Re-purposing 60s mic/audio tube amp for guitar

[Jennings]

Just picked up this old amp with the intention of overhauling, removing the superfluous elements, and modding for guitar use.  EF86 based preamp sections for each input, ECC82 PI, two EL84 for output and an EZ81 rectifier.  Excuse the dumb questions though...but is that a line only output transformer?  And should I just remove that and replace with one of the spare dedicated guitar amp OTs I have laying around from AC15 and 18w builds?

[PRR]

> is that a line only output transformer?

I have doubts about that schematic being your exact model. Two mike transformers?

However if the OT is right, half of the 50VCT winding, 25V, is 19.5 Watts in 32 Ohms and not the worst fit around. But who has 32r speaker cabinets?

[shooter]

 
a 4X12 8ohm gets you there

[Jennings]

  I might have 32 ohm headphones somewhere!


I’ll look to switch out the OT then rather than work with it. That schematic glued on the chassis cover seems to be a generic base design…they appear to start with something like that, then have added extra sections for extra inputs, and a bunch of additional power supply items to feed other equipment (presumably plugged in at the rear)…I’ll aim to remove that too. One of the EF86 is even labelled “optional” on the chassis. Probably that additional mic pre. I haven’t traced it out yet.


What would people’s suggestions be for preamp designs? I was thinking AC15 normal channel perhaps. Would anyone leave one of those mic transformer input sections as-is? Or what would people do with the extra two EF86?

[sluckey]

I like the sound of the AC-15 preamp. Lots of others like it as well. You have a perfect platform for a Dual Lite if that sounds interesting to you.

[Jennings]

Yeah great plan! It was the EF86 and EL84 combo that attracted me to the amp, with something like that in mind. My main gigging and recording amp for the last few years is a Dual Lite based on your design, and my last build married those preamps sections to a dual cathode biased KT66 output section. I tend not to use the 18w / Marshall side though, so I’m debating whether to do something else with the spare EF86/sockets on this one.

[Jennings]

I've re-named the thread as I start to plan the refurb and mods to the Type CS2414C (possibly GEC manufactured?) mic/audio valve amplifier.  Hope no one minds.


I drew a blank on finding a suitable 32 Ohm speaker just for the sake of keeping the original OT.  So as discussed, I'll switch out for an AC15 or 18w style with 4, 8 and 16 ohm taps, and use with the cabs I have already.  I've visually traced the schematic against the layout in my amp, and it seems to match, except I have the optional "Mic 2" / "See Note 3" on my amp...basically the additional components look to be a duplication of the "Mic 1" circuit, complete with mic transformer.


Given the tube lineup and the fact I love the AC15 Lite circuit, and my amp has the "optional" (as it's labelled on the chassis) second EF86 based mic input, I'll remove the Mic 2 input transformer and use the tube to build the AC15 Lite preamp with tone selector switch.  As I'll be removing the "power supply" socket I'm debating whether to mount an impedance selector on the chassis in the spare hole from the octal socket...might just omit that and put separate jacks on the rear though for ease of use.  As a mic input the "V7" "Optional" EF86 tube was mixed prior to the second EF86 gain stage, but I plan to move that (AC15 style) to the other side of the PI (annotated as point "A").  I'll take the HT from prior to the R22 15K dropper (annotated "HT"). The "Mic 2" volume pot will be switch to a 500K, and I'll add two more holes next to it on the front panel for the tone selector and Cut control.


I decided that since I've got a couple of Dual Lites currently, and I don't use much of the 18w channel, it might be interesting to keep the "Mic 1" and "Pickup" input un-changed for now.  I was thinking maybe just to see what they're like...perhaps as an amp with guitar / harp options.  Or do folks think that's not a great idea?


I'm going to use one single 1/4" jack input, but a TRS socket...which I'll link to a 2P3T switch so that I can toggle between mono jack cables for guitar and the "Pickup" channel (connection annotated "Ptip"), and a TRS cable for mics (connections annotated "M+" and "M-").  I'll likely keep the single input jack on the rear, removing the terminal block strip.  Does this all sound like a sensible approach?


I've annotated some red circles/ovals and shown the changes I'd need to make to the PI and output section to get it closer to the AC15 spec...should I do this, or leave them as-original to start with and see what the result is first? 


Can anyone see any issue or corrections in what I've put together?  I apologise for quality of the diagram...I haven't had a chance to draw a full fresh one while I was sketching out the ideas.

[Jennings]

Project update…I’ve completed the first phase. Namely reworking the V7 “optional” EF86 to a Steve Luckey style AC15 setup with rotary tone selector. I omitted the bright cap at the volume control. Also I routed the reworked channel into one side of the PI. OT changed to a Danbury unit with 8 and 16 Ohm speaker outs (1/4” jacks fitted to the rear). Input for all three channels via a single TRS 1/4” jack, with 2P6T rotary switch used to route appropriately to the preamp sections. Mains input change to an IEC socket…and I found the amp was originally wired with fused neutral rather than live, so I corrected that at the same time and earthed there chassis. PI tube switched to ECC83. All electrolytics changed for new ones.


All Works nicely on power up. AC15 channel sounds great, and the pickup channel sounds good too. Not tested the mic channel yet. A fair bit of hiss still though, so some work to do there. I’ve not replaced all the plate resistors and there are multiple chassis grounds etc which could be contributing to that. I also need to go through and confirm voltages and bias etc.


I’ve drawn up a new schematic of the amp as it currently stands…corrections and suggestions gratefully received as usual 👍

[sluckey]

V2 plate... consider replacing the two .005 caps with a single cap.

PS... need a dropping resistor between node B and node E.

[DummyLoad]

What's the issue with 2x 12" 16ohm speakers in series? Pair of alnico would probably sound really sweet.

[DummyLoad]

V1 - 22k R for cathode R, is that a typo? 2.2k?


--Pete

[Jennings]

V2 plate... consider replacing the two .005 caps with a single cap.

PS... need a dropping resistor between node B and node E.

Thanks Steve...much appreciated.  The double .005 caps was a typo (there's just one in reality), so I've updated the schematic.  When I checked the HT nodes I noticed I'd also mistakenly taken node E after B on my diagram when as=built it's after that subsequent 15K dropper...but you're still right too, as I realised that I still need to add a dropper to separate C and E at that point.  Updated on the schematic, and I'll update the amp shortly.

What's the issue with 2x 12" 16ohm speakers in series? Pair of alnico would probably sound really sweet.

Totally agree Pete...there's a bit of a story there.  I did consider a 2x12" or 2x10" with a nice pair of 16 Ohm alnico Celestions or Jensens.  But I actually don't own any dual speaker cabinets currently...I've only got 1x12" cabs these days.  So I'll need to find some space and make up a cab just for this head.  I noticed that Jensen offer the P12Q in 32 Ohms to dedicate one of my 1x12" to it, but couldn't find one anywhere.  So one of those options is still on the cards longer term, but to get the amp running I carefully removed and kept the original OT (all leads etc intact) and popped in that spare Danbury I had...I used existing chassis holes too as it could be temporary.

V1 - 22k R for cathode R, is that a typo? 2.2k?

Great spot, cheers...another typo.  Corrected now on the schematic.


Thanks for your help and advice...much appreciated, as ever.

[DummyLoad]

Tubes and more has the Jensen P12Q in 32Ω in stock. I'm assuming that your in the USA, please forgive assumption if otherwise.   

https://www.tubesandmore.com/products/speaker-jensen-vintage-alnico-12-p12q-40w

[Jennings]

I’m in the UK, but I’m half tempted to order anyway, or perhaps go for a custom Weber or similar. 32 Ohm speakers (for guitar) are like hens teeth here.

[Williamblake]

Give over. Nobody wants a one trick pony.

[Jennings]

Give over. Nobody wants a one trick pony.

😂 Very true…I could amass an army of unloved quirky line output amps to use with it? Or perhaps not 🤣 Either way, I need to find a use for that mic input. I don’t even play harmonica really 🤷🏻‍♂️

[Williamblake]

Reverb it is then?

[Jennings]

Reverb it is then?

That’s tempting actually…something different, and a tube reverb would be cool. None of my amps have reverb at all. Although I might be stuffed for space in that tiny metal head…plug in spring unit?!

[Jennings]

Last night I started doing initial voltage checks...Looks like I've got some adjustments to make.  HT Power supply voltage at node B is far too high, so clearly everything after that will be off anyway.  Thus all the downstream HT nodes are too high before we start, and plates and screens of the output section are way over limits etc.  I was a little surprised, as that's not a part of the circuit I've made any adjustments to, but it just goes to show what the original Mullard EL84 were putting up with!


I'll have to look to put an extra node in after the choke and before node B before I continue...also I noted that I didn't get the expected voltage drop at node C, so that's a point to check also.  What do people suggest as a resistor and cap value for my new node between A and B?  I was thinking another 16uF cap.  I'm estimating the current draw for three EF86, one ECC83 and two EL84 to be 12mA + 1.2mA + 55mA = 123.2mA, and if I'm looking to drop node B to ballpark 312V an online calculator makes that 500r and at least 7.7W...does that seem about right in practice?

[sluckey]

Are you using the 245VAC primary tap? If not, then connect mains to the 245VAC tap and see how much your DC voltages drop.

[Jennings]

Are you using the 245VAC primary tap? If not, then connect mains to the 245VAC tap and see how much your DC voltages drop.

Great call, thanks...I'm on the 230VAC tap, so I'll try that before anything else and report back.

[Jennings]

I did a wall voltage check...238VAC today.  And I moved the voltage selector from 220/230V to 240/250V accordingly.  Then retested.  The HT is still perhaps a bit high at node B, but nowhere near as much.  Plate and screen voltage for the EL84 is a good 38V higher than I'd like...


...would the best solution be to adjust the screen resistor values and add some plate resistors?  Or just go for the additional HT node like I first thought?


Ignore my current output bias checks...I just did that out of current curiosity, and note that I'm currently using 330r cathode resistors rather than 150r.  I'll check and adjust bias properly once I'm sorted with the HT line.

[pdf64]

Yes, the 225V tap was for mainland Europe, our typical mains is 240. The 230V harmonised euro norm is just a bodge to bridge the 2.
Never just assume the 6.3V heater voltage is fine, it needs measuring along with everything else, especially on a voltage chart.

[Jennings]

That makes sense...I do find a lot of imported gear seems to really be geared up for 220/230VAC, so 230VAC max and not our 230VAC plus some...and runs better if I use a 220V step down transformer.  A quick check of the filaments at the PT filament taps shows 3.17VAC on one side and 2.92VAC on the other.

[sluckey]

I'm currently using 330r cathode resistors rather than 150r.

Since you are using two separate cathode resistors you should probably stay with 330Ω. Changing to 150Ω will likely run the tubes really hot. Maybe too hot.

[Jennings]

I'm currently using 330r cathode resistors rather than 150r.

Since you are using two separate cathode resistors you should probably stay with 330Ω. Changing to 150Ω will likely run the tubes really hot. Maybe too hot.

Thanks, I will do.  I've got a dual 16uF can capacitor that might be convenient for me to chassis mount (there is a space) alongside the 16+32uF cap that's already there and I can make an extra HT dropping node with a suitable resistor...probably after what is currently node C, as node C is already at a suitable voltage for feeding the OT, and EL84 plates and screens.  I can then use my new node to drop what are currently nodes D and E to suitable levels.

[pdf64]

Here’s the Philips EL84 info https://tubedata.altanatubes.com.br/sheets/010/e/EL84.pdf

The suggested class AB cathode biased conditions on p4 may be interesting.

However, I’m unsure whether the 300V anode and screen voltages are with respect to the cathode, or to 0V common. As my understanding is that all voltages on such info is with respect to the cathode.
Does anyone know for sure?

[Jennings]

Here’s the Philips EL84 info https://tubedata.altanatubes.com.br/sheets/010/e/EL84.pdf

The suggested class AB cathode biased conditions on p4 may be interesting.

However, I’m unsure whether the 300V anode and screen voltages are with respect to the cathode, or to 0V common. As my understanding is that all voltages on such info is with respect to the cathode.
Does anyone know for sure?

Cheers…great datasheet 👍

[Jennings]

This is a good learning opportunity for me…first time I’ve had the situation of such high B+ Right at the front of the DC feed.


So, given the various options I might have to address it, what would be the learned experienced recommendations be? I’m conscious that most designs aim to keep the resistance as low as possible in the HT line before the power section feed. Would anyone go for the extra node I thought about and take the OT and power section feed from what is currently labelled node C on my schematic? Or would folks add some resistance one the EZ81 rectifier output? Or perhaps on the AC feed side of the rectifier? Or maybe leave the HT line as it is and either add some resistance to the OT centre tap DC feed, or before the EL84 plates? I’m confused as to which way to go for the “best”…or am I overthinking it?

[pdf64]

As the heater voltage is ok, another option is a beefy zener, either on the rectified high voltage, or between the winding CT and 0V common.

[Jennings]

As the heater voltage is ok, another option is a beefy zener, either on the rectified high voltage, or between the winding CT and 0V common.

I like that idea…fewer component changes too and it would dispense with the additional HT node idea. Is there any advantage of one location over the other, or does it not really make much odds performance wise? If I want to drop around 38V, would I need some kind of variable resistor to set the amount?

[pdf64]

There’s no adjustment possible, just have to source a suitable zener.
It’s makes life a bit easier to fit them in series with the HT winding CT. Much less voltage stress / high voltage terminals.

Another alternative is a power MOSFET to drop some voltage. I think that would have to be on the high voltage side.

[Jennings]

I think I’m sold on the zener approach. What do I need to look for when hunting for a suitable spec…for example, an HER508 seems to have a beefy 600V 5A rating.


One thing I just realised, and checked on the original amp’s schematic, is that the amp came with a bypassable 120k high wattage resistor to chassis ground on the HT CT. I say bypassable, as the HT CT was also connected to two jumperable “link” terminals on the rear of the unit to bypass the resistor and go straight to chassis ground. I removed that and hard wired to ground, mostly for safety and because I’m not used to seeing that setup. Now I realise it was intended as a sort of power boost/cut approach.


Would the diode approach be considered the better way to go, or would you try reinstating the resistor setup?

[PRR]

... ... ... ... node C is already at a suitable voltage for feeding the OT, and EL84 plates and screens.  I can then use my new node to drop what are currently nodes D and E to suitable levels.

Why are you looking at Zeners? BTW, HER508 is not a Zener.

Preamp -can- run from 300V... why not? Fender went near 400V. Or, since preamp current is small, a fistful of 10k 2 Watt resistors will both drop and filter.

It all worked for GEC. While you have changed the source, that's a small detail.

[Jennings]

Thanks 👍 Those diodes had appeared in my search for some reason and I’m not familiar with them.


I was more worried about the EL84 plate and screen voltages really, especially if I put any modern production tubes in there. I’m now thinking of seeing what the voltages work out like first if I add back the 120K resistor, per the option in the original GEC schematic.

[pdf64]

The amp couldn’t work with that 120k resistor restricting HT current to a couple of mA.
Dunno what it was doing. Maybe depopping a standby switch?

[PRR]

... already at a suitable voltage for feeding the OT, and EL84 plates and screens.

...worried about the EL84 plate and screen voltages really, especially if I put any modern production tubes in there.

I don't know what to say. 

[PRR]

The amp couldn’t work with that 120k resistor restricting HT current to a couple of mA.
Dunno what it was doing. Maybe depopping a standby switch?

R29 (right?) is bypassed by "H.T.Link".

I'm guessing "Standby".

Not a bleeder.

[Jennings]

... already at a suitable voltage for feeding the OT, and EL84 plates and screens.

...worried about the EL84 plate and screen voltages really, especially if I put any modern production tubes in there.

I don't know what to say. 

I know…I thought I might be over thinking it a bit. I was just curious whether there was a different preferred option than using node C for the output section and adding an extra node to replace it in the preamp feed. Plus reduce preamp screen volts to values approaching the schematic quoted values (in the GEC and for the Vox AC15.

The amp couldn’t work with that 120k resistor restricting HT current to a couple of mA.
Dunno what it was doing. Maybe depopping a standby switch?

R29 (right?) is bypassed by "H.T.Link".

I'm guessing "Standby".

Not a bleeder.

I was curious about the purpose of that…the “HT Link” wires went to the terminal strip on the back of the amp. So you’d have to connect them with a piece of wire or a plug (like on a voltage selector, I guess). There was a short piece of wire screwed into the terminals joining the link when I got the amp.

[Jennings]

HT Link connections are the twisted pair of orange and black wires, and R29 is the large brown resistor…I removed both and connected the CT directly to chassis ground at the tag strip where the resistor originally was.

[pdf64]

It may be for a remote switch. eg if the amp was used as part of a paging system for a factory etc. To save unnecessary valve wear in rather hot cathode bias, the operator would push a momentary switch connected to the HT link terminals before speaking on the mic to request ‘Mr Jones to Despatch’.
The 120k standby bypass resistor allows a trickle of anode current, to mitigate for cathode ‘sleeping sickness’ / poisoning / interface resistance.
See http://www.valvewizard.co.uk/standby.html
and p35 of http://tubebooks.org/Books/Atwood/Tomer%201960%20Getting%20the%20Most%20Out%20of%20Vacuum%20Tubes.pdf

[Jennings]

Ah, I see 👍 That makes sense. Thanks for filling me in…interesting learning curve on parts of a circuit I’ve not focused on much previously.

[Jennings]

Had a bit more time for the amp last night, so thought I’d share the current state of play.


A wise man once said “if it ain’t broke, don’t fix it”, and an even wiser man implied “if it’s good enough for GEC, who are you to question”. So with that ethos, I thought I’d pass on using this amp to learn about the various methods and pros/cons of trying to lower the HT voltage further. Instead I’ll just dial it in and enjoy it as is. It’s sounding good and running OK.


Measured voltages are as per the chart attached. I was a little surprised at how low V7 plate came out (especially vs screen), but the 220k measured fine, and the node voltage is OK. So I’m a little baffled why the voltage at the screen is higher after the 1M resistor…anyone see any value in lowering the 220k to bump plate V, or leave as is as it’s playing OK?


Checked bias using the cathode resistor method. Ballpark 70% idle and 8 and a half W dissipation. Anyone see any value in me lowering the cathode resistor value a little to up that, or again leave as-is?

[shooter]

any value in me lowering the cathode resistor value a little to up that, or again leave as-is?

What's the end game?
84's are real easy to drive, over-drive.  If a larger play range is wanted try some guitars with hot pickups, see how far/hard you can drive the PA and the resulting tone.
rotate as many speakers through as you can beg, borrow, or steal
even messing with impedance matching, IF 8ohm is "matched", throw 4ohms, then 16 ohms at it for testing, decide which way, if any gives you warm-fuzzies
Then, IF you still have the itch, bump the cathode to ~90%

[Jennings]

Wise words…thanks shooter 👍 I’ll close up and play a while before considering anything else. Got some band stuff approaching, and longer term I’m of half s mind to build a higher ohm cab in keeping with the metal aesthetic, then reinstate the original OT perhaps.


For anyone interested, here’re a few external shots of the amp post mods. I tried to keep to the original aesthetic. I also reused the empty pins of the Power Supply Socket inside as a series of HT node test points. Schematic with measured voltages etc above.


Thanks everyone who’s help with guidance, suggestions and advice…as ever all gratefully received 😊👍