Revox A77 Replay Amp - Why Use a 1600uF Electrolytic?

[Humber888]

I have been converting a couple of Revox A77 replay amplifier cards back from 7.5/15 ips to 'normal' 3.75/7.5 ips boards for use in a Dolby A77. This has been done but I got to wondering why Revox used a 1600uF electrolytic capacitor in the preamp feedback circuit rather than a more standard 1500uF. I attach a copy of the circuit and the item in question is C803 on the left.

I would guess that the original Frako type FPF capacitors were something like -50% +100% tolerance (can't find any data on them), so you would think that a 1500uF would be good enough. Maybe I'm missing something critical in the circuit. Nagravox replace the originals with two 3300uF radial ones in series, but my proposal is to use new Vishay/Philips 1500uF type 021 long life axials and just check their capacitance is not more than about 10% low. Any comments anyone?

Mike

[Restoration73]

The "old" system related to multiples of 2, e.g. 4, 16, 640, 32000 uF etc. I believe this is related to the capacitor time constant, 1 x RC = 64% charged, 4-5 x RC = 100% charge.

[Radio Wrangler]

It's in the DC feedback path governing the bias of the transistors. It's job is to prevent any signal passing down this route. A crude analysis takes the 68 Ohm resistor R806 as the source impedance, along with the 1600uF means that unwanted signal feedback is only attenuated by 3dB at about 1.5Hz, so there is a reasonable amount of attenuation by 20Hz so the bass response isn't robbed. That explains the surprisingly high value part.

Given the tolerances in the past, and the tolerances we get today a modern 1500 uF part will be fine, and the modern part's lower ESR will help, too.

David

[radiotechnician]

The thing that bothers me about that circuit, is the play head has a small amount
of DC current flowing through it. When new , probably too small to be important.

However in my experience in past years, on early solid state machines, problems
(ground loops) caused magnetization of play heads to add noise to a tape, every time
it was played. Quite difficult to diagnose !

Those BC 109s are decades old. Perhaps the original parts were selected for some
property, i.e. not turning into batteries.

[Humber888]

Thanks to all for the comments - looks like I assumed right that 1500uF would be OK. In response to Steve's comment about dc through the replay head, the manual does specifically acknowledge this. It states: 'The feedback current is less than 1uA which avoids magnetization of the playback head'. Would age have any impact on this statement, other than implying that you should demagnetise your heads every so often?

Mike

[Radio Wrangler]

The first BC109 runs at a bit under 100uA collector current. If it is ever replaced, the new part needs to be chosen not only fo low noise but also for Hfe >100 at this current

The bias currents and temperatures are too low for dopant or metal migration to be a worry, so I don't expect ageing to be an issue.

In circuits with electrolytic capacitors as DC blocks on inputs from tape heads people don't question whether there is any DC flowing, but then they don't check the capacitor's leakage specs, and capacitors do age.

David

[Humber888]

David is right about the first BC109. It is specified as a BC109C (high Hfe) in the parts list.

Mike

[radiotechnician]

Head , (and every item capable of magnetization in contact with the tape)
was routinely done back when tape machines were in vogue. It was usually
part of the cleaning kit chores.

These days careless use of the 50 Hz tool can induce very high voltages in
the head, so some thought is required.