Silvertone 1474........

[Jack_Hester]

I have a Silvertone 1474 that's supposed to be a '62.  The pots are CTS dated at '61.  One has 1376134, so that would make it a very early build for '62. 

Anyway, it's had much hands-on in it's life, plus the spring tank (if that's what they called it) is missing.  So, I'm thinking that after I clean up the wiring, and do some other housekeeping, I'll concentrate on getting the amp in good working order. 

Then, I want to see if I can configure a tube-driven substitute tank in place of what was originally used.  I have an Accutronices (Made in USA) 4FB3??? that may work.  Not sure if the 6CG7 is up to the task.  I have some Russian 6N2P tubes that are of the same pinout, but I believe that they have a mu of 100.  I wonder if they will work better. 

If not, maybe slip in an IRF820 to bump the gain up a bit, without having to add another tube. 

I'm nowhere near that point, but when the amp is restored to a good working condition, I want to explore this mod.  Is anyone aware of a transformer-less mod being done to one of these amps?  I've included a pre-amp drawing that an individual made of his own amp like this one.  Unlike the schematic most found, this one only has one Reverb pot, and two 6CG7's.  The commonly found schematic has two Reverb pots, and no 6CG7's. 

I have a 60's Reverberocket that has a transformer-less Reverb, but I haven't looked at the tank on it.  It's waiting it's turn to be put on the bench. 

I've also included some literature on spring tanks.

Please advise, and I will post my progress in bringing the amp back to life.  Thanks for all help and suggestions.  Have a good one. 

Jack

[Jack_Hester]

I stand corrected.  The other schematic has one 6CG7 being used as the PI.

Jack

[DummyLoad]

1474 schema here shows only one 6CG7 and that's the phase inverter. the reverb driver is 12AX7.

http://el34world.com/charts/Schematics/files/silvertone/silvertone1474.pdf

maybe the 6CG7 reverb driver/recovery in yours was a mod/revision?

in the 1485/1486 6FQ7/6CG7 is reverb driver - configured as a paraphase driver and another 6FQ7/6CG7 is final preamp and split-load phase inverter.     

--pete

[Jack_Hester]

1474 schema here shows only one 6CG7 and that's the phase inverter. the reverb driver is 12AX7.

http://el34world.com/charts/Schematics/files/silvertone/silvertone1474.pdf

maybe the 6CG7 reverb driver/recovery in yours was a mod/revision?

in the 1485/1486 6FQ7/6CG7 is reverb driver - configured as a paraphase driver and another 6FQ7/6CG7 is final preamp and split-load phase inverter.     

--pete

My amp only has one Reverb pot, like in the pre-amp schematic that I attached.  I don't remember where I found it, but I remember that fellow saying that he had a problem finding his configuration.  As his only has the one Reverb pot.  The wiring on this amp is very trashy.  Mainly I want to clean up the power cord connections, and the OT connections to the jacks.  Appears to be lamp cord on that one.  And, no care taken, as it was laying directly on tube sockets. 
 
If you will look on the schematic that you list, and see three 12AX7's next to a 6CG7, mine has 12AX7-6CG7-12AX7-6CG7, the last being the PI.  I made a quick comparison of my Reverb wiring and his drawing, and they appear to match.  A closer inspection, after I'm done tidying up the rest, will be in order to make a better verification.
 
Jack

[terminalgs]

here's an idea.  I was going to do this on 1484 that I thought had a failed reverb box, but it turned out to be something else,  so I never put the idea to actual work:

original reverb design:



modded reverb design to work with a "4F" tank:



change the 6CG7 to be a paralleled triode instead of a phase inverter. R40's 560K will need to be reduced (or eliminated).  R26's 560K values might need to be tweaked to get the wet/dry mix right.

[Jack_Hester]

Thanks for the drawings.  You plan would require an additional tube, which is why I was considering a MOSFET in the signal path between my driver and the tank.  Possibly as a CF, but I haven't given it any thought at this point.  Just don't want to add another tube.  Not much room to do so, as it's not a very well thought out design. 

But please, keep suggesting.  I'll come back and review all, when the amp is back up and running. 

Jack

[terminalgs]

Thanks for the drawings.  You plan would require an additional tube, which is why I was considering a MOSFET in the signal path between my driver and the tank.  Possibly as a CF, but I haven't given it any thought at this point.  Just don't want to add another tube.  Not much room to do so, as it's not a very well thought out design. 

But please, keep suggesting.  I'll come back and review all, when the amp is back up and running. 

no additional tubes needed.   All you do is take the 6CG7's two triodes that operate as a gain stage and para-phase inverter  and rewire the tube to be two parallel triodes.  you can even use the same value resistors and caps for the cathode and plates...   If you look at the schematics, one is the original, and the second shows the same tubes slightly modded.


I would suggest replacing the 1/2W R45 and R46's with 1W resistors.  Dano underspec'ed these as 1/2W, but they draw .6W ea. SO,, one or both resistors fail and the tank gets blamed and yanked out.

[Jack_Hester]

no additional tubes needed.   All you do is take the 6CG7's two triodes that operate as a gain stage and para-phase inverter  and rewire the tube to be two parallel triodes.  you can even use the same value resistors and caps for the cathode and plates...   If you look at the schematics, one is the original, and the second shows the same tubes slightly modded.

Looking at my schematic, my Reverb is driven by one 6CG7 triode, and recovery is the other half.  The second 6CG7 appears to be individual gain stages for each pair of power tubes.  One half of a 12AX7 appears to be a cathodyne PI feeding these triodes. 
 
I'm not seeing how to rework it to be like your second schematic.  But then, I can be a bit slow on such.  I'm at work, and will give it a closer look this evening if I'm not too late getting home.  But, thanks for staying on it.  I'm hoping to get to this soon. 
 
Got to go by Fasenal and get some washer-head screws to be rid of the crappy screws that was in it.  I'm sure the originals were not such, but washer-heads just look good.  Like someone cared to put it back better. 
 
Jack

[terminalgs]

My apologies Jack!!  You wrote 1474, and I read 1484....


ahem....  so.. anyway.

[Jack_Hester]

That's quite alright. 

Jack

[terminalgs]

You can drive a 4F tank with a single triode, but to make it a current driver,  the plate resistor needs to be really low, like 4K-22K.  this means you'll draw a lot more current through the triode.. right now that driver draws maybe 1-2ma?  in current driver format, it'll draw 10X that or more (so you'll need to decide if the Dano-spec'd PT has excess B+ capacity).  ampeg used 1/2 a 6CG7 to drive a 4F tank with the Gemini-I, Magnatone did it with the low-Mu side of a 6DR7 in the 480.


That Dano circuit steals the dry signal off the driver and then attenuates it down 75% before mixing it with the wet.  I'd skip all that and just use the input to the driver grid as the dry mix (like a fender, or that same ampeg gemini).

[DummyLoad]

use a 6DR7. power low-mu triode 4-10K 10W anode R with cap coupled to 600R tank driver, with the high-mu triode for recovery. better still, use 6DX8/6GW8/6BM8 with fender 6G15 reverb tranny and...it's reverb yamighty!


http://www.magnatoneamps.com/schematics/magnatone_480.pdf


--pete

[TIMBO]

 
Sorry guys, DL, are you saying that a 6G15 reverb tranny is a possible match to use with a 6BM8 driver into a 8 ohm tank.
These tranny specs says that they have a plate load of 7.5k. http://www.triodeelectronics.com/reverbbundle.html
From what I can find these tubes seem to range any where from 15k-25k. I looked up the tube data and I'm not sure what I am looking at, one bit says 20,000 ohms and the other is 7,000. 
 http://www.drtube.com/datasheets/6bm8-sed1996.pdf
 

The 6BM8 is part of the Fi Sonic circuit and I am having trouble finding a suitable transformer. Thanks

[jazbo8]

I looked up the tube data and I'm not sure what I am looking at, one bit says 20,000 ohms and the other is 7,000.

The replacement OPT for the 6G15 should work fine, there's not much a mis-match between 5.6k (typical Ra) and 7k (OPT).

[DummyLoad]

DL, are you saying that a 6G15 reverb tranny is a possible match to use with a 6BM8 driver into a 8 ohm tank.

yes. or find table-top junk tube phono player with SE output and cannibalize OT.

--pete

[Jack_Hester]

My intent is always to add with as little modification as possible.  That's why I was thinking an IRF820 that I can swing from the V2b tube socket, and drive the tank directly from it.  I went back and visited this article:
 
http://www.geofex.com/Article_Folders/mosfet_folly/mosfetfolly.htm
 
as a basis for my thoughts on this.  I'm sure that the little chip will put out some serious heat, but I can sink it and keep it insulated/isolated from the other components.  Maybe.  A very unkempt build, to say the least. 
 
Your thoughts, please.  I'll make a drawing, at some point, of how I think it should go.  Thanks.
 
Jack

[Jack_Hester]

Managed to get the line input wiring cleaned up way better than it was.  I now feel that part is safe.  Installed the 5U4G rectifier and began the slow process of bringing it up on the lamp limiter.  After that was taken care of, and full line voltage was on the PT, the unloaded B+ was measured at 520vDC.  Filament on all tubes sockets, but the rectifier (forgot to measure it) was 7vAC. 

The plate voltage (pin 3) on each of the 4 power tube sockets (no other tubes installed, but the rectifier) was 519.7vDC. 

Screen grid voltage (pin 4) on each was 519vDC. 

The control grid voltage (pin 5) was 12.5 to 13vDC, with the Tremolo Strength pot at minimum.  Around 55vDC with the pot at maximum.  I'm thinking that I have a leaking capacitor in the oscillator, .1uf cap connected to the plate of the 6AU6 tube socket. 

That's all for tonight.  Tomorrow, if I can, I will clean up the wiring from the OT's to the speaker jacks.  This is the stuff that looks like single conductor lamp cord. 

Then, I'm going to see if I can locate the suspicious .1uf in the oscillator. 

I may be at the point where I can install good tubes and connect 2 speakers.  If not, Friday evening.  At least by Saturday morning, I can tell what kind of amp I have, and if it requires more attention. 

Anyway, hopefully more progress by tomorrow.  Have a good one.

Jack

[Jack_Hester]

A little progress, since the last post.  I did get the internal speaker lamp cord replaced with a shielded cable, though I did not use the shield as such.  It's a 2/C with foil shield that has a bare third conductor that is used as the shield connection.  Used the bare conductor as the speaker common between the two jacks and the OT's.  Worked well to route the cable down against the chassis, and away from the other wiring.  Same for the power wiring.  Hugging the chassis, but it is shielded.

Removed the suspect .1uf cap that was connected to the plate of the 6AU6 oscillator.  Made a quick voltage check afterwards and the voltage disappeared from the wiper of the Tremolo Strength pot.  Soldered in a new .1uf-600v cap and all is well, there.

The tubes that came with the amp are such a hodgepodge mix that I had to replace all, but the 12AX7's.  Both of those checked 95% strength, and no other issues on my B&K 707.  Both 6CG7's were bad.  One filament never heated.  The other just barely made the needle move off of zero, on either triode.  Could not lay my hands on any old stock.  However, as good fortune would have it, I have 6N5P's.  The only difference that I can find in these and the 6CG7's is that pin 9 is used as an internal shield on the 6N5P, and has no connection on the 6CG7.

I checked and the 6CG7 (just before the power tubes) socket has no connection to pin 9, so I'm safe to use it there.  However, the Reverb socket uses pin 9 as a tie point, so I left that tube out for now, as I'm not troubleshooting Reverb.  At least, not until the rest of the amp is where I want to leave it. 

I used none of the 6L6/5881 tubes that came with it.  A mix of nothing that matches.  I put some 6L6 metal case RCA's in at first, but now have 6P3S-E's in.  More on those, in a minute.

So, with all tubes installed (minus the Reverb tube), I powered it up.  Note: A thermistor is connected between the 'hot' lead and the power switch.  The pilot lamp gave indication that this was working.  At some point, I'll take a voltage reading of the line-in, and the voltage after the thermistor.  Suffice to say, it's doing it's job. 

With all pots at minimum, and plate voltage on the power tubes right at 340vDC, I have bias currents anywhere from 49mA to 60mA.  Nothing consistent.  I did not record the readings with the 6L6 metal case tubes (need to do a better job of that).  But, I did not like the bias current on those, which is why I tried the Russian tubes.  These bias readings make me feel better, while I'm troubleshooting.  I can work on something better to get them closer when I'm finished finding any other issues that may be more important to the safe operation of the amp. 

With both Volume pots at minimum, the hum level is downright decent.  With the Standby on, it goes very quiet, subtle hum at the speakers. 

I have not identified the 12AX7 that is being used for the Driver/Cathodyne PI.  But, I want to maybe ground the grid of the driver, just for a test to see if the amp goes quiet, also.  If the Pre-amp schematic is correct, I should be able to ground the wiper of either Volume pot (check for voltage, first) and see what affect that has downstream. 

Anyway, I'm going to have to be very creative supporting this chassis (upside down), as there's not much support to keep it from twisting.  My wooden chassis cradle should do the trick, if I support the PT from underneath. 

More to come, later.  Have a good one.

Jack

[Jack_Hester]

the Reverb socket uses pin 9 as a tie point, so I left that tube out for now, as I'm not troubleshooting Reverb.  At least, not until the rest of the amp is where I want to leave it. 

I stand corrected.  Pin 9 tab is gone, and I made the mistake of counting backwards.  So, I'm safe to have both 6CG7's installed, no connection to either pin 9.

First, Channel 1 has a very loud hum that completely disappears when the input is grounded to the chassis.  Channel 2 has a loud hum, though not as much as 1, and too, disappears when that input is grounded.  This was a good time to test with a guitar.  Hum is not there with guitar connected, as I suspected to be.  I'm only running the volume on either channel at 09:00, as it is just way too loud, otherwise.  Plus, I'm standing too close, and the feedback starts just a little louder.  Will test it farther away from the amp, later on.  Lots of hiss, and background noise (not bad).  More like it's picking up external noise.  Could be the guitar.  Could be the poor layout (wires/components).  I'll work on that later, too.

Right now, I'm trying to get the Tremolo working.  No oscillation.  Replaced the .1uf, that leaves the plate and connects to the 330K in series with one leg of the Strength pot.  I also replaced the three .02uf caps with .022uf caps.  The drawing shows a .5uf cap between pin 6 and pin 7 (signal ground) of the 6AU6 tube.  That's missing and I don't have one. 

Anyway, that's where I am.  I'm using the schematic like Doug has in the library, for troubleshooting the Tremolo.  Mine appears to be like that one.  Suggestions welcome.  I'll post my continued findings.  Have a good one.

Jack

[sluckey]

First, Channel 1 has a very loud hum that completely disappears when the input is grounded to the chassis.  Channel 2 has a loud hum, though not as much as 1, and too, disappears when that input is grounded.  This was a good time to test with a guitar.  Hum is not there with guitar connected, as I suspected to be.

Schematic shows non switching input jacks. If you will replace all of them with Switchcraft 12As and wire them for a standard Fender hi-lo pair, that problem will go away.

Right now, I'm trying to get the Tremolo working.  No oscillation.  Replaced the .1uf, that leaves the plate and connects to the 330K in series with one leg of the Strength pot.  I also replaced the three .02uf caps with .022uf caps.  The drawing shows a .5uf cap between pin 6 and pin 7 (signal ground) of the 6AU6 tube.  That's missing and I don't have one.

A lot of pentodes use a .1µF in that position. I would too. Check the voltage on pin 6. It should be about 167.5vdc. If not close, check the two 330K voltage divider resistors connected to pin 6. Check the plate voltage and plate resistor also.

Could also be a weak tube.

[Jack_Hester]

Schematic shows non switching input jacks. If you will replace all of them with Switchcraft 12As and wire them for a standard Fender hi-lo pair, that problem will go away.

Got plenty of those.  Will save for last.

A lot of pentodes use a .1µF in that position. I would too. Check the voltage on pin 6. It should be about 167.5vdc. If not close, check the two 330K voltage divider resistors connected to pin 6. Check the plate voltage and plate resistor also.

Installed a .1uf across pin 6 and pin 7.  Got the oscillation.  If I were to guess (DMM doesn't make that easy, and I don't have my VTVM handy), I'd say the mid-point of the oscillation is about that voltage. 

Could also be a weak tube.

Swapped it out with a known good one.  Oscillation voltage swing is about the same.  So, I'd say the tube is ok.

With the Tremolo Strength at maximum, it's not very intense.  I'm thinking about reducing the size of the 330K resistor, in series with the Strength pot.  I can solder a 1M pot wiper to one side of the resistor, and a pot outside leg to the other end.  This would be a test, to determine a suitable replacement size.  What do you think? 

Jack

[sluckey]

I can solder a 1M pot wiper to one side of the resistor, and a pot outside leg to the other end.  What do you think? 

Just put a gator clip lead jumper across the resistor and you'll be able to get every ounce of trem from the oscillator.

[Jack_Hester]

Just put a gator clip lead jumper across the resistor and you'll be able to get every ounce of trem from the oscillator.

Jumpered the resistor and the Tremolo is now stronger.  Not as much as some of the amps that have, but this one is ok.  Soldered in a permanent jumper across it.  Thanks for the suggestion. 

So now, I think I'm ready to take a look at a fix for the Reverb.  I have a couple spring tanks with different impedances (Input/Output) that I can make a quick test of the recovery side of the 6CG7/6N5P.  This will be a new experience, somewhat designing a tube driven Reverb. 

I'll do some tweaks to the remainder of the amp, to see if I can eliminate some of the hiss.  I'll look into a Fender Hi/Lo configuration for the input jacks.  I'll report my progress.  More to come.  Have a good one.

Jack

[Jack_Hester]

If you will replace all of them with Switchcraft 12As and wire them for a standard Fender hi-lo pair, that problem will go away.

Decided to use up some of my Radio Shack closed circuit jacks.  Appear to be on equal quality with the ones that I removed.  Those were not what I would consider quality.

Anyway, pulled up the Bassman 5F6-A schematic, and wired like that one.  The original inputs were already using 68K's with a 1M.  So, I re-used those in the Fender configuration.  The amp is now quiet when nothing is connected.  And, each channel has a Hi/Lo input.  One thing that I noticed is that the input gain is a bit attenuated, but is definitely not a problem.  The volume feels more normal, now. 

That's all for tonight, as I got called to go into work at 23:00 hrs., and stay until 07:00 hrs. tomorrow.  So, got to get a nap.  I'll see what I can do tomorrow afternoon.  Have a good one.

Jack

[Jack_Hester]

When I say the amp is now quiet, with nothing connected, the fairly obvious hiss is there.  I'm going to have to address that, for sure.  The better the amp sounds, the more annoying the hiss becomes.  Just swapped out the Driver/PI 6CG7/6N5P with a 6N2P.  This tube has a mu of 100.  Just tried it for a bit, as I don't have much time before leaving for work.  Amp was super loud, again, with this in place.  The Tremolo oscillation is much stronger, but I swapped back to the lesser mu tube for now.  I need to look into stopping the oscillation, rather than grounding the wiper of the Tremolo Strength pot.  My Ampeg B-12-XY grounds the Speed.  I'm going to need to do this to take voltage readings on the oscillator, anyway.  I may just hang a temporary ground on the Speed, to take readings. 

Anyway, tomorrow afternoon will be a better time to think about it.  Have a good one.

Jack

[sluckey]

I need to look into stopping the oscillation, rather than grounding the wiper of the Tremolo Strength pot.

Why? I prefer to leave the oscillator running and kill the trem signal at the strength pot. This way when you step on the switch you will have instant tremolo. Killing the oscillator requires you to wait for it to start up after stepping on the footswitch. Some oscillators take a long time to start up and can be annoying. Yours may start quicker.

[Jack_Hester]

Good point.  I'll just hang a temporary ground on the Speed, for taking readings.  Then, remove for normal use. 
 
So, the next two fixes are Reverb and annoying hiss.  Thanks for your input.  Troubleshooting is fun.  Help with troubleshooting is 'funner'.  Have a good one. 
 
Jack

[Jack_Hester]

Ok.  I spent the afternoon thinking about how I wanted to make this tube driven Reverb work, with a conventional spring tank, and with as little mods as possible.  In order to do that, I needed to know exactly what I had.  So, I began the tedious process of drawing my amp. 

As per Sluckey's recommendation, I modified the inputs to match that of a Fender Hi/Lo.  Eliminated big time noise on the front end.  The resistors were already what I needed.  I swapped out the open circuit jacks with closed circuit jacks, and followed the Fender 5F6-A example. 

Next, I began tracing the circuit through the V1 (12AX7), and on through the Reverb circuit V2 (6CG7), which is all that I have drawn.  I stopped after the Reverb pot, which goes to the Channel 2 Treble circuit.  I will draw that next, but for now I'm posting the circuit that I want to work on.  The rest will come as I am able to decipher this very complicated layout. 

I have been studying this page:

http://www.dougcircuits.com/reverbtanks.html

to try and figure how I might use my spring tank that has a 1475 ohm input, and a 2250 ohm output.  My 6CG7 has a plate resistance of 6700-7700 ohms. 

Anyway, it's been a tiring night at work, and a not-so-restful morning.  I'm done thinking for tonight, and will pick back up on it tomorrow evening.  Have a good one. 

Jack

Note: See Reply #29 for current mods

[Jack_Hester]

This is a quickly drawn concept of how I would use an IRF820 as a Source Follower in the Driver circuit.  I'm on new ground, here.  So, the Source resistor value is only a guess. 

Please comment and make corrections.  I have just enough open space to attach a MOSFET and heat sink (insulated/isolated from the chassis).  Thanks for any and all help.  Have a good one.

Jack

Note: See Reply #29 for current mods

[Jack_Hester]

Good cold morning (19 DegF), and a good time to work on the Reverb. 

Made some mods and revised the proposed schematic for it.  First mod was to put a 2.2K resistor across the 1M at V2 - pin 2.  Powered it up.  Warmup is a bit slow, while the two 12AX7's are heating.  With the spring tank connected (2250 ohms output impedance of tank), I turned the Reverb pot to maximum and tapped the tank.  Man, what a difference.  Loud response.  I continued to tap, as I slowly rolled the Reverb pot to minimum.  Echo decreased, until none was apparent.  The volume for Channel 2 was set at max, while testing.  I'm satisfied with that simple mod.

Next, (power disconnected, and caps bled) I soldered in the IRF820, following my drawing.  Power on and amp warmed.  Immediately, I get a slow thumping that doesn't change with any volume setting.  Power off and Drain leg lifted from the B+.  Power on and thumping is gone.  I use my pliers (rubber glove on, as well) and touch the Drain leg to the B+.  Thumping came back. 

I am going to need some help with this mod.  As stated in the last post, I'm on new ground doing this.  Do I need to add a Drain resistor, to tame that voltage a bit?  And, one on the Gate, for the same reason?  Please advise.  Real happy with the little Recovery mod.  Thanks.

Jack

Note: Reply #38 has the complete schematic, plus the Reverb circuit mod that I hope to implement. - 16Nov14

[Jack_Hester]

At this point, I'm stumped.  Haven't been able to figure a way to drive my spring tank with the 6CG7 triode half.  I'm leaving the 2.2K resistor across the 1M on V2 - pin2.  The Reverb Recovery is strong, with my 2250 ohm spring tank output. 

And, the jumper around the 330K resistor in series with the 1M Tremolo Strength pot.  Other than those changes, the amp is stock.  I've completed drawing everything, but the power supply.  I'll post the whole drawing when that's done.  Maybe after Church, tomorrow. 

Jack

[PRR]

1.5K is *way* too small. Use Ohms Law. It is sucking more current than your power bottles!!

Also.... you have three cascoded stages on one B+ node. Move the Drain to another B+ node.

[DummyLoad]

try 10K ohm 5W under MOSFET.


--pete

[Jack_Hester]

1.5K is *way* too small. Use Ohms Law. It is sucking more current than your power bottles!!

Also.... you have three cascoded stages on one B+ node. Move the Drain to another B+ node.

Ok.  Will get the Drain moved to another node.  I haven't verified/drawn my power supply, but will probably do that this afternoon, after Church.  Then, I'll use another point closer to the PI, or closer to the source.

try 10K ohm 5W under MOSFET.

Thanks for the suggestion of value.  I've got some of those.

Thanks to both of you for your input.  I can make the mods in the early morning, if I don't get to it today.  Rather than swing these like the p2p of the rest, I'll see if I can put it all on a small board, for ease in mounting, when it's proven. 

By the way, do you think this MOSFET will need sinking?  I've got a small sink, but not sure how I'll insulate it from the chassis.  If I get creative (and have it all on a board), I may be able to let it swing in the opening that I think I have under the chassis.  It's very light, and the MOSFET will be fairly rigid. 

I'll also post the completed drawing with the modified Reverb circuit, hopefully later this evening.  More to come.  Have a good one.

Jack

[terminalgs]

At this point, I'm stumped.  Haven't been able to figure a way to drive my spring tank with the 6CG7 triode half.

Not to derail your mosfet effort, but what do you  think of Ampeg's 1/2 6GC7 4F tank driver?

[Jack_Hester]

I prefer to do it in tube, and that would be preferable for me.  However, I just don't have the space to add another socket.  That's why I'm going the MOSFET route.  But, thanks for bringing it up. 

A friend at work, who has completed his third tube build, is planning to add Reverb to his next build.  He has only used a non-spring tank Reverb.  And, when I carried a tank to work for him to see, and explained how it worked, he was exited about including it in his next. 

Anyway, I plan to show him the variety of ways to do such, and will include this one in what I show.  Thanks, again.

Jack

[terminalgs]

I prefer to do it in tube, and that would be preferable for me.  However, I just don't have the space to add another socket.  That's why I'm going the MOSFET route.  But, thanks for bringing it up. 

dont' add a tube.  Use you  current tank's "driver" a driver for the 4F box:





the blue box with a "D" is your driver.   Convert to the ampeg style, with a 10K 2W plate, and get fid of the series 27K resistor, coupling the tank directly to the .01 coupling cap of "D". (maybe up it .1 or .4 after seeing it works).




Right now, your "dry" from ch.2 comes off the plate of that same triode, however, you can  instead connect the dry directly from the vol. pot to the "mix bus" (orange arrow),  just like ch.1.  Although the dry comes off the plate of triode "D",  which is a gain stage (12ax7, 100K plate), that signal is attenuated  with the 560K/100K 'bal' pot, and then it has the larger mix resistor (red arrow) of 560K..,  so the dry, while 'gained up' by "D", gets busted down, maybe an equal about by those voltage dividers.


if you do that (dry connected to mix with a 330K like ch.1), that frees up that "D" triode to be converted to a current driver, which is what the 4F tank needs.  you still need an input to "D"'s grid, and I'd couple it to the plate of the 1st gain stage (blue arrow) with a .1uf, and add a grid leak of 1M or so.


What do you think?

[Jack_Hester]

I haven't finished with my drawing, but I have enough completed for you to take a look.  Zoom in on the Reverb of my configuration.  It doesn't match that schematic.  I only have a Reverb 'mix', and not the configuration of your schematic.  However, you may be able to describe what you mean, from mine.  I'm a bit tired to follow your description, tonight.  I'll give it a close look in the morning.  But, thanks for what you have presented.  Mods on this amp are messy, at best.  The MOSFET idea is the least messy.  My first driver test has been removed, and amp is back stock (except for Recovery circuit).  I'll proceed from there. 

Jack

[Jack_Hester]

Finally finished the complete drawing.  Components are labeled.

The power supply was a really big challenge, as the components are scattered all over the amp.  But, it turned out to be a very simple one. 

One thing I will do is change out the 1.5K PS resistor that I have labeled R67.  The original schematic calls for a 1 watt, and it is.  And, cooked.  It measures about 900R, and I just can see the color bands.  I don't have any larger than 1 watt, so I'll see if the local Radio Shack might.  Otherwise, I'll have some on the way, mail order.

I apologize for posting it so often.  Hopefully, only one more drawing post, to show a functional Reverb.  Give a look.  Have a good one. 

Jack

Note: See Reply #69 for complete schematic with modifications to the Reverb circuit.

[jazbo8]

Something seems odd about the reverb circuit as drawn, R25b is shunting the reverb signal nearly to ground (why is it there?) and the coupling cap from the reverb tank (0.0022u in the original schematic) to the input of the recovery stage (pin 2) is missing altogether.

[Jack_Hester]

Something seems odd about the reverb circuit as drawn, R25b is shunting the reverb signal nearly to ground (why is it there?) and the coupling cap from the reverb tank (0.0022u in the original schematic) to the input of the recovery stage (pin 2) is missing altogether.

R25b was an experiment that I tried, to match the impedance of the tank output (2250 ohms) to the Recovery stage grid that it was connected to.  Before, you could just barely hear the echo.  Everything had to be very quiet, Ch2 Vol at full, and the Reverb at full.  So, I clipped and looped the leads on a 2.2K resistor, soldered it across the 1M resistor, and performed the test again.  Woke it right up.  This will be a very good Reverb, when I am able to get the Driver output resolved.

As to the .0022uf cap, I see two of them in the original.  There's a 100K in parallel with the 100K Reverb pot, a .0022uf between it and the tank, and another .0022uf between it and the pot.  My Reverb pot is 1M, and is not connected to the grid of the Recovery stage, like in the original. 

My amp has a single wire connecting V2-pin 2 to the RCA Recovery jack center.  Originally there was (and still is) a 1M resistor from this jack center connection, to signal ground.  Now, my 2.2K is paralleled with this 1M.  Other than that, my drawing reflects the as-built amp. 

But, thanks for pointing out the differences.  I am curious as to why a cap (.0022uf or otherwise) is needed.  Is it there for frequency filtering? 

Jack

[jazbo8]

Here is the quote from the Prange article you linked to earlier:


"- The recovery circuit’s input impedance should be high enough to prevent roll-off due to the output coil’s inductive reactance. For example, the 10,000R at 1kHz coil appears as a 60,000R source at 6kHz.
- Use an output high-pass filter with a 50 Hz to 100 Hz cutoff to lessen the effects of rumble when the unit is mechanically shocked."

The 0.0022uF coupling capacitor (along with the 100k grid resisstor shown in the original schematic) were there to filter out the low frequency rumble transmitted through the reverb tank. I am surprised that the 2.2k grid load resistor "woke the reverb up", since it's a very heavy load for the reverb coil's output.

[sluckey]

R25b was an experiment that I tried, to match the impedance of the tank output (2250 ohms) to the Recovery stage grid that it was connected to.  Before, you could just barely hear the echo.  Everything had to be very quiet, Ch2 Vol at full, and the Reverb at full.  So, I clipped and looped the leads on a 2.2K resistor, soldered it across the 1M resistor, and performed the test again.  Woke it right up.  This will be a very good Reverb, when I am able to get the Driver output resolved.

Matching impedance is important when you want maximum power transfer. But that's not the case here. The tank puts out a very small amplitude signal similar to what come out of your guitar. Putting a 2.2K resistor at the output of that tank will decrease the amplitude of that signal due to loading. It's fine for a low Z output device to connect into a high Z grid when you want maximum voltage transfer.

[Jack_Hester]

Jazbo8 & Sluckey -

I removed the 2.2K resistor (R25b) and warmed up the amp.  Recovery works just fine.  Plenty loud.  My ears cannot tell the difference, with or without.  I have no idea why the Recovery was so weak, before.  The only difference is that I had to warm up two joints on each end of the 1M resistor to solder in the 2.2K.  Other than that, I made no other changes.  Maybe it was a cold solder that I corrected.  I have no idea.  But, the additional resistor is out and the schematic corrected to reflect its stock configuration. 

As to the rumble from mechanical shock, I can live with it, for now. 

On a lesser note, the amp did not like the MOSFET mod to the Driver circuit.  Much loud popping and static noise.  Even though this MOSFET is rated for 500v, I'm in the 350v range at B+2 on the Drain, and under 200v at V2-pin 6 for the Gate.  The Source had just a little over 100v.  I moved the Drain to B+3 and still the same popping and static.  Lifted the Gate and everything returned to norm. 

I'm thinking that I'll go dig out my little Radio Shack breadboard and wire some of this up.  A lot easier than soldering/desoldering. 

I'm open for suggestions on what to do next.  Thanks for your observations and help. 

Jack

[jazbo8]

So it turns out that this version of the 1474 does not have a "real" reverb tank afterall - instead it has a piezo element which does not sound too good according to many (and probably the reason that it did not come with the CR filter for the recovery circuit)... and one of the popular mods is to replace it with a proper reverb transformer and regular sping tank (expensive), or simply use the reverb I/O as your Fx I/O (~ free)...

[Jack_Hester]

So it turns out that this version of the 1474 does not have a "real" reverb tank afterall - instead it has a piezo element which does not sound too good according to many (and probably the reason that it did not come with the CR filter for the recovery circuit)... and one of the popular mods is to replace it with a proper reverb transformer and regular sping tank (expensive), or simply use the reverb I/O as your Fx I/O (~ free)...

Yep, that's what I got.  I'm satisfied that it had a piezoelectric type tank, but no tank was with this amp. 

I have my solid state breadboard handy, so I'm going to mock up the source follower circuit in my drawing.  And, see if it still gives the same problem as the one that I wired in place.  Just to be sure that I didn't have something crossed.  I'll work on that this evening, and report my findings. 

Jack

[Jack_Hester]

Ok.  Results not quite so noisy.  Driver still not working, but I did get some voltages measured.  Everything mounted on the Radio Shack breadboard, and alligator jumpers from there to the amp. 

Source: 130vDC
Gate:    130vDC
Drain:   263vDC (B+3)

No DC on the other side of the .01uf coupling cap. 

Back to studying. 

Jack

[Jack_Hester]

While I'm scratching my head on this Reverb dilemma, I have another question.  Not a problem.  Rather an understanding.

The two 12AX7 filaments are connected in series, with the open leg of V1 connected to signal ground, and the open leg of V3 is connected to the junction of a 39R-5W (R56) and a 220R-5W (R57).  The 220R is connected to signal ground. 

After voltages have had a few minutes to stabilize, I measure 21-22vDC at this junction. 

My speculation is that this DC filament supply is for noise reduction.  But, I'm curious as to the idea behind this design.  Makes for a very slow warmup time, but seems to work very well.

Jack

[sluckey]

You get the low noise benefits of dc heaters for the 12AX7s and at the same time you get a cathode resistor for all those 6L6s.

[terminalgs]

My speculation is that this DC filament supply is for noise reduction.  But, I'm curious as to the idea behind this design.  Makes for a very slow warmup time, but seems to work very well.

I believe it was old radio-days trick.  I recently saw a 1950 Magnatone with all the 12AX7 filaments in series underneath the cathode R of the power tubes.  slow warm up + DC heaters. 


Knowing Silvertone  and Danelectro, the PT's might not have have enough filament amperage to run all the tubes off the 6.3v.