms, so when they start to heat up, they do so at different rates.  The resistance rises with heating, so that one bulb increases its resistance relative to the other until they almost come to balance.  The circuit only passes audio with imbalance so the signal passed is reduced past the threshold point.  I hope all that made sense!

best
DaveP
 
DaveP said:
...designing a point to point layout (probably a dying art).

Dying...I don´t think so-
Art...most definitly!

Saying while I´m still very far away from what you´re doing here, I want to get there someday and I´m also far away from dying (knock on wood) and I´m sucking it all up to the last bit here thanks, threads like this are like presents for curious me ;D
 
DaveP said:
The gain reduction is done mainly by the Lamp Bridge, there is only a small amount done by the negative voltage on G3.  The main purpose of the G3 voltage is to provide some release control, gain reduction is minimal because 6J7's are not remote types.

Oh right  (penny drops)  :-[  Forgot that bit!

DaveP said:
The circuit only passes audio with imbalance so the signal passed is reduced past the threshold point.

Still pondering that bit.


Thanks Dave - that clears things up for me. Mostly.

Regards
 
Back
Top