Transformer wiring.

[AC4]

Hi all.

First time amp builder here with some noob questions about power and output transformer wiring - sorry!

I recently bought an Epiphone Valve Jr which I have gutted, planning to use the chassis to house a single ended EL84 circuit similar to the Bad Cat Mini II as a 'relatively' inexpensive way into amp building - he says £400 later...

I bought replacement power and output transformers with solder tags, rather than buying them pre-wired, as I wanted to wire everything myself. I think I have a fairly good understanding of how to wire the transformers up but I'd really appreciate it if someone could confirm whether or not I have it right. The amp will be running on 240v.

I've attached a couple of images of the transformers and the schematic I'll be working from. I've also attached a copy of the layout I'm planning on using. I'd be grateful if someone could cast an eye over it and see if I've made any rookie mistakes, particularly with the grounding. Happy to provide any details I've more than likely missed. Here to learn!

Many thanks.

Phil.

[Willabe]

Welcome!   

I bought replacement power and output transformers with solder tags, rather than buying them pre-wired, as I wanted to wire everything myself.

You bought the wrong transformers. They will work but how will you shield/cover the hot connections so no one gets shocked?

You should have bought the standard transformers with insulated wire fly leads.

When you say "I wanted to wire everything myself" either way if a power transformer (PT) and output trans. (OT) has fly leads or solder lugs they still have to be wired up to the circuit parts.   

Can you return those? It would be best to return them. There really not safe to use under normal construction methods.

[Willabe]

... and see if I've made any rookie mistakes, particularly with the grounding.

Here's a link to Merlin's page on grounding. Look at the bussed local multi star ground, works great. Starts at ~ page 270. Each B+ filter cap forms it's one local ground star with all the grounds from the circuit that B+ filter cap node feeds, then you run a buss wire from that local ground star to the next local ground star, etc.

And only 1 chassis circuit ground connection. Many guys here use 2 chassis ground connections, power amp and preamp. That works fine too. Pay attention to how Merlin shows how to wire the PT's CT connection to the 1st B+ filter caps ground lead. That's the noisiest, dirtiest ground in the amp.

https://www.valvewizard.co.uk/Grounding.html

[shooter]

I've used a couple like that, used insulated soldered spade plugs.  not much different than multi-caps in a can as far as "shock hazards" go.
put a sticker on the outside "WARNING only qualified techs may open this box"  worked for the big-dogs

[Willabe]

I've used a couple like that, used insulated soldered spade plugs.  not much different than multi-caps in a can as far as "shock hazards" go.

Yes they are, big difference, the cap cans connections are inside the chassis, those transformers have their connections outside of the chassis.

Yes, if you stack cap cans then the cans shell will be hot, but that's rare and he's not doing that here. (Who did that? Was it Sunn, Traynor or Hiwatt on a few models?)

[astronomicum]

(Who did that? Was it Sunn, Traynor or Hiwatt on a few models?)

Univox did it on a number of their 100 and 150 watt models: stacks and doublers.

[shooter]

connections outside of the chassis.

missed that

[AC4]

Welcome!   

I bought replacement power and output transformers with solder tags, rather than buying them pre-wired, as I wanted to wire everything myself.

You bought the wrong transformers. They will work but how will you shield/cover the hot connections so no one gets shocked?

You should have bought the standard transformers with insulated wire fly leads.

When you say "I wanted to wire everything myself" either way if a power transformer (PT) and output trans. (OT) has fly leads or solder lugs they still have to be wired up to the circuit parts.   

Can you return those? It would be best to return them. There really not safe to use under normal construction methods.

Ah. The rookiest of all rookie errors... First lesson learned, thankfully before I received my very own Darwin Award! I'll check with the manufacturer to see if they'll allow me to switch them for ones with fly leads. Thanks for pointing that out.

I've read Merlin's chapter on star grounding but, being a bit bamboozled by it all, decided to go with a simpler single ground buss grounded to the chassis at the input. Am I potentially entering a world of pain if I go this route? I get the impression that there are no guarantees where grounding is concerned, but I may have got that wrong. I guess its better to invest more time and effort early doors than chase my tail troubleshooting afterwards?

[AC4]

Hi all.

Finally got two, safer, replacement transformers for my first amp build (an AC4 style SE amp). The PT is a drop through and the OT has flying leads. No external live terminals now...    My Epiphone Jr chassis is repopulated with transformers, valve sockets, tagstrips, switches, power socket, fuse holders, jacks and pots and I'm ready to start soldering.

If someone would be kind enough to have a look at my layout and check I've got the transformers wired correctly I'd be very appreciative. The amp will be running on 240v (I'm in the UK) and I want to make sure I have the PT wired correctly for this, with the two centre lugs connected. I also want to make sure the blue and brown wires are connected to the right lugs. Likewise with the OT, I'd like to be sure that the 5k and 0k leads are going to the correct parts of the circuit.

I've attached a copy of the schematic I'll be working from and photos of the transformers I'm using also. Looking forward to getting stuck into my first amp build.

Many thanks.

Phil.

[tubeswell]

The dc-coupled cathode follower will work better with a 100k load resistor (to match the 100k plate resistor in the driving stage)


As it’s a single ended amp, you make want a CLC filter between the rectifier and the EL84 plate node (for better hum reduction)


You also may want to consider adding global NFB (to make the most out of the blackface tone stack) YMMV

[DummyLoad]

Use some heat-shrink over those connections and then put the exposed barrier connections behind some silicon rubber.

Never mind - I missed that last post where you bought transformers with flying leads and end-bells. If you do use that PT in another build follow what I opened with, or mount the PT inboard.

BTW, it looks like you can use .110" (2.8mm) insulated fast-ons with that original PT.

https://www.digikey.com/en/products/detail/te-connectivity-amp-connectors/2-520080-2/289209



--Pete

[DummyLoad]

Single diode on HT - Are those 1.6KV PIV diodes? If not, you need a pair of 1KV PIV rated diodes in series for each HV secondary.

What's with the 100Ω resistor from 6.3V CT to ground? Just ground the CT or leave it unconnected build a virtual CT with a pair of 100Ω 2% from each 6.3V winding end to K of EL34 or grounded.

Typically we ID connections with wire color - Green/Yel are Filaments (Yellow 5V and Green 6.3V or 12V) Red is B+, Orange is Screens, Blue or Brown for plates, Yellow for cathodes. Violet for negative supplies.

10uF for B+ filter by itself to plate of EL84 is going to make lots of 100Hz buzz. Tubeswell pointed that out already. Passing on from experience, your plan will not have pleasing results in the buzz/hum realm as presented. Suggest use of a CRC filter there with 32uF at least, most preferable is use of a CLC with a choke with a current rating equal to the sum or greater than the sum of the draw for all.

--Pete

[AC4]

Single diode on HT - Are those 1.6KV PIV diodes? If not, you need a pair of 1KV PIV rated diodes in series for each HV secondary.

What's with the 100Ω resistor from 6.3V CT to ground? Just ground the CT or leave it unconnected build a virtual CT with a pair of 100Ω 2% from each 6.3V winding end to K of EL34 or grounded.

Typically we ID connections with wire color - Green/Yel are Filaments (Yellow 5V and Green 6.3V or 12V) Red is B+, Orange is Screens, Blue or Brown for plates, Yellow for cathodes. Violet for negative supplies.

10uF for B+ filter by itself to plate of EL84 is going to make lots of 100Hz buzz. Tubeswell pointed that out already. Passing on from experience, your plan will not have pleasing results in the buzz/hum realm as presented. Suggest use of a CRC filter there with 32uF at least, most preferable is use of a CLC with a choke with a current rating equal to the sum or greater than the sum of the draw for all.

--Pete

Hi Pete

The diodes I have are UF4007, 1kV. I'll put two in series as you suggested. Thanks for that.

I got the 100ohm CT resistor idea from the Valve Wizard website, which says:

Transformer Centre Tap
The traditional way to reduce hum is to use a transformer with a centre tap, and connect it to ground. A refinement is to ground the centre tap through a small flame-proof resistor (anything up to 100 ohms). This will act as a fuse if there is a short between the anode and heater pins of the power valves, thereby reducing collateral damage.

For someone as inexpereinced as I am in amp building this seems like a sensible and easy modification to do? I could be wrong!

Thanks for the info regarding colour coding of wires, good to know for future reference.

Phil.

[AC4]

The dc-coupled cathode follower will work better with a 100k load resistor (to match the 100k plate resistor in the driving stage)


As it’s a single ended amp, you make want a CLC filter between the rectifier and the EL84 plate node (for better hum reduction)


You also may want to consider adding global NFB (to make the most out of the blackface tone stack) YMMV

Hi Tubeswell.

In what way will the 100k resistor work better? What would the difference in sound be between 100k and 56k?

I take it that a CLC filter involves having a choke? How do I determine what kind of choke I need?

Thanks.

Phil.

[Merlin]

The diodes I have are UF4007, 1kV. I'll put two in series as you suggested.

The transformer is only 190-0-190V, so single 1N4007s are fine.

What would the difference in sound be between 100k and 56k?

You'll get a bit more distortion with 56k.

I would suggest one minor change to the layout to improve grounding:

[tubeswell]

The dc-coupled cathode follower will work better with a 100k load resistor (to match the 100k plate resistor in the driving stage)

As it’s a single ended amp, you make want a CLC filter between the rectifier and the EL84 plate node (for better hum reduction)

You also may want to consider adding global NFB (to make the most out of the blackface tone stack) YMMV

Hi Tubeswell.

In what way will the 100k resistor work better? What would the difference in sound be between 100k and 56k?

I take it that a CLC filter involves having a choke? How do I determine what kind of choke I need?

Thanks.

Phil.

Partly to do with the biasing of a DC-coupled CF and partly because making both resistors equal normally yields optimum gain, and the (further) load of whatever comes after the CF (i.e., the tone stack) is in parallel with the CF load resistor and so loads the CF down even further - so why make the CF resistor smaller than it needs to be?


DummyLoad already told you about the choke.

[AC4]

The diodes I have are UF4007, 1kV. I'll put two in series as you suggested.

The transformer is only 190-0-190V, so single 1N4007s are fine.

What would the difference in sound be between 100k and 56k?

You'll get a bit more distortion with 56k.

I would suggest one minor change to the layout to improve grounding:

Thanks for the advice and info Merlin.

So, if I'm reading that correctly, don't ground the CT and the first filter cap node to chassis, just the CT of the heater supply?

[Merlin]

So, if I'm reading that correctly, don't ground the CT and the first filter cap node to chassis, just the CT of the heater supply?

Yes. It doesn't really matter where you ground the heater, but you do want to isolate ripple current away from all other grounds, which is what my suggestion does.

[AC4]

As it’s a single ended amp, you make want a CLC filter between the rectifier and the EL84 plate node (for better hum reduction)

I take it that a CLC filter involves having a choke? How do I determine what kind of choke I need?

Thanks.

Phil.

DummyLoad already told you about the choke.

Cards on the table, I'm a complete novice when it comes to amp building and the associated math so I'm not sure where to start calculating the current draw/required choke specs etc... but I can follow a schematic and I'm competent with a soldering iron, so just need pointing in the right direction.

The 5E2 Princeton had a 4H at 50ma DC power choke. Hammond make the 194B which is 4H at 90ma. Would that be suitable?

https://www.mouser.co.uk/ProductDetail/Hammond-Manufacturing/194B?qs=hGvGUDq83YECcv5eZzVbqw%3D%3D

Hammond also make the 194D which is 5H at 150ma. Would that be an option? is there such a thing as an over-specced choke?

https://www.mouser.co.uk/ProductDetail/Hammond-Manufacturing/194D?qs=hGvGUDq83YGkPv75q6145g%3D%3D&_gl=1*gphj7v*_ga*dW5kZWZpbmVk*_ga_15W4STQT4T*dW5kZWZpbmVk*_ga_1KQLCYKRX3*dW5kZWZpbmVk

Here to learn and appreciative of the help.

Phil.

[tubeswell]

You can work out the total idle current draw on the High Tension winding by adding up the cathode current (Ik) of all the tubes (because all the current in the tube passes through the cathode). For peak current (under signal conditions), multiply that by 1.2 x (which is a handy rule of thumb for a single-ended amp operating in Class A)

Use Ohms Law (V = I x R) to work it out by rearranging the equation to solve for current (I), so: I = V/R

e.g., for Vox AC4:
V1 = 2.7V/1,500R = 0.0018A (1.8mA)
V2a = 163V/100,000R = 0.00163A (1.6mA)
V2b = 1.7V/3,300R = 0.00051A (0.5mA)
EL84 = 8.5V/150R = 0.056A (56mA)

so that's 59.9mA.

x 1.2 (for estimate of peak current) = 72mA

so you'd use the 90mA choke for a 'whole-of-supply' CLC filter (between the rectifier and the OT primary)

[AC4]

You can work out the total idle current draw on the High Tension winding by adding up the cathode current (Ik) of all the tubes (because all the current in the tube passes through the cathode). For peak current (under signal conditions), multiply that by 1.2 x (which is a handy rule of thumb for a single-ended amp operating in Class A)

Use Ohms Law (V = I x R) to work it out by rearranging the equation to solve for current (I), so: I = V/R

e.g., for Vox AC4:
V1 = 2.7V/1,500R = 0.0018A (1.8mA)
V2a = 163V/100,000R = 0.00163A (1.6mA)
V2b = 1.7V/3,300R = 0.00051A (0.5mA)
EL84 = 8.5V/150R = 0.056A (56mA)

so that's 59.9mA.

x 1.2 (for estimate of peak current) = 72mA

so you'd use the 90mA choke for a 'whole-of-supply' CLC filter (between the rectifier and the OT primary)

That's great, appreciate the detailed response Tubeswell. Thanks!

[AC4]

Just figuring out where to place the choke I've just ordered and how to orient it when I realised that my new drop through PT might not be oriented the same way as the previous clamped version. I think I need to reposition the OT 90 degrees and orient the choke the same, am I correct?

Many thanks.

[tubeswell]

As your PT is the Z-mount (‘lay down’) type, and the choke will most likely be x-mount (‘stand up’) type, the coils will be unlikely to be aligned. So it won’t matter where you put the choke.

[AC4]

Hi all.

I've re-configured my ground scheme based upon the info on bus bar grounding at Merlin's website. Due to the chassis being a repurposed Epiphone Valve Jr I'm having to work around a pre-existing set of holes to a certain extent, so my layout is less than optimal. That said, is it likely that despite longer lengths of cable being used to connect each gain stage to ground this new scheme will be an improvement, or will the distance between the filter cap nodes and each of the valve sockets negate any benefit?

Many thanks.

Phil.

[Willabe]

If node D is grounded at the terminal strip chassis bolt connection, then you have a ground loop.

You should have moved the filter caps to being right next to the circuit they feed.

And need a schematic to be able to go over the layout grounding.

[AC4]

Hi Willabe.

If node D is grounded at the terminal strip chassis bolt connection, then you have a ground loop.

It isn't.

You should have moved the filter caps to being right next to the circuit they feed.

I realise this. Like I said, the layout is less than optimal because I didn't start with a blank chassis.

And need a schematic to be able to go over the layout grounding.

Schematic posted above. Here it is again.

Thanks.

Phil.

[Willabe]

You should have moved the filter caps to being right next to the circuit they feed.

I realise this. Like I said, the layout is less than optimal because I didn't start with a blank chassis.

Doesn't matter.

There's room in that layout drawing to move the C node cap and the D node cap right next to the circuit they feed. Merlin calls them local ground stars.

And if you moved the C and D node caps, then you could have moved the B and A node caps closer to the EL84 tube socket.

You are using terminal strips, you can put those where ever you need to. 

[AC4]

You should have moved the filter caps to being right next to the circuit they feed.

I realise this. Like I said, the layout is less than optimal because I didn't start with a blank chassis.

Doesn't matter.

There's room in that layout drawing to move the C node cap and the D node cap right next to the circuit they feed. Merlin calls them local ground stars.

And if you moved the C and D node caps, then you could have moved the B and A node caps closer to the EL84 tube socket.

You are using terminal strips, you can put those where ever you need to.

I should point out that the layout I posted here doesn't entirely represent the reality of the situation, but point taken. As things are currently, the A and B node caps are only 6cm and 4cm away, respectively, from the EL84 tube socket, the C node cap is 5cm away from the V2 tube socket and the D node cap is the furthest away at 15cm to the V1 tube socket.

My follow up question would be, would those lengths of connecting ground wire potentially make a significant difference to the noise floor?

If I've understood correctly, it would be better to attach the negative side of the filter cap to the closest ground to the tube socket and use longer lengths of wire to connect the positive side of the filter caps back to the rectifier?

[Willabe]

My follow up question would be, would those lengths of connecting ground wire potentially make a significant difference to the noise floor?

No, I don't think there very long. Fender had longer runs than those. But that's not the thing.

If I've understood correctly, it would be better to attach the negative side of the filter cap to the closest ground to the tube socket and use longer lengths of wire to connect the positive side of the filter caps back to the rectifier?

That's not it either.

Each filter cap's ground lead goes to the grounds of the tube circuit that cap feeds. So node D's cap's ground lead should go to the 1st preamp tubes grounds. That will make a ground star. Then you run a wire (ground buss wire) from that ground star to the next ground star. It's a bussed multi local ground star system. It's wired ground not a random chassis ground.

You have fixed this but, you had the OT's secondary ground grounded to the 1st preamp tubes B+ node D filter cap ground. That OT ground has a lot of current, the 1st preamp tube ground star has the least current in the amp. That OT current can modulate the preamp tubes ground, that will cause humm.

So we keep the filter cap ground lead wired to the circuit grounds of the circuit that cap powers. 

You put the cap close so you don't end up with long wire runs all over the chassis. Big mess.

Look at the drawings I posted from Merlin.  You have your grounds jumbled together, they should be organized together with the B+ nodes filter cap ground lead that feeds the circuit, local ground star, then bussed together, then grounded at only 1 chassis point. 

It looks like, from your layout drawing, it's the only thing I have to go by, that the D node cap would fit nicely on the far right side with the cap ground lead facing up towards the top of the chassis/input jack, which is where the tubes and input jack grounds are. And then you could put the + lead on the terminal lug just below that preamp tube. Nice and neat.

[Willabe]

I've re-configured my ground scheme based upon the info on bus bar grounding at Merlin's website.

I didn't see it before, but your re-configured ground scheme is much better. Not jumbled anymore. 

But, I'd disconnect the buss wire going from node A over to the input jack ground lug. Very high charging current in that ground going to the lowest current (signal) in the amp, from guitar PUP's. 

Then the ground wire that goes from the input jack to the chassis ground, disconnect at the chassis ground and run it to the 1st preamp tube terminal lug. Then run a wire from that 1st preamp tube star ground to the chassis ground.

What this does is takes the chassis ground point connection from the 1st preamp tubes ground star, instead of the node A ground.

1 last thing, move the OT ground off the node A ground and move it over to the node B ground. You want nothing but the PT B+ CT and the node A filter cap ground together, so they can charge up/discharge, cycle all they want without disturbing anything else.     

I think that will be much better.

[Willabe]

Oh, the volume pot's ground should go to the 2nd triodes K ground, 22uF/1K5 R ground. It gets grounded with the tubes other grounds and the B+ filter cap nodes ground. That's the local star.

[AC4]

Oh, the volume pot's ground should go to the 2nd triodes K ground, 22uF/1K5 R ground. It gets grounded with the tubes other grounds and the B+ filter cap nodes ground. That's the local star.

Thanks for the advice Willabe. I think this is better?

[Willabe]

Yes, that's much better with the B+ filter caps moved close to the tube circuit their feeding.

But still a couple things;

Then the ground wire that goes from the input jack to the chassis ground, disconnect at the chassis ground and run it to the 1st preamp tube terminal lug. Then run a wire from that 1st preamp tube star ground to the chassis ground.

You want the input jack grounded with the other grounds from the 1st preamp tube. You have it going to the buss closest to the 2nd preamp ground star. Make a local ground star, by having all the grounds from the 1st preamp tube with the D node filter caps ground. Then run that local ground star to the chassis ground connection. 

1 last thing, move the OT ground off the node A ground and move it over to the node B ground. You want nothing but the PT B+ CT and the node A filter cap ground together, so they can charge up/discharge, cycle all they want without disturbing anything else.

[Merlin]

It's such a simple circuit you could probably use random grounding and it would be fine. Sometimes you don't need to sweat the small stuff.

[Willabe]

It's such a simple circuit you could probably use random grounding and it would be fine. Sometimes you don't need to sweat the small stuff.

Easy circuit to learn on for later builds.

[AC4]

Yes, that's much better with the B+ filter caps moved close to the tube circuit their feeding.

But still a couple things;

Then the ground wire that goes from the input jack to the chassis ground, disconnect at the chassis ground and run it to the 1st preamp tube terminal lug. Then run a wire from that 1st preamp tube star ground to the chassis ground.

You want the input jack grounded with the other grounds from the 1st preamp tube. You have it going to the buss closest to the 2nd preamp ground star. Make a local ground star, by having all the grounds from the 1st preamp tube with the D node filter caps ground. Then run that local ground star to the chassis ground connection. 

1 last thing, move the OT ground off the node A ground and move it over to the node B ground. You want nothing but the PT B+ CT and the node A filter cap ground together, so they can charge up/discharge, cycle all they want without disturbing anything else.

Fixed. Thanks for your help Willabe, I appreciate it.

[Willabe]

That looks right to me. And you have isolated input and speaker jacks, so you have only 1 chassis circuit ground.

Should be very quite once you get it in it's cab.   

[AC4]

Just a quick update. I finally finished my amp build and I'm really pleased with it.

I just want to thank everybody here who offered me guidance and advice, it really helped and I appreciate it. The amp is quiet as a mouse, even when cranked, and it sounds as good as I hoped it would. It might not be the greatest amp ever built but its an improvement on what was originally in the Epiphone Jr chassis I used, which isn't saying much, admittedly!

I'm already planning my next build, a 15w Top Boost style preamp into a pair of EL84s (Z Wreck Jr, Matchless Lightning etc...). I'm going to do this one from scratch so I'm sure I'll be back with even more questions.

Thanks again!