Revox A77 biasing question

[Robert Stevens]

I have a modded 1/2 track A77 that runs at 7.5/15ips. It used to be a normal A77 but the capstan was replaced for high speed operation. All this happened before it came into my posession.

When I try to set the bias, I run out of trim pot before the level starts to drop. Basically, it just keeps going up until i run out of trimpot. I've tried normal bias tapes (RMGI) and high bias (499). Same results.

Both my other machines (PR99, A800) behave normally (i.e. I can overbias) with both these formulations.

Not sure what to make of it. Any ideas?

[[email protected]]

On Monday, May 8, 2017 at 1:48:46 PM UTC-4, Robert Stevens wrote:
> I have a modded 1/2 track A77 that runs at 7.5/15ips. It used to be a normal A77 but the capstan was replaced for high speed operation. All this happened before it came into my posession.
>
> When I try to set the bias, I run out of trim pot before the level starts to drop. Basically, it just keeps going up until i run out of trimpot. I've tried normal bias tapes (RMGI) and high bias (499). Same results.
>
> Both my other machines (PR99, A800) behave normally (i.e. I can overbias) with both these formulations.
>
> Not sure what to make of it. Any ideas?

Anything?....

http://www.tapeproject.com/smf/index.php?topic=1445.0

Jack

[Robert Stevens]

Not really, I set up playback with 250nwm MRL and then record as per service manual, then I start biasing with my console vu at -10db at 10K and bias trim pot fully anti-clockwise. As I turn clockwise, the vu level keeps going up, up, up until I am out of trim pot. I'm pretty good with setting up all my other machines, so just don't know what the deal with this one is.

> Anything?....
>
> http://www.tapeproject.com/smf/index.php?topic=1445.0
>
> Jack

[PStamler]

> Not sure what to make of it. Any ideas?

Which RMGI tape are you using? 911 requires a higher bias level than 468.

Peace,
Paul

[Mike Rivers[_2_]]

On 5/8/2017 1:48 PM, Robert Stevens wrote:
> When I try to set the bias, I run out of trim pot before the level
> starts to drop. Basically, it just keeps going up until i run out of
> trimpot. I've tried normal bias tapes (RMGI) and high bias (499).
> Same results.

Have you considered the possibility that there's something wrong with
it? It could be that the bias oscillator isn't putting out enough level
or the capacitor that couples it to the heads is shot, or the speed or
record switch contacts are dirty. Or maybe the waveform is distorted.

It's been a long time since I've been inside an A77, but I know I set at
least one up for Ampex 456 or 3M 226 and there was no problem setting
bias and EQ for good performance.

How are adjusting the bias? At 1 kHz? at 10 kHz? You should be adjusting
it at 10 kHz, for about a 2 dB drop over peak. Do you have a service
manual and an oscilloscope? You can download a service manual here:
https://elektrotanya.com/revox_a77.pdf/download.html You can't download
an oscilloscope. <g>



--

For a good time, call http://mikeriversaudio.wordpress.com

[Scott Dorsey]

Robert Stevens > wrote:
>I have a modded 1/2 track A77 that runs at 7.5/15ips. It used to be a normal A77 but the capstan was replaced for high speed operation. All this happened before it came into my posession.

Okay, the A77 could be ordered as a high speed or a low speed machine. If you
had a low speed machine, there were two ways to do the modification, one of
which involved totally replacing the motor assembly and the other just involved
replacing the motor and capstan and kind of jamming it into place so it sort
of works.

In either case, the electronics need to be changed so that the low speed
time constants are replaced with high speed ones.

Now... most of the low speed machines were equipped with quarter-track heads
instead of half-track. So in most cases when people did the modification
to high speed they would replace the heads as well.

So, there are a whole lot of things that can go wrong with this sort of
modification. People don't use the right heads, they get the head wiring
wrong, they screw up the time constants so that it lines up on three tones
but when you play the tone ladder back the levels are all over the place.

>When I try to set the bias, I run out of trim pot before the level starts to drop. Basically, it just keeps going up until i run out of trimpot. I've tried normal bias tapes (RMGI) and high bias (499). Same results.

I don't know about 499, but the A77 should have no problem setting bias for
RMGI 468 if the heads are clean and the gaps not clogged up, if the heads
are correct, if there is good tape to head contact, and if the bias oscillator
is working properly.

So, check the part numbers on the heads and make sure they are the original
A77 heads. Scrub the heads hard with xylene. Make sure there is no ridge
on the tips of the heads. Try pressing the tape against the heads with your
thumb; if the meter levels jump you have a head contact issue.

>Both my other machines (PR99, A800) behave normally (i.e. I can overbias) with both these formulations.
>
>Not sure what to make of it. Any ideas?

It's either electronic or mechanical. Some poking and prodding with thumbs
can rule out mechanical issues. Electronic ones mostly center around the
heads as well.

Check the voltages on the bias oscillator against the manual. Check the
waveform on the erase head and make sure it's a nice clean sine wave.

But _because_ this is a Frankenstein machine, you can't assume that it ever
worked properly, so you have to spend some time going step by step trying to
figure out what the guy doing the modifications did.

There is a Studer mailing list. They are likely to look down on the lowly
A77, but you might find some people there. If this were 1975, I'd be sending
you to Jon Hall at Hall Electronics but I don't know if he is even still
alive.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

I do have service manual, and a 20mhz scope. My tech skills, however, are suspect. I bought a couple different oscillator cards but none of them fixed the problem. You are most likely right that there is something wrong. Thx.

[Robert Stevens]

I got the machine modded, but I had to go through all the IEC modifications to get the EQ right. It was super trebley when i got it. Much better now, but I have never got the bias right. The machine sounds good to my ears, but still have never been able to overbias it like i can on my PR99 and my A800. It just annoys me i guess.

[Robert Stevens]

On Monday, May 8, 2017 at 3:46:07 PM UTC-4, PStamler wrote:
> > Not sure what to make of it. Any ideas?
>
> Which RMGI tape are you using? 911 requires a higher bias level than 468.
>
> Peace,
> Paul

I'm using RMGI 911 standard bias tape for this machine.

[Robert Stevens]

On Monday, May 8, 2017 at 9:57:01 PM UTC-4, Robert Stevens wrote:
> On Monday, May 8, 2017 at 3:46:07 PM UTC-4, PStamler wrote:
> > > Not sure what to make of it. Any ideas?
> >
> > Which RMGI tape are you using? 911 requires a higher bias level than 468.
> >
> > Peace,
> > Paul
>
> I'm using RMGI 911 standard bias tape for this machine.

And while I will defer to your experience, the website for the current manufacturer says the 468 is high bias, whereas the 911 says normal bias.

http://www.rmgi-usa.com/page4/page4.html

I'm guessing, they don't really know what they are producing...

[Trevor]

On 9/05/2017 5:55 AM, Mike Rivers wrote:
> How are adjusting the bias? At 1 kHz? at 10 kHz? You should be adjusting
> it at 10 kHz, for about a 2 dB drop over peak. Do you have a service
> manual and an oscilloscope? You can download a service manual here:
> https://elektrotanya.com/revox_a77.pdf/download.html You can't download
> an oscilloscope. <g>

Wrong, you can download plenty of oscilloscope programs you can run on
your PC for simple audio purposes like that!

Trevor.

[Mike Rivers[_2_]]

On 5/9/2017 12:41 AM, Trevor wrote:
> Wrong, you can download plenty of oscilloscope programs you can run on
> your PC for simple audio purposes like that!

I figured that someone who never used an oscilloscope for
troubleshooting would say that. What are you going to do about probes?
How about input level adjustment? Calibration? How about looking at the
waveform of a 100 kHz bias oscillator with your 44.1 kHz sample rate?

There are, indeed, functional computer-based scopes, but they don't use
your computer's sound card as the input device.

You can buy an old oscilloscope for less than $25 that will work just
fine for troubleshooting analog audio equipment.


--

For a good time, call http://mikeriversaudio.wordpress.com

[geoff]

On 9/05/2017 10:49 PM, Mike Rivers wrote:
> On 5/9/2017 12:41 AM, Trevor wrote:
>> Wrong, you can download plenty of oscilloscope programs you can run on
>> your PC for simple audio purposes like that!
>
> I figured that someone who never used an oscilloscope for
> troubleshooting would say that. What are you going to do about probes?
> How about input level adjustment? Calibration? How about looking at the
> waveform of a 100 kHz bias oscillator with your 44.1 kHz sample rate?
>
> There are, indeed, functional computer-based scopes, but they don't use
> your computer's sound card as the input device.
>
> You can buy an old oscilloscope for less than $25 that will work just
> fine for troubleshooting analog audio equipment.
>
>


..... and hook up to signals from 1mV per segment to 5v, or more with a
BNC 10X probe.

But a computer soundcard input can at a pinch for many audio
applications. Just get your grounding and coupling right !

geoff

[Scott Dorsey]

Robert Stevens > wrote:
>I do have service manual, and a 20mhz scope. My tech skills, however, are suspect. I bought a couple different oscillator cards but none of them fixed the problem. You are most likely right that there is something wrong. Thx.

If you swapped the whole oscillator card and it didn't do any good, now you
have ruled the oscillator card out.

So, all that is left are the heads and the power supply. And the power supply
would be causing plenty of other issues that you'd be noticing too.

See if improved tape-head contact will do the job and make sure you have the
correct heads for the machine. If they grafted some Tandberg 64 heads in
there, it will kind of work, but not really....
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Scott Dorsey]

Robert Stevens > wrote:
>I got the machine modded, but I had to go through all the IEC modifications=
> to get the EQ right. It was super trebley when i got it.

If you had the machine modified by some guy, and he sent it back without the
equalization constants changed, then he was a butcher and you should be very
suspicious of the rest of his work.

What did you have this guy modify?

>Much better now, =
> but I have never got the bias right. The machine sounds good to my ears, =
>but still have never been able to overbias it like i can on my PR99 and my =
>A800. It just annoys me i guess.

Annoys you? It makes the machine useless because you can't calibrate it
properly. Whoever did this modification for you should be ashamed.

What exactly did he do and where did he get the parts? If he didn't change
the motor AND the motor mounts, just scrap the machine because it will never
work properly and it is not worth the time and effort to undo the butchery.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Scott Dorsey]

Robert Stevens > wrote:
>On Monday, May 8, 2017 at 3:46:07 PM UTC-4, PStamler wrote:
>> > Not sure what to make of it. Any ideas?
>>
>> Which RMGI tape are you using? 911 requires a higher bias level than 468.
>
>I'm using RMGI 911 standard bias tape for this machine.

There is no such thing as "standard bias" tape. 911 was designed to be
"bias compatible" with Ampex 456 but really it's not exactly since no two
batches of 456 had the same bias point anyway.

But 911 should bias up fine on a clean and stock A77.
911 is slightly higher bias than 468, but still within the range that
an A77 can handle.
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Phil Allison[_4_]]

Mike Rivers wrote:
--------------------
> Trevor wrote:
>
> > Wrong, you can download plenty of oscilloscope programs you can run on
> > your PC for simple audio purposes like that!
>
>
> I figured that someone who never used an oscilloscope for
> troubleshooting would say that. What are you going to do about probes?
>

** Yep - you need a 10:1 probe with no more than 10pF capacitance.




> How about input level adjustment? Calibration? How about looking at the
> waveform of a 100 kHz bias oscillator with your 44.1 kHz sample rate?
>

** That could very well be a deal breaker.....


> There are, indeed, functional computer-based scopes, but they don't use
> your computer's sound card as the input device.
>

** 100% correct.


> You can buy an old oscilloscope for less than $25 that will work just
> fine for troubleshooting analog audio equipment.
>
>

** Ain't that damn shame....





...... Phil

[Robert Stevens]

"What did you have this guy modify? "

I do not know who made the modification. I bought it this way pretty cheap. I've got it working quite well after the EQ mods and it works fine for a playback only machine as is. But I'd like to get the record side working on it too.

To me, it looks like the only thing modded, btw, was the capstan motor was replaced completely. Which to my understanding, is the only real correct way to do this mod. Getting a fatter capstan and glueing it on causes all sorts of tape wrap issues.

But perhaps, I am mistaken and that is actually what the problem is. I'll futz around with the tape wrap tonight and see what happens.

Thanks for all the ideas!

[Robert Stevens]

Also, sorry for the bad english. When I said, "I got it modded", I meant "When I got the machine, it was already modded".

[[email protected]]

On Tuesday, May 9, 2017 at 10:15:39 AM UTC-4, Robert Stevens wrote:
> Also, sorry for the bad english. When I said, "I got it modded", I meant "When I got the machine, it was already modded".

also check that the part number for the erase head is correct.

the wrong Z erase head can place an excessive load on the bias osc lowering the bias level making it not able to peak.

or more basic, check the DC supply voltage to the bias osc.

m

[[email protected]]

On Monday, May 8, 2017 at 1:48:46 PM UTC-4, Robert Stevens wrote:
> I have a modded 1/2 track A77 that runs at 7.5/15ips. It used to be a normal A77 but the capstan was replaced for high speed operation. All this happened before it came into my posession.
>
> When I try to set the bias, I run out of trim pot before the level starts to drop. Basically, it just keeps going up until i run out of trimpot. I've tried normal bias tapes (RMGI) and high bias (499). Same results.
>
> Both my other machines (PR99, A800) behave normally (i.e. I can overbias) with both these formulations.
>
> Not sure what to make of it. Any ideas?

Hope you get it rectified. Just interested about this "bias", it was totally different than what I once thought, so I value your question that gets me curious.

Jack

[Robert Stevens]

So with a digital voltmeter, I put the leads across the record head left channel 1. With no audio and bias signal at full pot I get 1.9V. On the right channel 2 at full pot I get 2.4V. On the 468 tape, I can see it just about wanting to overbias on channel 2. I get about 1/4 to 1/2 db drop. On channel 1, long way to go before it starts to overbias.

So I think I know it is an electronic issue now. I really need to get both channels a little higher than 2.4v.

I did try pressing the tape into the head with a Q-tip, that yielded no real visible results.

I have some more tests that have been suggested to run tomorrow, thanks for all the ideas.

FWIW, On a similar 4track a77 running at normal speed 7.5ips, the voltages are about 2.2V on each channel. I set it up about 3 months ago or so and overbiasing it was easy with 468 tape. The 1/4 track uses a different bias card though. Yellow dot vs red dot. Not sure the internal differences.

Also, I have 3 A77 1/2 track red dot bias cards. Voltages are the same with any of them, so not likely the oscillator cards themselves.

Looking forward to finally figuring this out, it's been bugging me for quite some time now.

[Scott Dorsey]

Robert Stevens > wrote:
>So with a digital voltmeter, I put the leads across the record head left ch=
>annel 1. With no audio and bias signal at full pot I get 1.9V. On the rig=
>ht channel 2 at full pot I get 2.4V. On the 468 tape, I can see it just ab=
>out wanting to overbias on channel 2. I get about 1/4 to 1/2 db drop. On =
>channel 1, long way to go before it starts to overbias.=20

What kind of digital voltmeter and what is its response at 100 KHz?

>So I think I know it is an electronic issue now. I really need to get both=
> channels a little higher than 2.4v. =20

Oh, I am sure it's higher. You need a VTVM with an RF probe to measure that,
or just use a scope.

The fact that it's different between the two channels is alarming though
and makes me worry about the head condition. Head gaps not starting to
open up, are they?

>I did try pressing the tape into the head with a Q-tip, that yielded no rea=
>l visible results.

Don't use a Q-tip, use your thumb. You're going to need some pretty heavy
pressure.

And scrub the hell out of the heads with xylene. Again, don't use a dinky
little q-tip, use a big rag and a lot of elbow grease.

Inspect the heads with a magnifier to make sure the gaps are clean and open.

>FWIW, On a similar 4track a77 running at normal speed 7.5ips, the voltages =
>are about 2.2V on each channel. I set it up about 3 months ago or so and o=
>verbiasing it was easy with 468 tape. The 1/4 track uses a different bias =
>card though. Yellow dot vs red dot. Not sure the internal differences.

Don't trust RF measurements made with a cheap DMM.

>Also, I have 3 A77 1/2 track red dot bias cards. Voltages are the same wit=
>h any of them, so not likely the oscillator cards themselves.

Yes, I said that. So are the heads original?
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Phil Allison[_4_]]

Robert Stevens wrote:
>
> So with a digital voltmeter,

>

** Giant HUH ??

You have a 20MHz scope, why are you not using it ?

You also need a 10:1 probe ( to minimise capacitance loading ) for that job and the voltage you should see is more like 25V rms than 2.4V.



...... Phil

[Mike Rivers[_2_]]

On 5/9/2017 8:26 PM, Robert Stevens wrote:
> So with a digital voltmeter, I put the leads across the record head
> left channel 1. With no audio and bias signal at full pot I get
> 1.9V. On the right channel 2 at full pot I get 2.4V. On the 468
> tape, I can see it just about wanting to overbias on channel 2. I
> get about 1/4 to 1/2 db drop. On channel 1, long way to go before it
> starts to overbias.

Hmmm . . . you really should be measuring with an oscilloscope so you
know what you're actually measuring. The bias frequency is 120 kHz and
most digital multimeters poop out above a few kHz.

If you look at the service manual (if it's the same one I have) you'll
see that you should have roughly 500 mV measured from the low side of
the head to ground - the green wire on channel 1 and yellow wire on
channel 2. If you're measuring around 2 V with your meter on AC, that's
way too much, and it's a pretty amazing meter. If you're measuring
around 2 V DC, well, I really don't understand that.




--

For a good time, call http://mikeriversaudio.wordpress.com

[Robert Stevens]

It's a fluke. Not sure the model. It's on AC. I do have a scope, I'll dig it out tomorrow.

I was measuring across the head. Both wires on each channel. I'll try the green to ground and yellow to ground tomorrow. Thx.

I don't actually see that info in the service manual. What page was it on.?

[Phil Allison[_4_]]

Scott Dorsey wrote:

----------------

> Oh, I am sure it's higher. You need a VTVM with an RF probe to measure that,
> or just use a scope.
>

** Oddly enough, an old fashioned analog multimeter will likely work fine at 100kHz.

They normally use a single germanium diode rectifier and my own 20kohms/V meter is good to over 250kHz.


...... Phil

[Trevor]

On 9/05/2017 8:49 PM, Mike Rivers wrote:
> On 5/9/2017 12:41 AM, Trevor wrote:
>> Wrong, you can download plenty of oscilloscope programs you can run on
>> your PC for simple audio purposes like that!
>
> I figured that someone who never used an oscilloscope for
> troubleshooting would say that.

:-) :-) I have 3, and used hundreds over the last 40 years or so!


> What are you going to do about probes?

Standard probes work fine with adapters.

> How about input level adjustment?

An external amplifier or attenuator is required.

>Calibration?

Use the PC output and a voltmeter if necessary.

> How about looking at the
> waveform of a 100 kHz bias oscillator with your 44.1 kHz sample rate?

Well you can't do that, I don't remember saying a PC solution replaced
ALL uses of a CRO though? What the PC solution gets me is lower
distortion, less noise, and higher resolution measurement capabilities
than any CRO I can afford at home myself, for AUDIO. Plus Spectrum
analyser, storage, print out capabilty and lots more cheap CRO's don't
have.
And 192kHz sample rate actually becomes useful. :-)

>
> There are, indeed, functional computer-based scopes, but they don't use
> your computer's sound card as the input device.

And then there are, but yes you usually need a buffer amp.

>
> You can buy an old oscilloscope for less than $25 that will work just
> fine for troubleshooting analog audio equipment.

Yep, and you can even buy a brand new one for about $100 that will do
some jobs. But not as good for many others. And vice versa. See it
really helps when you actually know your instrument capabilities and
requirements and which to choose for what purpose.

However you do seem to miss my *ONLY* comment completely, and that was
you were wrong when you said you can't download an oscilloscope! The
rest of your post is simply a straw man.

Trevor.

[Mike Rivers[_2_]]

On 5/10/2017 6:50 AM, Trevor wrote:
> However you do seem to miss my *ONLY* comment completely, and that was
> you were wrong when you said you can't download an oscilloscope!

What you're describing is a series of things that you might or might not
need in order to create some of the capability of an oscilloscope using
a downloaded program.

I stand by my statement that you can't download something and have an
oscilloscope unless you add other things to it that, if you had to buy
them, would probably cost more than a surplus oscilloscope that's
perfectly adequate for analog tape recorder maintenance.

I suggest that you take another look at the subject of this thread. It's
not about building a pseudo oscilloscope out of digital audio parts and
a program. It's about troubleshooting a Revox A77.

By the way, when it comes to measuring the timing between events - for
example, to measure latency in a digital I/O system - I find that an
audio editing program is more useful and more accurate than an
oscilloscope. But that doesn't mean I'm using a "downloaded oscilloscope."

--

For a good time, call http://mikeriversaudio.wordpress.com

[Mike Rivers[_2_]]

On 5/9/2017 11:47 PM, Phil Allison wrote:
> ** Oddly enough, an old fashioned analog multimeter will likely work
> fine at 100kHz.
>
> They normally use a single germanium diode rectifier and my own
> 20kohms/V meter is good to over 250kHz.

In this case it would probably work, though resolution and perhaps
accuracy may not be very good around the 500 mV level that's expected at
the test point. However, the voltmeter isn't involved in the actual
adjustment of the bias current, it's only a ballpark check.

However, a multimeter won't tell you anything about the waveform of the
bias signal. If it isn't a decent sine wave, the recorder won't work right.

Why all the arguments? If you're going to work on a tape recorder, you
really should have the right test equipment at your disposal. Otherwise,
you'll waste a lot of time asking on forums how to make do with
something else. And, besides, the guy who asked the initial question
says he has a scope, he just hasn't dug it out yet.




--

For a good time, call http://mikeriversaudio.wordpress.com

[Trevor]

On 10/05/2017 9:21 PM, Mike Rivers wrote:
> On 5/10/2017 6:50 AM, Trevor wrote:
>> However you do seem to miss my *ONLY* comment completely, and that was
>> you were wrong when you said you can't download an oscilloscope!
>
> What you're describing is a series of things that you might or might not
> need in order to create some of the capability of an oscilloscope using
> a downloaded program.
>
> I stand by my statement that you can't download something and have an
> oscilloscope unless you add other things to it that,

Depends on WHAT you are trying to do.


> if you had to buy
> them, would probably cost more than a surplus oscilloscope that's
> perfectly adequate for analog tape recorder maintenance.

Well I made a buffer amp for a lot less than any storage CRO. And gained
an audio spectrum analyser for free. :-)
(was before I bought a proper digital scope when they became more
affordable, which still doesn't have a spectrum analyser anyway)


> I suggest that you take another look at the subject of this thread. It's
> not about building a pseudo oscilloscope out of digital audio parts and
> a program. It's about troubleshooting a Revox A77.

I suggest you read the bit you wrote and my reply again. IF you meant
you can't download a CRO program to measure a 100kHz bias oscillator,
you should have specified that. So now you are just wasting your time
and mine with pointless argument.

>
> By the way, when it comes to measuring the timing between events - for
> example, to measure latency in a digital I/O system - I find that an
> audio editing program is more useful and more accurate than an
> oscilloscope. But that doesn't mean I'm using a "downloaded oscilloscope."

Right, and doesn't mean others can't use a downloaded oscilloscope
program for SOME purposes. Nobody said *either* can do everything.

Trevor.

[Scott Dorsey]

Robert Stevens > wrote:
>It's a fluke. Not sure the model. It's on AC. I do have a scope, I'll dig it out tomorrow.

Most of the better flukes with the true RMS detector are usable up to 20 KHz
but beyond that they go very wonky. You are seeing what happens when they go
wonky.

>I was measuring across the head. Both wires on each channel. I'll try the green to ground and yellow to ground tomorrow. Thx.

Use the scope. You can't measure the bias supply with a DMM.

>I don't actually see that info in the service manual. What page was it on.?

What info? The schematic should have some voltages on it. But at this point
you haven't even verified that the heads are original and correct.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Scott Dorsey]

Phil Allison > wrote:
>Scott Dorsey wrote:
>
>> Oh, I am sure it's higher. You need a VTVM with an RF probe to measure that,
>> or just use a scope.
>>
>
>** Oddly enough, an old fashioned analog multimeter will likely work fine at 100kHz.

Hmm.... I just tried my Simpson 260 on a signal generator and it agrees pretty
well with the fancy Agilent probe meter. That's cool.

Sure loads the source down a lot at 100kc though.

>They normally use a single germanium diode rectifier and my own 20kohms/V meter is good to over 250kHz.

I can believe that!
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[[email protected]]

OP
check the erase heads too,

that is a major part of the bias system.

Even if it erases fine, there may be a problem overloading the bias generator.

m

[Robert Stevens]

Scope says referenced to ground:

Right:
500mv on green wire,
20V at full pot on red wire

Left:
400mv on yellow
15V on purple

I pulled heads off, the record head is very flattened out. It is part number 022, which I believe is an A77 record head. I have no way of knowing if it is the machines original head, but it appears to be an genuine revox a77 head. The play head seems better, which makes sense since I have been able to get it to reproduce quite nicely.

Sending both off to JRF to be relapped today. Hopefully they will find them relappable.

I still think there may be an electronics issue, but for sure that record head needs some work.

Thanks,

Rob

[Robert Stevens]

On Tuesday, May 9, 2017 at 9:06:10 PM UTC-4, Phil Allison wrote:
> Robert Stevens wrote:
> >
> > So with a digital voltmeter,
>
> >
>
> ** Giant HUH ??
>
> You have a 20MHz scope, why are you not using it ?
>
> You also need a 10:1 probe ( to minimise capacitance loading ) for that job and the voltage you should see is more like 25V rms than 2.4V.
>
>
>
> ..... Phil

Basically because I am not very smart and my eyesight is bad, so nice big digital readout vs my tiny scope (3 inch screen). But after dusting off the scope and relearning how to use it, i'm getting 15 and 20 volts. 20V just starts to overbias. The 15 doesn't get close. Need those extra 5 and 10 volts. Anyway, sending heads off for service today. The saga will continue when I get them back I'm sure.

[Scott Dorsey]

Robert Stevens > wrote:
>Scope says referenced to ground:
>
>Right:
> 500mv on green wire,=20
>20V at full pot on red wire

That's about right.

>Left:
>400mv on yellow
>15V on purple

That's a little low. These are peak readings I assume?

>I pulled heads off, the record head is very flattened out. It is part numb=
>er 022, which I believe is an A77 record head. I have no way of knowing if=
> it is the machines original head, but it appears to be an genuine revox a7=
>7 head. The play head seems better, which makes sense since I have been ab=
>le to get it to reproduce quite nicely.

022 and 027 are half-track heads, so that is good. The guy that did the speed
modification likely put those heads on in the process and he used the right
ones. If the tip is flattened out and the gaps aren't clogged or worn,
pressing with your thumb should bring the levels up. If the gap is open wide
because the head is too worn, JRF will tell you.

> Sending both off to JRF to be relapped today. Hopefully they will find the=
>m relappable.

What about the erase head? If the erase head is bad, the bias levels will
be off because the erase heads are part of the bias oscillator circuit.
I would definitely include erase if you're shipping them off.

>I still think there may be an electronics issue, but for sure that record h=
>ead needs some work.

If you can get one track up to a normal level, the electronics are good.
You have one track at something close to a normal level.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

On Wednesday, May 10, 2017 at 9:37:17 AM UTC-4, wrote:
> OP
> check the erase heads too,
>
> that is a major part of the bias system.
>
> Even if it erases fine, there may be a problem overloading the bias generator.
>
> m

Erase head:
9 volts on the center pin, 13 volts on the left pin.

Pots have no effect on these readings. I'm guessing they are fixed.

Same 4-5 volt difference.

I also took some more readings on the bare wires for the rec head with the heads out. Readings are the same.

[Scott Dorsey]

Robert Stevens > wrote:
>On Wednesday, May 10, 2017 at 9:37:17 AM UTC-4, wrote:
>> OP
>> check the erase heads too,
>>
>> that is a major part of the bias system.
>>
>> Even if it erases fine, there may be a problem overloading the bias generator.
>>
>> m
>
>Erase head:
>9 volts on the center pin, 13 volts on the left pin.
>
>Pots have no effect on these readings. I'm guessing they are fixed.
>
>Same 4-5 volt difference.

That's nice, but doesn't tell you anything about the head condition. If
you don't have a bridge, send it to JRF with the others.

>I also took some more readings on the bare wires for the rec head with the heads out. Readings are the same.

That also really tells you nothing. Get the heads checked.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

On Wednesday, May 10, 2017 at 10:39:22 AM UTC-4, Scott Dorsey wrote:
> Robert Stevens > wrote:
> >Scope says referenced to ground:
> >
> >Right:
> > 500mv on green wire,=20
> >20V at full pot on red wire
>
> That's about right.
>
> >Left:
> >400mv on yellow
> >15V on purple
>
> That's a little low. These are peak readings I assume?

Correct. Peak readings. Probe on either 1X or 10X.

[Robert Stevens]

> That also really tells you nothing. Get the heads checked.

Fair enough, took me an hour to figure out how to get the rec/play heads off, I don't think I see how to get the erase head off. But I will figure it out.

Thanks.

[Robert Stevens]

>
> or more basic, check the DC supply voltage to the bias osc.
>
> m

21V.

Thx.

[Scott Dorsey]

Robert Stevens > wrote:
>
>> That also really tells you nothing. Get the heads checked.
>
>Fair enough, took me an hour to figure out how to get the rec/play heads off, I don't think I see how to get the erase head off. But I will figure it out.

The manual should tell you how to do the R/P heads.

The erase head is comparatively easy since the alignment on it isn't very
precise.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

Thx, I got it off. All 3 are off to JRF.

[Robert Stevens]

On Wednesday, May 10, 2017 at 4:23:05 PM UTC-4, Robert Stevens wrote:
> Thx, I got it off. All 3 are off to JRF.

Heads are back, installed, aligned. Still cannot overbias :(.

[Scott Dorsey]

Robert Stevens > wrote:
>On Wednesday, May 10, 2017 at 4:23:05 PM UTC-4, Robert Stevens wrote:
>> Thx, I got it off. All 3 are off to JRF.
>
>Heads are back, installed, aligned. Still cannot overbias :(.

And what did the head report say? Do the record heads now have the same
inductance top to bottom? Was the tape contact pattern on the heads even?
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

Erase: 0.144, 0.141
Rec: 8.15, 7,80
PB: 221.55, 241.04

Tape contact looks even. PB aligned nicely with MRL. Azimuth is very good at 8K and 16K. Rec Azimuth is also very good at both 8K and 16K.

[bruce seifried]

In article >,
Robert Stevens > wrote:

> Erase: 0.144, 0.141
> Rec: 8.15, 7,80
> PB: 221.55, 241.04
>
> Tape contact looks even. PB aligned nicely with MRL. Azimuth is very good at
> 8K and 16K. Rec Azimuth is also very good at both 8K and 16K.

Were these an upgraded set of heads from Revox? I seem to recall that
Revox had a half-track upgrade kit that included an additional
capacitor meant to be placed in parallel with the bias adjust trimmer.
This would supply the additional bias current that you'd need for a
wider track width head. It would probably be the 100 pF range. I'm a
little fuzzy on the details after all these years, but maybe John
French would know, or a quick check of the schematics might be helpful.
Does a Studer/Revox service dept still exist? I doubt that there's
anyone still on the payroll who'd remember, but that might be worth an
email or phone call.

[Scott Dorsey]

Robert Stevens > wrote:
>Erase: 0.144, 0.141
>Rec: 8.15, 7,80
>PB: 221.55, 241.04

Yeah, that record head has a little more metal on top than on bottom, so if
you are really trying to get the absolute most bias flux, you'll need a little
more voltage on one track than the other. So this explains why you can get
the peak on one channel but can't quite get to it on the other. Still, you
need to be able to get a little past that peak.

I'm still in Berlin and don't have the A77 manual in front of me. But if
you get the manual and go to the page with the oscillator board, it will
show you test points A and B and you have verified that the levels on those
are a bit low. Check the +20V input on pin 8 of the board and make sure that
is correct. If that is low, everything is going to be low.

You can try measuring at points A and B with the record board removed from the
machine, to see if the record board is pulling the bias down. I have never
seen that kind of failure before but there is a first time for everything and
it's a thing you can check without too much difficulty.

Now, take a look at the inductor on the oscillator board and compare it with
the one that you have. It's different for the 2-track and 4-track machines
and if you have a quarter track to 2-track conversion, they may not have done
that correctly.

Oh, and...
Replace the 250uF cap on the oscillator board. It has nothing to do with
your problem but it frequently fails and while you have the board out,
you might as well do the job. Putting a 330uF long-life axial in there is
a good plan.

>Tape contact looks even. PB aligned nicely with MRL. Azimuth is very good at 8K and 16K. Rec Azimuth is also very good at both 8K and 16K.

These machines have remarkably stable and clean transports considering how
flimsy they are and how much static stuff they have in contact with the
tape. They are not professional machines, but they were likely the best
consumer-grade machines ever made. You should have a rock solid azimuth
pattern, better than on some professional decks.

Part of what makes them not professional machines you are already figuring
out... it is nearly impossible to get to the control pots while looking at
the meters... daily alignment is not a 10-minute touchup like it is on the 440.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Scott Dorsey]

Bruce Seifried > wrote:
>In article >,
>Robert Stevens > wrote:
>
>> Erase: 0.144, 0.141
>> Rec: 8.15, 7,80
>> PB: 221.55, 241.04
>>
>> Tape contact looks even. PB aligned nicely with MRL. Azimuth is very good at
>> 8K and 16K. Rec Azimuth is also very good at both 8K and 16K.
>
>Were these an upgraded set of heads from Revox? I seem to recall that
>Revox had a half-track upgrade kit that included an additional
>capacitor meant to be placed in parallel with the bias adjust trimmer.

The idea being that instead of changing out the inductor you'd just change
the effective value of the trimmer?
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

> Were these an upgraded set of heads from Revox

The Oscillator board is the red dot version for half tracks. I have a spare which i get same results with. I also have a yellow dot version spare (meant for 1/4 tracks) for another machine. As you would expect, results with this card are even worse.

The supply voltage to the oscillator card is 21V with the card out. I'll try to get a reading tonight with the card installed.

I strongly suspect that this has always been a half track machine, since the markings on the pinch roller assembly also indicate it is a 2 track version. But of course that could be a swap out as well.

[bruce seifried]

In article >,
Robert Stevens > wrote:

> > Were these an upgraded set of heads from Revox
>
> The Oscillator board is the red dot version for half tracks. I have a spare
> which i get same results with. I also have a yellow dot version spare (meant
> for 1/4 tracks) for another machine. As you would expect, results with this
> card are even worse.
>
> The supply voltage to the oscillator card is 21V with the card out. I'll try
> to get a reading tonight with the card installed.
>
> I strongly suspect that this has always been a half track machine, since the
> markings on the pinch roller assembly also indicate it is a 2 track version.
> But of course that could be a swap out as well.

Alright, scrap that theory then...

As Scott points out, the difference between the half- and quarter-
track oscillator boards is in the inductor, which for the half-track
heads is labeled L703. Also, the bias adjust trimmer is a variable
resistor, not a variable capacitor as I wrongly remembered.

Sounds like the problem exists external to your oscillator card anyway,
since you say both of your cards act the same. You might try measuring
the resistance of the 100 ohm resistors located in the ground leads of
each record head winding. Seems unlikely that they would drift in
value, but it's simple enough to check.

If all else fails, and you can't find the actual cause of your problem,
you could try bridging a small value cap, say 30-50 pF or so, across
C710 and C711 on the osc pcb. That -might- give you a little bit more
bias signal to work with.

[[email protected]]

>
> If all else fails, and you can't find the actual cause of your problem,
> you could try bridging a small value cap, say 30-50 pF or so, across
> C710 and C711 on the osc pcb. That -might- give you a little bit more
> bias signal to work with.

again,
don't forget the erase head is part of the circuit. The problem could lie there.
m

[Scott Dorsey]

In article >,
> wrote:
>
>>
>> If all else fails, and you can't find the actual cause of your problem,
>> you could try bridging a small value cap, say 30-50 pF or so, across
>> C710 and C711 on the osc pcb. That -might- give you a little bit more
>> bias signal to work with.
>
>again,
>don't forget the erase head is part of the circuit. The problem could lie there.
>m

No, that was actually my first guess, but the erase head checks out okay
with JRF.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

> Sounds like the problem exists external to your oscillator card anyway,
> since you say both of your cards act the same. You might try measuring
> the resistance of the 100 ohm resistors located in the ground leads of
> each record head winding. Seems unlikely that they would drift in
> value, but it's simple enough to check.
>
> If all else fails, and you can't find the actual cause of your problem,
> you could try bridging a small value cap, say 30-50 pF or so, across
> C710 and C711 on the osc pcb. That -might- give you a little bit more
> bias signal to work with.

Okay, ground pins of record head to sheild of head are about 100ohms. But this is odd, the hot pins to shield are not infinite. Come in around 60 ohms. I may need to unsolder the hot wires and see if it stays that way.

[Scott Dorsey]

In article >,
Robert Stevens > wrote:
>
>> Sounds like the problem exists external to your oscillator card anyway,
>> since you say both of your cards act the same. You might try measuring
>> the resistance of the 100 ohm resistors located in the ground leads of
>> each record head winding. Seems unlikely that they would drift in
>> value, but it's simple enough to check.
>>
>> If all else fails, and you can't find the actual cause of your problem,
>> you could try bridging a small value cap, say 30-50 pF or so, across
>> C710 and C711 on the osc pcb. That -might- give you a little bit more
>> bias signal to work with.
>
>Okay, ground pins of record head to sheild of head are about 100ohms. But this is odd, the hot pins to shield are not infinite. Come in around 60 ohms. I may need to unsolder the hot wires and see if it stays that way.

That's probably fine. Pull the record and erase boards out and that resistance
will disappear most likely. Check the schematics on the record and erase
diagrams on the schematic.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

Thanks for all the help everybody, I've dug into the machine as deep as I dare. I'll get a tech to take it from here.

[Scott Dorsey]

Robert Stevens > wrote:
>Thanks for all the help everybody, I've dug into the machine as deep as I dare. I'll get a tech to take it from here.

I think you're doing better than any of the techs I know these days. If
you actually find someone who knows the A77 well, PLEASE let me know because
there is a real demand for work on those machines. Nobody likes working on
them because they really are kind of cantankerous.

You've got the hard part down.

I _may_ have a test jig somewhere... if you send me an oscillator board I
might be able to see that it meets spec, but I have to admit that I last
used the thing in the mid-eighties....
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

All the mods I made were on the removable cards plus 2 resistors on the main connector board which I made without physically removing the board. I'm afraid experience tells me that once I start removing major components, I break more than I fix.

I may however, have to replace the connector board as some of the card headers are starting to crumble. The Relay and Bias card headers are not among them though. 1 Rec card is now hot glued in place at this time. A gift from 2 nights ago.



On Friday, May 26, 2017 at 10:30:48 AM UTC-4, Scott Dorsey wrote:
> Robert Stevens > wrote:
> >Thanks for all the help everybody, I've dug into the machine as deep as I dare. I'll get a tech to take it from here.
>
> I think you're doing better than any of the techs I know these days. If
> you actually find someone who knows the A77 well, PLEASE let me know because
> there is a real demand for work on those machines. Nobody likes working on
> them because they really are kind of cantankerous.
>
> You've got the hard part down.
>
> I _may_ have a test jig somewhere... if you send me an oscillator board I
> might be able to see that it meets spec, but I have to admit that I last
> used the thing in the mid-eighties....
> --scott
>
> --
> "C'est un Nagra. C'est suisse, et tres, tres precis."

[[email protected]]

On Friday, May 26, 2017 at 11:23:42 AM UTC-4, Robert Stevens wrote:
> All the mods I made were on the removable cards plus 2 resistors on the main connector board which I made without physically removing the board. I'm afraid experience tells me that once I start removing major components, I break more than I fix.
>
> I may however, have to replace the connector board as some of the card headers are starting to crumble. The Relay and Bias card headers are not among them though. 1 Rec card is now hot glued in place at this time. A gift from 2 nights ago.
>
>
>
> On Friday, May 26, 2017 at 10:30:48 AM UTC-4, Scott Dorsey wrote:
> > Robert Stevens > wrote:
> > >Thanks for all the help everybody, I've dug into the machine as deep as I dare. I'll get a tech to take it from here.
> >
> > I think you're doing better than any of the techs I know these days. If
> > you actually find someone who knows the A77 well, PLEASE let me know because
> > there is a real demand for work on those machines. Nobody likes working on
> > them because they really are kind of cantankerous.
> >
> > You've got the hard part down.
> >
> > I _may_ have a test jig somewhere... if you send me an oscillator board I
> > might be able to see that it meets spec, but I have to admit that I last
> > used the thing in the mid-eighties....
> > --scott
> >
> > --
> > "C'est un Nagra. C'est suisse, et tres, tres precis."

Looks like someone modified this one, too....

http://www.ebay.com/itm/REVOX-A77-1-4-TRACK-STEREO-TAPE-RECORDER-FACTORY-VARI-SPEED-CONTROL-amp-15-IPS-/222519962937

Jack

[[email protected]]

OK here is another thing to check. The record amp has a bias trap at its output
to keep the bias out of the amp and to keep the amp from loading down the
bias. This could be out of tune or defective. Try swapping the record amps
and see if the problem follows one amp.
Mark

[Robert Stevens]

So one final "Hail Mary". I replaced the 100 ohm resistors in the bias card with 10 ohm resistors. This seems to have done the trick for 456 tape. Still cannot get enough bias for 499. But I'll take it!

[PStamler]

On Friday, May 26, 2017 at 10:26:56 PM UTC-5, Robert Stevens wrote:
> So one final "Hail Mary". I replaced the 100 ohm resistors in the bias card with 10 ohm resistors. This seems to have done the trick for 456 tape. Still cannot get enough bias for 499. But I'll take it!
Congratulations. If it'll bias 456, it'll do 468 and probably 911. My recollection is that the A77 was never up to 499, so I wouldn't worry about that too much.

Peace,
Paul

[Scott Dorsey]

In article >,
> wrote:
>OK here is another thing to check. The record amp has a bias trap at its output
>to keep the bias out of the amp and to keep the amp from loading down the
>bias. This could be out of tune or defective. Try swapping the record amps
>and see if the problem follows one amp.

Yes. This is why I asked earlier about whether the voltage measurements
change when you pull the record amp boards.

I am not sure how much a bad bias trap would affect the bias point, I would
think it would cause record distortion that is pretty severe, but it's a
thing you ought to check.

The problem is clearly on both channels but it seems to be worse on one than
the other... but since it's so close to the edge that may not indicate more
than just a normal channel difference.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Scott Dorsey]

In article >,
Robert Stevens > wrote:
>So one final "Hail Mary". I replaced the 100 ohm resistors in the bias card with 10 ohm resistors. This seems to have done the trick for 456 tape. Still cannot get enough bias for 499. But I'll take it!

This isn't fixing the problem, though, this is working around it. You still
need to find it.

And no, a brand new and perfectly functioning A77 won't bias up 499 so do
not expect that.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

I found what appear to be a pair of 2nF caps across the wires that send signal to the record heads and ground. I removed them and now biasing works great. No idea what they were there for, but they are gone now.

Thanks everybody.

[Robert Stevens]

On Wednesday, June 14, 2017 at 4:54:35 PM UTC-4, Robert Stevens wrote:
> I found what appear to be a pair of 2nF caps across the wires that send signal to the record heads and ground. I removed them and now biasing works great. No idea what they were there for, but they are gone now.
>
> Thanks everybody.

I should also point out they are not on the schematic, so probably a mod someone did in the past for some unknowable reason.

[Robert Stevens]

On Wednesday, June 14, 2017 at 4:59:25 PM UTC-4, Robert Stevens wrote:
> On Wednesday, June 14, 2017 at 4:54:35 PM UTC-4, Robert Stevens wrote:
> > I found what appear to be a pair of 2nF caps across the wires that send signal to the record heads and ground. I removed them and now biasing works great. No idea what they were there for, but they are gone now.
> >
> > Thanks everybody.
>
> I should also point out they are not on the schematic, so probably a mod someone did in the past for some unknowable reason.

And I put the 100ohm resistors on the oscillator card back in where they go and still plenty of overbias!

[Phil Allison[_4_]]

Robert Stevens wrote:

-------------------------

> I found what appear to be a pair of 2nF caps across the wires that
> send signal to the record heads and ground. I removed them and now
> biasing works great. No idea what they were there for,
> but they are gone now.
>
> Thanks everybody.
>


** Were they fitted on the PCB ??

How come you never spotted them before ?

This sort of thing demonstrates the UTTER FOLLY of attempting remote diagnosis.



..... Phil

[Scott Dorsey]

In article >,
Robert Stevens > wrote:
>I found what appear to be a pair of 2nF caps across the wires that send signal to the record heads and ground. I removed them and now biasing works great. No idea what they were there for, but they are gone now.

That is very, very weird.

I can't think of any reason why anyone would do that.... if the machine had
sel-sync maybe I could think of a misguided attempt at dealing with radio
interference maybe but even that doesn't make sense.

Everything that you have done to this machine and figured out about it... write
it down and tape it to the inside back cover. It will save you, or someone
else, a lot of time someday.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Stevens]

O
> ** Were they fitted on the PCB ??
>

Sort of. They were soldered to the header pins where the signal goes from the VU meter board to the heads. As for why I never noticed before, probably because they did not look out of place. I only noticed because I was going through the BOM to buy parts for a proper recap and could not find them listed. Then started looking at the schematic closer and could not find them there either.