Ferrograph F307 Stereo Amplifier

[DMcMahon]

Very recently purchased a Ferrograph F307 amplifier and it is quite reasonable condition. Gave it a quick test yesterday and it sounds good but have so far only tested on headphones.

Opened it up to double check it was a Mk. 2 which it was, loads of dust inside so cleaned.

I see online reviews where some people say it is light on bass especially if not driving into 15 ohm speakers and saying that bass performance can be improved by increasing the output capacitance, I doubt I would to do that unless the bass performance was really poor.

It will match my wood cased Ferrograph Series 7 R2R and also provide an incentive to find/make time to get back to progressing my Series 7s.

[Ted Kendall]

Electrolytic coupling capacitors cause excess distortion long before causing bass roll-off - see Self for a blow-by-blow account, but if the signal voltage across the cap is greater than about 80mV at the lowest frequency of interest, distortion rises.

[Radio Wrangler]

They may well have gone a bit high in ESR by now as well. Output coupling capacitors often were a bit parsimonious on the microfarads.

David

[DMcMahon]

A good word parsimonious, I had to Google it

[DMcMahon]

Originally Posted by Radio Wrangler

They may well have gone a bit high in ESR by now as well. Output coupling capacitors often were a bit parsimonious on the microfarads.

The output capacitors are 800uF & 250uF wired in parallel, so giving a theoretical value around 1,050uF for each channel, all rated at 40 volts.

Would you have any typical ballpark figures for ESR for these sort of values of capacitors, i.e. what ESR figure might be considered not good.

Later out of interest I will measure the capacitance (both parallel and individual) and the ESR.

[DMcMahon]

Have now tested on a pair of speakers (B & W 683 S2) sounds good with plenty of bass.

Intermittently would lose one or both channels, seems to be a dodgy push button selector switch for Stereo/mono left/mono right, intermittently bad in all 3 positions, will try switch contact clean later.

[DMcMahon]

Measured the output capacitors:-

In circuit parallel combination of left channel C700 + left C500 = 1, 189uF, ESR 0.09 ohm.

In circuit parallel combination of right channel C701 + right C500 = 1, 106uF, ESR 0.08 ohm.

Out of circuit (off load) left channel C700 = 944uF, ESR 0.09 ohm, left C500 = 225uF, ESR 0.33 ohm.

Out of circuit (off load) right channel C701 = 860uF, ESR 0.09 ohm, right C500 = 228uF, ESR 0.37 ohm.

So overall nothing obviously worrying. However the left channel C700 has a small rupture to the top can, photo shows it after cleaning up a small amount of leaked electrolyte.

[DMcMahon]

After a spray of contact cleaner to the stereo/mono left/mono right selector switch the switch is now behaving itself, no more loss of sound being experienced.

Planning to replace the left channel ruptured C700 with a new 1,000uF 50 volt, difficult to find new ones rated at 800uF.

Will first check out the frequency response at the bass end and stability/distortion, then recheck with the new capacitor fitted.

[DMcMahon]

Before replacing C700 decided to test the low frequency response so that could compare pre and post results.

Checked both channels, all looked good down to 20Hz, signals clean and stable, I think there is > 80mV voltage drop across both C700/C500's but unable to check distortion as do not have suitable equipment.

While disconnecting the scope lead from C700 on the right channel I dropped the lead and its earth lead shorted out on C700 and took out the 1 amp output fuse !!!

Replaced the fuse but right channel is dead. Checked the associated two 0.35 ohm wire wound resistors R501 & R715 (which could act as fuses) but they are good. Next to check the two 16020 power output transistors and the two 40360 driver transistors .

[Radio Wrangler]

Transistors are much faster than fuses. Famous for it.

This is a solid-state power amplifier. There is a big pitfall for the unwary repairer. You take one transistor out at a time and test it. You find a bad transistor and replace it. You try the amplifier... no joy.... so you resume testing transistors where you left off. You continue until you find another dead transistor and replace it... you try the amp again, no joy..... back to where you left off resting transistors. Eventually you find you've tested them all. You may even have replaced them all, one at a time, and still the wretched thing doesn't work.

If you go round testing the transistors again, starting from the beginning, you find dead transistors that were OK last time. Even one of your new ones is now dead.

Huh?

The DC coupling between stages allows one dead transistor to kill others. When you found your firs dead one and replaced it, it's likely that a dead one you hadn't got to yet killed your new one at the instant of switch on. You assume the new one must be good because it's new, right?

In this way you can get stuck in a never-ending loop changing the same things over and over again.

The way out is to test all the transistors (and being a fully paid up suspicious *******, measure the resistors and diodes as well) and only when you're really sure you've found and replaced all the fatalities dare you try it with power on.

The urge to take short cuts is strong, but don't.

This is why solid state amplifiers have a bit of a reputation for being hard to fix.

David

[Boulevardier]

Originally Posted by Radio Wrangler

Transistors are much faster than fuses. Famous for it.

This is a solid-state power amplifier. There is a big pitfall for the unwary repairer. You take one transistor out at a time and test it. You find a bad transistor and replace it. You try the amplifier... no joy.... so you resume testing transistors where you left off. You continue until you find another dead transistor and replace it... you try the amp again, no joy..... back to where you left off resting transistors. Eventually you find you've tested them all. You may even have replaced them all, one at a time, and still the wretched thing doesn't work.

If you go round testing the transistors again, starting from the beginning, you find dead transistors that were OK last time. Even one of your new ones is now dead.

Huh?

The DC coupling between stages allows one dead transistor to kill others. When you found your firs dead one and replaced it, it's likely that a dead one you hadn't got to yet killed your new one at the instant of switch on. You assume the new one must be good because it's new, right?

In this way you can get stuck in a never-ending loop changing the same things over and over again.

The way out is to test all the transistors (and being a fully paid up suspicious *******, measure the resistors and diodes as well) and only when you're really sure you've found and replaced all the fatalities dare you try it with power on.

The urge to take short cuts is strong, but don't.

This is why solid state amplifiers have a bit of a reputation for being hard to fix.

David

And that is the "curse" of DC-coupled circuitry. I remember, from when it first became common in consumer equipment, thinking "how on earth are you supposed to fault-find this stuff?" Given the time involved in testing all or many of the components on a PCB, plus the damage that can be done by repeated removal and reinsertion of parts on an increasingly fragile PCB, it is nearly always more realistic to just replace the whole board. And that is always assuming that the root of the fault isn't on another connected board.

Mike

[DMcMahon]

Thank you guys, duly noted, I had a feeling that it was not going to be as easy as it ought to be.

Have removed the power amp board for full passive component testing.

The two 16020 power output transistors (on chassis heatsink) check out OK.

David

[Herald1360]

Something (a resistor?) looks very fried in the middle of that PCB!

At least there's not too many active devices to check.....

[Radio Wrangler]

Originally Posted by Boulevardier

And that is the "curse" of DC-coupled circuitry. I remember, from when it first became common in consumer equipment, thinking "how on earth are you supposed to fault-find this stuff?" Given the time involved in testing all or many of the components on a PCB, plus the damage that can be done by repeated removal and reinsertion of parts on an increasingly fragile PCB, it is nearly always more realistic to just replace the whole board. And that is always assuming that the root of the fault isn't on another connected board.

It can make economic sense, depending on what you pay for replacement boards. Probably so for a manufacturer's service operation, less likely if faced with massively marked up prices from some 'profit centre'

Certainly not worthwhile nowadays.

Also not very educational.

You may just want to fix it properly as a matter of personal pride.

If you're involved in the development of a genuinely new circuit, you can run it from current limited supplies, you can break it down to the minimum and add stages until running. You can add test jig circuitry to cover from stages taken out of circuit.

I've developed a few solid state audio power amps and had to develop methods to run them in parts and to really deduce back to root cause problems. Also on a newly designed proto you don't have the help of knowing one ever worked. This leaves you feeling high and dry.

But it can be done. You merely have to get inventive.

Is this extent of trouble worth it to fix one amplifier? Yes if you want the education. Yes if you can't simply replace it. Yes if you want to know the cause so you can take precautions next time. No if you expect someone to pay a fair rate for your time (and at a level appropriate to the skills).

David

[Radio Wrangler]

Not many active devices. Room to swing a cat. Eeeee! Luxury, lad!

David

[DMcMahon]

Originally Posted by Herald1360

Something (a resistor?) looks very fried in the middle of that PCB!

At least there's not too many active devices to check.....

Yes saw the burnt resistor a couple of weeks back, it is R506 (160 ohm) and is in parallel with RV518 which is the output quiescent current setting pot. It measures around 180 ohms, its value is not really critical as in parallel with the pot, I have set up the quiescent current recently and will replace the resistor when I get one then recheck the current.

So far everything looks OK apart from driver transistor VT501 a 40360, which has shorted emitter to collector junction, have ordered a couple.

[DMcMahon]

Everything on the right channel power amp board checks out OK apart form the shorted driver transistor VT501, hopefully a new replacement will get it up and running again.

[DMcMahon]

Originally Posted by Radio Wrangler

You may just want to fix it properly as a matter of personal pride.

David

Yes would always do everything possible to fix it myself without resorting to replacement board etc.

David

[DMcMahon]

Replaced the shorted driver transistor VT501 (40360) but the right channel is still dead. Disappointing as that was the only thing that measured bad, well back to the drawing board.

David

[Boulevardier]

Originally Posted by DMcMahon

Replaced the shorted driver transistor VT501 (40360) but the right channel is still dead. Disappointing as that was the only thing that measured bad, well back to the drawing board.

David

Has the new transistor survived the assignment intact?

Mike