TR-2025

Tandberg TR-2025 is the smallest receiver in the late 70’ies TR-20xx series.

This TR-2025 was bought at an auction for 60 euro. It was supposedly not working.
My own first TR-2025. After restoring the first TR-2025 for someone else I wanted one for myself. Project started in 2021.

Unfortunately I didn’t take any photos before opening it up and starting working on it. But you know how a TR-2025 looks. This one was in very good cosmetic condition, maybe even mint condition. It just didn’t play.
Having recently worked on a that first TR-2025 (which you can find here), I wanted to take on another one. This very 2025 however would pose a completely different set of challenges.

Throughout this project I (Jens is my name) had assistance via remote from my friend Stein in Norway, who is a very experienced electronics guy. This restoration was a teamwork. We talked via Messenger. I executed the troubleshooting. It worked very well and was lots of fun. Stein is working on a “Vintage Hifi Troubleshooting Manual” and some of the questions that arose here will probably make it into this manual.

Unfortunately you now can see how incredibly primitive my lab is. This is (was) the dinner table, which has been an electronics workshop for over one year. Hifi IS more important than food. This also means – you don’t need fancy stuff to do this. You just need to prioritize away less important things like eating and inviting friends home.

On first inspection, the cap on the incoming power had blown. Replaced it with a safety cap of the same value.

Resoldered the transformer to 240V, as seen below. Actually, this was done later. Easy to see the different wiring for 220V and 240V.

Next, testing with current limiter. Right channel is weak, Left is working. But that joy was shortlived – suddenly lots of current being consumed, hum in test speakers followed by relay shut down (the protection circuit on the output).
This is where the real troubleshooting began, when me the blog owner Jens and Stein in Norway communicated via Messenger.

First, connecting the power amp through the lab supply. Left channel ok, but Right channel consumes too much.
Checking the output transistors. They were ok.

Don’t run it with the output transistors like this. The middle little transistor need to be thermally coupled to the bigger siblings or the bias current will run away and cause excessive current draw and potential problems.

This was an odd repair. Strange things happened. Now we have sound in Right channel power amp. Probably bad solder joints somewhere in the power amp. No bad parts. Checked drivers and and some other transistors too, all well.
Now however Left channel has a problem. Or rather, it is apparent now that Right is functioning rather well that Left was not functioning so well. It is weak.

Tracing the weak signal from Left channel back to the pre amp. There is a problem in the volume pot. Stein taught me the trick, resoldering with spraying contact cleaner in the pot first. The electronics cleaner works better with heat applied. It solved it. We now have channel balance!

Dummy loads on the speaker outputs. Signal from vintage Thandar signal generator.

But no time to celebrate yet. The DC detect relay goes on and off.
Looking for DC on the output, but there is not that much, about 140mV on Right channel and 50mV on Left.
Checking the power supplies, the DC offset.. Something must be wrong with the protection circuit itself.

Desoldering the outputs does not even help (to make sure the protection circuit surely does not detect anything).
This was a tough problem to solve and the process was not completely methodical because of some impulsive actions on my part, so I can’t remember all the steps. Lesson learned – be methodical! But all components in the circuit was checked and none of then faulty, or maybe one transistor was, or possibly it failed on probing or soldering. My experience with the other TR-2025 showed quite a few faulty small signal transistors, in this unit however they were fine. Testing the protection circuit back and forth, how it reacts on desoldering the 150ohm resistor R776 for example. Here the discussion ran deep with the remote expert, on the workings on this circuit and even on physical properties of transistors and their behaviour. A simple circuit, but deep questions.

In the schematics above you see the protection circuit at the bottom.
The explanation of the workings of this circuit, brilliantly explained by Stein:
Q771 takes care of positive offset. Q772 of negative offset. With no detection on the output Q772 will conduct, it’s collector will be zero, so Q773 will not conduct. This makes Q774 and Q775 conduct so that the relay is on and the power amp signal is sent to the speaker outputs.
The slow start works so that on power upp C773 is charged, which takes a few seconds during which Q774 is not conducting and the relay is off.

The protection circuit/slow start circuit has two power supplies, the +31V and +40V above. The +40V comes from the regulated pre amp supply, on it’s 1000uF capacitor before the regulator. The +31V supply comes from two diodes straight after the power input, before anything else. See below. The diodes were checked and even replaced, but to no avail. A meaningless replacement on my initiative, the ripple was 100Hz and the diodes were ok.

Green circle is the +31V tap for the protection circuit.
One of the experiments – smooting the +31V supply. The cap was tested here in the protection circuit, on the base of Q772, and it was tested on the +31V supply itself. The best effect was with the cap between the diodes, this is where it is now (and in the photo below).
The capacitor above is 100nF, or 0,1uF, but later this was changed to 47nF. The reason was “bump” in the speakers on power off. With the smaller cap this is almost eliminated, but that was done later.

In the troubleshooting process of the DC detect/protection/slow start circuit, all the transistors were replaced and of course the capacitors. No capacitors were bad but they were at first the prime suspects.

It seems this power supply makes the DC detect circuit act like there is DC on the outputs. With a capacitor this ripple is reduced but it makes you wonder – did this circuit ever work?

Steins thoughts/explanation on why the circuit was acting like it had detected DC.

Perhaps this very TR-2025 had problems from the beginning, from factory, and because of this was never used? It looks like it, not a scratch. All lamps working and looking original. Maybe it was stored away in a repair shop or retailer, waiting to be sent back to Norway but this never happened? We can’t know. But it is rather odd.

The erratic behaviour of the protection circuit probably had several causes, a bad solder joint, the +31V supply as mentioned and also the +40V supply. The regulated pre amp +40V behaves strangely, it works and it does not work. The fault was found in bad solder joints in the regulated power supply. Touching and pressing the board with the regulated power supply provoked this behaviour and thus the suspicion was confirmed.

I found several bad solder joints. First one in the power amp, or at least one. There were bad joints in the pre amp regulated power supply, and in the volume pot and possibly other pots. Easy to fix. Not so easy to see though and these caused us some unnecessary troubleshooting time. There are lots of joints and it isn’t feasible to redo them all just in case. I think this unit had an above normal number of bad joints. Considering also that it seems it had not been much in use and not worked on earlier this was surprising.

Ok, we have a working protection circuit. We have decent channel balance. But we have horrible noise and pops, it comes and goes. Tracing the origin of this noise it seems to come from the pre amp. Another solder joint problem. So the whole pre amp board was resoldered.

Doing a test run on “real” speakers. It delivers. Also tested output on the oscilloscope, 26,2W, good power considering also that power in the wall is about 226V rather than the 236V or so it should be.

But – no time to celebrate yet. Thought maybe it was fixed, but no.
Fiddling with the tonecontrols and increasing the bass a horrible, speaker-murdering low frequency noise! It just happens, suddenly, total freak out.
Am I a magnet for messed up Tandberg gear? Wanted a challenge so no complaints..

What is this?

Troubleshooting of this noise began, tracing it back through the pre amp. Very odd, it seems to affect everything in the pre amp. After a long time of checking, head-scratching and probing, tested to put more capacity on the the bypass capacitors, on top of the already existing 100uF + 470uF. Amazingly this stopped the ‘oscillations’. I call it oscillations, because that’s my theory of what it was. It seemed the pre amp supply dipped when it started freaking out. Only the bass pot could provoke it. The low frequency cut filter stopped it.

Late at night, the revolutionary discovery that a big increase in the pre amp rail bypass capacitors would inhibit the explosive oscillations. Soldered out the now replaced 100uF for 470uF. But this was not enough. So in with 820uF per channel. This stopped the oscillations completely.

In the search for culprits all the caps in the pre amp and regulated power supply were replaced, but ultimately they were all good and I kind of regret replacing them now. Wanted to keep this 2025 as original as possible.
Partly recapped pre amp board.

However in the pre amp, DC measurements were a little bit on the low side. I had suspicions about the “silent switch” IC in the pre amp.

TR-2025 was made in different versions. Later versions looked like this, they had a “silent switch” IC in the pre amp. Earlier models lacked this, or so the Tandberg lore goes.

A close up of the silent switch IC below. It is powered from the pre amp rail and the pre amp output passes through it. The switch controlling it seems to be connected to the radio station selector switches. I found this a bit odd and maybe not a totally necessary thing. Without it, which was tested, the radio station buttons were not really that noisy.

The IC close up. On Stein’s advice, tested to short pin 6 to ground, which should silence it (shut of the pre amp outputs), but it doesn’t. The IC is kaputt. It is posssible, theoretically, that I ruined it when troubleshooting the pre amp and soldered out the power supply to the IC. But, I don’t think so. The very odd thing is that it seems this IC consumed too much power and taxed the pre amp. It should theoretically be highly unlikely that it could draw that much, but measurements show that it caused something to happen, internal short?

So I soldered the IC out. Shorted it’s in and out holes on the PCB with a cut of Neotech hookup wire.

This stabilized the pre amp supply a bit and voltages increased about 1V all over.
After this the large 820uF caps in the pre amp rail were removed, because they were too large in size to fit the front, so 470uF caps were put in place instead. This caused oscillation before, so now with the IC gone it was interesting to see if we could provoke oscillation. We could not. It seems, no matter how strange it sounds, that the silent switch IC caused the pre amp to oscillate on increasing bass. It was like the power supply could not keep up and a low frequency oscillation was the result. The tone control section was carefully investigated, including feedback system, but there was nothing wrong anywhere.
In my search for answers the 100uF caps were removed temporarily, on my own silly initiative, to see what would happen – this was a terrible idea, the whole amp fell in runaway oscillation mode and maxed out the outputs, don’t try that at home. Now at least I know how incredibly important those caps are. I saw a bit of smoke! Quick switch off, but would something have been damaged? Checking the power amp, everything works, the components (in particular resistors) suspected to have overheated were checked but nothing wrong. Curiosity could have killed the amp.

Anyway, I liked the idea of the overly beefy bypass caps and put back 820uF caps again but of a smaller type from Wurth. Wanted to see what results, if any, it may have on performance.

Should the IC be replaced? That is the question. But my conclusion was no. First, it is not really necessary. And although I want to try to keep things as original as possible, the less ICs and SMD components the better. The world has enough integrated circuits. I cringe at the word “digitalization”, instead I encourage analogization! And should the signal really go through a switching IC? No, on this 2025 let’s keep it out, keep it purist.

It is absolutely normal to have at least 7 receivers set up in your listening room.

Ok, it works. Now some cosmetic care and adjustments. Bias was checked. One channel bias, or “hvileström” as it is called in Norwegian, was a bit low. Then took out and reapplied thermal paste on the other channel.

Original thermal paste.
How necessary is it to reapply thermal paste? Not sure. But since the other channel had already been done out of necessity it would feel odd to leave this one untouched.

The usual cable cleaning process. Maybe not essential on this one, very clean. But the fuse holders were dirty.

New Wima cap and diodes in bottom left, for the protection circuit supply.

Cleaning the outer chassis parts as well.

Do you also shower with your hifi gear?
A lamp went out. But it wasn’t burnt out, just oxide in the socket. Cleaned it with a little toothbrush.
The black plastic sheet in front of the switches and behind the front panel was attached again with double sided tape.
Little bolts on the front, where the side panels attach. They fell out, forgot to put them back before screwing back the front panel, after discovering that the side panels screws didn’t want to attach – the frontsection had to be removed again.

By the way, the mail just came. Ordered from ebay, from China, 6-4mm motor axle adapters. To be used when replacing potentiometers because modern pots usually have 6 mm axles and the tandberg knobs take 4mm axles.

Rather clever idea, isn’t it? No need for them on this TR-2025 though.

Putting it together. After cleaning it looks like new.

Almost there. Pristine, mint+++ (the TR-2025, the room is a disaster)

Combining it with the newly restored TL-5010 speakers I am really impressed how good this receiver is!
Different from the TR-1000 and Huldra series, it is more modern perhaps in that it sounds ‘cleaner’ and not so meaty, creamy and lush. Light is measured in Kelvin, if the TR-1000 is a super warm tropical colour, the TR-2025 is a more blueish tint, does that make sense? The sound is crisp and almost gentle. Sometimes I think it has a tone of melancholy, more anyway than the extroverted and bold TR-1000 and Huldra 10 relatives. This is the thoughtful younger sibling. Because of this I personally found it goes well with acoustic music and jazz. I find it more dimensional and nuanced than mass produced receivers of late 70’ies and 80’ies.
The amount of power it puts out is impressive, it sounds more, much more, powerful than the specs suggest (I think it is specified to 2x25W in 8 Ohms).

Looks like it stands on a mirror, but here the TR-2025 stands on top of it’s bigger (not physicallly much bigger) brother the TR-2080.

During this restoration I was communicating with my friend Stein in Norway via Messenger. Troubleshooting via remote so to say. It was lots of fun. Without ‘the remote assistance’ and Steins advice I would not have come this far.
The 2020-2021 plandemic made us find new ways to do things. Some good things came out of it, like that it is possible and can be fun to do things together through a communication platform, in spite of the physical distance it can feel almost like you meet IRL and do something as complex together as restoring a late 70’ies receiver.
Besides during that terrible lockdown, what better hobby to have than being at home fixing vintage audio gear?