The Teac A3440 is a classic reel to reel tape recorder from the late 1970′s, early 80′s significant in that you could use it to make 4 track multitrack recordings at 15 inches per second, the professional tape recording speed. At the time there was precious little else around at the price to do this so this machine was perfect for small bands and studios who didn’t have the wallet busting amounts needed to buy a larger format 16 or 24 track recorder.

While the A3440 isn’t the last word in high quality analogue recording it has had some significant users in the past none less than Lee Perry who’d used the earlier but similar A3340 on his Heart of the Congos album.

the problems

This machine didn’t initially look heavily used, the heads had little wear but it hadn’t been used in a long while and there was a heavy build up of tape residue on the whole tape path. The capstan was very dirty and i these cases xylene is a more effective cleaner than IPA but don’t get it anywhere near plastic!

This A3340 powered up but the right hand tension arm was hanging at an odd angle and it wouldn’t play or wind.

Time to take it apart – the fake wood sides and rear panels get removed and it’s pretty easy to see what one of the problems is.

A very nasty ‘melted’ rubber capstan belt that took a fair while to scrape and clean off with IPA.

A new belt has purchased online – Teac parts in the UK are a big pain to get hold of through the official channels and I’ve given up contacting them unless I’m desperate for a part I can’t find anywhere else.

The other lack of drive was caused by broken micro switches. When the right hand tension arm is moved up, two micro switches behind the front panel switch power on for the capstan and reel motors. As this arm can get a los of use / abuse it’s common for the micro switches to crack. Both on this example were broken as was the small plastic piece that stops the arm moving too far down.

the repairs

To replace the capstan belt:-

1. Remove both screws holding the capstan flywheel against the front panel.
2. Make sure the flywheel is cleaned of all old belt debris.
3. Make sure the motor wheel is cleaned of all old belt debris.
4. Refit new belt over small motor wheel, then flywheel.
5. Replace bracket remove in part 1, making sure you’ve cleaned off the old grease and regreased where the end of the capstan shaft can run.

Although not essential I took the opportunity to remove the whole capstan shaft, clean, check for wear and reoil before putting back. If you do this you will need to reset the endfloat though.

microswitches

It’s not possible or worthwhile trying to repair the microswitches as the modern equivalent that fits perfectly is very cheap. Two were purchased from Farnell and replacement is just a case of:

1. Unscrew and move away the control PCB to get more space
2. Make a note of or photograph wiring connections for switches.
3. Unscrew and carefully desolder the existing microswitches.
4. Connect wires and solder the new switches in.

The arm end stop was repaired easily with strong super glue and after many hours is still holding.

is it working?

In a word, kind of! The belt and microswitches got the deck and transport moving. It will pull tape and make a noise which is great but an annoying intermittent problem started to appear after some initial testing.

When play or wind are selected, large solenoids clunk and release the reel brakes and move the pinch wheel. This was working BUT occassionally and only in play the right reel brake solenoid didn’t move, leaving the brake on, causing the tape speed to slow, back tension to increase and wow to go crazy!

See our next post for the repair of this problem…

We’ve been a bit quiet since last year on our blog here primarily because we have been processing large, ongoing audio archiving digital migration job for Mood Media Ltd

In essence audio archiving, digital migration, transfer or digitisation, whatever term you prefer, is conceptually simple, one analogue or digital format is moved to another, usually digital, format.  It is in the details of this process that things can get complex requiring experience, specialised tools and often custom-built solutions.

This job has many factors not just attaining the best analogue transfers but also addressing the organisation of such a large amount of 10.5″ reels, their digital management once migrated and the creation and management of metadata.

The archive, dating from the 1950′s up the early 90′s, even though it had been stored in less than ideal conditions, is in generally good condition. An interesting and often fantastic sounding collection of recordings and music are evident all recorded at 15ips in mono, stereo and some encoded with Dolby A noise reduction.
There are specific issues with some tapes that are not uncommon with older audio tape:

• Binder hydrolysis or sticky shed syndrome (SSS) has affected a significant proportion of the archive in particular Ampex branded tapes from the 1980′s onwards.
• Some glue used in spices from the 1980′s spreads over time and tends to stick layers of tape adjacent to the splice together. This can be a very frustrating problem which rarely damages the tape but affects the tape tension on certain sections so that head to tape contact is momentarily lost and a very obvious audio glitch noticed.

The Analogue side is important to get correct initially but the management of the data is critical with such a large number of reels and individual artists and tracks. Each batch of tapes would be catalogued in a .csv file. The digital audio transfers were split into individual tracks which then had to be renamed based on the data in the .csv file. This was handled by some php coding which matched the metadata supplied with the files, checking for inconsistencies then renaming the .wav files on one of our servers.

This type of customised workflow can be tricky to setup initially and find the balance between accuracy and speed but once done a smooth workflow is possible maintaining high quality, accuracy and value.

We specialise in tape transfers, especially cassette tapes.

Tascam, Fostex and Yamaha sold cassette multitrack recorders in the golden days of home recording in the 1980′s and 90′s. The 4 track format was especially popular but an 8 track format was also developed that squeezed even more out of the small tape width of the cassette.

We love these cassette formats and their accessibility helped start many musicians’ careers. Unfortunately one of the best 8 track machines, the Tascam 238 Syncaset also suffers from a common and frustrating problem that renders most of these machines useless over time, the dreaded direct drive capstan motor failure…

The 238 and other 8 track and high quality stereo tape decks, the 688 and 122 MkII and III, used a direct drive capstan motor for precise speed control and reduced speed variation or wow and flutter (w/f). The circuit that controls this motor fails in certain ways causing lack of speed control and in our case the capstan motor wizzing away at a crazy speed, not the 9.5 cm/sec that it should do.

This here is the culprit  – you can see the attempted repairs which didn’t ultimately work.

One common failure is that the surface mount electrolytic capacitors fail or their capacitance changes to such an extent to cause speed problems. These can be changed for standard through hole caps but you do need to be very careful as the tracks are damaged very easily – good tools are essential.

The other point of failure in the circuit is the BA6304F SO16 IC – we even changed this but the motor still didn’t turn!

There was some suggestion from previous repairers that the grease at the end of the capstan flywheel hardens over time. increasing the friction and causing problems with the circuit.

This can become frustrating quickly, especially when you have a large archive of cassettes to digitise.

When we can’t repair we reluctantly do the next best thing and buy the whole replacement part but this is another exercise in frustration. Teac parts and Teac UK don’t have any european supply of this capstan motor (part no. 53700075-01) anymore as of November 2011. Interestingly about 6 months ago they did at around £60 GBP, then about 3 months ago they had one left at £160 GBP!

Lots of emails later to Teac US, Teac Canada and Teac Japan there seem to be a nice stock still on shelves somewhere and at reasonable prices UNTIL you ask them to ship to the UK when you discover they can’t do this and I’d need to go through Teac UK!!! I’m pretty persistent but I gave up finally even though some of the support staff tried to be pretty helpful.

We find support for older machines from the original manufacturers is not good generally and unreasonably expensive when you can find it. This is similar across audio and video, semi-pro and professional products. Some companies are easier to deal with and have better parts situation than other but stockpiling machines, parts, manuals and obsolete knowledge is the best course of action.

What we finally did that worked and was a good solution was purchase 3 Tascam 122 Mk III stereo cassette decks which use the same but a later revision of the capstan motor, (part no. 53700121-00).One machine was donated for the cause and the capstan motor removed, modified and refitted in the 238. The 122 motor has a few factory extras, such as these resistors, shown here:

You also need to solder / desolder the speed pads, to change the motor speed from 4.8 cm/s to 9.6 cm/s that the 238 needs to run at.

It’s also a good idea once you’ve got the capstan motor apart to clean the old grease from the seat, check the end float which can be adjusted using the screw shown on the left and apply new grease to the capstan end.

Put it all back together – be careful to solder the wires to the motor to the correct pads – they’re different on the 122 and test… Ours worked almost perfectly.

As the transport hadn’t been used for a while the reel motor would intermittently stop as if sensing the tape end. This can sometimes be loose counter belts but on the 238 it’s a digitial counter. We cleaned up the leaf switches on the transport top and also sprayed a small amount of deoxit into the inside of the reel motor. A bit more use and it finally worked to spec…

For 4 track and 8 track multitrack cassette transfer please contact us for more information.

Digital formats are great and the ease of access and distribution that an Audio CD and a Video DVD offers is fantastic at the moment with cheap players, recorders and personal computers. In the audio and video tape transfer business, getting your business website to rank well in the search engines for key phrases such as ‘copy tape to cd’ or ‘convert video to dvd’ should get you a steady stream of enquiries from people who, primarily, can’t access their tapes any more. The preservation of the valuable recording is usually perceived to be CD or the DVD.

My  concerns with the ‘…to CD’ and ‘…to DVD’ process are:

• Optical Media that can be burnt using computers, CDr and DVDr, are not as reliable as commercially produced or ‘stamped’ discs and their quality and longevity can vary between different manufacturers.
• The digital video stored on a DVD Video is a lossy compressed format that depending on the original footage and if pushed for too much running time can easily make the digital transfer look much worse than the original – not a good reason to transfer the tape!
• The everyday nature of audio and video discs and some of the domestic technology we now have has conspired to make what is and can be a complex, time consuming and often problematic process too simple in the eyes of many people.

Since the early marketing of the compact disc in the 80′s by spreading jam on it to highlight it’s resilience compared to vinyl or tape many of us have been taken in by this and have a skewed belief on the longevity of CD and DVD optical media. In fact optical media can degrade and become unreadable and unplayable in a very short space of time even with good storage practices. Tape based storage formats of course have their numerous problems but they can often be incredibly resilient over time even in under poor storage conditions. When tape does degrade it is rare event when it cannot be treated and made playable .

CDr and DVDr Longevity

Optical media that can be burnt form CD / DVD writers uses a dye that the laser alters to create pits that are read back as digital data. CD and DVDr discs use different dyes due to the different data storage needs. The dye can be affected by bright sunlight and compared to commercial or ‘stamped’ optical media it is not as reliable.

In an exhaustive Library of Congress report, http://www.loc.gov/preservation/resources/rt/NIST_LC_OpticalDiscLongevity.pdf  ’good quality’ CDR media was expected to last ‘several decades’ and be more reliable than DVDr media:

It has been speculated that the archival quality of CD is superior to that of DVD. The basis for this lay with the physical differences in CD and DVD optical media and the maturity of the technology. In particular, the size of the bit markings in CD media not only reduces the relative effect of media or dye degradation, but also means that more stable, less sensitive dye may be used.

Out with old, in with the new

The march of technology can cost us money while making huge profits too. This of the CD reissues of old back catalogue albums, the pressure to replace vinyl collections with CD’s, video tape collections with DVD’s or CRT televisions with LCD or plasma screens. New technology that is marketed as better in many ways yet it can be worse too

One of our audio and video archive customers has a large collection of AVCHD video files that are stored in 1.9GB ‘chunks’ as xxxxx.MTS files. All these files are of 60 minute and longer duration and must be joined, deinterlaced, re encoded to a suitable size and bitrate then uploaded for online access.

This is quite a task in computer time and file handling. These small domestic cameras produce good HD movies for a low cost but the compression to achieve this is very high and does not give you a file that is easily edited. The .MTS files are MPEG transport stream containers for H264 encoded video.

There are some proprietary solutions for MacOS X and Windows that will repackage the .MTS files into .MOV Quicktime containers that can be accessed by MacOS X or re-encoded to a less compressed format for editing with Final Cut Pro or Premiere. We didn’t need this though, just a reliable and  quick open source workflow.

1. The first and most important issue is to rejoin the camera split files.
   These cameras use FAT32 file systems which cannot handle individual files larger than 2GB so they split the .MTS video file into chunks. As each chunk in a continuous sequence references the other chunks these must be joined in the correct order. This is easily achieved with the cat command.
2. The rejoined .MTS files can now be reencoded to a more manageable size using open source software such as Handbrake. We also needed to deinterlace our footage as it was shot interlaced and it would be accessed on progressive displays. This will increase the encoding time but without it any movement will look odd with visible artifacts.
3. Finding the ‘sweet spot’ for encoding can be time consuming but in this case was important as projected text needed to be legible but the file sizes kept manageable for reasonable upload times!

Cassette tapes run at a very slow speed of 1 7/8′s inches per second (ips) with a very small track width of 1.59mm

Cassette decks when they left the factory or a service centre should have been aligned to a standard reference for the position of the record and play heads. Unfortunately they often weren’t all the same and over time the alignment can get drift, get knocked out or manual ‘fiddled with’ by an owner.

What this means is that unless you’re playing back your tape on the machine it was originally recorded on you may be getting the maximum quality as the angle of the head to the recording or azimuth will not be optimal.

Without calibration tones recorded at the start of the tape which is very unlikely on most domestic cassette tape recordings you must set the playback azimuth manually. A few high end tape decks, namely those made by Nakamichi, either had a easily accessed Azimuth adjust or could even automatically adjust this throughout the tape. The Nakamichi Dragon was one such tape deck and could be the best, if working well, for high quality playback.

If you want to transfer or convert a cassette to CD and adjust the azimuth yourself this is the an easy way to do it:

1. Look at the tape path (everything the tape will move across) and if it looks brown and dirty get some isopropyl alcohol and give it a good clean with a cotton bud.
2. If you haven’t demagnetised your deck for a while now would be a good time to do it..
3. Power up your cassette deck, which hopefully works correctly and doesn’t have too much speed instability!
4. Pop your tape in the cassette well and start to play.
5. Turn your amplifier’s volume up and if you can put it in Mono.
6. Now, look under the tape machine’s playback or combined record and playback heads you should see a small screw or nut possibly with anti tamper paint on it.
7. Using an appropriate tool, turn this nut or screw a little left or right while listening to the audio.
8. You should hear the recording, especially if it has a lot of high frequency content such as cymbals etc get bright and dull sounding or more technically get more in or out of phase.
9. Your aim is to get the most in phase or bright sounding playback.
10. Sounds better now?? Great, start to record using you favourite computer audio software. We like SoX for the control but there’s a huge range out there.

We often get sent Digital Audio Tapes or DAT’s for transfer to .WAV computer files. As these recordings are already digital or ‘born digital’ the process should be straightforward. Our audio interface cards accept the SPDIF or AES digital audio stream from the DAT machine and record this as a WAV or BWAV file. This file can then be burnt as a CD or delivered digitally on a hard drive or removable media.

The big problems though come with the tape that these digital recordings are made on. The tape is only 3.81 mm wide and moves at a very slow 8.15 mm/sec. The tape is also very thin at 13 microns. The recording system and transport used is helical scan just like in video recording but with the very slow tape speed and small tape dimensions any defects or problems with the tape can result in many errors which may not be correctable by the error correcting system of the DAT machine.

One problem we’re starting to see more and more are tapes that snap. The tape pictured above was a D120 which was never recommended by the DAT machine manufacturers but was still often used for it’s extended recording time. This tape snapped without warning a quarter of the way through the recording. There were no outward signs or potential problems just a sudden clean break on a diagonal.

To recover this tape it could have been spliced with splicing tape of the correct width like in analogue recording but there is a high risk if not done perfectly of unrepairable damage to heads on the drum. Even with this type of repair some of the material would have been lost. A safer solution is to rehouse each spool in another shell this lets you recover as much as possible from the tape without the risk of head damage.

Whichever solution you decide the DAT shell must be disassembled. A small crosshead screwdriver needs to be used to remove all the case screws. There are two hidden ones, accessed by sliding part of the cassette shell down:

You can now carefully lift both halves of the DAT shell apart, making a note of the tape path inside the shell. Be careful not to touch the tape with your bare skin as fingermarks and grease can cause head to tape contact problems and audio errors and dropouts.

We have several of these large, wonderful machines. It’s not often we need or want to get involved in DAT repair as generally they are not easy to service machines and many key transport parts are becoming unavailable. The Sony 7030 DAT though has been designed with easy servicing in mind. There’s alot of room in these things and each section is clearly marked and separated into distinct boards much like Sony Broadcast video machines.

These are timecode DAT machines and were once common in video post production houses and the more well funded recording studios. The problem with some of this well built kit though is exactly that it works too well and gets left on for long periods through it’s life and this can take a toll on certain components, especially electrolytic capacitors. Heat builds up in electronic circuits, especially in switch mode power supplies that larger broadcast items often use. Capacitors have a rated life at 85°C or 105°C at several thousand hours. With hotter environments, substandard parts and long operating hours these capacitors can soon outlive their original design life.

Our 7030 DAT had started behaving oddly and at first the display would flash on and off after a short while powered on. Another machine would power up for 30 secs then just die. Before delving into the enormous service volumes it’s always worth replacing the Switch Mode Power Supplies (SMPS). These like many broadcast machines use supplies that are sometimes generic made by other companies and which can be bought at Farnell or RS. We did it the harder was and desoldered all the old capacitors in the power supply and replaced these with high quality low ESR Panasonic ones which should give us another 6000 hours of running time. So far this machine has worked perfectly although you do need good soldering and desoldering technique on these boards. A powered air desoldering station is a good idea, much, much better than a hand solder pump.

In amongst a batch of very mouldy quarter inch master tapes we were recently asked to look at was this unreleased recording by Shoes for Industry, the Bristol band on Fried Egg Records.

Like much late 1970′s and 80′s studio recordings, this was recorded on Ampex branded tape that suffers badly from binder hydrolysis or ‘sticky shed syndrome’ that must be addressed before the tape can be successfully played and digitised. This was in addition to the mould growth that was evident on the tape pack edges, and cardboard box. Storage in damp conditions and high humidity causes this type of mould and increases the breakdown of magnetic tape generally, sometimes to the point where de-lamination occurs, that is, the binder breaks away from the polyester structure of the tape. When this happens, which is luckily quite rarely, the magnetic information is damaged and mostly lost beyond repair.

Thankfully this tape, whilst it looked in poor condition was relatively straightforward to restore but time consuming. Careful hand winding, and mould cleaning is necessary as is awareness of the potential health effects of some mould spores so good ventilation and protective masks are necessary.

We have recently worked on probably the worst looking tapes but with some of the best sounding music recordings we’ve seen for a while! A batch of 10.5″ NAB studio masters had bad tape mould growth.

Andy Leighton, owner of Bolex Brothers and music publisher of the Rocky Horror Show, found a batch of studio masters on quarter inch tape that had been growing mould for some of them over 30 years. All of these recordings had been made at the renowned Sound and Recording Mobile studios better known as SARM, later creative home of Trevor Horn.

Among these tapes was the first ever recording made at SARM in 1972 by Richard o’Brien, writer of the Rocky Horror Show, in addition to rare tracks by artists such as Kimi and Ritz.

When the tape mould was finally cleaned from the tapes and some of them baked for binder hydrolysis the quality of the recordings was very high and testament to the high quality available from analogue recording and even though tape can be vulnerable to physical problems it is also robust. If these had been tape based digital recordings, in the same condition, I doubt we’d have been able to achieve the same results.