Tandberg TD20 A Tantrums

[af024]

A few years ago I was kindly donated a Tandberg TD20 A – my first (and only) high end reel-to-reel tape recorder. I was over the moon as you can imagine.

From then until now I had purely been using it to archive recordings – I had not attempted to try any recordings. Increasingly though I was getting intermittent playback on one and sometimes both channels. The fault reminded me of this thread:- https://www.vintage-radio.net/forum/...d.php?t=182997 (which I think is pretty much the same problem).

Interestingly, I think the left channel on my machine also seemed to cut out more than the right. This really started to niggle at me as I never knew when it was going to happen and when archiving the likes of a large 2500ft NAB reel tape at say 3¾ i.p.s., multiple cut-outs were just not funny!

Wiggling the ‘Playback Mode’ (track(s) selector) would restore things, sometimes for the duration of the tape, other times not. I also noticed that the Monitor (source/tape) selector switch was also giving a bit of grief.

For weeks now I have been meaning to take a look to see what was going on, but with such a large and heavy lump of a beast, and minimal working area available, it just kept going to the back of the pile. That was until an ex-work colleague asked to me to see what was on a load of fully loaded 10.5” NAB reel tapes.

At that point, I thought I’d better bite the bullet and try to resolve the issue and whilst inside, check for any other problems. Of course, this was also an ideal time to attempt my first recording (yes I know 5+ years down the line since it was given to me, and I’m only just doing the first recording now). Well it was a disaster – nothing but very low level distortion on both channels. OK, it erased the content of my test tape ok, but that was about it. The meters deflected fine, as expected, showing between 0 and +3db after I set the levels up. Switching the Source/Tape switch to monitor what was coming off the tape resulted in very little meter deflection and hardly any audio in Tape mode. I did wonder whether the Playback Mode switch had finally given up the ghost totally. At this stage though I wasn’t really sure whether it was a record or playback issue. I was unable to try my test recording by playing the tape on another machine because I have nothing else that can accommodate a large NAB 10.5” diameter reel. I did try to play a supposedly known good tape (recorded on an Elizabethan LZ34) on the Tandberg and I was greeted with another very low level and distorted playback. So perhaps it was just a playback issue caused by the Playback Mode switch, then again, looking at the circuit where the meters are fed from suggested that there might be something wrong with record downstream of the meter take-off point (see to the right of the red line). Confused? You will be, read on …

[af024]

I thought a useful starting point was to try and eliminate the dreaded Playback Mode switch and perhaps the Monitor selection switch too, so out of the box the chassis came – four screws (1 at each side and 2 underneath). I was feeling brave (for once!)

It’s a heavy lump for sure. I quickly came to the conclusion that I shouldn’t be laying the machine down on its front to pull the box surround off – not even on a soft large block of foam, so the only sensible way was to have the machine sitting vertically on its feet. It’s then possible to push the deck out of the box by pushing on the four feet at the back.

I must say that it was rather a daunting prospect to clear the way – all the knobs, threaded plastic nut/covers, facia, more pot nuts, earthing spring, multiple screws, side plate etc had to come off.

Eventually it was possible to gain access to all of the left hand side cluster of controls and having removed the pot retaining nuts, put the switches in a position to give max clearance and slightly bent the pot wires. Having prepared it in such a manner it was possible to jiggle the complete daughter board PCB out from its socket without damage.

Somewhat against my better judgement, but with and ‘in for a penny, in for a pound’ dare nagging at me, I thought I’d unsolder all of the switches (in turn) - 2 from the daughter board PCB and 2 from the main board PCB, and take them apart to see what was going on. Once a switch becomes liberated, the four metal tabs shown underneath have to be bent out of the way before th guts can be extracted (and you also have to hope that they don’t break off upon reassembly!).

I’m pleased that I decided to go that far, because every single one of the contacts and every one of the sliding wipers was totally black – just take a look at the photos.

Just for fun I tried my good old trusted Servisol 10 and a cotton bud, but this crud was not for turning - no wonder the switches were playing up! Even my next trick with Brasso/Duraglit wadding was a miserable failure. Now that always works for me, not this time, well not in its own anyway!

I know it sounds a bit brutal, but I was reduced to taking the crud off using a fibre pen. It’s probably taken off some kind of coating that might have been there, but I figured that the coating or no coating debate was purely academic if the black stuff couldn’t be got rid of. Following the fibre pen treatment I then polished the contacts up with Brasso/Duraglit followed by a buff up using cotton buds. Finally I applied some Servisol (and some Molykote to the white latching plastic located between the contact strips).

As you might expect, the sliding spring contacts are very fragile. I can imagine that they can be very easily bent out of shape - perhaps irreparably, so extreme caution is advised here. I certainly was not going to attempt to open them up to gain access to the hiding contact points, so I soaked some Brasso/Duraglit into the edge of some paper, carefully placed the contact to straddle the paper and swept it along the paper multiple times until the black mark on the paper started to subside. I then moved the contact to an unsoaked part of the paper to polish things up with the same sliding motion.

[af024]

Any black gunge that was accessible on the outer surfaces was removed with the fibre pen. I figured that even though the contact points had now been restored, it made sense to remove the black gunge, just in case it decided to creep back to the contact points in due course. Again a good spray of Servisol 10 was applied.

Then came the reassembly. Again, extreme caution is needed here. Not only do the spring contacts have to be straddling the switch contacts properly, but they also have to be seated in their correct cut out position in the black plastic holder/slider (in the case of the two switch types that reside on the main PCB at any rate) – the ones on the daughter board can only go in one way. The photos show which of the positions is correct.

At one stage I thought I’d broken off some kind of position restrictor, as one of the switches (used for Sync selection) should only have two switch positions (off or on). It ended up with three! The Playback Mode switch should of course three positions (Left/Stereo/Right) but not the Sync switch. I couldn’t find a stopper or any missing/broken part anywhere. So, I had a bit of panic at this point.

I took the decision to carry on with the re-build regardless. Later on, I realised that the two switches in question (on the daughter board) had the same part number and that the position selection restriction is achieved via the clever use of a plastic insert that’s in the facia panel, and has nothing to do with the switch itself.

The lower two switches are mounted on the main PCB, and access seemed fairly easy, and since one of said switches (the one for Monitor) had been playing up, I thought I’d tackle those ones also. Note that the switches on the main PCB are different from those on the daughter board.

The two pots (Output Left and Output Right) on the same daughter board panel hadn’t been giving me grief, but since I had ready access, I thought I’d at least given them a good spray whilst I was there.

At this point I decided to re-assemble the machine before I forgot where it all came from – it was all a bit fiddly, especially re-aligning the left metal chassis metal piece and the top facia panel, but I got there with loads of jiggling about (good old Granville eh). One observation I should point out is that the left cheek of the chassis doubles up as a heatsink for a 7824 regulator (the right cheek provides the same for a 7805 regulator). I had to unscrew the regulator temporarily to remove the metalwork. Once I’d cleared the way, I noticed that the regulator legs weren’t soldered directly onto a PCB, but were rather pushed into a socket. I had knocked the regulator slightly sideways whilst removing the chassis metalwork, but I repositioned it, applied some heatsink compound and re-affixed it to the metalwork with the proper screw on reassembly.

I decided not to venture towards the other toggle switches (on the right hand side of the machine), or those pots, as nothing was playing up there. That might have been a mistake of course, but so far so good. Fingers crossed.

At this point, I tried to investigate why, on record, the meters were nicely deflecting, but there was nothing going to tape. Logically I thought it would be some fault downstream of the meter take-off point and I concentrated on the left channel as I had that part of the circuit on a piece of paper in front of me. I started looking the dc voltages (in record of course) downstream of the coupling capacitor C124 (see red-lined circuit). All looked well. Also, he head wasn’t open circuit and it was a clean as a whistle, as were all the other heads. I suddenly remembered at this stage that the fault seemed to affect both channels, so it was probably unlikely that there were two identical faults present – one in the left channel with the other in the right, so it was more likely going to be something common.

[af024]

Rather irritatingly, there’s not much that can be concluded here because when I tried a test recording, all was working ok, as if by magic. The only thing I can think of was that there was perhaps something amiss with the 24Vd.c. rail which was ‘put right’ after I’d jiggled the regulator out of position and back again. Does that sound credible? I’ll never know for sure of course.

So I boxed it all up and decided to do some new recordings on a decent new old stock tape. I must say it’s impressive at 7½ i.p.s.!

I then decided to attempt to play a tape that I had recently recorded on an Elizabethan LZ34. The Tandberg TD20A is a quarter track (or stereo) 3¾ or 7½ i.p.s. machine depending upon switch settings, and with the Elizabethan LZ34 being a quarter track, I thought it would be a valid check.

Well interesting is the word.

There was hardly anything coming off the tape, so I fetched the Elizabethan down from upstairs and tried to play said recording back. All was fine, loads of signal off the tape, so much so I could drive the Elizabethan machine into clipping with the volume right up.

I decided to put the NAB tape back onto the Tandberg (just in case the machine had gone wrong again) - all was well (phew!!).

Of course I couldn’t fit the NAB reels on the Elizabethan, so I couldn’t do any testing that way round.

So what’s going on? Tracking perhaps? Well here’s the thing, the Elizabethan ¼ track recording plays fine on the Elizabethan and a Pilot/Thorn/Ferguson/Ultra ¼ track machine that I have, so what’s the chances that two machines are wrong and the Tandberg is right? No screws/locking paint look disturbed in and around the Tandberg heads and I wouldn’t want to fiddle about anyway – there are three heads in there, and I don’t have an alignment tape, so that’s not an option.

Then I remembered that I’d had some fun with the Elizabethan LZ34 many moons ago when one track would trample backwards over another (I forget which tracks). Right back then, having unwittingly destroyed a few good recordings, I looked closely at the REC/PB head and found that it was severely warn – the worst wear I’d ever seen actually. After years of searching, I eventually I came across a genuine Marriot replacement at one of the NVCF dos at the NEC. It was the correct high impedance version for the machine and I duly changed it out. I set the azimuth to maximise the top end and left it at that thinking ‘that’ll sort it’. Well I’m not so sure that it has (I perhaps need to do some more testing with the Elizabethan). If the Elizabethan tracking is off, this would no doubt explain why the Tandberg reproduces the Elizabethan-recorded tape so badly. That doesn’t explain why the Pilot machine plays the Elizabethan tape back ok though does it?! Also, I can see no method to raise or lower the head in the Elizabethan – it just sits clamped inside a pot mental assembly at a predetermined height. So even if the tracking is off on the Elizabethan, there’s no way to correct it that I can see. It’s probably all down to manufacturing tolerances of the mount and the head itself?

I had always assumed that machines purporting to be ¼ track compliant would have the same track positioning and width, so what’s going on here?

[af024]

Photos

• Extract from circuit showing left channel record amp
• Tandberg TD20 A overview (misty – sorry!)
• Underside chassis - location of daughter board PCB in question
The 7824 regulator position (sorry, I forgot to take one, but it’s just off shot top
right on photo ‘(3) Underside of chassis (with daughter board removed)
IMG_0897.JPG’
• PCB in close up showing daughter board nearly extracted
• Daughter board extracted showing one switch type
• Disassembled switch
• Blackend contacts
• Polished-up contacts
• The other switch type (Sync and Edit/Cue) removal from main PCB
• Cleaned and polished contacts – switch ready for re-assembly
Shows spring slider slot positions (Sync and Edit/Cue switches) – the sliding
contacts must be located in the upper positions of the black plastic carrier/slider
• Close-up of left hand side of the front facia showing plastic stopper inserts (don’t
mix them up if they pop out!)
• Elizabethan LZ34 ¼ track machine
• Pilot (unknown model) ¼ track machine

Sorry about the misty-eyed look of some of the photos – not sure what happened there!
Also, and you’ll probably be able to see it one some photos (eg ‘(11) Switch stopper plastic insert IMG_0914.JPG’), the front facia seems to have some staining on it. It’s a strange brown marking here and there, almost rust colour but not quite. It won’t come off. If anyone has any hints and tips that might help me remove it, then I’d love to hear from you. Thank you.

Sorry, this has been a long one with not much to say! If nothing else, it might help others with the switches?

Photos follow:-

[af024]

Photos ..

[af024]

More photos ..

[af024]

Last set ..

[ben]

Great work! A for the tape, here is where you need a stereo quarter track tape that is pre-recorded or done on a known decent deck. I would not trust the one you made on that Elizabethan with a replacement head that may or may not be kosher.

Shame I'm not in the UK right now or I could send you a tape. Maybe look on ebay for some of those world record club or EMI 5" stereo prerecorded ones. Even though most were cheesy easy listening stuff like tin pan alley, they would do the job here.

[Ted Kendall]

Most of the WRC tapes are half track mono. EMI did market quarter track stereo for a year or so, but these are harder to find and, in my experience, of distressingly variable quality, largely because the signal is put on pretty hard, to keep hiss down, I suppose. Still, the track positions should be correct.

[af024]

Thank you Ben. You are right about a test tape for sure. I am somewhat intrigued about the fact that the Elizabethan and the Pilot seem to agree with each other (and not the Tandberg) though, but my gut feeling is telling me that the Tandberg is most probably more likely to be correct, simply because of the build quality. Who knows though. I've seen test tone tapes, but nothing obvious in quarter track. As you say, perhaps just some prerecorded easy listening stuff (in the hope that that there was some care taken over tracking). Of course even if I identify the wrongly set one(s), there's then the dilemma as to whether or not to make it good, on the basis that tapes previously recorded on the bad set up won't be able to be played properly from that point on. All good fun eh Ben.

Regards,

Andy

[af024]

Ah Ted, right, they may be a tough find then. Oh dear ...

[Ted Kendall]

7.5 ips tapes of American origin might be easier to find - there are some up on eBay from UK sellers.

[ben]

You are absolutely sure the Tandberg is quarter track? If it were in fact half-track, that would explain the lossy signal off the quarter track tape and why the NAB reel stuff played back well (most machines using that type of reel were half track, semi pro kit). Check what heads are installed, you may be surprised

[TIMTAPE]

This may have been covered but have you confirmed the playback amplifier chain is working normally in play mode?

[af024]

Hello TimTape, yes all seems to be working fine now, so what went on is a bit of a mystery.

Ben, yes, that's a good thought. I was going on what I was told, so perhaps that wasn't correct. Here's an extract from the manual (the Word document below). It shows the two options that the machine came with - four track and two track, so it's certainly a possibility.

Taking a close look at the head isn't perhaps conclusive, but it does look like the spacing suggests four track.

The number on the ERASE head is 6484/E, the REC head is 6482/R and the PB head is 64800/P. This site suggests four track:-
https://picclick.com/Tandberg-Reel-T...270993303.html

Just to be sure, if it was two track, and if I made a stereo recording then flipped the tape over and played it in reverse, I should hear the recording backwards I guess, well I didn't, so I guess it's four track after all. In a way, this confirmation is a bit of a shame as that would have solved the tracking comparison mystery. Good thinking though!

Regards,

Andy

[af024]

Sorry, I forgot the Word document..