Roberts R707 circuit oddity. - Page 2 - UK Vintage Radio Repair and Restoration Discussion Forum

AC/HL · 30th Aug 2014, 1:14 am

The term "Lockfit" refers to the barbs on the leads used to hold the device in place during assembly & soldering.
Perhaps some machines, or operatives, were more heavy handed than others?

mhennessy · 30th Aug 2014, 12:27 pm

I hadn't spotted the reduced value of R27, but I guess it makes sense as the loudness network presents a lower impedance to that stage.

Regarding Lockfits, the failure mode that I encounter most often is increased noise. However, whether this is noticed depends on where the transistor is in the circuit.

So, for example, the first transistor in the amplifier circuit used in common Hacker sets is quite vulnerable to this, and problems here are very common indeed. But taking that faulty transistor and putting in the T2 position doesn't result in anything like the same noise problems because there isn't so much gain following on from T2. If the same transistor was being used as a Vbe multiplier, or as a simple voltage regulator, then its noise wouldn't be noticed at all.

With that in mind, it's hard to come up with a sense of how reliable or unreliable they are. I've often thought about making a test jig that biases up a suspect transistor and follows it with a large amount of AC gain using an op-amp. The result could be rectified and displayed on a simple meter. Then I could grade transistors, and use the noisier ones in positions where the noise isn't an issue. But life is too short

The second problem I notice concerns the BF19x series, which tend to cause the set to have low gain. If the transistor is reasonably easy to get at, I check by substitution. OK on the IF boards, harder on FM front ends. This is mostly a problem encountered on FM sets - I don't recall having this problem in the AM sets. Again, I've often thought about a test jig to measure the gain at 10MHz or similar. But of course, it could be increased noise that causing the apparent loss of gain...

Neither of these problems cause a change in the DC behaviour, so you can't use DC checks to find them. And low RF or IF gain is hard to pin down at the best of times - in many cases the transistors have been blameless.

Given that these problems are almost unique to Lockfits, I can't help suspecting that the package might be the problem. I wonder if it allows moisture ingress? When you have a faulty sample, heating and cooling has a big effect on the noise (or signal gain) - good transistors are relatively immune to temperature.

Lastly, complete failures - the sort that can be found with an Avo - do happen. I wouldn't like to say if these are any more commonplace with Lockfits than anything else used at the time, apart from the BC464/465, which were often used as output transistors with minimal or no heat sinking (e.g. Hacker Aviemore, Sovereign IV, SP80)

Philips210 · 9th Sep 2014, 6:02 pm

Hi

I wonder if the OP managed to locate the cause of the distortion.

Regards
Symon.

G6Tanuki · 9th Sep 2014, 6:14 pm

I'm still working on it.

The voltages round the second preamp-transistor are all wrong; I'd expect rather more than 3.2V at the collector.

When I get back to it I plan to replace both the base-bias resistors and see if that makes an improvement. [if the base-to-ground resistor has gone high or is O/C it could explain a lot].

30th Aug 2014, 1:14 am	#21
AC/HL Dekatron Join Date: Nov 2003 Location: Heckmondwike, West Yorkshire, UK. Posts: 9,642	Re: Roberts R707 circuit oddity. The term "Lockfit" refers to the barbs on the leads used to hold the device in place during assembly & soldering. Perhaps some machines, or operatives, were more heavy handed than others?

30th Aug 2014, 12:27 pm	#22
mhennessy Dekatron Join Date: Jul 2007 Location: Evesham, Worcestershire, UK. Posts: 4,244	Re: Roberts R707 circuit oddity. I hadn't spotted the reduced value of R27, but I guess it makes sense as the loudness network presents a lower impedance to that stage. Regarding Lockfits, the failure mode that I encounter most often is increased noise. However, whether this is noticed depends on where the transistor is in the circuit. So, for example, the first transistor in the amplifier circuit used in common Hacker sets is quite vulnerable to this, and problems here are very common indeed. But taking that faulty transistor and putting in the T2 position doesn't result in anything like the same noise problems because there isn't so much gain following on from T2. If the same transistor was being used as a Vbe multiplier, or as a simple voltage regulator, then its noise wouldn't be noticed at all. With that in mind, it's hard to come up with a sense of how reliable or unreliable they are. I've often thought about making a test jig that biases up a suspect transistor and follows it with a large amount of AC gain using an op-amp. The result could be rectified and displayed on a simple meter. Then I could grade transistors, and use the noisier ones in positions where the noise isn't an issue. But life is too short The second problem I notice concerns the BF19x series, which tend to cause the set to have low gain. If the transistor is reasonably easy to get at, I check by substitution. OK on the IF boards, harder on FM front ends. This is mostly a problem encountered on FM sets - I don't recall having this problem in the AM sets. Again, I've often thought about a test jig to measure the gain at 10MHz or similar. But of course, it could be increased noise that causing the apparent loss of gain... Neither of these problems cause a change in the DC behaviour, so you can't use DC checks to find them. And low RF or IF gain is hard to pin down at the best of times - in many cases the transistors have been blameless. Given that these problems are almost unique to Lockfits, I can't help suspecting that the package might be the problem. I wonder if it allows moisture ingress? When you have a faulty sample, heating and cooling has a big effect on the noise (or signal gain) - good transistors are relatively immune to temperature. Lastly, complete failures - the sort that can be found with an Avo - do happen. I wouldn't like to say if these are any more commonplace with Lockfits than anything else used at the time, apart from the BC464/465, which were often used as output transistors with minimal or no heat sinking (e.g. Hacker Aviemore, Sovereign IV, SP80)

9th Sep 2014, 6:02 pm	#23
Philips210 Nonode Join Date: May 2007 Location: Redruth, Cornwall, UK. Posts: 2,580	Re: Roberts R707 circuit oddity. Hi I wonder if the OP managed to locate the cause of the distortion. Regards Symon.

9th Sep 2014, 6:14 pm	#24
G6Tanuki Dekatron Join Date: Apr 2012 Location: Wiltshire, UK. Posts: 14,007	Re: Roberts R707 circuit oddity. I'm still working on it. The voltages round the second preamp-transistor are all wrong; I'd expect rather more than 3.2V at the collector. When I get back to it I plan to replace both the base-bias resistors and see if that makes an improvement. [if the base-to-ground resistor has gone high or is O/C it could explain a lot].