Speccing power transformers for valve amps

Keefy

A couple of related questions on the above:

(1) When looking for a suitable PT it’s usually easy enough to work out the secondary voltage needed for the HT, but how can you work out the required current rating?

(2) I have an unidentified PT I picked up off eBay. The secondary gives about 300-0-300 V AC off load but how can I work out its current rating? I was thinking of graphing voltage against calculated current through a series of chunky resistors and finding the point at which it delivers maximum power.

Danny1969

If you assume the power transfer efficiency is around 90 % then an increase in voltage leads to a decrease in available current ... power is preserved it's just the format of it changes. So basically 240V 1amp primary could give 500V at 380mA secondary etc. I'm sure @ICBM could explain it better 

Keefy

Just been reading a description by Chris @RiftAmps in this month’s Guitar mag of how he worked out the specs for his Gibson EH-185 build, which has helped.

RiftAmps

From Merlin

Design Example 
To design the power supply we need to know how much average DC current the amplifier circuit will draw. The preamp stages will normally be single-ended (class A) so they draw constant average current. If the preamp valves are ECC83 / 12AX7s then they will usualy be biased around 1mA per triode, or less. You can work this out from load lines. 
If the output stage is class AB (e.g nearly all push-pull amps) then its current will increase with signal level. The output valves will therefore usually be biased to around 70% of the data sheet max anode dissipation, thereby allowing some power headroom, so they don't red-plate at full drive. 

Let's suppose a 50W amp uses three ECC83 / 12AX7s and a pair of EL34s. 
The preamp valves will consume around 6 x 1mA = 6mA (there are two triodes per bottle, remember). 
The EL34s are rated for 25 watts maximum, so they will probably be biased around 0.7 x 25W = 17.5W. However, we should work with the maximum average figures, i.e. full drive when the average dissipation increases to 25W. Thus if the raw DC supply voltage is 400V then they will each consume about 25W / 400V = 62.5mA, or 125mA for the pair. 
The data sheet suggests a screen-to-anode current ratio of 6.5, so we can expect the screen currents to amount to 125 / 6.5 = 19mA for the pair. 
The total for the whole amp is therefore 6 + 125 + 19 = 150mA. 

Keefy

RiftAmps said:

From Merlin

Design Example 
To design the power supply we need to know how much average DC current the amplifier circuit will draw. The preamp stages will normally be single-ended (class A) so they draw constant average current. If the preamp valves are ECC83 / 12AX7s then they will usualy be biased around 1mA per triode, or less. You can work this out from load lines. 
If the output stage is class AB (e.g nearly all push-pull amps) then its current will increase with signal level. The output valves will therefore usually be biased to around 70% of the data sheet max anode dissipation, thereby allowing some power headroom, so they don't red-plate at full drive. 

Let's suppose a 50W amp uses three ECC83 / 12AX7s and a pair of EL34s. 
The preamp valves will consume around 6 x 1mA = 6mA (there are two triodes per bottle, remember). 
The EL34s are rated for 25 watts maximum, so they will probably be biased around 0.7 x 25W = 17.5W. However, we should work with the maximum average figures, i.e. full drive when the average dissipation increases to 25W. Thus if the raw DC supply voltage is 400V then they will each consume about 25W / 400V = 62.5mA, or 125mA for the pair. 
The data sheet suggests a screen-to-anode current ratio of 6.5, so we can expect the screen currents to amount to 125 / 6.5 = 19mA for the pair. 
The total for the whole amp is therefore 6 + 125 + 19 = 150mA. 

Thanks Chris, that's exactly the sort of thing I was looking for!

PeteC

The Blencowe books are definitely worth getting hold of.  Tons of great information in them 

Keefy

PeteC said:

The Blencowe books are definitely worth getting hold of.  Tons of great information in them 

His info on grounding schemes was a revelation and helped me sort my Rocket build. I'll stick 'em on my Christmas list.

Keefy

Just found this.

robinbowes

PeteC said:

The Blencowe books are definitely worth getting hold of.  Tons of great information in them 

Some very useful info here: http://www.valvewizard.co.uk/

Nice guy, met him in York once to pick up something I'd bought form him - I forget what...

R.

Keefy

robinbowes said:

PeteC said:

The Blencowe books are definitely worth getting hold of.  Tons of great information in them 

Some very useful info here: http://www.valvewizard.co.uk/

Nice guy, met him in York once to pick up something I'd bought form him - I forget what...

R.

Yes I have that site bookmarked

PeteC

I think there’s quite a lot of material he has written for another book that hasn’t yet been published.   Some great info on his site.   

Keefy

I've been testing a PT that I bought from eBay a while back. My idea is to use it at some point for either a 10W Weber Maggie (based on a 5F2-A Princeton but with a 6L6), or failing that, a regular 5W Champ with a 6V6. I set up a test rig to allow me to (briefly) measure CT to one side of the HT with different resistive loads in place. Here are my results:

Load (Ω)	V AC (V)	I (mA)	P (W)
OPEN	327	0	0
22100	325	14.71	4.78
17900	323	18.04	5.83
12170	319	26.21	8.36
9910	315	31.79	10.01
8060	314	38.96	12.23
5700	312	54.74	17.08
3470	300	86.46	25.94
2290	292	127.51	37.23

I used this calculator to get the current draw for 1x12AX7 and 1x6L6GC assuming the PT is working at 300-0-300, and it gave 96mA so it looks like it will work. How far is it normal for the voltage to be pulled down without caning the transformer? It's a fairly chunky unit as you can see from the pic below.

Keefy

I also tested the other secondaries, which I assume are filament supplies. Both read over 7V off-load. The green pair still put out 6.84V when drawing 7.6A - that's a lot of filaments! With the same load the brown pair's output has dropped to 6.33V. I'm going to treat them as nominal 6.3V and 5V filament supplies respectively, but I'm going to need resistors or diode pairs to drop the voltages down to thew desired values when I get round to a build with this transformer.

ICBM

I notice your mains voltage is reading 249V - that's very high. What voltage tap is the primary?

7.6A is way too high for a filament winding in a transformer that size - it won't be rated for anything like that, more likely about 3A max.

Keefy

ICBM said:

I notice your mains voltage is reading 249V - that's very high. What voltage tap is the primary?

7.6A is way too high for a filament winding in a transformer that size - it won't be rated for anything like that, more likely about 3A max.

Wall voltage here is 247-249V, hence I tested it at 249V off my Variac for worst case. Just tested it again (with my decent Megger AVO) and at present it's 247V.

The transformer primary has an old typewritten note stuck to it advising you how to connect the two primary windings in series for 210-250V or in parallel for 105-125V.

Those output current readings were done with the mains power on for a couple of seconds at a time. I wanted to see how much current would pull the outputs down to 6.3V or 5V. I'm only going to be running a 5Y3 (2A of 5V filament current), a 12AX7, and a 6L6GC (combined filament current 1.2A), so I'll drop the voltages to bring them into range.

My main concern is whether the HT has enough beans to run the 6L6GC.

ICBM

Keefy said:

Wall voltage here is 247-249V, hence I tested it at 249V off my Variac for worst case. Just tested it again (with my decent Megger AVO) and at present it's 247V.

The transformer primary has an old typewritten note stuck to it advising you how to connect the two primary windings in series for 210-250V or in parallel for 105-125V.

In that case I think you'll need to drop the voltages - those are much too high for filaments.

Keefy said:

My main concern is whether the HT has enough beans to run the 6L6GC.

I think it should probably be OK - that size of transformer looks like it should be for something like a 2-EL84 15W amp in near-Class A, so a single-ended 6L6 won't pull any more than that.

Keefy

ICBM said:

Keefy said:

Wall voltage here is 247-249V, hence I tested it at 249V off my Variac for worst case. Just tested it again (with my decent Megger AVO) and at present it's 247V.

The transformer primary has an old typewritten note stuck to it advising you how to connect the two primary windings in series for 210-250V or in parallel for 105-125V.

In that case I think you'll need to drop the voltages - those are much too high for filaments.

Keefy said:

My main concern is whether the HT has enough beans to run the 6L6GC.

I think it should probably be OK - that size of transformer looks like it should be for something like a 2-EL84 15W amp in near-Class A, so a single-ended 6L6 won't pull any more than that.

Thanks, that’s good then, just need to keep my eyes open for a 5.5k:8 OT that will handle 10W...

PeteC

I can recommend the hammond txs 
or if you want to try high spec
from a small maker then AskJanFirst in germany is particularly good.