Rock-Ola 40218-1-A Transistor Amplifier Problems
 

Hi all, been a while since I've worked on vintage audio equipment so I'm a little rusty!

Finally decided to finish off the Rockola 425 I've been restoring for the past year. Everything is good and working except for the amplifier which is giving me a hard time.

I replaced the major capacitors a year back and left it at that until I got to the stage in the restoration when it was needed.I am now in the stages of testing the amplifier and so far this is how it has gone:

Initial test of the amplifier, I turned it on and a very loud hum came from the left hand channel speaker for a few seconds before blowing the 1A slo-blo fuse. I replaced the fuse and the same thing happened. I replaced the fuse again this time with a 1.6A fuse and removed the pre-amp, the volume control and one of the 2N2869 transistors (The ones in the left hand channel have been replaced by AU103 transistors) I tested the amplifier again brief enough as to not blow the fuse and the loud hum was at about half volume. I removed the second 2N2869 (AU103) transistor and the hum was gone.

I assumed these transistors were the cause of the fault but to check I left them out and turned the amplifier on. It lasted about 15 seconds before a large cloud of smoke and hissing confirmed a capacitor had blown its top.

Upon inspection it was one of the 1200uF capacitors and had accidentally been installed in the wrong polarity by me earlier, I replaced the capacitor with the right polarity this time and tested the amplifier again for half a second with all transistors fitted expecting the capacitor was the cause of the problem but again an extremely loud hum is heard from the left hand channel only.

This is where I am currently at, attached is a picture of the bad capacitor (excuse the poor soldering!) and the relevant half of the circuit diagram, I may post the other half of the diagram if it becomes necessary later on.

Thanks in advance!
Peter

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

It could be a short circuit somewhere. I'd advise re checking all the caps you replaced at least twice, then check for shorts with an ohm ,meter starting with the OP trannies, then try across the power rail/s, then across the IP.

Before you power it up, either build a lamp limiter or power it up on a current limited power supply, then take voltage measurements between the good channel and the faulty one. Check power rails on AC volts range to see if excessive AC ripple, which could indicate a SC.

Andy.

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

Just to remind you this is a Germanium PNP transistor unit. All the chassis references (and therefore cap returns) are Positive, not negative.

Check you have not cooked the driver transistor that feeds the 'failed' 1200uF cap.

Are AU103 suitable replacements for the original transistors? 10W vs 30W for example.

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

Curious circuit, it seems to have a lamp (limiter?) in the emitter of each output transistor. It’s unlikely that the reverse connected cap would have damaged the output transistors, but might have damaged the driver. The loud hum suggests one of the output transistors or one of the bulbs has failed and need checking.

As previously suggested it’s worth also checking the driver and all the caps before applying power again.

Repeated blowing of the fuse is not good, and if you don’t fix all the faults before applying power, you could instantly blow any new parts you fitted. A lamp limiter, or increasing the supply slowly on a variac whilst monitoring supply current, could show if you still have uncorrected faults without blowing replaced parts.

Stuart

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

I would be surprised if the driver transistor has survived so I would test those first. I would also test the four output transistors. The driver transformer should make this a relatively easy fix.

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

I have checked the 2 AU103 transistors with my analogue multimeter and they are both completely shorted between the collector and base. However upon checking with my (old) digital multimeter it shows they have around 1k of resistance between them although on one of the transistors it initially measures 8k before dropping semi-rapidly to 1k

Not too experienced with transistors so I'm not sure if this is good or bad

Upon researching I found they are not entirely suitable. The main difference is the AU103 is 10W compared to the 2N2869's 30W. Also the emitter to base voltage is only 4V compared to the latter 10V

If these do need replacing, finding suitable replacements is very expensive and I was wondering if a silicon transistor with equivalent values would do instead if it came to it?

Thanks

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

I think there was a later version of this amp that used silicon transistors. It was short lived as they went back to valves because of the high failure rate. It was transformer driven so if you can find the circuit it may be possible to adapt it for your use.
As others have said the driver stage is suspect because of the cap error. Maybe an AD149 might do for the outputs as they were more common in the UK.

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

When testing the transistors, are you looking for two diodes, C-B and E-B, by testing each of those junctions with your ohmmeter both ways round? In the past, I’ve found that testing power germanium transistors C-E with the base open circuit gave inconclusive results, probably because of leakage currents to the base.

If you stick with (expensive) germanium power transistors, I’d also invest in a 270ohm 5W resistor, and connect it between the power transformer centre tap and the first big electrolytic ie in series with the power feed to the amp. If all is well, the amplifier will work, but with reduced output. If you still have a fault, eg in the biassing chain, the resistor will get hot (maybe very hot!) but your new transistors should survive.

Stuart

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

Put the negative meter lead on the base and measure the resistance to the collector and emitter, repeat with the positive lead on the base, and post the results here. Do this for the 2 drivers and 4 output transistors.

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

As the above mention, you need to check with the meter 'both' ways. Use a high resistance range too.
I would start with the C to E test first. If it is very low or short in both directions the transistor is dead. As Stuart says Ge transistors can throw odd results.

Do you have any history of the amplifier? Has it ever worked?

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

If your DMM has a diode test facility you can test the transistors as described here:-

https://www.vintage-radio.net/forum/...ad.php?t=42194

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

2N2869 seems pretty rare.

The AU103 was designed for TV's but should be OK despite the 10W dissipation restriction. The quiescent conditions are 33mA and 22V so about 0.7W. The extra speed should not be a problem as there is no feedback.

The OC28 should also work and there are a few at reasonable prices on ebay. The OC25 Vce is marginal.

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

I totally agree - resistance readings across transistor junctions can be very confusing, whereas the diode test setting on a DMM quickly tells you whether there's a junction present. Look for around 0.7V across a silicon junction and around 0.3V across a germanium junction. If there are two working junctions present in a transistor, then it's probably OK. I've managed to faultfind successfully this way on several occasions even without the benefit of a circuit diagram.

Martin

Martin

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

Borrowed a DMM from work that has the diode function. Here are the results for the left channel (positive lead = first letter)

AU103 (1)
C-B - 0.119v
B-C - No cont
E-B - 0.139v
B-E - No cont
E-C - 0.29v (decreases from initially 0.9v)
C-E - No cont

AU103 (2)
C-B - 0.099v
B-C - No cont
E-B - 0.132v
B-E - No cont
E-C - 0.152v
C-E - No cont

2N2147
C-B - 0.120v
B-C - 1.806v
E-B - 0.120v
B-E - 1.807v
E-C - 0v (Short)
C-E - 0v (Short)

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

The 2N2147 is dead. The AU103's are probably OK.
The 2N2614 may also have been killed by the 2N2147 so you need to do the resistor measurements on that too.
The 2N2147 is a Germanium NPN power transistor. As these look to be difficult to find I would substitute a silicon one, a 2N3055 should do.


What is the position with the other channel amplifier? Is that working?

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

2N2614
C-B - 0.256v
B-C - No cont
E-B - 0.261v
B-E - No cont
E-C - 0.309v
C-E - 1.471v

A 2N3055 is an NPN transistor, is there a PNP equivalent?

I have not been able to test the other channel with an input just yet due to the fuse blowing issue. Admittedly I could de-solder some wires to isolate the left channel but I haven#t got round to it yet

Thanks

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

2N2614 is probably OK.
2N2147 is NPN hence why I suggested a 2N3055. If you only have different NPN power transistors to hand then post the details here so we can advise if any would work.

I would disconnect all four power output devices AU103/2N2869 so you can get the driver stage voltages correct. There is DC feedback from both the output and driver stages but the output bias circuitry should maintain the -22.5V so the feedback should not be a problem.

Clip your meter on the collector and emitter of the 2N2147 to measure voltage and power up but only long enough to take a reading. According to your circuit it should read about 10V. The collector current would be around 140mA.

Whilst you have the other channel transistors disconnected do the diode tests on them.

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

Slow down guys!
2N2147 is a PNP transistor as are all the ones shown in the part schematic.
https://www.radiomuseum.org/tubes/tube_2n2147.html

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

Ooops! You are right, comes from reading the circuit on it's side. 2N2955 would do then.

Re: Rock-Ola 40218-1-A Transistor Amplifier Problems
 

Further to my post - They stayed with PNP outputs but changed the drivers to NPN before finally getting rid of the transformer driver.