Help PSU Voltage way off

[colin.hepburn]

SE Headphone Psu rebuilds

Can someone help with this I built the headphone amp as a kit which I got from http://hollowstate.netfirms.com/ear+purist.htm
My first valve build pleased to say that it worked first time
I then upgraded the amp with black gate cap and tantalum resistors

I then thought the Amp warranted a better psu
I designed the psu with Duncan’s psu free software
Using the mains transformer from the kit I also added a choke and solen poly caps I posted my results hear

James Ds comments

Psu looks very nice. residue ripple is 3mV and 500uV and the psubehaviour is good, nice job One small question is it a mono psu as the current s seem to be low for a stereo pair?

James Comments
Gave me the confidents too order the parts

Transformer from the kit
Secondary 230 volts 12 VA 50mA
Diodes MUR 860 8 Amp 600 Volt
Choke 30H 40mA 595 ohms
C1 4.7uf 630 volts
C2/3/4 22uf 630 volts
R1 2.2k 12 watt
R2, 2.7k 12watt

Duncan’s psu came out good with correct Voltages @ C3 212 volts and C4 202 Volts
Problem voltages are way out # under load amp connected
After Bridge Rectifier is 384 Volts should be 304 volts
C3 is @ 376 should be 212 Volts
C4 is @ 372 should be 202 Volts

Voltages @ Pin 1 drive ecc83 207v should be 99 Volts
@ Pin 9 12B4A 371v should be 212 Volts

Heater supply has its own TX@12.6 volts

Can I add a resistor between the bridge rectifier and first cap too pool down the voltage if so can you give the value required sorry I’m **** at this maths think I do have to ask if you would not mind working out the maths stuff and give me just the value of the parts required
The Headphone schematic And PSU can be seen @
http://uk.msnusers.com/MyAudioPics/shoebox.msnw
I do thank you all for the help

[A Stuart]

I notice the transformer in the diagram is labelled 276V, not 230V. Is this helpful?
Alastair

Further thoughts - If these were measured rather than theoretical voltages, it appears that the 28mA current was not being drawn, or the 2.2k resistor was not 2.2k. 28mA x 2.2kOhm = 62V drop.

Looking at the diagram using a transformer yielding 276V at rated current,
Volts at C1 should be arround 395V,
17V dropped across L1 ( (28+2)mA x 595Ohm =~17V) leaving 378V at C2,
66V dropped across R1 ( (28+2)mA x2200Ohm =~66V) leaving 312V at C3,
5V (2mA x2500Ohm = 5V) dropped across R2 leaving 307V at C4.

So unless I have misunderstood badly, you have the wrong transformer AND either a) are not drawing 28mA or b) do not have a 2k2 resistor at R1 position. You have only dropped 8V across L1 plus R1. The 2mA current alone would drop almost 6V through L1 plus R1.

[NickG]

The different secondary voltage would certainly explain a higher voltage, but you seem to have almost twice. Whats the primary of the TX like, is it a 110-110 primary, you have got it wired for 240v I hope?

[colin.hepburn]

yes 115-115 wired 240v its a Hammond 229B230v
http://www.hammondmfg.com/229.htm

[A Stuart]

A 230V rms transformer would yield 230 x 1.41 = ~322V peak voltage out of the rectifier,

but only when the transformer's rated current was being drawn. As mentioned above there is the suspicion that the 28mA load is not being drawn. (Were the valves heated up and running at the time of measurement?) Under minimal load, as I suspect is the case, if we suppose 10% transformer regulation, this increases by 10% to 352V.

If we also suppose your mains is 240V when the transformer rating is for
230V this multiplies the rectified peak output to 366V.

This is not a such huge distance from what you measured.

[colin.hepburn]

the transformer is 230V secondary but isint the no load voltage higher making it 276v R1 correct @2.2k

[colin.hepburn]

valve heaters are a bit hight @13.89 v

[A Stuart]

The rated output of a transformer is the no-load output minus a percentage to allow for internal losses under load. So, for example 100V no-load, 10% regulation, rated at 90V.

Secondly the output is rated as rms volts.
"rms" means "root-mean-square". This is a sort of average of the output as the output goes through zero to maximum to zero through the 50Hz cycle. The rms voltage is the peak voltage divided by square-root-of-two.

So a 1V rms voltage peaks at 1.41V , 100V rms peaks at 141V etc, conversely 1V peak = 0.707V rms etc)

The rectifier snatches current at the peak voltage, tops up the capacitor to this level, and the capacitor loses little between cycles. (complicated further by the choke, but you didn't want the maths, and I would struggle there anyway).

If you don't have 66V across R1 (and as you confirm, it is 2200Ohms) then you are not drawing 30mA through it.

Hope this helps, Alastair

[colin.hepburn]

Thanks for you're help Alastair
not sure why its not drawing the 30mA though

[A Stuart]

Further further further thoughts -

I notice your recommended voltages drop 10V across R2 (2k5) drawing 2mA. This will not happen across a 2k5 resistor, so I am further baffled
Inspection of the diagram (New SE Headphone Amp PSU3) agrees with the load being 1mA per channel = 2mA, and the output valves drawing approx 15mA each channel = 30mA total.

We would still need to ascertain what voltage was being yielded under correct load, before calculating any resistor to allow further adjustment.
We would need to drop a total of 110V from, 322V (230Vrms under load x1.41 = 322V peak) to 212V as recommended at C3.

We also still need to clarify whether your measured values are under load. I still think the primary discrepancy is that little current was being drawn during measurement.

Alastair.

and more thoughts .... 7k5 resistors R4A & R4B (Large 5 Watt ones) will both have 114V across them if each of the output valves is drawing current, according to the diagram. ... and they will get hot.