Sound Sales PX4 Amplifier

[hmv1944]

I have been working on a power amp from a large radiogram made by Sound Sales of Farnham (see pictures) which dates from around 1947 but have failed to find any service information or schematic diagram despite many months of searching on the web. The valve line-up is: 6J7, 6SN7GT, 6SN7GT, PX4, PX4, 5Z4G. I have created my own schematic from inspection of the amp which I hope is correct and have uploaded it for the information of others. (Do ask me if you would like a higher res image.) The component values shown have been taken from the components in the existing unit. There is a large carbon resistor running from a centre-tap of the PX4 heater winding to ground which has lost almost all of its paint due to sixty years of heat dissipation. It does, however, retain a little mauve paint on one end and measures 52.5 ohms (with DMM) so I have assumed it was originally of nominal value 47R.

After electrolytic capacitor reforming and replacement of wax paper caps the amp sounds wonderful using an FM radio as signal source and using the existing speaker in its large padded cabinet. There is just the barest suggestion of a 50Hz hum. (Happily SWMBO loves the sound as well!) The two PX4s are not identical, one having extra fins on its anode plates, the other not. The DC voltage drop across the large carbon 47R resistor that has been hot in the past is around 28 volts and after about an hour of running is only warm. Is this resistor really dissipating 15 watts? If it needs replacing what could be used instead? There is an adjustment pot from the grid of one PX4 to ground which I assume is there to balance the two output valves. My question is what measurement should be used to make the adjustment? Using the DMM on DC there is 0.04 V on the grid of V4 and 0.13 V on the grid of V5. By measuring the voltage drop across the 100R resistors in the anode circuit of the output valves it shows there is a difference in the current flowing in each valve; around 20mA for V5 and 36mA for V4. Should these values be the same or does this just reflect a difference in the two valves? I saw with alarm that used PX4s retail for £80-£90 each! In order to be gentle to the old lady I have reduced the mains voltage to give an HT of around 225 volts.

Alan.

[PJL]

56mA in a 47ohm resistor gives 2.6V so it must be a 470ohm. This gives a better dissipation of about 1.5W. The balance control is to match gm. Stick a scope across the 470 ohm, put in a sinewave and adjust for minimum amplitude (an AC meter should also be OK). The DC characteristics of the PX4 are a bit different which will increase distortion at higher volumes but should be OK...Peter

[Peter.N.]

Hi Alan

A very neat and nicely made amp, takes me back a few years! The current flowing through the valves should be pretty well the same, try adjusting the pot to give similar curents in both valves. I have never seen PX4s used with octal valves before, I would have thought that 6L6s would have been of the same ilk, unless they particularly wanted triodes. I dare say someone will put me right on that.

Peter.N.

[jjl]

Hello Alan

As others have stated, your cathode resistor must have a value of around 470 or 500 ohms.
The rated anode current for a PX4 is 30mA to 60mA depending on anode voltage, bias etc, this means that one of your PX4s has a fair bit lower emission than the other and also means that your amplifier appears to be running the PX4s at fairly low anode dissipation. What anode voltage do you measure at pin 1 on each PX4 with normal main voltage?
In principle the anode currents of the PX4s should be equal, this is more important where there is a shared cathode resistor as in your circuit. The PX4 with lower emission current may need to be replaced one day but will probably carry on working for a long time.
I'd advise aginst reducing the mains supply voltage to the amp as this will underrun the heaters of all of the valves and is likely to shorten their life.


John

[screengrid]

Very, very nice!
All the above is very good advice.
I would add that the heater voltage should be measured at the PX4 pins with a DVM to make sure it's as near to 4v as possible.
PX4s do not like being over-run, it reduces their life quite drastically.
Measure the hum at the speaker terminals with a DVM, what is it?
Converting to DC heaters will reduce the hum somewhat.
One of the PX4s is clearly near the end of it's life, as stated previously, when you feed your sinewave in, measure the (AC) voltage on the anodes will show how much.
Lovely kit, have fun!

[hmv1944]

Thanks everyone,
Yes, that large cathode resisitor for the output valves should be 470R and not 47R. (Just another case of misreading the decimal point- will I never learn?) I will add an amended schematic diagram. Will take another look at voltages and see what I can do. Is it safe to keep the output valves as they are and run the amp even if one is starting to fail?
Alan.

[DIN Plug Pete]

Hello Alan,

A very nice amplifier, but if it is exactly as your diagram then I think there might have been a wiring error when it was made? I can't understand how V4 could get any audio drive? The anode pin 5 of V3 feeds a 0.1 uF coupling capacitor, which drives V4 pin 3 and also a divider resistor arrangment. The output of this divider is fed straight back into the same triode section (pin 4 of V3). I can't see how any audio drive is applied? In my opinion, the capacitor connections to V3 anodes (pins 2 and 5) should be swapped. Any others agree with me?

Peter.

[Richard]

Hi

I wondered about that as well, but assumed I was not understanding something........

Richard

[PWH]

looks like a phase splitter for a push pull output to me

[PJL]

Peter is right, the audio coupling caps are drawn to the wrong 6SN7GT anodes. The 6SN7GT is a phase splitter and takes a proportion of the output from the first half to feed the 2nd half via the pot that allows the gain to be balanced. Hence my earlier comment that the correct way to balance the amplifier is to put a scope across the 470ohms cathode bias, inject a sine wave and trim the pot to show the minimum AC amplitude which will occur when the PX4 are delivering the exact same output.

There is no method for DC bias adjustment so you just have to put up with the difference in anode current which will cause it to enter class B at a lower output than it should. (or redesign the output to use seperate cathode bias with bypass caps)

[hmv1944]

Thanks guys. Another silly mistake on my part. The coupling caps for the output valves should be as Peter describes. I must admit I thought it looked a bit odd but my electronic theory was not up to working out why. And I should have checked the hardware again!
Using an 800Hz sine wave and measuring the AC voltage across the 470R cathode resistor with a DMM I wasn't able to find a clear minimum reading. I'll try again with a lower frequency. No spare time over Christmas!
Alan.

[msaunby]

Hi,

I've just joined the forum, because I was searching for information on Sound Sales and found this thread.

My interest in the company is largely as a commercial manufacturer of the 1934 Wireless World Quality Amplifier. http://mike.wepoco.com/retro-geekery...ity-amplifiers This amplifier was redesigned many times until eventually replaced in 1947 by the famous Williamson design. For the Quality Amplifier PX4 or PX25 were the output valves of choice even when 6.3V octals were used for the rest of the amplifier.

In advert in March 1946 Wireless World the "DX Plus 7" Quality chassis is launched featuring a "paraphased PX4 push-pull output stage". It seems you have an amplifier of this design.

Hope this is of some use.

Michael

[Synchrodyne]

The updated (and simplified) Wireless World Quality Amplifer designs were also described in an article in the January, 1946 edition. There were 4 and 8 watt (both PX4) and 12 watt (PX25) versions.

Cheers,

[Valvepower]

Hi All,

I’ve just stumbled on this thread.

I repaired and tidied up one of these amplifiers about 8 years ago.

There was a question in a previous post regarding the setting of the balance pot feeding one of the PX4 via the 6SN7 pin 4. The method I used was to adjust this pot by measuring the THD (set to get a minimum) at a few dB below the clip level in to a 15-Ohm load. This was perfomed after roughly matching the positive and negative half’s of the output wave form using an oscilloscope.

The other aspect I liked was the tone control; where the signal via a high and low pass filter network on the o/p of the 6J7, are added together via level controls, and fed to a mixer stage, where anodes of the 6SN7 are coupled together. The 1M2 provides a signal path for the wide band signal, and the two controls adjust the LF and HF levels.

Sadly I sold this amplifier, its one of those sales I regret - ho hum .

Regards.
Terry.

[msaunby]

As there is almost no information about Sound Sales on the web I've created a couple of pages on one of my sites. I'll put what information I have as I get around to scanning it. Not much yet, but here's the link -

http://mike.wepoco.com/retro-geekery/sound-sales

[rayemond444]

Hi, I have one very similar to this with pointed knobs on front, case in very nice condition, shame i have no space for it.

[Station X]

rayemond444. If you want to sell your amplifier please post in the Sets and Parts Offered section.

[RogerWalker]

Sorry to come to this thread so late - but in the picture, it looks like one PX4 and one PX25 are fitted - this would explain the different currents?

Roger Walker