Revox A77 MK3 frequency response

[Gerdy Revox]

Looking for help with the restoration of a Revox A77 MK3 van Oct 1973 (ser. nr. G72835):
-2 track
-9,5 & 19cm/s
-extremely low playing hours (tens of hours), confirmed by owner, machine has been stored for decades in a garage under the owner’s house.
-no visible head wear.
-fully recapped, new trimmers, new tape path bearing, voltage levels and capstan speed adjusted.

-Syslevel set (Aux input -> Cinch Output) to 1V = 0dB.
Note: Syslevel was originally set to 1,55V but to support the newer tapes with higher magnetisation it is set to 1V.

-Problem 1: repro with test tape 320nW/m.
Playback Amps adjusted with test tape for Cinch output 1V/0dB (1000Hz).
Output level checked with test tape tracks at -10dB at 1KHz/10KHz/14KHz/18KHz but ....
At 19cm/s NAB I get the following results for CH1 / CH2 (these are relative dB results vs. the expected -10dB):
400Hz: +1 / +1
1000Hz: +0,5 / +0,5
3150Hz: 0 / -0,5
10000Hz: +0,5 / +1
14000Hz: -1 / 0
18000Hz: -3 / -1 (CH1 lets me down a bit)
A roll-off at higher frequencies on CH1 only.

With the same test tape (recorded at 19cm/s but played at 9,5cm/s which is not 100% correct but I wanted to see how the playback goes as frequency rises, I get half the recorded frequency of course):
500Hz: +0,5 / +0,25
1575Hz: 0 / -1
5000Hz: -1 / 0
7000Hz: -3 / -1,5 (goes in the wrong direction)
9000Hz: -5 / -3 (and even more ... )
A serious roll-off at higher frequencies.

Notes:
-Playback head azimuth is adjusted, heads cleaned and demagnetised.
-I then slightly adjusted the heads a little bit up and down with the 2 adjustment screws, no improvement so it seems the factory setting was correct and I put it back there.

Then put 2 other working Playback Amps in (of my working MK4), I get a slightly better result but nothing spectacular (on these cards, as opposed to the MK3 Playback cards, the factory did not assemble C807/ R812 and R811 is 3K9 (instead 4K7), according to the remark in the Playback Amp schematic).
Also Cx (across the Playback head at 9,5cm/s) is not assembled, not in the MK3 and not in the MK4 machine.

So I’m thinking direction Playback heads, but these are as good as new

Resistance measured (in the circuit): 13,6 Ohm recording head & 87 Ohm playback head. This looks OK. I then demagnetized the heads again to continue.
I then did the same mod on the MK3 cards as on the MK4 cards, nearly no improvement.
Running out of options I decided to try to adjust the recording.

-Problem 2: recording.
Recording level (NAB output) on the Recording Amps adjusted for 0dB/1000Hz (=1V).
Then generator to -20dB, checked 1000Hz and level OK.
Next Bias adjustment at 10KHz/-20dB for both speeds (then a couple of times back and forth between the 1000Hz and 10KHz because the settings influence each other).
Also azimuth recording head adjusted.
For both speeds I do not even reach -20dB/10KHz, whereas all recorders I adjusted until now can be adjusted to even -15dB which allows for overbias to get the output back to -20dB.
I lack 2dB at 19cm/s and 4dB at 9,5cm/s, even more at higher frequencies (CH1 is always worse).

I adjusted the EQ (was set to min. for both speeds) in an attempt to reach -20dB/10KHz.
The EQ trimmer at 19cm/s is at 1/3 but at 9,5cm/s it is at 80% for CH2 and even at 100% for CH1. I never saw that before.
A bit worried I checked the THD (400Hz) and to my surprise it is well within spec (and this without overbias??).

Listening before/after tape gives very little difference at 9,5cm/s and none at 19cm/s.
But my gut feeling tells me something is wrong, I just can’t put my finger on it.

This extremely little used machine should have super specs but it does not.
I’m stuck, anybody who had a similar experience?

[knobtwiddler]

If you posted this info about certain Japanese machines that have 'hard ferrite' heads, I'd suggest getting a microscope and looking for 'gap erosion' along the gap. This can occur with very little use and it tends to be specific to certain types of head. I am not aware of it happening on the A77 (or any Revox / Studer deck), but it might be worth digging around the Internet and learning a little about what to look for.

nb - just to be clear: gap scatter / erosion can be seen on all head types. However, due to the extremely brittle nature of some 'hard ferrites' ("hot pressed ferrite" was the Technics type, ironically developed in order to make heads more immune to wear from general use), you can see it with very low use.

and another thing to be aware of: heads can be 'lapped', making them look like new, but having lower inductance and thus less HF extension. Not suggesting this is the case here, but it's worth knowing.

and another edit: a simple test would be to disconnect the PB head, measure inductance and compare against a new A77 head (A77 is popular, so inductance value will be easy to find).

[Audiophile_AU]

Hi Gerdy,
I had a thought - what are you using to measure the audio output voltages? I have found that digital multimeters have a limited frequency response for audio. They are accurate at power line frequencies, but 10kHz, 20kHz
May I suggest that you connect the measuring voltmeter directly to the audio generator and see if you get consistent results up to 20kHz and beyond.
A conventional (analogue!) audio mVolt meter (high input impedance) would be my recommendation.
Them's my thoughts on the start of a 40 deg C day down under...
best regards
David

[John_BS]

Looking at the half-speed results I'd be a bit suspicious of the test tape.