Today I am fixing another Fostex R8 from a customer. The problem was that the plastic capstan wheel was slipping on its axis, and the spooling motor wheels did the same. This is a common problem with these decks. The Fostex will not play and/or spool correctly then.

Fortunately, John from https://www.tascamninja.com/ makes aluminium replacements. These fit tightly on the axis and can be fastened with a screw.

I also replaced the belt. Now, the deck plays and spools perfectly.

A small problem

But unfortunately, one problem remains: the arm on which the right tension roller is mounted, has got a fair amount of play, maybe caused by a blow or something.
This is causing the right tension roller to guide the tape in a skewed position. It is not a show stopper, but it is certainly not ideal for the tape path.

So I could purchase a Fostex R8 cheap. The unit was supposed to be in a fairly good working condition, maybe a few scratches on the front. The price was that good that I could not walk away from it. Besides, I always wanted a Fostex. Just because.

Overview

This deck is squeezing 8 audio tracks onto a standard 1/4 inch tape. To put that into perspective, normal home audio deck use 4 tracks (2 -> and 2 <-) and the professional world uses just 2 tracks onto that same width of tape. So the tracks are super small. To compensate for the loss of fidelity and the added tape hiss the speed at which the tape runs is fixed at 38 cm/s or 15 IPS (cassettedecks run at 4,75 cm/s). Also, the deck has Dolby C. Dolby C comes from the cassette world, it is an enhanced version of the omnipresent dolby B and it works quite well on this deck. But I’m getting ahead of myself.

The deck, being one of the very latest generation of tape deck being made, is heavily computerized. Not for the calibration electronics, like the Studer A810 and all Studer models beyond that, but in the controls. There are a LOT of possibilities and functions that I have not even tried tried yet.
When the deck is turned on, the display shows a scrolling text of ‘FOSTEX’.

First impressions

When I got the deck I first cleaned the outside. I like to work on a clean deck. This one was not dirty, but sometimes you can get a lot of dirt off decks you collect. After that, I cleaned the heads and the tape path in general. I noticed the very plasticy rollers and the flimsy arms. Also, the roller covers were coming of way too easy. Hmmm, not so good.

I put a tape on and the deck played. So far, so good. Rewind, fast forward, no problem. The capstan was noisy however. I played around with it and decided to try recording. I was NOT disappointed with the sound quality. I was surprised how well it sounded on playback. I tried several channels and it seemed ok. The dolby C noise reduction really took away tape hiss in quiet passages and in between tracks. Just like on cassette decks.

Capstan wheel

After a while I noticed that the tape speed was dropping a little. It was ok after a while, but just after hitting PLAY it was not good.
And after a little more testing, the tape was not moving at all anymore. Even pressing the pinchroller to the capstan did not work. I noticed the capstan was not spinning anymore. When in STOP, the capstan would slowly start spinning again. Only to come to a quick stop once I pressed PLAY. But I could hear the capstan motor spinning constantly. Weird.
I had already ordered a new capstan belt, as is almost always necessary when old rubber belts are involved, so once that had arrived I put it on. But, unfortunately, that was not a solution. I found out that the wheel attached to the capstan motor shaft was slipping. It is a simple plastic wheel that had come loose from the shaft, probably because it had worked itself loose after being clamped on in the past. But now the wheel was just stationary and the shaft of the motor was spinning because of too much room inside. (blowfly pun intended)

Tascamninja to the rescue

I tried to glue the wheel to the shaft a couple of times using instant glue and some stuff that had worked great in the bathroom glueing shelves, but after a short while the wheel came loose again. I had to bite the bullet and order a replacement wheel made out of metal from the website https://www.tascamninja.com/.
I could not play any tape, or do some more tinkering while waiting for the arrival of the metal wheel, because the slippage was such that the capstan did not move at all anymore.
It arrived in about 2 weeks time, and as soon as I put it on, the tape speed was super reliable and super constant. Very happy with this solution!

recapping audio cards

I did do something in the meantime however. I recapped some of the audio boards. Although this deck is one of the last reel to reel decks to be made, it still has 40 year old capacitors in it, so the cards could use some fresh ones. But first, I had to get these cards out. Noone on the internet that I found managed to get them out, or, at least, did not mentioned it at all. Also the manual and the Service Manual had no mention of this removal. I was beginning to suspect that maybe they weren’t supposed to be removed.

But, eventually I managed to pry them out. They are just on connectors on a backplane, so definately supposed to be removed..

Knowing that I had to recap 8 boards, I made a document to use as a Quick Reference Guide with some handy info on the location of the caps and their value, the location of the pots for calibration, and the complete summary of the calibration procedure. I made it along the way as I worked on the deck.
The document is available at my site at: https://www.reeltoreel.nl/QRG/
You can find the latest version there.

After recapping I started to go through the calibration procedure. It is very labour intensive, doing everything 8 times, but the 0-locate-and-auto-play function of the Fostex is a very handy feature. But then I found an even better feature: you can repeat a section of the calibration tape endlessly. To achieve that, do the following:
– press CLR
– type timecode of beginning, like 0000 and press STO and then the number of the position slot you wish, like 1
– press CLR
– type timecode of end of the fragment, like 29 and press STO and then the number of the position slot you wish, like 2
– press CLR
– type ‘1-2’ and STO. display shows 2-1.
– make sure you play the section and the deck will automatically repeat the section.
I type this from memory so I hope this is correct 🙂

Testing after calibrating

I tested the sound quality when I was done and it sounds very good!

This is a sample from playing the calibration tape. But recording a sweep to 18kHz resulted in an equally straight frequency responce! (I forgot to take screenshots)

I recorded a youtube video where you can listen to a sample. You can find it here:
https://youtu.be/W2jRQmiAHzQ

TO BE CONTINUED

When you receive one or more DCC tapes of unknown origin, you have to assume that they have not been used for a very long time. This means that you will have to treat them in a certain way, even before putting them in a deck or playing them.

Why, you ask? Because the tape has been sitting still in its shell for a long time in the position that it is in now. It has become deformed at the place where it is guided along the heads. And even worse, there is a sticky residu that has formed on the back of the tape where there are two internal felt pads inside the cassette. I am not talking about the felt that is opposite the head. These felts are inside and can not be seen, unless you break open the case, and this can not be undone.
These two felt pads have started to ‘leak’ a sticky oily substance onto the backside of the tape in the place where they touch the tape. That sticky bit will prevent the player form playing beyond that point, because well, let’s say it slips and cannot continue. It will either stop or go to the other side continuously. Until you fix this, the tape will be unplayable.

So now you know why you will always have to resist the urge to play them right away to hear what is on them, and follow this procedure.

The steps.

1. So you will have to carefully pull the tape out. Be careful not to crease or rumple the tape, because that will surely create a dropout on that spot. To be able to do that first you must get the protective aluminium slider out of the way. Open it manually and place the wood part of a cotton swab into the round hole that has just opened up. It will fit almost perfectly and it will hold the slider open.
2. Then, carefully pull out a length of tape of about 30 cm or more, making sure that you get tape from both reels. I use a pair of bent tweezers for that. Lay the tape flat on a white clean soft surface, like a piece of paper, with the shiny side down. The shiny side is the side that the DCC head touches, where the magnetic information is. But the gunk is on the backside. Lay the tape so that you are now looking at the matte back coated side. You may need to use some tools to hold the tape in place, I use two erasers, they are heavy and soft enough to hold the tape down.
   
3. Now, and this is hard, look for some sticky residu. It will be hard to see, I use a desktop light and/or a handheld flashlight to find them. These patterns may also repeat at a ~10cm interval, so look up and down the length of the tape that you have pulled out the casing.
   I tried to make a picture:
4. Once you are able to see the problem area clearly, take a cotton swab and put some alcohol 96% on it and gently rub the backside of the tape. Do it gently as not to crease the tape. Rub very carefully but firmly up and down the length of the tape. Once you have done this a few times, you will get the hang of it. Make sure all the gunk is gone. Let the alcohol on the tape dry for a minute and check again and repeat if necessary.
5. Once you are satisfied with the result, you should see an even back-coated surface with no spots on it, and you can carefully manually rewind the tape into the housing using a pencil or the end of the cotton swab or as I do with the tweezers. After the tape has been carefully stored back into it’s shell, only then you can remove the wooden stick and close the lid, not sooner.
6. The tape has been cleaned, but the third felt pad (yes the one opposite the head) needs your attention now.
   I can not explain it better then Ralf did in this video: https://youtu.be/m3DcMtor_Ck

Now you understand why it is vital to not wind DCC tape when you first handle them and not put them in a deck first, because the deck will spool the tape a bit and after that you will have a hard time finding the greasy spots.

The 2 felt pads will of course still continue to shed oil a little bit, so always rewind tape after use. You will never know when you are gonna use it again, could be years from now.

See also my older post: https://www.reeltoreel.nl/blog/?p=779

So, the inevitable has happened……….again.

I have bought myself another piece of studio equipment to live next to my Studer. I got myself an Otari MX-55N-M

This one I discovered in Belgium. Yes, down there. It came originally from a guy with a small studio who sold all his stuff, but I got it from someone who bought some records from that guy and got the Otari as a side catch.

Getting it from Belgium required a road trip. In about 2,5 hours we were there, it was actually not too far across the border. The place was a 2^nd hand vinyl shop with a lot of records, but it had CD’s and DVD’s too. Oh, and comic-books, like Suske en Wiske.
The Otari was all dusty and just a little bit dirty, but not much. I brought a test tape that I had recorded with my Studer, as this is also a 2-track studio machine. I loaded the tape, plugged in my headphones and after fiddling with the knobs (how to put this thing in playback mode, and how to get signal from it) I got some sound from it and boy, did it sound good! Within a few minutes I knew that the sale was going to be made. I just had to do a quick check for any defects that I could find and if they would be a showstopper. I could not find any.

So then came the problem of transport. My car has a loading space with a flat floor that is 70 cm high and 100 cm wide. But the machine is with a cart and wheels and a meter bridge, as you can see from the pictures. We measured the Otari and thought: well, I may fit just laying on it’s back. And it did. So I didn’t have to dismantle the meterbridge or take the machine from the cart. That was good.

After we got home I cleaned it up. From beneath the layer of dust came a beautiful and almost pristine looking deck that has no marks on it of prior use. I started to play with it and explore it’s possibilities.

• the deck is tilt-able in various positions.
• it has a selection switch for 2 standards: NAB and IEC.
• it has another 3-way switch for selection of the reference fluxivity. I found that mine is calibrated to the following setting:
  • LOW – 250 nWb/m
  • MED – 320 nWb/m
  • HIGH – I don’t know, I haven’t got a calibration tape in that range. It will probably be around 510 nWb/m.
• high and low speed is now set to 19 & 38, but can be adjusted to 9,5 and 19.
• has a built-in oscillator with test tones of 0,1-1-10 kHz for bias calibration.
• reel size is selectable for LEFT and RIGHT reel separately.
• Slack tape can be picked up by pressing STOP.
• in- and outputs are fully XLR and +4dBu.
• the speed is adjustable from veeeerrryyy sllllooowww to very fast.
• Startup is very quick, 1-2 cm at 38 cm/s and its up to speed.
• Digital tape counter hms with goto and 3 memory positions.
• Recording and playback levels are adjustable at Standard Reference Level (SRL) or user controlled with a knob.
• Tapedump function, where the takeup reel is switched off so the tape is dumped.
• Button for listening (cue-ing) while spooling.

This is what I found so far. It came with a remote control with 10 meter cable, 2 Otari NAB adapters (which are very well built), and 2 aluminium reels.

I noticed that the left tape roller is making more noise than I would have liked, but it works ok and someday I will replace it. The heads are ok.

I am very fond of the smooth tape transport and the overall robustness of the machine. This Otari really is meant for heavy duty studio use.

I have not measured anything except for the fluxivity levels yet, but I expect the results like wow/flutter, distortion, frequency response etc. etc. will be very good. This will make a nice younger brother to the Studer.

Oh and by the way………..

I am looking for the manual, service manual and schematics for this MX-55 N-M.
I have found manuals for other models of the MX-55 family, but not specifically for the N-M. 

I found some manuals online, but they only contained the first two chapters.

So, if anyone know of these manuals, either in printed or electronic form, I would be happy to get hold of one. Please leave a message in the comment section.

Just a quick post about my latest acquisition: an AGFA Bezugsband 9,5 DIN 45513.

Some pictures here:

I think it was made in 1967, which is a good construction year 🙂

I came with some documentation, which I will post here for reference and archival:

Funny the documentation mentions 250 pico Weber per millimeter track width, where nowadays we use NanoWeber per meter.

Also, much use is made of time constants like 3180 µs, which corresponds with 50 Hz.

The tape appears to be in very, very good condition.

Sometimes, for instance when there is a sticker on top of the text that says whether it is a DCC90, DCC75 or DCC60, you have to guess the tape length.

Not anymore. I have created this diagram that show you just how long the tape is by looking at the casing (backside):

Handy!

It is really silly, some brands (hello BASF!) mention the length of the tape only once on the casing, and that is exactly where normally a title sticker would be placed. By looking at the diagram, you will no longer have to guess. The black circles in the drawing are holes in the casing.

DCC90 would mean 45 minutes on side A, and 45 minutes on side B, totalling 90 minutes. In practice, there will always be a few extra minutes of tape in the cassette. This goes for all the formats.

DCC75 is a bit strange, it was also not used in normal analog compact cassettes, but it would gave you 2 times 37 1/2 minutes. Not quite sure what the reasoning behind that was.

DCC60 is 2 times 30 minutes.

Have fun with it!

A box of DCC tapes

I have lying around a box with +/- 50 used DCC tapes that I purchased a while ago. DCC is the Digital Compact Cassette, invented by Philips. I have gone through them all, I listened to some of the tracks on each tape and sorted which ones I wanted to keep because of what was recorded on it. But most of them could be recorded over.

history

In my youth I had the privilege to experience the evolution of the compact cassette (CC), also a Philips invention btw. It had just made its way from dictafoon-like devices to the hifi/consumer audio market. And I witnessed its rise to a high end audio device (Nakamichi et al.). I have lots of CC tapes, probably my entire youth is recorded on those.

Then, in 1992, DCC came. But by that time, I had lost interest in recording music, or audio as a whole for that matter. So I never really noticed at that time. Also the Sony Mini-disc and DAT recorders went completely past me.

But when I rediscovered my audio hobby around 2010, I got very interested in these three now obsolete recording techniques. And fortunately you can get these devices rather cheap these days.

My Philips DCC600

I have all three devices now.

DCC surprised me in a very positive way. I got the same, familiar feeling playing cassettes that I had back then and that I had gotten used to. Also the deck that I got feels very solid and tape handling is direct. The DCC cassettes themselves feel very, very well thought about. Even the casing and the jewel box is outstanding well designed, even more so the prerecorded tapes.
On top of that, you get CD like audio quality. I say CD like, but I dare you to hear a difference. I was very, very impressed with the whole concept. In hindsight I wish I had used it back then.

squeaking

I noticed when I went through the lot of old tapes however, that some of the DCC tapes were making a squeaking sound when they were played. And some of those would not play at all. The deck, just kept clicking, the drive mechanism reversing a few times, before finally giving up. So I thought that these were bad tapes, or maybe they had some shedding or sticky tape or whatever and that they could not be relied upon. I was resolved to throw them away.
It was clear that it was the tapes that were the problem and not the deck, because most of the tapes would play perfectly and the problem ones would not.
I did some research into cleaning the head of the deck, but found out that this was not a trivial task, and that the fragile head could easily get damaged. The general conclusion was that you don’t clean it, unless it is absolutely necessary because all tapes play bad and it is clearly visibly dirty. So I didn’t.

forum

I researched some more and came across a post on a Dutch forum dedicated to Philips equipment. You can find the post here:

http://www.mfbfreaks.com/forum/viewtopic.php?f=5&t=10492

although I think you need to register first. And it is in dutch 🙂

gunk

It is described there that there are actually 2 problems: the squeaking and stuck tape. The squeaking problem is related to two pieces of felt that are inside the DCC cassette, not visible from the outside. There is also a third piece of felt, this is the familiar one that presses the tape on the head. The 2 problematic felts are located near the left and right rollers located inside the cassette:

These felt pads leave behind a greasy spot on the backside of the tape when the tape is not used for a long time. The deck has trouble running the tape past these spots and finally gives up. This spot is somewhat greasy, but also a bit sticky. This can be observed by pulling the tape somewhat from the cassette and checking the backside. It is not necessary to open the tape casing for this.

In order to do that, you must pull back the sliding cover and pull the tape out very, very gently with a small tool and make sure that you do not damage or crinkle the tape. You can see the gunk if you look very carefully. Let light shine on the surface. Make sure that the tape is at the problem spot when you stopped the deck and pulled the tape out, from BOTH reels.
It was impossible to make a good photo of this, so you will have to look for this yourself.

cleaning

In order to clean the tape, make sure you gently use some non residu cleaner, like IPA or alcohol or videospray cleaner 90 or whatever fits you, and gently rub the problem spot using a cotton swab. Hold the tape down while you swab, careful not to damage or wrinkle the tape. That happens so easily! Remember it’s the backside of the tape.

After that, let the tape dry for a moment, simply wind the tape back inside the housing and play the tape. Voila! It’s working again!

squeaking

For the squeaking problem: I fixed this by cleaning the felt located opposite the head as well. It won’t get completely clean, but just try to clean it as much as possible. I did it with a cotton swab with IPA. Just rub it a lot and also twist the cotton swab over the surface of the felt. Let the felt dry, this will take some minutes, and wind the take back inside.

After that, the squeaking was gone.

more tips

• Always rewind the tape completely before you unload and store it for a longer period. The (clear) leader tape, that’s what the felt presses against when the tape is rewound, doesn’t seem to have as much problem with the grease.
• Also don’t demagnetize the head! Never! It will ruin it forever. The head does not require demagnetization.
• Never ever use a cleaning cassette. The head is way too fragile and it will ruin it forever.
• If you absolutely have to clean the head, use alcohol or IPA . Preferably with natural chamois leather, but a cotton swab will work as well.
  Update: since I wrote this article, I *have* cleaned the heads of my DCC decks. Several times. No problem. Just use a cotton swab with IPA and be very gentle. I have taken out the mechanism however. See also my video:
• Avoid playing normal (analog) compact cassettes in a DCC player. Normal cassettes are not as clean as DCC cassettes because they have no tapeprotection and will make the head and the drive mechanism dirty. But most of all the tape formulas used in the analogue tapes are not as good as in the DCC and will shed oxide on the head. Better safe than sorry!
• See an other post of mine https://www.reeltoreel.nl/blog/?p=809 where I explain the tape length identification marks on the DCC cassette.

All info and the picture of the open DCC cassette are from the forum mentioned.

See also my new post here:

https://www.reeltoreel.nl/blog/?p=949

recording

Since I was very young I was intrigued by recording studios and the equipment therein. Can you tell?? (cue Studer A80 blog entry). A few years ago I had the opportunity to visit the famous Wisseloord studios here in Holland and needless to say that that was a thrilling experience. I am intrigued by the recording process, and mainly the tape part. But recording techniques in general have my interest. Like how to record a solo acoustic guitar, or how to record a symphonic orchestra. And as pure as possible. In order to do that, in the old days, you need a tape recorder, and there lies the base of my fascination.

tape hiss

Tape has a very nasty flaw: it introduces tape hiss into the signal. Tape hiss is clearly noticeable when you play around with consumer decks, giving a nasty hiss in quiet sections when you replay your recordings. This wasn’t so much of a problem when recording from LP or FM radio, which has a background noise that would make the tape hiss less noticeable. It became a bigger problem since CDs came around and the source of the recordings were becoming quieter or even dead silent.

If you wanted to minimize the tape hiss, you had to resort to quieter, and more expensive tape. You had to use the more high-end decks, and have them calibrated properly to that tape. Thorough maintenance became even more important.

studio – multitrack

If you thought that you had a problem with tape hiss at home, imagine the problem professional recording studios had with it. They have to conform to the very highest of standards because they are at the very beginning of the audio path, a path that would result eventually with you playing the record or CD at home. So they want to introduce as little noise (in general) to the signal as possible. Adding to the problem was the fact that multi-track recorders had 16, 24 or 32 tracks that all brought their hiss to the final 2 channel stereo end-mix.
Without proper noise reduction, even in the most expensive studio with the high end tapemachines and studiotape running at high speed the end result would suffer quite substantially from tape hiss.

noise gate

To circumvent this problem, studios often reverted to using devices like noise gates. When the audio level of a track would drop below a certain level, the noise gate would kick in and mute the channel on playback. This was of course done so that the downmix to 2 tracks would be more silent in the quiet parts. There is an interesting article on the use of Kepexes used by Alan Parsons on Dark Side of the Moon by Pink Floyd:

Dolby A

Dolby SR

The test

A very good acquaintance gave me some recorders recently. Among them was this beauty: The AEG Magnetophon AW-2.

As you can see, this recorder is from 1951 and is more than 65 years old! This particular device that I have was used in cinemas for sound reproduction. It sits in a suitcase for easy carrying around. It is quite heavy though. It uses platters and pancakes.

aufname und wiedergabe switch

on/off-volume and play/rewind switch

It is in a very reasonable condition. So far nothing seems to be broken beyond repair. My goal is to try to restore it so that it plays tapes again. I have not turned it on yet, it first needs a lot of TLC.
It appears that when used with ‘modern’ tape, this unit will nowadays produce stunning results that would exceed what would have been achieved back then.

For instance, there are three very heavy duty belts in this one motor machine. And with belt I mean like belt from a car engine. They are not in a condition to be used again, so these will have to be replaced. Also some of the electronic components, like capacitors and rectifiers are way beyond their normal operating life so they will have to go too. And then everything else that I will run into. Hopefully the tubes, there are a few, still work.

I have started to open the device up and clean it.

Immense belt

fltr: erase, r/w head capstan and pinchroller

the heads:

somewhere in this shell is the head

erase head before cleaning

oil’s well!

There are of course a lot of electrical components as well that need to be replaced. The first and foremost candidates are the capacitors, as there are certain types that don’t last for so long.

The device contains some paper-oil capacitors, like this one:

This will certainly have to be replaced before I can switch it on. I am told that ones like this will probably be ok:

The device is really beautifully made. When I ventured inside I fell in love with it more and more. But there are a few ‘minor’ issues to be resolved, like: where is this supposed to be connected?

I was told that this rectifier could go up in (probably very nasty) smoke:

So I will need to figure out how to replace it.

Fortunately I have got the electrical diagram:

I recently went on a road trip and came back with:

• a Philips N4522 2-track. This machine is in very good condition and has recently been serviced in Germany.
• another N4522 2-track, but this one is going to need some TLC. There are some issues with it, although none of them are serious problems. But I just like to fix it as much as possible.
• another N4520 4 track. I already have one, but this new arrival is not recording well. This has probably something to do with the levels, but I have not looked into it yet.

So now a have a total of 4 Philips N452X machines and that looks like this:

The two on the left with the white VU meters are the 2 track stereos, and the two on the right are the 4 track stereo machines.

Why speed measurement

For some reason I got interested in the speed of my tape decks. Why, you ask? I really don’t know. I was fiddling around with my vintage Philips multimeter, which is capable of measuring the frequency of a tone, when I thought it would be cool to use that meter to measure the 1kHz tone which is a the beginning of my (generic) calibration tapes. If the tape runs too fast, the measured frequency would be over 1000Hz, and if the speed was too low, it would be under 1000Hz. Suddenly I was fascinated by the fact whether my decks were running at the exact speed or not, or if not, how much deviation it had.

Now, you have to understand that I do a lot (well, actually not that much) calibrations of decks, but I never ever calibrated the speed. I never bothered before. I never had a problem with speed before. I never had the need to adjust the speed of a deck. So whatever I measured now, it would be the ‘default’ speed, untouched.

So I loaded the test tape onto all of my decks, connected the multimeter, and did my measurements.

The results were sometimes surprising, sometimes more or less expected.

And a graph says more than words, so:

What can we conclude from this?

• The Akai GX-77 really runs too slow

• almost all decks run within +/- 0,5% of the correct speed, which is within the DIN norm

• only 3 decks run a EXACT the correct speed: the Philips N4520, the Tascam 34B, and the Studer A80.

  • the Philips N4520 is know for its quartz/PLL like control, so this is no surprise

  • the Tascam 34B is a professional deck, but this result still surprised me

  • the Studer A80…..well…….need I say more?

enter the wow and flutter

So that has been a very interesting experience. But then I started thinking about variations of the speed, also known as ‘wow’ and ‘flutter’.

Wow is the slow variation in speed, with a period below 6 Hz.
Flutter is the fast change in speed, with a frequency above 6 Hz.

One can experience wow then playing a record where the hole is not centered, and the speed goes up and down all the time.

Flutter is more difficult to hear because it is so fast. It can be measured however. Flutter is created by faulty mechanics: a capstan here, a roller there, dirt, bearings, the overall construction etc. It would appear that less flutter makes for a more tight and clean sound.

There is always wow and flutter. No mechanical system is without them. It is just a matter of reducing it as much as possible.

causes of W&F

what are the causes of W&F? There can be many. The capstan or axis involved might not be perfectly round. The bearing idem. The system may not be perfectly lubricated. There may be dirt on tape guides, causing jerky tape motion (on a microscopic scale). Tape may not be slit evenly. Etc. etc. etc. There really are too many to mention here.

Remember: There is always wow and flutter. No mechanical system is without it. Realize that.

measuring wow and flutter

Now there I had a problem. To measure W&F you need a special device called a ‘wow and flutter meter’. Of course, you can rely on Philips that they had such a product back in the days. They did. It is this one:

But unfortunately I don’t own one. It is on my wanted list though. They are relatively rare, and sometimes expensive. If you have one, or one like this, and you would like to sell it to me, please let me know.

software to the recue

What to do now? Fortunately I found an interesting program on the web that is supposed to measure W&F. It’s called WFGUI. Here is a screenshot of the program in action:

That looks impressive, right? It can measure DIN 45507, wow < 6Hz, and flutter >6Hz. According to the documentation provided with the program it has been tested against some real life hardware W&F meters and the results were very very comparable. Good!

Now I had a different challenge with this program, as it runs on windows, and I only have Linux. But, I soon found out that it runs perfectly fine in Wine! Yeah!

You can select between the 2 most used frequencies in calibration: 3000 and 3150 Hz. (It also shows the measured frequency for your convenience.) Then there are 3 values to take notice of:

1. RMS (%): this shows the root mean square value in % of the current measurement
2. Peak: this shows the current peak in % using a nice needle display
3. maximum value of RMS and Peak in % during the last 10 seconds
4. and it also produces a nice waveform of the current fluctuation measurement (this is NOT the input or output sinewave)

Now, this is where it gets interesting. The program has different settings so it can measure different values. It can measure:

1. DIN (Deutsche Industrie Norm), or more specifically DIN 45507
2. wow: measure variations between 0,3 Hz and 6 Hz
3. flutter: measure between 6 Hz and 200 Hz

Now I will not go into too much detail here, but take it from me that the DIN norm is a sort of average between the 2nd and 3rd setting, i.e. it takes a little bit of both. If you want to know more, you can read the documentation that comes with the program, that has some nice graphs that show it all. That will explain everything. For my measurements I used all three settings (DIN, wow and flutter), and I measured all 3 values (RMS, Peak and Max RMS)

calibration tape

The problem I had is that measuring these kinds of W&F is traditionally done with a tone of either 3000 or 3150 Hz. And, more specifically, you need a special (and expensive) speed calibration tape with that tone on it recorded in a very precise manner.
Guess what, I do not own such a tape. They are quite rare actually, although you can still buy them at MRLtapes.com I presume. My regular Bezugsband (German for calibration tape) has a 1kHz tone at the beginning, but that is used foor reference level measurements.

making my own test tape (huh?)

So I thought about it a bit and I knew I was never going to buy that special speed calibration tape (>€130). What if I took my trusted tone generator and used that to record and immediately playback a 3kHz tone? Of course that would be an imperfect recording, because it would contain W&F, but wasn’t that just what I wanted to measure? And if you play back that recording on the same deck, doesn’t that give a good indication of the performance of that deck? The more I thought about it, the more sound (pun intended!) the idea seemed to be.
Doing that would mean that I could not refer to a know calibrated source anymore. But I’m not interested in those figures. I can now compare the performance of my decks between themselves, see who stands out, either way.

Some people may argue that one can not make his or her own test tape. Of course that is true. But that is not the purpose. I want to measure the W&F of a deck. I was going to use the deck itself to record a tone and play it back at the same time.
Some may argue that that is not a good way to measure W&F. I disagree. I will explain.

When you record the test tone on the DUT (device under test), it wil will be recorded with wow and flutter. That is clear. So it will be an imperfect recording. So what. That is what I want to measure, so that seems ok to me. Let’s assume that the recorded tone will contain at the maximum an amount of W&F of n. If you would examine the tape in a laboratory you would find max n W&F. The DUT obviously has a W&F value of n. So at some points in time the value would be 0 (zero), at some points in time the value would be n (the max value) and most of the times some value in between. This will vary constantly, and it will vary very quickly.
Now I play back this tape on that same deck that has that W&F reading of max n. So again, some of the time that tape will play back with no W&F, and some of the time it will play back with the maximum value of n, and most of the time some value in between. The resulting W&F will be 2*n at the most, to be exact that would occur when a piece of the tape that was recorded at that moment with n W&F value will be played back at a moment that the deck plays back with a W&F value of n.

testing

So again I dragged all my decks to the test bench and hooked them up one by one. Now reading the values from the program proved to be quite a challenge, because the readings vary a lot and change very quickly. So i took a sort of ‘observed average’ , as I would call it.

I did the measurement all at 19 cm/s, because that is the common speed all my decks have. I have noted where I used an other speed, like 38 cm/s.

 So what are the first results?

• the kind of tape used for the test has a great influence on the end result. At first, I used an older tape and got very different (much much worse) result that when I later used a brand new tape of the same type. This was kind of shocking!
• tape speed has a great influence on the stability as well. All deck produce much better results on 38 cm/s than on 19 cm/s. A few quick test at 9,5 cm/s were even worse.
• I did not find any difference when using the beginning, the middle or the end of the tape, but I have not tested that on every deck. I suppose if a deck has too much or too little tension, the position on the tape would matter

After a lot of testing I had a spreadsheet full of values that were very difficult to interpret. So I made a graph out of all the numbers and this is the end result:

This may seem overwhelming and confusing at first, but it really isn’t. Let me clarify for you.

• On the X-axis (lower) you see the three different measuring methods for each deck: DIN, wow and flutter.
• Each deck has an corresponding line in the graph for these 3 values.
• Lower values (lower lines) are better.
• Legend is on the right, showing which line corresponds to which deck. It is a bit crowded but bear with me.

Here are the conclusions:

• the most upper line, meaning the worst deck, is the Teac A-3440 at both speeds. This is one of my oldest decks.
• for fun I included one turntable, my Technics SL1900. That is the blue dotted line. I found that I had a test record with a 3kHz tone on it, so I could actually do the measurement. I was surprised to see that the values produced were in the same range as the tape decks. However, unlike all decks, the wow is the worse value of the three.
• the best deck is the Studer A80 @ 38 cm/s. Of course, Studer was famous for their mechanics and always paid special attention to the tape path. Studer used the best available bearings and even had special lubricant that had to be used.
• a very positive measurement is the Philips N7300, the light blue line in the lower part.
• in general, the newer decks (Philips N7300, AKAI GX-77 and Teac X-2000R) are in the better part of the graph. Maybe it is because when they were manufactured, the W&F was better under control, or it is simply because they are fewer used.
• the ReVox PR99 (also a Studer brand), the dark blue line, is giving the A80 a run for his money.
• the results for the dual capstan decks that I have (Teac X7, X7-R, and X2000-R) are not as I expected. Yes, they are good, but they are beaten by single capstan decks. The whole idea of the dual-capstan system is that that would eliminate (well, minimize) W&F values. I think the system fails to deliver. And on top of that, it introduces a lot of unwanted negative side effects, that are most notable after all these years when the decks age.
  The result of playing in the reverse direction were mostly comparable to the forward direction, so no real diffrerence here.

There you have it. I hope you liked my article. If you have any comments, please leave them below.

-edit  31 may 2016-

I have recently purchased a ‘real’ wow and flutter meter, the B+K Precision wow and flutter meter 1035. 
I have used this meter to check some results from the softwareprogram above, and the results fortunately match. Great!

Today I picked up a mint condition Akai GX-77

This photo is from the internet; I will upload a real photo later.

The unit does not work; the two pinch rollers won’t come down all the way and the tape loader roller in the middle is stuck at the top so I will have to fix that. The middle roller is necessary for tape loading and unloading.

auto reverse

This unit is a real beauty. It is an auto-reverse deck, and is capable of playback AND recording in both directions. Therefore it has 6 heads and two capstans. The capstans are always spinning, in opposite directions and only the pinch roller associated with the current playing direction is engaged. The result of this is that switching tape direction takes only 0,4 seconds.

other features and specs

• It has user adjustable bias through a dial on the front
• capable of using EE tape (chrome)
• L+R recording level knobs have a 3rd ‘master’ knob which controls both channels
• output level is adjustable from the front
• digital tape counter with optional backup adapter

Here are the full specs:

• Track system: auto reverse, 4-track, 2-channel, stereo system
• Heads: 2 x GX record, 2 x GX playback, 2 x erase
• Motor:
  • 1x FG servo direct-current capstan motor
  • 2x direct-current reel motor
• Reel size: up to 7 inch reel
• Tape speeds: 3 3⁄4 7 1⁄2 ips
• Wow and flutter: 0.03% (7 1⁄2 ips)
• Fast forward rewinding time: About 80 seconds
• Frequency response:
  • 25Hz to 33kHz ±3dB (19 cm/s) 
  • 25Hz to 25kHz ±3dB (9.5 cm/s)
• Signal to Noise Ratio: 63dB
• Total harmonic distortion: 0.5%
• Input: 70mV (line), 2mV (DIN)
• Output: 0.775V (line), 0.3V (DIN)
• Dimensions: 440 x 244 x 227mm
• Weight: 17kg
• Power consumption: 28W

work begins

The first thing I did after opening up the unit was checking the belts. It has three: one capstan belt and two for the spooling motors. The capstan belt was ok, it was still feeling like it should, but it was probably a little stretched because it didn’t feel very tight to me. I ordered a new set which included the 2 spool motor belts for €12,-

Once I had the belts replaced, it still would not engage the middle roller. It is supposed to go up and down.

akai gx-77 with the roller stuck in top position

Searching on the internet I found that there is a large white gear wheel located behind the 2 flywheels which controls the up/down movement of the roller. When I manually engaged this wheel, everything worked. Tape was loading fine. But I had to help it every time. So what could cause this?

old dried out grease

The answer is: old grease. Old grease is grease that, over the course of 35+ years, has started to dry out and instead of lubricating, it starts clogging things. So I had to take the unit further apart to the point where every gear wheel is removed and cleaned.

gear

akai gx-77 plunger arm dried out grease

This plunger arm was very slow returning to it’s default position and probably missed the next turn when the hook had to engage. So I fixed that.

Also the right flywheel (as seen from the back) has a gear on it. This was also very loose on the spindle and could be turned quite easy. I put it in a ultrasonic bath and tried to clean it with IPA. Now it moves a lot rougher.

flywheel with gear

So after almost completely removing all moving parts and cleaning and re-lubing them I carefully put the unit back together to try a cautious test. Here is the result:

As you can see the unit works smoothly now. At the moment the deck is still mostly disassembled but I will put it back together again and then start the mechanical adjustments and followed by the electrical calibration.

As you can see the beauty is put together again and is recording and playing like new. Very happy with it!

The JVC, my first 8-track

I possess one JVC 8-track player, the JVC ED-1240, a very nice and featureful deck.

The main reason I am interested into the 8-track world is, of course, the fact that aside from 4 track reel to reel tape, this is the only format that has discrete quadraphonic. I realise that the quality of the sound is not up to par like the Studer, or even a normal tape deck, but it will have to do 🙂

The first time I heard this 8-track deck i was very disappointed with the sound. So I found myself a challenge at hand. I have cleaned it, repaired it, refurbished the motor, changed the belt, demagnetized it, changed all the caps on the power board etc. etc. and eventually it performs quite well.

tapes

I cleaned and repadded the 8-track cassettes that I have. I even found it necessary to lube the inside axis a little bit to prevent squeaking and to lower the rumbling and the mechanical noises from the cassettes while playing. The end result is quite pleasing I must say.

But there are still some issues with the player, so if anyone has the schematics, or the service manual, please contact me!

Quadraphonic 8-track

But this post is not about this deck. The JVC is just stereo. This post is about a new one I just bought. A few weeks ago I ran, quite to my surprise, into a quadraphonic 8-track player from National Panasonic, the RS-845us.

Here it can be seen on top of my JVC. At the moment the picture was taken, it was playing a quad tape as you can see by the ‘4 channel’ indicator light. In this mode there are only 2 programs, not 4 as you would find with a stereo tape.

problems ahead

But, alas, this player also came with its problems. I tested it before I bought it, and it was ok as far as I could test it at the time. Coming home, when I could test it for a longer period, I found the same issues I had with the JVC, like irregular tape speed etc. Also the sound was not as good as I was used to from the JVC.

So I did my usual fixup things, like cleaning everything, oiling etc. No good. I quickly found out that the voltage over the motor was dropping several volts at times. I suspected the power circuit board. So i tested the diodes, they were ok. I tested the caps and replaced them all nevertheless, although they were not really out of spec. Unfortunately, that did not help much. New belt. There was some improvement, yes. But still the voltage drop at times, from 12-13V all the way down to 6-7V. The motor would almost stop turning then. And it would squeak a lot.

motor fix

The solution would be to open up the motor and fix it there. That is always a bit tricky but I did it before and I therefore have some experience with the procedure. It went surprisingly well. There was some dust (powder) from the magnets and from carbon brushes. I also scraped clear the spaces between the 3 parts of the rotor (I think it’s called) so there would not be any more shortage. I reassembled a clean motor and lubricated it. When I tested it it it ran very smooth and silent. Cool! The tape speed is also a lot more constant than before. Also the sound seems to have some more punch to it. Maybe it has to do with improved tape-head contact, or with a more stable power supply to the audio board due to the changed caps.

audio fix

But I also have the capacitors ready for a complete overhaul of the audio prints. So somewhere in the next few days I hope I will find time to replace them and see if that will fix the muffled sound. I don’t expect a lot from that action, but hey, that’s what this hobby is all about.

[Update:]

Yesterday I did the recapping of one of the 2 audioprints, i.e. the print for the front channels. This is the result:

When I tested the audio after the process, I found that there was very little, if any at all, change in sound quality. I had made a ‘before’ recording so I could compare the sound.

So I am a little disappointed about that, but I guess the 8-track format was never intended for good audio quality, At least I didn’t break it. 🙂

So while I have all the caps here to replace the rear print also, I will not do that; too much trouble for little or no result…..

Anyway the unit is reassembled now and it playing happily.

My new addition to the collection is already a few months old, but here is the blogpost anyway.

It is an AKAI 1730D-SS Quadraphonic tape deck that I received from the first owner.

[picture shamelessly stolen from the internet]

When picking it up I tested the deck and only 2 of the 4 channels worked. Only the REAR. Strange. And one of those 2 channels failed intermittently.  When I got home I opened up the deck and had a look at it. I put on one of the many quadraphonic tapes that came with the deck. Quickly I found out that cleaning and fixing the 2ch-4ch switch solved that problem. That was easy.

But the FRONT channels still did not work. At all. No sound, no noise, no click when the deck was powered up, no nothing. I sure hope the playback amp hasn’t died. It appeared there are 4 audioprints in the machine, see picture. At the left of the picture you see the 2 playback prints, at the right the record prints. Each print is stereo. FRONT at the top, REAR at the bottom.

[picture showing the 2 right audio prints removed from their slots]

By turning the pot on the prints I found out wich print was unresponsive (I am going to recalibrate it anyway, so that did not matter). I took it out of its slot for further inspection.

Unfortunately I could not find anything wrong with the print or the components. No burned components, fused caps, burn marks, loose contacts, nothing. I was beginning to feel that I was reaching the end of my knowledge because I have not got expert knowledge of electronics.

When I was going to put the print back into its slot, I noticed something odd about the slot. The slot is a kind of very crude ISA-style slot type, like in the first IBM PC. I saw that the plastic of the slot was broken! That caused the print to not make good contact with the pins in the slot. See the pictures.

When I carefully with my hand pressed the slot so that the print made contact, everything started working! Eureka! ‘There, I found it.’ 

Now how to fix it. Glueing was no option. I soon thought of a tight tie-wrap, that might do the trick. I pulled the tie-wrap as tight as I could. See the pictures.



Then came the next issue. The deck did not want to record at all. I soon found out that the other slots were damaged too (duh). So another set of tie-wraps to the rescue. I put tie-wraps around everything and reconnected it all.

It worked ok, but not reliable. Even when just looking at the recorderprints weird effects on the signal were noticable. First I doubted my tie-wraps, but that solution DID work reliably at the playback prints. Just to be sure I sanded the contacts on the prints and voilá: everything worked flawlessly! So I put the print in and secured them thoroughly. ‘There, I fixed it again’

This is a machine from the early seventies, that has been idle for who knows how long, and I’m surprised at the audio quality.

Now I need to demag it, and calibrate it. Then it will probably sound even better!

[edit]
Although it sounded good at first hearing, further test revealed there was noise in the left front channel ,even when it was not playing. Same on the right rear.

So I replaced the 2sc458 transistors on all the audio prints. After that, I eagerly turned on the machine.

No signal at all!

Remember that I am only doing this as a hobby, I am not a professional guy and I know very little about electronics.

Searching and searching and searching again I found out that the new transistors I had ordered, had a different pin-out than the ones that came of the print. I had put the new ones in with the same orientation as the old ones, that seemed logical but that was incorrect.  They simply needed to be turned around, so that was relatively simple. I did not need to cross their legs.

After that it worked again, fortunately! Apparently these guys can withstand some rough handling haha. And the best part is, that the deck is silent again now, the noise is gone. I recalibrated it and it sound great!

Recently I obtained, for a fair price, a Teac X-2000R. It is a stereo reverse deck, with 6 heads, which means it will record also in reverse. It has dbx built-in, and it can use (chromium) EE tape. A digital counter in hours, minutes and seconds and bias fine tuning on the front panel is also provided. It has 9,5 and 19 cm/s speeds.

Dual capstan

Another prominent feature that is has, however proved to be a very problematic one. The deck is equipped with 2 capstans. This feature was introduced into the later tape deck models, almost towoards the end of the tapedeck-era, to further improve tape-head contact and wow&flutter. These assumptions are correct, and they work well when the deck is in new condition and everything is well calibrated and up to specifications.

On this deck however, things were not new and not up to specs. The dual capstan design is a challenge in itself, but things are further complicated by the fact that it has to function in reverse as well. On this unit that is accomplished by a belt that ‘travels’ across different parts of the flywheel that varies in diameter.

Problems

After the initial cleaning, optical and technical/mechanical, it soon became apparent that the tape handling was not up to par. Sometimes the tape would ‘run away’ from the heads a bit, resulting in loss of audio or worse: tape jam. Things were even worse when playing in reverse. I soon found out (through the service manual) that there were adjustments to be made to the tape tension.

Tape tension

But the increased tape tension was not the solution. When it was ok in forward play, it was bad in reverse. Or vice versa. Or it differed when loading large reels (26 cm) compared to 18 cm reels. Or it was different with different brands of tape. And that wasn’t even reproducible across sessions. And the next day it would be different again.

Solution

Of course the first thing I did was to change the belt. I have a blog of that action, but it is in dutch. So here it is, sorry only in dutch:

====== BEGIN OF DUTCH TEXT ======

Nou, de siem-snaar was snel binnen, en vanochtend was het toch regenachtig, dus aan de slag gegaan met het vervangen van de Teac X-2000R snaar.

Ik heb de uitbouw gedaan volgens de foto’s in deze thread, daar heb ik heel veel aan gehad:

http://www.audiokarma.org/forums/showthread.php?t=59729

Daar wordt behandeld een X-1000R, en die blijkt toch niet helemaal hetzelfde van binnen. Ik kwam meteen al een grote print tegen die in de weg zat, deze is er niet in een X-1000.
Hierdoor kon ik niet bij de bovenste schroeven komen van de plaat waar de capstanmotor op vast zit. De print moest dus los en daarna omhoog gelift worden, er zitten helaas te veel kabels aan vast om hem compleet uit de weg te krijgen.







Na wat kabelbomen losgeknipt te hebben, kon ik de print opbeuren en omhooghouden met een plakbandrol.





Hierna kon ik er wel goed bij.





Nog even de plaat losknippen van de motor zelf:



Er zitten alsnog wat kabels in de weg, maar die zitten gelukkig vast met connectoren. Loshalen die hap dus.





Eindelijk uitgebouwd. De 2 vliegwielen in de achtergrond:



De 2 vliegwielen verwijderen. Goed links en rechts uit elkaar houden.



Overigens viel me meteen op het grote verschil tussen de snaren: de oude is een stuk langer dan de nieuwe. Dus of de oude was wat uitgerekt, ondanks dat het rubber nog vers aanvoelde, of de nieuwe is ietsje te krap. Of allebei.

Toen heb ik alles losgehaald en schoongemaakt en opnieuw gesmeerd: de vliegwielen, de aandrukrollen en het aandruk mechanisme. Oud vet verwijderd, nieuw vet er op. 

Alles zeer grondig schoongemaakt. Vliegwielen er weer op, meteen de washers aan de voorkant weer over de capstans gedaan. Nieuwe snaar er op:



En toen alles in omgekeerde volgorde weer vast. Het was een beetje pielen om de snaar weer over de motor te krijgen, maar het ging uiteindelijk wel. Als laatste de print weer laten zakken en vastgezet.

====== END OF DUTCH TEXT ======

Unfortunately this was not the solution that fixed it completely, although the situation improved a bit. Even increasing the tape tension to incredible heights (or incredible lows) did not fix it. So I went online to find a solution.

Several fora messages suggested that everything in the tape path was relevant to my problem. From tape guides, to the rubber idle rollers, to lubrication of the tension rollers, to the state of the rubber of the pinch rollers. From the tape tension to the smoothness of the tape in question. It seems that everything was related to weather the dual capstan principle would function correctly.
So I soaked the pinch rollers in detergent for one night. The rubber was much softer and grippier after that treatment. Again, a little improvement, but no permanent fix.
I switched the pinch rollers from left to right, to see if that would improve the situation. No way.
I lubricated all the moving parts, I installed the little O-rings that were required for the rollers to function correctly (expensive little buggers!). Still there was no definitive solution that worked all the time.

Meeting

In April I went to a meeting of fellow tape deck enthusiasts where there was an opportunity to work on decks. I tagged the deck along, and fortunately there was an expert who was willing to take a look at the Teac. Immediately he found one problem: I had put the tape tension way too high. We measured more than 100 gr. when 50 was required. He used a Tentelometer, a very rare instrument which he posesses. It looks like this:

After we adjusted the tension to 50 grams, the tape handling was still not good. In fact, it was worse. Then, the guy helping me thought the problem might be in the capstan motor, of which the carbon brushes would have degraded over the years resulting in less powerful drive. So, on the fly he opened the capstan motor, removed the old brushes and (he had them available) installed new ones. After this, still no good. But I had new carbon brushes! Jeey!

At the meeting we came to the conclusion that maybe the new belt I had put in was not an original Teac spare part but a fake. So I went to an official Teac repair center and got myself an original Teac approved belt. After I installed that belt, still not any better. Grrrr.

The solution!

Eventually, as a sort of last resort, I took the Service Manual and went through all the mechanical adjustments mentioned there. Some I could do, some I couldn’t. For instance, I haven’t got a spring scale necessary to measure torque. And then, finally, after I adjusted the pinch roller pressure to a much lower value, it suddenly all seemed to come together.
When I adjusted for a very, very low pressure of the pinch rollers on the capstans, even to the extend that I could very easily stop the tape when playing, the tape would indeed run fine along the heads and reverse play was no problem. So I adjusted for a little more pressure but anyway now it works great and I haven’t got any more tape issues.

Great sound

After that, I planned to calibrate the deck. This also proved to be not as easy as I thought. The service manual mentions a lot of steps and I think I did them all 4 or 5 times. The problem was that the record calibration required EE or Chrome tape, which I did not have. Fortunately I could borrow a Maxell XLII tape from a friend.
Bu after I calibrated the deck using the EE tape, it did not sound good. The level was way too low. And when I recorded at a higher level, the amount of distortion was unbearable. Something was not right here. After advice given to me on my favorite forum, I used normal tape to do the calibration. It still wasn’t to my expectation. I’m not sure what fixed it, but I repeated the calibration procedure several times, and after that I found the recording to be very good!

EE (chrome) tape

So the deck was performing as it should, reversing as it should, handling tape as it should and now I was to test it with EE tape and with the dbx noise reduction and dynamic expander on.
The result was, well, as close tot the original sound as you can hope for. The high tones were beautiful, and the sound was quiet as a CD. Out of nowhere the music starts. This is tape technology at it’s finest. I am very happy with my new tape deck.