Rotel (Rank Audio) RA-210 Clean Up

[OldTechFan96]

A few weeks ago I bought this early '70s Rank Audio branded amplifier from a member. It looked like it had spent a considerable amount of time in a garden shed.

The front bezel, knobs and cabinet were filthy but looked like they would clean up well.

Internally, the amplifier and cabinet was full of cobwebs, dirt and all sorts of miscellaneous garden debris.

I concluded that my normal cleaning methods would not go far enough in tackling the large amounts dirt inside.

Since I had little to lose, I decided to fill the kitchen sink with warm water plus a small amount of Fairy and submerge the amplifier in the water. I let it sit for an hour before I started giving the insides a thorough cleaning with a paintbrush. After all the muck was removed, the amplifier was rinsed off then blown dry with a hairdryer to remove excess water. Finally, the amplifier was sat in front of a fan heater for a few hours.

The aluminium (?) bezel was gently cleaned with a soft cloth. Once it was dry, I gave it a once over with Brasso. There was some staining that was removed with Autosol. The Knobs also cleaned up well.

The 'simulated wood-grain' cabinet was cleaned with Cif and a cloth. A rub down with T-Cut gave the cabinet a bit of a shine. To remove the T-Cut that gets ingrained in the crevices of the wood-grain, I sprayed the cabinet with Mr Sheen and scrub it down with a toothbrush. This left the cabinet looking quite good.

Here are some 'before' photographs.

[OldTechFan96]

More 'before' photographs:

[OldTechFan96]

A few more:

[OldTechFan96]

Once everything was clean and dry I decided to tackle the mains transformer. Since it was dunked in water there was no way it would dry completely sitting in front of a fan heater.

I checked the insulation resistance of the two windings with my Megger at 500V. They both measured about 100KΩ; a long way away from my target of an infinite reading!

I decided that the best way to drive moisture out of the transformer was to connect my benchtop PSU to the primary winding at 20V. This drew about 300mA and 6W was dropped over the winding. The transformer was warm to the touch which gave me confidence that this would work well.

When cleaning all of the switches and potentiometers, I noticed that the on/off rotary switch did not feel quite right. It felt like it was on its way out. Luckily, a strip-down and rebuild sorted it out. This was a real pain to do.

After five days on the PSU, I rechecked the insulation resistances and they were both reading infinity. A good result!

Now I could power the amplifier up. I connected a CD player and some 2 1/2" 8Ω speakers that were salvaged from a TV some years ago.

Once fired up, everything worked as expected after a few minutes powered on. A soak test of a few hours was in order before any extensive checks.

The output of both channels were checked with a function generator at 1000Hz and an input signal of 180mV. Probing the outputs with my oscilloscope I saw an unstable, jittery waveform. My first guess was mains hum from a leaking smoothing capacitor. I measured something like upwards of 200mV of AC on the positive side of C616 (1000uf, 35V).

I expect that C616 is the culprit here. I don't have a suitable replacement at the moment. I probably couldn't hear the mains hum due to the limitations of the test speakers I was using.

A new mains lead will need to be attached in the future.

Here are some 'after' photographs:

[OldTechFan96]

I rechecked the audio output with my function generator and found no issues with jitter or ripple. It turns out I had connected things up wrong due to confusion over some grey and black phono connectors.

This amplifier is rated at 8W P/C at 8Ω or 6.5W when both channels are driven.

My final test was to check the power output of each channel with my wattmeter.

RHC measured 7W and the LHC measure 8.8W. There is a bit of an imbalance here. I assume that the RHC should be higher? Hopefully voltage measurements shed some light on this issue.

[Jez1234]

Looks quite Amstrad internally.... From shed to skip would have been my route!

[samjmann]

What a lovely job. Top marks!!!

I remember being involved with one of these that the customer ran into a 2ohm load. It came back to us 3 times before we found out what the the user was doing. All that ever happened was that the fuse had blown due to overload. I ran it for 8hrs at max power into a dummy load - fault free.

I wonder if this is one of the Arena designs designs when Rank took Arena over?

I've still got a Bush-Arena BH-4000 in the shed on the to-do list!

SJM.

[samjmann]

Just had a look at the Arena F-220 amp. This is 8Wpc rated and uses driver transformers in the output stages. Is this similar to the RA-210?

SJM

[OldTechFan96]

I have been trying to find out why one channel has more output than the other. R=7.6W vs L=9W.

I've measured the voltages of the first two transistors in each channel:

B C E
TR501: 2.9V 14.2V 2.35V
TR601: 2.8V 15V 2.2V

TR502: 0.7V 6.8V 0.13V
TR602: 0.8V 4.6V 0.17V

The voltage on the collectors of TR502 and TR601 are both too high. The other voltages are about right.

What explains the high collector voltages? The voltage rails from the PSU are correct and all of the resistors around the two transistors measure bang on. Is the issue further down the circuit?

Any help would be much appreciated!

[OldTechFan96]

I've taken measurements from every transistor and a certain pattern emerges. Each L/R transistor pair has a slight variation in voltage measured.

The C of TR502 is the only voltage that is way off. The transistor tests fine. I don't know what is causing this.

[samjmann]

Looking at the voltages, it looks if the base voltage on 502 is slightly low. Are you sure the 1meg resistor is spot on?
You could also swap Tr502/602 and see if the fault 'moves' if it does then the transistor would be at fault.

Also, check for leakage in the base coupling electrolytic. With a 1meg acting as a pull-up, the circuit will be quite high impedence in this area.

SJM.

[samjmann]

Just another thought before using the soldering iron. Measure T502 collector, and at the same time rotate the volume control. The voltage should not move. If it does, then there is a dc leak into the control itself.

SJM.

[OldTechFan96]

I swapped Tr502/602 and the fault did move! The offending transistor (2SC644) was replaced with a BC549C and the C voltage now measures 4V. I'll do some more checks soon.

[samjmann]

Sounds like a result! The new BC149C must have quite a bit more gain as the collector voltage has dropped 2v.

See what the Watt meter says next!!! 😀

[OldTechFan96]

With the BC549C (TR502) installed the voltage measured at the C is now 4.2V vs 4.3V on the other channel.

I'm measuring 9.4W O/P for the LHC. The increase makes sense considering the higher gain of the BC549C. The O/P of the RHC is 7.6W. Replacing TR602 would presumably bring the O/P into parity. I only have one BC549C so I can't replace TR602.

I do have two BC547Bs. AllTransistor.com lists the BC547C as an equivalent to the 2SC644. The BC547B has a lower minimum hfe than the BC547C. 200 vs 420. Compared to the 350 of the 2SC644.

I'll experiment with the BC547Bs. Hopefully the lower gain does not change O/P power too much.

[OldTechFan96]

I tried the BC547Bs and the C voltages were off. I'll blame the low hfe. I have found two BC184KCs with a hfe of 450. This should be more promising.

[OldTechFan96]

With the two BC184KCs installed the transistor voltages look sensible.

The O/P of the RHC is 7.6W at 1000Hz, 180mV sine input. No distortion at maximum volume.

The O/P of the LHC is 9W at 1000Hz, 180mV sine input. Slight distortion at maximum volume. When the input signal is reduced by 20mV the O/P reduces to an undistorted 7.6W.

Things aren't 100% yet. I could try the pre-driver adjustment.

[samjmann]

Yes, the pre-driver preset should affect output. Perhaps worth checking C511/611 in the output stages. Down to the Lily-gilding stages now. ��

SJM.

[OldTechFan96]

I tested C511/611 and they both check out fine. Adjusting the driver preset did not change much.

The LHC is still a watt higher than the RHC. It's an issue I want to solve but the transistor voltages I measured all seem fairly close.

I measured the amount of audio signal before R501/R601. I should be measuring 120mV here. LHC was 107mV. RHC was 104mV.

It makes sense that the LHC is stronger since there is more signal at its input. Between this point and the input of the amplifier there is a 68K resistor, one in each channel. Both were a bit high. Replacing these did not even things out.

I hooked up some Warfdale speakers to the amplifier and it sounded good. You could not tell that there was a slight difference in power O/P. A hypothetical layman probably wouldn't notice at all.

It might have been like this from the factory.

Thanks to those who helped!

[samjmann]

I would do the same at this point, a lot more work for a diminishing return.
Would make the basis for a compact Bluetooth 'office' or study-room system, more than enough power for that I would have thought.

SJM.