GA-5T Skylarck Mod - Is this idea planned correctly ?

[kagliostro]

This is a begin of an idea and starts from something that Greg told in one other (my) thead about GA-5T

..... If you have an ECL82 with transformers, you could always euro-gibsonize with a single EF86 up front and keep the tremolo circuit as is without negative feedback

I've enough current for 3 tubes on mt PT (no Rectifier Tube will be used) so I was thinking to the 5879 CH1 on GA40 Les Paul ....

at the moment it is only an idea ...

Stock GA-5T Schematic


Modified GA-5T + GA40 Les Paul CH1 Schematic


Thanks

Franco

[Greg]

You will loose about 40% of the gain in the mixer network that form a voltage divider with the grid resistor of your power tube. You could always plan a simple SPDT channel switching solution. A powered relay might look like an anachronism. Also, the speaker will play backwards in the pentode channel which will not make a big difference. I would suggest both tubes being fed from the same and last B+ node. Of course the single pentode stage have less gain than the two 12AX7 triode stages. You could try to boost the 5879 a little with a 220k plate, 820k G2 and 1200 Rk combo and it will still have less gain than the 12AX7. My idea was only for a simple one channel amp less-is-more philosophy. But of course, more could be more. 

[kagliostro]

Ciao Greg

Thanks for the answer

I was thinking to a loose in efficency, but 40% seems really too inefficent

About the fact that the 5879 will have less gain respect the 12ax7 gain, this will not be a great problem

thinking to add the 5879 the intention was to have a different flavour channel, not a higher gain one

(I rejected the option to insert the signal coming from the 5879 on G2 of V1b, also if this is a feasible option)

I'm still thinking about what to do (to add or not to add the 5879 channel)

however the mixing resistor solution is scrapped, may be a switch will solve the problem
(I must do a search for an old thread here on the forum where this thing was discussed)

a relays, as you say, seems excessive

--

OK for the node supply suggestion, but that is the standard PS schematic and I'll submit to you and friends a slightly modified version (your councils and Pete's councils)

I forgot also to mention that my OT has on the primary winding an intake with choke function to be used with G2 of the Power Tube

The OT I've is similar to those used on this Grundig Radio



Franco

[Greg]

Hello Franco!

a relays, as you say, seems excessiv

Well I said anachronism, but that depends on the vibe you want to give that amp. Kind of practical though (just one input jack and a switch)

however the mixing resistor solution is scrapped, may be a switch will solve the problem
(I must do a search for an old thread here on the forum where this thing was discussed)

Well you plug your guitar in either channel and you flip the spdt to assign it.

I forgot also to mention that my OT has on the primary winding an intake with choke function to be used with G2 of the Power Tube

That will be a good idea to use! A choke-shared primary...very clever...never saw that before. You can use the same 2k/50µF for screen decoupling for the ECL82 and 100 µf for the first reservoir cap (like the Grundig) to minimize ripple since you are going solid state for the rectifier. You can scrap the choke idea that was discussed on the other thread. It was appropriate for use with the 5Y3 that can only take 20µF.     

 

 
 

[kagliostro]

Well you plug your guitar in either channel and you flip the spdt to assign it.

I'm still wondering about what to do, also if to add or not the second channel

Other meat on the fire was put by my son, I discuss with him the arrangement of the amp (the amp is for him) and he asked

a pair of things (that previously I've think to avoid, Low and High inputs and a Tone Control ...... )

The chassis I want to use is small, only 12cm (4.70") large, so there is few space on the faceplate

(however the chassis will be rebuild, so no problem for the position of controls on faceplate)



I'm redrawing the schematic (instead to modify an image) so it will be easier to submit ideas for mods

Franco

[DummyLoad]

was thinking something similar to the 5F6A channel mix network...see attached.


--pete

[shooter]

was thinking something similar

DL you're way too fast   I was sketching up a frontend just like yours!  only difference, I just had 1 input, 4 input choices would confuse me

[DummyLoad]

channel switched. note: R6 & R7 skipped. you could add a 22M from V2B g1 (pin 2) to ground for to protect the triode if the switch network fails, and for the momentary loss of bias during the switching transition time which is insignificant.


--pete

[kagliostro]

Thanks Pete & Shooter

Nice, the switch version will have best performances

Pete, which is the purpose of the 56K resistor from the input (after 47K resistors) to G1 of 5879 tube ?

This is what I've draw before to see your version



switch (before power tube) and two shared inputs

I'm not sure if to short to ground the exit of the unused tube will be enough

or if it will be better to short to ground also the input of the unused channel

(and this will become in separated inputs with ground switch because I don't want to use relays)

Franco

p.s.: PS is to be completely revised, this is not the definitive version

I forgot to insert the volume control of the 5879 tube

[DummyLoad]

k, the switched 5879 as you posted will likely not have enough overall gain. if you use a EF86 you'd push at least 50% more gain. maybe lower gain is what you want.

56K is grid stopper for RF. you need one on your 5879 - don't share the grid-stopper. split the input at the 1M grid leak and add 56K for the 5879. pentodes require a larger grid stopper than triodes.

--pete

[DummyLoad]

just killing vacation time...hot-rod GA-20.   


--pete

[kagliostro]

Pete

the more I look to your version the more I like it



Franco

[kagliostro]

This is the new version



I've think to arrange that way the inputs (10K resistor on the pin of tubes, 39K resistors on jack - the original GA40 has 51K resistors on jack), the PS is arranged to use the material I've (PT - OT - Choke)

I've also a pair of ON-ON-ON Switch (similar to C&K 7211 used by Steve), so I've think to use one of it (instead of a DPDT Switch grounding the unused channel), using this ON-ON-ON Swich, is still necessary the use of the 22M resistor ? Can be a problem to connect in parallel A+B channels ?

I lowered the 1K resistor on the PS to 500R as I added the choke and my B+ is low (may be I can lower also the 22K and 10K resistors that feed the V1a/V1b and V1bis tubes), I added a 220R resistor on G2 (as from Peter's council) and feed the G2 via the "Choke" incorporated on the OT

Now I miss a Tone Control (will the 5F2A be adequate ??), how can I insert it  ? and, in general, what comments do you have ?



Many Thanks

Franco

p.s.: Peter, the resistor that feed G2 on the 5879 tube, on your schematisc is 820K, I put the 750K value of the standard GA40 on the schematic

which is your council about the value to be used (with my B+) ?

[DummyLoad]

K, i used 820K as it is a standard value and more common than a 750K. if you have a 750K then by all means use it instread.

you could use lower values for the PS rail drop resistors, however, if you halve the value of the resistance, then double the capacitance.

if you do use the ON-ON-ON switch, you should not need the 22M POP suppression resistor - try it and first without it then add it if you need to.

the 10K grid-stopper for the triode is not needed: the 2x39K are sufficient. install it if you so desire, as it won't cause any issues. i'd suggest using a 33K for the pentode though. too large of a value grid-stopper shaves highs off. if you pick up radio, then increase the value of grid-stopper or add a 10-22pF RF shunt from grid to ground or grid to cathode.

running both triode and pentode in parallel should not be a problem and it makes for some unique harmonic overtones when they are blended. fiddle with the NFB resistor value for even more complexity.

--pete

[DummyLoad]

adding a tone-stack will suck some gain. try it. the 5E3 style mixing network changes HF response at lower volumes. you can compensate by using a large value coupling cap. using a larger coupling cap may cause farting low frequency response (excessive bass) with a smaller speaker. the 5E3 values seem to be a good compromise.


--pete

[kagliostro]

if you pick up radio, then increase the value of grid-stopper or add a 10-22pF RF shunt from grid to ground or grid to cathode.

Good, I didn't remembered the capacitor trick 

... fiddle with the NFB resistor value for even more complexity.

That is easy to be implemented (only I must have space on the faceplate for a pot    - I hope I'll have enough)

.... the 5E3 style mixing network changes HF response at lower volumes.

Are you referring to the Tone Control as arranged on 5E3 ? (you know, sometime my translation isn't adequate to the require)

Many Thanks Pete

Franco

[Greg]

Unfortunately in your power supply, both of your chokes are pretty much useless in your updated version. All the current must pass into the big choke and your midget choke isn't decoupled. Pete's plan is obviously good, but you might want to modify it to include your own ideas, which would look more like this.

[kagliostro]

Ciao Greg

Usually those "choke" windings on OT are rated only for G2 current, the OT is recovered, so I've no more data about it, I think I can't use that part of the winding to supply the whole amp

What I can do safely is here



But .... we are talking about an SE amp that runs at idle only a bit lower than max roar, so in my poor mind the two versions of the connection of the 7H choke to the circuit will have near the same practical result, or I'm completely wrong about this ? 

Thanks

Franco

[Greg]

Usually those "choke" windings on OT are rated only for G2 current, the OT is recovered, so I've no more data about it, I think I can't use that part of the winding to supply the whole amp

Not the whole amp, just the preamp, and that will only be 2.5ma if you stick with high plate loads. The screen won't go that much higher if it is properly decoupled at a lower potential than the plate. Also the added series resistor could have a limiting effect if large enough. The whole amp current must pass through the first choke if you want to minimize ripple because there is no humbucking effect in a S.E output stage, that's why. Your midget choke will go further in knocking some more but now that you are using a pentode followed by a triode, the hiss is likely to overpower any small amount of hum.

[Greg]

And now you have plenty of gain to add a tonestack. It could be assembled around a 250k/250k concentric pot if you got limited chassis space.

[kagliostro]

Ciao Greg

The whole amp current must pass through the first choke if you want to minimize ripple ................

OK, I understand, is my last version doable ?

--

About Tone Control which will be your insertion point, after V1b, before G1 of the power tube will be a bad choice due to the NFB circuit, immediately before G1 of V1b ?

Franco

[DummyLoad]

And now you have plenty of gain to add a tonestack. It could be assembled around a 250k/250k concentric pot if you got limited chassis space.

i'd leave it alone. gain structure in that amp without the tone stack is a significant contributor to the tonal character of the GA5. the pentode is an added bonus. add the tone stack to the pentode if you desire. without a tone stack on the triode channel it stays true to the original GA5.

--pete

[Greg]

OK, I understand, is my last version doable ?

Yes but you still have to decouple the screen stage with a capacitor after the midget choke to make it work as a choke. However I would add a 1k resistor to bring down the screen voltage a little. You can now scrap an extra RC section in your B+ rail.

[kagliostro]

Yes Greg, you are right

I forgot to recreate the original situation (but one reason is I've a bit less B+ than the original GA-5T and this distract me in some manner)

About the Tone Control I agree with Pete, it will affect the original tone in some way

we must remember that the GA-5T didn't use one Tone control and use high value plate resistors on V1a - V1b, sure this is due to achieve a specific sound, now there is also a pentode on the equation, but all we know is that pentodes didn't like Tone Control on theyr exit because of the excess of load (only few storical amp has a Tone Control that follows a pentode), may be V1b can be of some help as it will recover and increase the signal before to send it to the power tube, but I'm not sure because of the heavvy load on V1bis

May be something of positive for the use of a Tone Control will happen when V1a and V1bis are paralleled but also here all is experiment

One possible option will be to put a Tone Control between V1b and the power tube and insert it using a switch that at the same time disconnect NFB, but all this is very intriguing but intricated

Franco

[kagliostro]

I corrected the PS and remembered something that may be can solve the Tone Control issue



The High Cut Tone Control is borrowed from here, as far as I have no idea if the values of the components can be right

and this will be something that requires higher abilities than I've, so I think this will remain an idea ....



Franco

[Greg]

You could install the control before the power tube. If you install one with a switch, it will remove itself from the circuit at zero. If you want to «preserve» more highs at low volume you could bypass one (or both) of the volume control with a 120pf cap (to make humbuckers sound brighter). You might be tempted to use the tone control when plugging a tele to roll-off some highs. It all depends on the speaker.

[kagliostro]

You could install the control before the power tube. If you install one with a switch, it will remove itself from the circuit at zero

if placed there the Tone Control will interfere with the NFB Loop, this will be a problem, so, the switch that insert the Tone control must also disconnect the NFB

I must define other details befor to decide what to do

--

The standard NFB resistor is 47K and, as suggested by Peter, I would like to modify the situation as to have control on the NFB loop

I think to use a pot and, may be, an in series resistor

any idea about values and if it will be interesting to have a switch to disconnect the NFB (despite the Tone Control presence or not) ?



Franco


p.s.: Editable FidocadJ file (.fcd) on the attached .zip file (FidocadJ is a free CAD for schematics and PCB)

[Greg]

Ok, I would suggest that you build your amp in a way to facilitate component change in critical areas for some A/B testing. I think you are ready now. Do some tweaking when you hear the amp through its speaker. I think that's where the fun is. When you are settled, replace all the hanging trimmers and switches and drill some extra hole(s).

[kagliostro]

Thanks Greg

I go on with the layout

Franco

[DummyLoad]

.... the 5E3 style mixing network changes HF response at lower volumes.

Are you referring to the Tone Control as arranged on 5E3 ? (you know, sometime my translation isn't adequate to the require)

Many Thanks Pete

Franco

no, am referring to the way the volume controls are used as a mixing network- the resistive load to the triode/pentode plate changes as you sweep the wiper, therefore the -f3dB knee changes. e.g., the knee moves up as you decrease resistance (volume): bass is shaved.


--pete

[kagliostro]

Ciao Pete

OK, translation problems solved, but now begins my poor knowledge problems

This is the 5e3 Tone Control



that is the same on 5f2-a



but not in the 5f2



does the 5f2 arrangement solve the problem of shaved bass and lows the interactivity on controls ?

Thanks

Franco

EDIT: I've just read that 5F2 was build from 1955 to 1956 and 5F2-A from 1957 to 1960

I supposed that the 5F2 Tone Control was an improvement respect the 5F2-A control, but now I'm furtherly confused


       

[PRR]

> the knee moves up as you decrease resistance (volume): bass is shaved.

This is mitigated by the very large coupling cap used: 0.1uFd. If Vol is turned to 10%, the 1Meg acts like 100K, the -3dB is 20Hz, you got lots of bass still. The drawback is that a 0.1uFd cap is "much" more expensive than 0.05 or 0.02. Realizing this may be why Leo turned the Volume pot around: save pennies, and pennies matter in mass production.

[kagliostro]

Many Thanks PRR

Franco

[kagliostro]

Done a little work (not much)

Here the part of layout I've draw (miss all the PS section)



Here the schematic with revised input and PS with virtual CT and elevated DC for Heaters



and here the idea to add a DPDT switch to bypass V1b and so have only one tube before the Power Section (like in the GA40)

but I think anyway it must be revised because when V1b is out of the path his G1 didn't has a leak resistor





Please give your opinions on the whole thing

Thanks

Franco

[kagliostro]

Here is what I've done till now



The chassis is small so, after a long time spent to think what to do, I decided to build a second chassis (to be put on the bottom of the cabinet)

where to place the PS, but now I need help

How to arrange the ground connections and ..... Which will be the better way to connect the PS with the amp chassis ?

I've think to put the ON/OFF switch on the PS chassis as to avoid to have 220V AC across the cabinet, pilot lamp will be a LED feed with the 6.3v of the heaters and it will be on the Top chassis






On this link you can download the .vsd files (the first file is too big to be loaded on the forum)

http://www.mediafire.com/file/w6a398q78kwu0m0/Gibson_GA-5T_%2B_GA40_CH1_-_Chassis_Samick_Smart_2_.vsd

http://www.mediafire.com/file/1w18l1ezjsrg2e7/Gibson_GA-5T_GA40_CH1_-_Chassis_Samick_Smart_2_-_PS_.vsd


Thanks

Franco

p.s.: Layout must be revised

[DummyLoad]

Maybe this would work...

Use 2 x 1mm conductor with shield cable for mains AC from PS chassis to AMP chassis for power switch and use shield to ground both chassis together: Use a second 2 x 1mm conductor with shield for B+ and ground only the PS end of the shield: Twist the filament wires, then encase the whole works in heat-shrink tubing or nylon braid. See drawing page #3.

Use one end Octal Male plug with wire cover for chassis wire to Female Plug with cover for other end to AMP chassis for wire to wire,  Octal Male and Female Plug and Wire Cover like these...  So, Make Plug Socket on AMP chassis and Female Socket on PS chassis. 

Dropbox link to files here...

--Pete

[DummyLoad]

Source for Plugs and Covers.


http://www.acoustic-dimension.com/hardware/bayonet-connector-page.htm


--Pete

[kagliostro]

Ciao Pete

Many thanks

I'll see what I've in my stock, I'm sure I've something that is usable to follow your indications

--

Did you had problems with the file hosted at mediafire ? (please, let me know)

Have a nice day

Franco

[DummyLoad]

Ciao Pete

Many thanks

I'll see what I've in my stock, I'm sure I've something that is usable to follow your indications

--

Did you had problems with the file hosted at mediafire ? (please, let me know)

Have a nice day

Franco

Franco, you're most welcome. That is what came to mind. It may/may not work out. Hoping others with experience in building split chassis designs will chime in.


no issues with media-fire - they did want to feed me "malware" - had to close browser to clear the annoying pop-up.


--pete

[kagliostro]

OH, sorry, I've no problems with mediafire (     ) and I use it because they didn't cancel file after a certain time

This also give you problems ?

http://www.filedropper.com/gibsonga-5tga40ch1-chassissamicksmart2

http://www.filedropper.com/gibsonga-5tga40ch1-chassissamicksmart2-ps

--

I did a quick search and I finded something, but octal connectors and socket are too big to be positioned on the chassis (on front and rear plate there is enough space, but this will be not comfortable), so my choice (at the moment) go to some 9 pin items I've

I've a Male and one Female cable connector and to have 9 pin socket is simple, what I don't have is a Male panel connector, I must rearrange a Male cable connector for this purpose or, may be I can find something different going on with my search



Franco