Sony BVH-2180-PS: Servo Error

Want to know how something works? Harness the power of the Brainiacs!
User avatar
DolbySR
Posts: 22
Joined: Sun Jan 03, 2016 4:48 am
Location: Switzerland
Contact:

Sony BVH-2180-PS: Servo Error

Post by DolbySR » Sun Jan 03, 2016 5:15 am

Hello guys!

I'm new here and I found this forum by searching for information about the BVH-2000/2180 1" C-Type Video Tape Recorder. Finding information about this piece of hardware isn't that easy, I have to admit. So I thought I might give it a shot here?!

There's one thing that boggled my mind for quite some time now. I have a Sony BVH-2180-PS machine which is working okay so far - more or less at least. Before I go further though, I have to admit two things: I made the major mistake to forget the extender card at the location where I got the machine from (stupid me...) and the machine is still at my old place (I moved recently and didn't get the chance to move this heavy monster as well until now)... And I don't have an alignment tape. I only have a NTSC tape somewhere, but since I require a PAL tape, this doesn't help me at all... :?

Now to what boggled my mind and still does:
Every time I play something, the red Servo Error LED comes on and stays lit all the time. No matter whether I play a pre-recorded tape (such as a Submaster Tape I got from TV stations) or I play a tape I recorded myself. The video output is good as long as I am using the DT head and leave the TBC doing its magic. As soon as I switch to the regular R/P head, the image is heavily distorted, it periodically fades from a more or less acceptable picture to a black screen and in beween it is distorted, has black lines all over it and of course the TBC is unable to make anything useful with this chaos. When using the DT head without the TBC, the image is clear but very unstable on my reference monitor.
I noticed that the CTL track is read using the same head as used for the audio tracks. I have perfect audio, so a major misalignment of those heads doesn't seem to be a very likely problem (?).

What I have done so far:
I cleaned everything thoroughly and tried to check whether the CTL track is even being read using a pre-recorded tape which, by my understanding, should have such a CTL track recorded. As far as I can understand, this track is used to synchronize the drum servo to the tracks on the tape (?). The machine allows me to see the CTL track signal directly on the oscilloscope output but there I didn't see any signal at all. I then tried to hook up my oscilloscope directly to the AP-Board which contains the reproduce amplifier for the CTL track head. There I couldn't see anything as well.

Well, I hope to find someone who has more experience with this machine than me and that someone might be willing to lead me towards a possible solution. Although I have a Master's Degree in Electrical Engineering and specialized in signal processing, information and communication technologies, I am in need of some help here. This machine is really something else and I want to see this red LED extinguish when playing back tapes. This would be so great!

Concerning me not having the machine at my new place: If there is someone around here willing to help me, I will organise a transport so that I can carry out whatever there is to be done or checked right here at my new place.

I really hope that someone reads this post and thinks "yes, I will help this fellow!". This would certainly make my day!

Regards from Switzerland,
Patrick

User avatar
NECRAT
Site Admin
Posts: 2911
Joined: Sat Nov 17, 2007 9:13 pm
Location: Taunton, MA
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by NECRAT » Sun Jan 03, 2016 11:03 am

This may be a simpler fix. Pull the boards out and inspect the caps. I've had more than one Sony BVH machine that had a bad series of caps on them, that needed replacing.
http://www.necrat.us

"Arguing with an engineer is like mud wrestling with a pig. After a couple of hours, you realize the pig likes it"

User avatar
DolbySR
Posts: 22
Joined: Sun Jan 03, 2016 4:48 am
Location: Switzerland
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by DolbySR » Sun Jan 03, 2016 11:10 am

Hello NECRAT,

I heard about this issue as well in the past few days. Is there a simple way to determine whether they are bad? Other than testing them with the LCR-Meter? Are they really visibly bad? If so, I would have to get my eyesight checked :lol:

Regards,
Patrick

w9wi
Posts: 808
Joined: Tue Feb 05, 2008 11:40 am
Location: Pleasant View, Tennessee
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by w9wi » Sun Jan 03, 2016 11:34 am

Yes, sometimes they're visibly bad. He's referring to the electrolytics and yes, they often visibly bulge at the top when they fail.

The fact that one *hasn't* bulged doesn't necessarily mean it's good, but if one *has*, it definitely needs to be replaced.
--
Doug Smith W9WI
Pleasant View, TN EM66

User avatar
DolbySR
Posts: 22
Joined: Sun Jan 03, 2016 4:48 am
Location: Switzerland
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by DolbySR » Sun Jan 03, 2016 12:33 pm

So replacing all the electrolytic caps would be a good idea anyhow? I did this with analog audio equipment a few years ago and it was worth it there.
The big question remains: Do I need to recalibrate something afterwards? Without the extender card and the proper alignment tape, this might be a bit of a pain I assume.

User avatar
PID_Stop
Posts: 475
Joined: Thu Apr 01, 2010 11:58 am
Location: Syracuse, New York
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by PID_Stop » Mon Jan 04, 2016 9:33 am

I assume that the servo light you refer to is the one on the control panel, right? That LED is driven by a lot of different systems, so the first step is to narrow down which servo is producing the alarm. While the machine is playing, lift up the control panel and take a look at the RL-12 board (it's the reel servo controller); you will see a row of eight LEDs, four on the top and four on the bottom. If you have a front-panel servo alarm, one or more of the red LEDs on the RL-12 board will also be on:
D7 (SYS) means a system controller interface error (usually the SY-79 and SY-80 boards aren't communicating properly with SV-43)
D6 (DR) means the drum servo isn't locked
D5 (TTP) is a Tape Transport error; usually it's a problem with the reel tension sensors
D4 (SV) is a system error, generally within the SV-43 servo board.

The green lights below should also be lit during proper playback:
D3 (MODE) is a very general indicator that everything is proper
D2 (CF) means the capstan and drum servos have locked to the correct color frame (this is relevant to NTSC,and I imagine it would also be relevant to PAL and SECAM versions as well)
D1 (DR) means the drum servo is locked
D0 (CAP) means the capstan servo is locked

A couple of quick things to check, some of which can also lead to this alarm: make sure the front panel servo reference switch is in the AUTO position, and that the machine is getting a valid reference signal. On the RL-12 board, also make sure the SKEW switch is set to AUTO. On the CD-17 board, set the top toggle switch (ADV) according to the type of time base corrector you have: if it's an external BVT-2000, it should be set up to BVT; if it's the four-card internal TBC, it should be set down to BKH. You can make the machine lock more easily by setting the CAP LOCK switch to 2F. On the SR-15, try setting the SC ADJ toggle switch up to the MANU position, and turn the knob above it to make the SC PHASE meter's needle move to the center mark.

Here's the nickel tour about the control track: it's a series of pulses recorded by the same headstack as the audio channels, and its function is analogous to the sprocket holes in film: the pulses delineate where the video tracks are on the tape, in the same way that a sprocket hole has a definite mechanical relationship to the position of a frame on the film. During recording, the drum servo positions the head so that it begins a new track toward the end of a vertical interval period; the control track pulses are also generated to correspond to a particular point in the incoming video. As a result, the control track pulses winds up in a definite position that relates to the position of the video tracks.

During normal playback with the R/P head, the capstan servo works to make the control track pulses synchronize with the incoming playback reference; the drum servo then positions the drum according to where the control track pulses on the tape wind up. The end result is that the rotating head should line up with the original recorded track on the tape. Inevitably, there will be variations from tape to tape, so the tracking knob (you have to pull it out for it to have an effect) slightly advances or delays the drum servo so that you can position the head in the center of the recorded track, and maximize the playback level. If you look at the front of the MD-14 board (toward the left), flip the METER switch to the up RF(V) position; the meter above the tracking knob will show you the RF level coming from the video head, so you just rotate the tracking knob to maximize the meter reading.

Playback with the DT head is very similar, except that the tracking knob is taken out of the loop, and the drum servo moves the drum 120° around so that the DT head is lined up with the track on the tape instead of the fixed R/P head. The DT head is mounted to a quartz piezo element that is driven by the DT-04 board; the short version is that the DT-04 board adds a small sinusoidal wobble signal to the piezo element, and watches the recovered RF level from the head; it adds a DC control voltage to the wobble so that the RF is at its peak when the wobble signal is closest to zero. This automatically compensates for tape variations; better still, it has enough correction range to make up for a ±2x tape speed variation (that is, ranging from –1 to +3 speed).

The fact that you are seeing video on the DT head suggests that your signal system is working properly, and that the servo is pretty close to normal.

Do you have an eccentricity gauge to measure the head tip projection? Absence of playback RF is often a result of the tip having worn down. A new R/P or DT head typically measures around 90 microns; you start having this sort of problem when it gets down around 50. Heads that have been refurbished sometimes come back with really weird tip projection... and one thing I've encountered is that refurbished heads that have worn down will start having the tip gap start to widen, which kills the high frequency response on playback. It's not such a problem during recording: whereas playback response is strongly limited by the gap width (the induced voltage is a function of the differential flux density across the gap), the signal imparted to the tape during recording is a function of the flux density at the trailing edge of the head as it moves across the tape.

Okay, that was long winded... but it brings back several weeks spent in the Sony training facility in San Jose, back in 1985. In my younger days, I had my hands on most of the BVH machines in the Central New York area at one point or another... a pretty good freelancing gig for a fellow starting a family. We still have a working machine in our control room, and from time to time I amaze the younger operators who have literally never seen a working reel-to-reel machine.

-- Jeff
Last edited by PID_Stop on Mon Jan 04, 2016 9:44 am, edited 2 times in total.

User avatar
PID_Stop
Posts: 475
Joined: Thu Apr 01, 2010 11:58 am
Location: Syracuse, New York
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by PID_Stop » Mon Jan 04, 2016 9:37 am

DolbySR wrote:So replacing all the electrolytic caps would be a good idea anyhow? I did this with analog audio equipment a few years ago and it was worth it there.
The big question remains: Do I need to recalibrate something afterwards? Without the extender card and the proper alignment tape, this might be a bit of a pain I assume.
If you replace caps in the servo system, you will very probably need to so some major recalibration. That's not to say that it isn't worthwhile... but you will need an alignment tape and the BVH-2000 tool kit (in particular, the precision weight set and spring scales), and on oscilloscope. For the most part, it's not crushingly difficult... just time consuming. You're making a combination of electrical and mechanical adjustments, and some of them interact a bit.

-- Jeff

User avatar
DolbySR
Posts: 22
Joined: Sun Jan 03, 2016 4:48 am
Location: Switzerland
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by DolbySR » Mon Jan 04, 2016 12:55 pm

Hello Jeff

Thanks a lot for your highly interesting reply! I read your posts earlier when I was on the train back home from work and noticed once more, how this machine catches my interest over and over again.

Yes, you are right. It's the servo light on the control panel to the left. Unfortunately, I don't recall by heart which LEDs on the RL-12 are lit in playback and which aren't. But there are some red ones and some green ones... But I see, I simply have to get this machine to where I live now. I really want to solve this issue :)

I mostly don't supply a video reference signal and just leave the machine run on its own. But I remember that when I did, the stop button stopped flashing. The servo problem remained though.

What does this SC ADJ do? There are so many abbreviations that I sometimes struggle to keep track of what they all mean... :roll:

Something I didn't mention: When playing back something with the DT head, I see that there is a periodic vertical jump in the video picture. The entire picture shifts up, let's say "half a line" and gets slightly unsharp. It will then shift back down again and will be sharp as before again. The RF(V) level dips each time this shift happens by about 10-15%. There's one switch on one of the cards (called something with "...ADD...", maybe on the internal TBC?) which stops this and leaves the image in its shifted unsharp position. But again, I really have to get this machine over here now.

I thought as well that there cannot be much wrong with the system itself when record and replay using the DT head works okay. Does this mean that the CTL signal is being read as well or not? I wasn't able to see anything on the CTL setting when I had my oscilloscope connected to the respective output.

No, I cannot measure the head tip projection but everything I recorded looked as crisp as the original, usually a DVD, when played back. So no real problems with a widened head gap I assume.

I guess I'll leave the recapping for now until it either becomes really inescapable or I have the required tools and tapes to do it.

But anyway, I'd really love to read some more from you, Jeff :wink: There's still tons I can learn about this marvelous piece of tech!

Regards,
Patrick

User avatar
PID_Stop
Posts: 475
Joined: Thu Apr 01, 2010 11:58 am
Location: Syracuse, New York
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by PID_Stop » Mon Jan 04, 2016 2:50 pm

DolbySR wrote:I mostly don't supply a video reference signal and just leave the machine run on its own. But I remember that when I did, the stop button stopped flashing. The servo problem remained though.
Okay, that sounds right... that's how the stop lamp is supposed to behave. I have encountered cases where an inappropriate reference signal will satisfy the detector enough to make the lamp stop flashing, but drives the servo nuts. Doesn't sound like what's going on in your case, though.
DolbySR wrote:What does this SC ADJ do?
"SC" stands for subcarrier; in an NTSC system, there's a 180° phase shift in subcarrier between successive frames; if you're doing editing where exact horizontal position is important, you need to make sure your edits always match up frames with the same subcarrier phase. That's called color-framing (which is also the significance of the green "CF LOCK" indicator on the front panel). The adjustment is helping the servo system detect the subcarrier phase reliably. Incidentally, there's a switch on one of the boards marked "2F / 4F" that determines whether the servo will take the extra effort and time to force a color frame lock, or if it will lock to the nearest frame regardless of subcarrier phase.
DolbySR wrote:Something I didn't mention: When playing back something with the DT head, I see that there is a periodic vertical jump in the video picture. The entire picture shifts up, let's say "half a line" and gets slightly unsharp. It will then shift back down again and will be sharp as before again. The RF(V) level dips each time this shift happens by about 10-15%. There's one switch on one of the cards (called something with "...ADD...", maybe on the internal TBC?) which stops this and leaves the image in its shifted unsharp position. But again, I really have to get this machine over here now.
That rings a bell... I've seen that before, but it's been a quite few years. I'll have to go back into my notes and see what that turned out to be.
DolbySR wrote:I thought as well that there cannot be much wrong with the system itself when record and replay using the DT head works okay. Does this mean that the CTL signal is being read as well or not? I wasn't able to see anything on the CTL setting when I had my oscilloscope connected to the respective output.
Yes, if the machine wasn't seeing the control track, it wouldn't begin to know where to send the DT head.

To see the control track pulse, hook your scope to TP24 on the CD-17 board (it's by the front edge) and play the tape. You should see a succession of very narrow spikes, most of which go negative, but a few go positive. The peak-to-peak amplitude of the spikes should be 3 volts, ±0.2v. An NTSC machine's control track looks very different, more like a distorted square wave. If you have volume 1 of the manual, you can see a picture of this on page 11-10.
DolbySR wrote:No, I cannot measure the head tip projection but everything I recorded looked as crisp as the original, usually a DVD, when played back. So no real problems with a widened head gap I assume.
Ahhhh... one thing I didn't point out. Remember, analog video isn't recorded the same way as audio. Whereas an audio signal is fed directly to the recording head with some high frequency bias thrown in to overcome tape hysteresis, video is recorded as an FM signal. The incoming video is clamped to force the video to have a specific relationship to ground, and modulated to produce a carrier that ranges from 7.06MHz at the sync tip, to 10.00MHz at 100IRE white (for an NTSC system). So frequency response issues don't directly influence picture resolution (which would be the more intuitive result); instead, you start seeing breakup at high luminance areas of the picture. With a refurbished head, though, the progression from normal to unusable is astonishingly rapid... and playback is the first thing to go.

For now, I would defer re-capping of circuits that are part of calibrated servo paths; regulator filter caps and signal decoupling caps should be pretty safe to change out.

Regards,

-- Jeff

User avatar
DolbySR
Posts: 22
Joined: Sun Jan 03, 2016 4:48 am
Location: Switzerland
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by DolbySR » Tue Jan 05, 2016 12:27 am

Hello Jeff,
PID_Stop wrote:Okay, that sounds right... that's how the stop lamp is supposed to behave. I have encountered cases where an inappropriate reference signal will satisfy the detector enough to make the lamp stop flashing, but drives the servo nuts. Doesn't sound like what's going on in your case, though.
Do I need to feed a certain video signal for best results? Color bars or something? I usually feed a black video signal from one of my other recorders.
PID_Stop wrote:"SC" stands for subcarrier; in an NTSC system, there's a 180° phase shift in subcarrier between successive frames; if you're doing editing where exact horizontal position is important, you need to make sure your edits always match up frames with the same subcarrier phase. That's called color-framing (which is also the significance of the green "CF LOCK" indicator on the front panel). The adjustment is helping the servo system detect the subcarrier phase reliably. Incidentally, there's a switch on one of the boards marked "2F / 4F" that determines whether the servo will take the extra effort and time to force a color frame lock, or if it will lock to the nearest frame regardless of subcarrier phase.
When should this indicator be on? When playing back, I didn't see this one to be lit. But maybe that's because I never really adjusted this using the SC ADJ knob? Or is it due to me maybe not forcing color frame lock?
PID_Stop wrote:That rings a bell... I've seen that before, but it's been a quite few years. I'll have to go back into my notes and see what that turned out to be.
One thing that popped into my mind: When I view the video image without the TBC and, for example, play the color bars part of a tape, the vector scope should ideally show a steady image with the single "dots" of the individual colors within their respective boxes. This works with the TBC in the signal path. Without TBC, the vector scope only shows a heavily rotating image of, what I could make out, is the same image as I get with the TBC on, but simply rotating. Maybe this helps?
PID_Stop wrote:Yes, if the machine wasn't seeing the control track, it wouldn't begin to know where to send the DT head.

To see the control track pulse, hook your scope to TP24 on the CD-17 board (it's by the front edge) and play the tape. You should see a succession of very narrow spikes, most of which go negative, but a few go positive. The peak-to-peak amplitude of the spikes should be 3 volts, ±0.2v. An NTSC machine's control track looks very different, more like a distorted square wave. If you have volume 1 of the manual, you can see a picture of this on page 11-10.
That's a relief... I never really saw a signal like that though. But it might just have been a wrong measurement with my digital oscilloscope. I will get into that as soon as I am standing in front of the machine again and probe at TP24 directly instead of at the oscilloscope output at the rear of the machine.
I have volumes 2 and 3 of the manual series, but I saw an image of how the track signal should look like. There are also hand written notes including the description of the CTL signal on the schematics for the CD-17 card. Maybe someone already had a go there, but then I cannot be sure that this manual has always been with this machine only.
PID_Stop wrote:Ahhhh... one thing I didn't point out. Remember, analog video isn't recorded the same way as audio. Whereas an audio signal is fed directly to the recording head with some high frequency bias thrown in to overcome tape hysteresis, video is recorded as an FM signal. The incoming video is clamped to force the video to have a specific relationship to ground, and modulated to produce a carrier that ranges from 7.06MHz at the sync tip, to 10.00MHz at 100IRE white (for an NTSC system). So frequency response issues don't directly influence picture resolution (which would be the more intuitive result); instead, you start seeing breakup at high luminance areas of the picture. With a refurbished head, though, the progression from normal to unusable is astonishingly rapid... and playback is the first thing to go.
Oh, right! That explains a lot. I always asked myself why there were tiny black "fish" in the video image of my SVHS recorder when something was really bright. This might then be caused by a worn head. Fortunately, I could not see anything like that with my 1" recorder. At least using the DT head... No matter how bright an image gets.
PID_Stop wrote:For now, I would defer re-capping of circuits that are part of calibrated servo paths; regulator filter caps and signal decoupling caps should be pretty safe to change out.
This might indeed be the better choice for the moment.

As for now, I'm planning to drive to my old place this week-end. I doubt though that the machine will fit my 2 door sports car :wink: but I'll take my oscilloscope and other measurement tools with me to get some measurements done and document them in images.

Regards
Patrick

User avatar
PID_Stop
Posts: 475
Joined: Thu Apr 01, 2010 11:58 am
Location: Syracuse, New York
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by PID_Stop » Tue Jan 05, 2016 9:12 am

DolbySR wrote:Do I need to feed a certain video signal for best results? Color bars or something? I usually feed a black video signal from one of my other recorders.
If you have an internal TBC, you don't strictly need to feed the machine reference at all. With the external BVT-2000, the system worked better with external reference. Generally, analog equipment likes to lock to color black, though a signal with active video will often work. The key thing is that it needs to be stable, or else the VTR will unlock every time the reference shifts. I'd be concerned about using another VTR as a reference source... a camera would be better. If your switcher has an internal sync generator with reference outputs, that would be better still. A dedicated master sync generator is the ideal.
DolbySR wrote:When should this indicator be on? When playing back, I didn't see this one to be lit. But maybe that's because I never really adjusted this using the SC ADJ knob? Or is it due to me maybe not forcing color frame lock?
On an NTSC machine, the green CF LOCK lamp will come on after the servos have settled down, and the playback field matches the reference field. This presupposes that you're feeding the machine a valid reference where the subcarrier burst has a specific phase relationship to the falling edge of horizontal sync (that's called SC/H phase, and is part of the RS-170A specification). Also, the switch on the CD-17 board has to be in 4F, not 2F position. It's possible that this isn't relevant to a PAL machine. I've only worked on two PAL machines (Eastman Kodak had a video facility in Rochester for a number of years where they did instructional video production for both US and European markets). Unfortunately for this instance, they only needed minimal maintenance, so I never had occasion to fully compare the differences between NTSC and PAL versions.
DolbySR wrote:One thing that popped into my mind: When I view the video image without the TBC and, for example, play the color bars part of a tape, the vector scope should ideally show a steady image with the single "dots" of the individual colors within their respective boxes. This works with the TBC in the signal path. Without TBC, the vector scope only shows a heavily rotating image of, what I could make out, is the same image as I get with the TBC on, but simply rotating. Maybe this helps?
Well, you've just discovered the real purpose of the TBC. Here's a bit of history: in the early 1950s, there were a number of roadblocks to recording broadcast-quality video. We touched on two yesterday: first, the frequency response of a recorder depends heavily on how fast the head moves past the tape (and on the size of the magnetic particles). The earliest approach was to have a fixed head, like an audio recorder, and move the tape insanely fast; a 14-inch reel might last a minute if you're lucky. This didn't work for a lot of reasons, and several engineers at Ampex came up with the idea of mounting heads on a spinning drum that scanned arc-shaped tracks on the tape, while it moved at a more reasonable speed. That head design wasn't terribly practical, and they replaced it with a headwheel that made a succession of straight tracks perpendicular to the edge of the tape. This is how the early "quad" machines worked: the tape ran at 15 inches per second, and the headwheel (with four tips positioned every 90°) spun at 14,400 RPM. We were still using these quad machines up through the early 1990s; in our shop, they overlapped the Sony BVH machines by about ten years.

The second challenge was also related to frequency response: good video doesn't just need to extend to high frequencies (on the order of 5-6MHz for color)... the range of frequencies needs to extend from virtually DC all the way up. If you think in terms of octaves, that's an impossibly stiff challenge for direct recording, especially on the low end. The solution was conceived by Ampex's Charles Anderson: have the video modulate an FM carrier, and record that. Instead of trying to directly record a video signal spanning nearly 20 octaves, the FM carrier deviates less than one octave. The young fellow who came up with the practical FM circuitry was none other than Ray Dolby, still a student who started out hanging around Ampex and who got pulled away to serve during the Korean War. The video recording project was where he picked back up after his discharge.

Getting the head-to-tape speed up and using FM recording solved the biggest problems, and the first Ampex VTRs demonstrated in 1956 and rushed to the networks for time zone delay were examples of this first generation of technology. The machines were not terribly stable: most of the electronics were tube-type, and constantly drifted; the more basic problem was mechanical. Even if you assume ideal conditions for a given recording: the capstan is moving the tape at precisely the same speed with no flutter or other variation, the headwheel is spinning at a perfectly uniform rate and is perfectly locked to a completely stable video source, and that the geometry of the tape path is perfect, you still have the reality that the machine playing back the tape is not ideal. The capstan and drum servos will wander, the tape stretches, the head tips won't have exactly the same projection as the recording head (changing the headwheel's effective circumference), and the guides -- especially the vacuum guide that forms the tape into the concave shape it needs in order to fit against the headwheel -- won't be in precisely the same position as they were in the recorder. The result is that the recovered playback signal will contain considerable jitter and other time-based errors. In the early black-and-white machines, this was most noticeable when one of the rotating tips would end its track and the next would begin: quad machines took four headwheel revolutions to scan one video field (your helical scan BVH records a complete field in one revolution). That meant that each video field was broken into sixteen horizontal stripes as each head tip recorded a piece of the picture... and mechanical discrepancies would create very noticeable delay problems where these stripes met. Incorrect tip projection would make vertical lines in the picture appear to be a series of arc-shaped segments; incorrect guide height would skew the segments diagonally like a venetian blind.

These sorts of geometric distortions are called time-base errors; for black-and white video, careful adjustment could make the picture reasonably good... but if you're trying to record color, even very small time base errors become enormous phase shifts up at color subcarrier frequencies.

The first fix for time base errors was rather clever: pass the playback RF through a variable delay line, whose delay is controlled by a DC voltage. Use your house reference signal to start a ramp generator, so you have a sawtooth-shaped signal that resets every horizontal line. Watch the raw playback video coming off the FM demod, and as soon as you see the horizontal sync pulse, use that to capture a sample of your sawtooth. The earlier the playback sync shows up, the lower your sawtooth sample will be; as the playback delay grows, the higher the sample voltage becomes. So now you have a sample voltage that corresponds to how much the playback is varying in time... apply that to the delay line, and bingo: the time base errors get cancelled out. Ampex's system for this was called AMTEC, and it worked really well... in black-and-white.

Even with AMTEC, there was still too much error to make color stable... so a very similar corrector was added to make much finer adjustments, based on comparing reference subcarrier phase against playback subcarrier burst phase. This second corrector was called COLORTEC, and could produce a color playback nearly indistinguishable from the original. This is impressive stuff, when you consider that the technology was completely pre-IC analog.

Quad machines managed to be workable mainly because they were overbuilt to minimize the limitations of mechanically-induced errors... and even though early remote production trucks might have several on board, they really weren't portable. Practical portability wouldn't come until helical-scan recorders, that wrapped the tape diagonally around a spinning head, came about. The upside to helical scan was vastly reduced size and weight, and the elimination of head switching during the picture; the downside is vastly increased mechanical error, even if the machine isn't moving. Where an Ampex quad machine took sixteen head tip passes to make one field of video, each pass scanning a fairly wide track just over an inch long, a BVH-2000 scans an entire field in one track that is thinner than a human hair, and stretches something like sixteen inches along the tape. Chuck Hintz, the instructor at the Sony training school, remarked that the precision needed to scan a 1" tape was similar to driving your car across the United States without touching the steering wheel, and without leaving your lane. There's just no way to make up for all of the error that results using analog delay lines; it needed the development of fast analog-to-digital converters and sufficiently large memory ICs to become practical... and digital time-base correctors were born.

Even so, the early TBC's were pretty large: the Sony BVT-2000, the BVH's original time base corrector, was a box about a third the size of the tape machine itself, and had to be connected through a large umbillical cable so that all of the servo systems could interact with the TBC (especially the dynamic tracking system). After a few years, Sony managed to fit the TBC into four cards that could fit into the VTR itself. Looking back, the people who designed the BVH-2000 clearly were looking well ahead of available technology, since they designed the machine for an accessory that wouldn't be possible for a number of years. For some reason, that never dawned on me until just now.

So... what you're seeing when you look at the raw demod output of your machine is quite normal, and shows just how big a problem mechanical factors become in video recording... and how impressive a feat digital time base correction really is!
DolbySR wrote:I have volumes 2 and 3 of the manual series, but I saw an image of how the track signal should look like. There are also hand written notes including the description of the CTL signal on the schematics for the CD-17 card. Maybe someone already had a go there, but then I cannot be sure that this manual has always been with this machine only.
That's going to be a real problem: all of the alignment procedures, visual guides, and theory of operation are in volume 1. My marked-up version is sitting on the shelf next to me, and it's about two centimeters thick. One of these days I should think about scanning it -- not just for its own sake, but also because it has all of my notes from school, and from years of working on various machines.
DolbySR wrote:Oh, right! That explains a lot. I always asked myself why there were tiny black "fish" in the video image of my SVHS recorder when something was really bright. This might then be caused by a worn head. Fortunately, I could not see anything like that with my 1" recorder. At least using the DT head... No matter how bright an image gets.
A worn head will definitely cause this; but then, so will trying to record excessively high video levels. The FM modulator has an upper limit on how high its frequency will deviate; some systems will clamp at overly high levels, others go to pieces and you wind up with black or other values breaking into the peaks. (Remember too, SMPTE type C recorders like the BVH series are so-called direct color recorders: that is, they record the composite video signal intact, with luminance and chroma subcarrier. Machines like the 3/4" U-matic used a "color under" system that used a heterodyne scheme to reduce the upper frequencies, and machines like Betacam recorded luminance and chrominance on separate tracks altogether. Anyway, a direct-color recorder like a BVH-2000 don't produce the greatest FM deviation with white, but with very light saturated colors like yellow, where you have a lot of color subcarrier sitting on top of high luminance. So if you're going to see this sort of tearing in the video, it will probably show first in really bright yellows or hot pinks, not white.
DolbySR wrote:As for now, I'm planning to drive to my old place this week-end. I doubt though that the machine will fit my 2 door sports car :wink: but I'll take my oscilloscope and other measurement tools with me to get some measurements done and document them in images.
Yeah, physical reality really stinks sometimes! About 20 years ago I had an opportunity to buy a 24-track Scully audio recorder at an auction for $50, but it wouldn't fit in my Chevy Chevette. Probably just as well: my wife, despite being quite competent technically, would not have been amused.

Regards,

Jeff

User avatar
PID_Stop
Posts: 475
Joined: Thu Apr 01, 2010 11:58 am
Location: Syracuse, New York
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by PID_Stop » Tue Jan 05, 2016 10:21 am

If you are interested in such things, I just ran across an account of the birth of video recording by Charles Ginsburg, part of the early Ampex team.

User avatar
DolbySR
Posts: 22
Joined: Sun Jan 03, 2016 4:48 am
Location: Switzerland
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by DolbySR » Tue Jan 05, 2016 11:15 am

Hello Jeff,
PID_Stop wrote:If you have an internal TBC, you don't strictly need to feed the machine reference at all. With the external BVT-2000, the system worked better with external reference. Generally, analog equipment likes to lock to color black, though a signal with active video will often work. The key thing is that it needs to be stable, or else the VTR will unlock every time the reference shifts. I'd be concerned about using another VTR as a reference source... a camera would be better. If your switcher has an internal sync generator with reference outputs, that would be better still. A dedicated master sync generator is the ideal.
I really don't have a great reference source. That might be a bit of a problem, but I'll certainly keep my eyes open. One thing on my list of "things to buy" is a reference pattern generator. I could have used one of those a few times already. It doesn't generate super fancy signals, but a simple white image or basic color bars would have been nice to have from time to time. Definitely something to track down soon.
PID_Stop wrote:On an NTSC machine, the green CF LOCK lamp will come on after the servos have settled down, and the playback field matches the reference field. This presupposes that you're feeding the machine a valid reference where the subcarrier burst has a specific phase relationship to the falling edge of horizontal sync (that's called SC/H phase, and is part of the RS-170A specification). Also, the switch on the CD-17 board has to be in 4F, not 2F position. It's possible that this isn't relevant to a PAL machine. I've only worked on two PAL machines (Eastman Kodak had a video facility in Rochester for a number of years where they did instructional video production for both US and European markets). Unfortunately for this instance, they only needed minimal maintenance, so I never had occasion to fully compare the differences between NTSC and PAL versions.
What a pity that they worked so well :mrgreen: On PAL machines, this switch on the CD-17 board can be in the 2F, 4F or 8F position, where the 8F position is marked as "default" I guess.
PID_Stop wrote:Well, you've just discovered the real purpose of the TBC. Here's a bit of history:
Consider my mind blown by your insight into history and the evolution of TBC :o Oh dear, I sometimes think that I was born some decades too late. This is so interesting and I could read on about such things for days.
PID_Stop wrote:Even so, the early TBC's were pretty large: the Sony BVT-2000, the BVH's original time base corrector, was a box about a third the size of the tape machine itself, and had to be connected through a large umbillical cable so that all of the servo systems could interact with the TBC (especially the dynamic tracking system). After a few years, Sony managed to fit the TBC into four cards that could fit into the VTR itself. Looking back, the people who designed the BVH-2000 clearly were looking well ahead of available technology, since they designed the machine for an accessory that wouldn't be possible for a number of years. For some reason, that never dawned on me until just now.
It was really miles ahead. I mean, they reserved the space for an internal TBC with 4 cards from the beginning. Presumably without knowing whether it would be possible to do this sort of thing on 4 cards in the near future. Great engineering!
PID_Stop wrote:So... what you're seeing when you look at the raw demod output of your machine is quite normal, and shows just how big a problem mechanical factors become in video recording... and how impressive a feat digital time base correction really is!
Yeah, I'm able to understand the full potential of this TBC now. It is truly amazing what 1980's technology was capable of and what they did with the limitations of that time. Today, it wouldn't be a big deal to fit something like this into one single IC, an FPGA for instance. But back then... Where so many things were done purely analogue and no signal processing on a DSP was used to simplify things. Sure, the digital domain isn't purely all-good. There are a lot of other problems and limitations to consider. But even so. Amazing. Just a-mazing.
PID_Stop wrote:That's going to be a real problem: all of the alignment procedures, visual guides, and theory of operation are in volume 1. My marked-up version is sitting on the shelf next to me, and it's about two centimeters thick. One of these days I should think about scanning it -- not just for its own sake, but also because it has all of my notes from school, and from years of working on various machines.
I was looking all over the place for the 1st volume at the place I got the machine from but couldn't find a copy of it. You are right though, I often would have liked to know more about some simple basics. Until now, I know how to record and play back a tape, but those editing controls and capabilities this machine has... no clue. They primarily look nice (I have the version with the "fully equipped" control panel with tons of buttons) :wink: I tried to track down a copy of this 1st volume on one of those PDF manual sites, but so far didn't succeed. I mean, I would even pay for one, this isn't the problem. But it seems as if it's just a bit of a rare item online. Understandable, since every machine came with its own set of manuals.
PID_Stop wrote:A worn head will definitely cause this; but then, so will trying to record excessively high video levels. The FM modulator has an upper limit on how high its frequency will deviate; some systems will clamp at overly high levels, others go to pieces and you wind up with black or other values breaking into the peaks. (Remember too, SMPTE type C recorders like the BVH series are so-called direct color recorders: that is, they record the composite video signal intact, with luminance and chroma subcarrier. Machines like the 3/4" U-matic used a "color under" system that used a heterodyne scheme to reduce the upper frequencies, and machines like Betacam recorded luminance and chrominance on separate tracks altogether. Anyway, a direct-color recorder like a BVH-2000 don't produce the greatest FM deviation with white, but with very light saturated colors like yellow, where you have a lot of color subcarrier sitting on top of high luminance. So if you're going to see this sort of tearing in the video, it will probably show first in really bright yellows or hot pinks, not white.
I recorded a concert DVD onto 1" tape and there are quite a few sections with yellow and hot pink (I really like this description). Fortunately, I didn't see anything wrong with it. That's good :)
PID_Stop wrote:Yeah, physical reality really stinks sometimes! About 20 years ago I had an opportunity to buy a 24-track Scully audio recorder at an auction for $50, but it wouldn't fit in my Chevy Chevette. Probably just as well: my wife, despite being quite competent technically, would not have been amused.
Yes, I know how it is. I once worked for STUDER (maybe a known brand in the US, too?) here in Switzerland and just at the end of my "time" there, they sold pretty much every machine they had in their museum because they moved their location a few hundred meters to a different building and didn't have the space to have one piece of every single machine and, especially, version (!) ever built. There, they had 48 channel DASH recorders for sale.

Image

Hours on the machine: less than 100. Aligned and serviced just a few weeks earlier. The price: about 1000$. The problem: Pick it up until 4 pm or they will be destroyed. So anyone with the money PLUS the suitable vehicle could have taken one. Of course, no one did. A bit sad. But hey, if something goes wrong, even though there still are replacement parts around, it would be a financial disaster. Don't get me started about the tapes...

I chose to buy a A807 MK2 tape recorder instead. This one fitted perfectly into the trunk of my car with enough room to spare for some other audio gear :mrgreen:

And about the wife... I still have a complete 35mm projection equipment including all available digital formats and fancy-pantsy stuff back at my old place. Unfortunately, there is no space for it in the living room here at our new place :lol:

Regards
Patrick

User avatar
DolbySR
Posts: 22
Joined: Sun Jan 03, 2016 4:48 am
Location: Switzerland
Contact:

Re: Sony BVH-2180-PS: Servo Error

Post by DolbySR » Tue Jan 05, 2016 11:20 am

PID_Stop wrote:If you are interested in such things, I just ran across an account of the birth of video recording by Charles Ginsburg, part of the early Ampex team.
Lab Guy's World rings a bell. I visited this site quite some time ago by chance. I'll definitely spend my evening reading through this page :mrgreen:

Just as I read about your TBC-story with those Quadruplex machines, I checked out a video on YouTube about it. Very interesting stuff. I even have a 2" reel, empty, here. Once used as a door-stop in my old apartment. A fancy one :wink:

grich
Posts: 293
Joined: Fri Sep 02, 2011 11:19 am
Location: MP89.5, Mason City Subdivision

Re: Sony BVH-2180-PS: Servo Error

Post by grich » Tue Jan 05, 2016 2:24 pm

DolbySR wrote:...I once worked for STUDER (maybe a known brand in the US, too?)...
Definitely known in the US...even in the small towns in Northwest Missouri where I went to college, you could find Studer B67 and Revox A77 tape decks. One of my stations still has a A77 in storage.

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest