Bush TR82C - some musings on component spec anomalies.

[David G4EBT]

Some musings about what to me seem to be unusually high working voltages of the caps in this radio, and other anomalies, the logic of which eludes me.

I have several Bush TR82Cs in various states waiting for attention, so have got one on the bench now. (AF117 Mixer/IF version ). It’s some years since I last restored one, so it’s a while since I’ve looked at the circuit. In looking at the component specifications, I note that some of the working voltages are very high, which seem puzzling for a battery powered transistor radio. Just musing on why that might have been so. It may reflect that towards the end of the valve era, to some extent, they used what standard components were available at the time, but I’m not sure that is so.

Some examples (Numbering is from the Bush Data Sheet not the Trader Sheet):

C3: 10pf padder in the front end, (subject to less than 1 Volt) is 750V. Huh?

Most of the other non-electrolytic capacitors are 125 – 250V, but not all.
For example, C17 .01uF 12V, yet C18 .01uF is 125V, and C28 (0.1) is 250V.

Several .04/.047 are rated at 150 – 250V.

As to electrolytics: two are 8uF 12V and one is 8uF 6V. (Why not all three 12V?). -20%/+100% tolerance, hence, from 6.4uF to 16uF.
The 6V one is much smaller physically than the other two, but doesn’t need to be as space isn’t at a premium.

100uF: two are 3V, two 12V. (Also -20%/+100% tolerance, hence, from 80uF to 200 uF). Why not just standardise on 12V?

Some might say: "It will be the accountants overriding the wishes of the design engineers to say a penny here and there". I somehow doubt that. If accountants had a hand in it, why would they want to extend the inventory by having similar values of capacitors at differing working voltages, and why a 750V for a cap with less than 1V across it, and 250V caps elsewhere? I guess we'll never know the rationale, but it seems a bit if a mish mash to me, with no apparent logic to it.

As it’s as much effort to lift one end of each electrolytic cap to test them in turn as it is to just replace them, given that they’re 60 years old and have had a good innings, I’ve replaced them (using 10uF in place of the 8uFs). I’ve checked the capacitance and ESR of the removed ones, with the following results:

100uF 12V on test: 108 uF, ESR 1.6 Ohms
100uF 12V on test: 112 uF, ESR 5.8 Ohms.
100uF 3V on test: 94 uF, ESR 38 Ohms.
100uF 3V on test: 88uF, ESR 32 Ohms.
8uF 12V on test: 19uF, ESR 3.4 Ohms.
8uF 12V on test: 20 uF, ESR 2.6 Ohms.

(I’ve yet to remove and test the diminutive 8uF 6V one, C21).

Despite two having high ESRs, no doubt had they all been left in situ, these caps would have not greatly detracted from the performance of the radio.

As an aside, for the most part, the electrolytics I’ve most often found to be leaky and morphed into low value resistors have been the blue Philips ones in Roberts radios such as the R505. Always a giveaway when at switch on, the radio seems to work for a couple of seconds then stops. A check on the consumption with a mA meter in line with the battery supply show maybe an Amp or so being drawn due to leaky caps in the output stage. The output transistors won’t stand that for long.

I noticed too, that the tone control of the TR82C (ganged with the on/off switch) is 10K inverse log, (the replacement of which would have to be a Blore Edwards special if they could do one). I’ve never had one fail, but if ever one did, I wonder if there would be any discernible difference if a linear law track was used?

The radio has been got at before it fell into my hands and some caps and resistors had been changed. It does work after a fashion, but with distorted audio. (Changing the electrolytics had no effect on that). That seems not to stem from the audio stage but from the front end/IF stage. A 1K tone injected into the audio stage sounds fine from the speaker, though I haven’t checked the waveform on the ‘scope yet. Using my signal tracer at the second transistor into IFT2, the signal sounds undistorted, but distorted at the input to IFT3 from TR3 so I’ll focus on that area of the radio.

I’ve not checked any voltages as yet or to see if any resistors are out of spec.

From a design perspective, I notice that the tone control, ganged with the on/off switch is 10K inverse log, (which would have to be a Blore Edwards special if they could do it). I’ve never had one fail, but if ever one did, I wonder if there would be any discernible difference if a linear law track was used? (I think that generally, tone control pots are linear if not inverse log).

Pic 1: Six of the removed electrolytic caps which have been replaced.
Pic 2: The diminutive (white) 8uF 6V cap (C21) still in situ, ringed in yellow.

[Graham G3ZVT]

I've always seen the TR82 as a valve set with transistors shoe-horned into it, and in a sense this is perfectly true.

You didn't mention the mains voltage double pole on off switch.

[crackle]

No doubt a matter of what they had in stock. I know for a fact that if I was replacing a .01uF 150v cap I would fit a .01uF 630v one. I just don't stock all the values and all the voltages.
Mike

[Graham G3ZVT]

That's what is so interesting about these.

Transistor sets from a factory still using valve tooling and techniques.

[Refugee]

The earlier OC series transistor ones had output transformers that were wound on lams and frames of the type usually seen in valve record players.

[Boulevardier]

I’ve often wondered about such things, and reflected that manufacturing, eg padders that had a very low max working voltage, might actually have been more expensive than making ones with higher voltage - making very thin mica insulation might be a more expensive process overall than making mica that would withstand 700V. Perhaps even more so back in the 1950s and 60s, when manufacturing was still more geared up to valve voltages and the processes were less developed for miniaturised production. Also, with a radio like the TR82, I shouldn’t think space for padders was a huge problem.

Mike

[Tractionist]

And ..... the on/off switch is a dirty great 250v DP unit too! I've changed the tone controls on these - but using a standard log [which fools me occasionally when I use one of these sets!].

They are soooo easy to work on ............ once one gets the dirty great tuning knobs off!!!

[crackle]

Its one of the few transistor radios that had the on/off still on the tone control. Again a legacy thing from the valve days.
Why did most valve sets have the on/off on the tone switch?
Mike

[Graham G3ZVT]

Originally Posted by crackle

Why did most valve sets have the on/off on the tone switch?
Mike

Well one reason would be you don't have to guess how far to advance the volume it while the set is warming up, it just "remembers" the last setting.

I never advance the tone control of my TR82 any further from the "on" detent. Perhaps when they were new the MW/LW transmissions had more "top" and benefited from some top-cut, I'm not sure when the current restrictive "brick-wall" transmission filters became mandatory.

[PsychMan]

At a guess, because the tone control would click to the ON position at the bassier and more acceptable end. So people could switch it on and walk away while it warmed up, rather than waiting to adjust the volume to their desired level?

Seem to have cross posted as I didnt refresh the thread from yesterday

[David G4EBT]

Originally Posted by David G4EBT

As it’s as much effort to lift one end of each electrolytic cap to test them in turn as it is to just replace them, given that they’re 60 years old and have had a good innings, I’ve replaced them (using 10uF in place of the 8uFs). I’ve checked the capacitance and ESR of the removed ones, with the following results:

100uF 12V on test: 108 uF, ESR 1.6 Ohms
100uF 12V on test: 112 uF, ESR 5.8 Ohms.
100uF 3V on test: 94 uF, ESR 38 Ohms.
100uF 3V on test: 88uF, ESR 32 Ohms.
8uF 12V on test: 19uF, ESR 3.4 Ohms.
8uF 12V on test: 20 uF, ESR 2.6 Ohms.

(I’ve yet to remove and test the diminutive 8uF 6V one, C21).

Despite two having high ESRs, no doubt had they all been left in situ, these caps would have not greatly detracted from the performance of the radio.

The radio has been got at before it fell into my hands and some caps and resistors had been changed. It does work after a fashion, but with distorted audio. (Changing the electrolytics had no effect on that). That seems not to stem from the audio stage but from the front end/IF stage. A 1K tone injected into the audio stage sounds fine from the speaker, though I haven’t checked the waveform on the ‘scope yet. Using my signal tracer at the second transistor into IFT2, the signal sounds undistorted, but distorted at the input to IFT3 from TR3 so I’ll focus on that area of the radio.

I’ve not checked any voltages as yet or to see if any resistors are out of spec.

Pic 1: Six of the removed electrolytic caps which have been replaced.
Pic 2: The diminutive (white) 8uF 6V cap (C21) still in situ, ringed in yellow.

Just an update on the above comments.

Re the audio quality and apparent 'distortion':

I checked the waveform on my 'scope and with a 2mV p-p 1kHz sine wave (from my Velleman K7000 injector/tracer) injected at the volume control, with the volume almost turned up to maximum, the sine wave at the speaker was 2V p-p and undistorted. Only when the VC was turned to maximum volume did the sine wave start to clip.

So if the audio stage was fine, why the distortion?

I checked the alignment and firstly, the IFs needed to be peaked, and the RF stage was also off tune. After realignment, (clearly outlined in the datasheet and only takes a few minutes), there was no distortion and excellent signal strength. Here in East Yorks, on LW, in addition to R4, RTE from Ireland was loud and clear, and a French station perfectly audible too. Lots of stations on MW, notably the shouty blowhards on R5 Live, banging on about sport. I bet they’re a pleasure to live next door to. (not!).

All back together and sorted.

Second TR82C (also MK2):

This radio was dead.

As with the first radio, I injected a signal into the audio stage, which produced a loud tone at the speaker, and a nice and clean waveform on the 'scope so the audio stage was fine. As well as AF117s in TR82Cs being suspect due to the tin whisker problem, the OA 90 detector diode can sometimes fail, so working back from the audio stage I checked for signals at the input to the diode (which is hidden in the can of IFT3), and there were no signals. Had the OA90 been defective, (open circuit), if the IF and RF stages were working, signals would have been heard at the input to the diode on pin 5 of IFT3, but not at the output on pin6. If the diode does fail, there's no need to go to the trouble of removing the IFT and disgorging the diode from the can to replace it. A new diode can be soldered across pins 5 & 6 on the underside of the can, leaving the dud OA90 to rest in peace in its little metal 'coffin'. I keep a stock of 1N34 germanium diodes as a universal replacement for any OAxx germanium diodes. (Typically ten for £2.00 post free from UK suppliers on eBay).

So, the silence pointed to TR1/2/3.

Working back towards the RF stage with the signal tracer, I checked at the collectors of TR3, TR2 and TR1, with nothing heard. If the RF stage (TR1) isn't working, nothing will get to the IF stages, so I checked TR1 first. I snipped the screen of the AF117 to see if it was the 'tin whisker phenomenon' to no good effect, so I removed and tested the transistor, which came up as a diode. I had a spare NOS AF117 (which is no guarantee that it would be any good), and it tested fine. When fitted, the set came to life, but I snipped the screens on TR2 & 3 in case at some later stage they grew whiskers. I checked the alignment, which didn't need tweaking, and the radio was lively on both bands.

As with the first TR82C, though the radio was working fine, given the small amount of effort involved, I decided to change the electrolytics, with the following results:

100uF 12V: 120uF, ESR 1R
100uF 12V: 148uF, ESR 2.3R
100uF 12V: 96uF, ESR 17R
100uF 3V: 76uF, ESR: 48R
8uF 12V: 14.5uF, ESR: 14.5R
8uF 12V: 20uF, ESR 6.6R

There were two 0.01uF waxy paper caps, so I changed those too. One measured 0.02uF, the other .017uF, so they would have been fine left in place.

I've always thought those Plessey electrolytics which are red at one end, black at the other, with the yellow label peeling off, looked a bit suspect and time expired, but based on these findings, that’s a mistaken view. Also we tend to view waxy paper paper caps with scorn, and not without good reason in valves sets such as audio coupling caps. And in the above two radios, it takes almost as much effort to test each cap by lifting one end (always remembering to make sure it's discharged so as not to damage the test-meter), as it does to remove and replace them. That said, despite their age (at least 60 years), all of the above capacitors could have been left in place without any adverse consequences.

A third TR82C - a MK1 with all OCxx transistors was working well.

The voltages around the transistors were close to spec, and the audio waveform looked fine on the scope, so I left all the electrolytics in place except one - a 500uF (6V) Plessey tucked away behind the speaker, (replaced with 470uF). I also replaced a 0.25uF waxy across the tone control, one end of which was swinging loose.(Replaced with 0.22uF. On testing those capacitors, the 500uF one measured 517uF, with an ESR of just 0.96 Ohms, so in effect, it was as good as new. The 0.25uF measured 0.95uF so was well out of spec, but I doubt it mattered much.

In conclusion, before I bother to check and replace electrolytic capacitors on transistor radios, I'll first check the current being drawn by the radio. On the TR82C, the datasheet says it should typically be between 20 - 30mA with a station tuned in at listening volume, though signal peaks will of course be greater on music than speech. With each of these radios, the consumption varied from 15 - 20mA.

The transistor radios I've had most issues with due to electrolytic capacitors going leaky (in the electrical sense) have been Roberts (EG RT1, R505) with the little blue Philips caps, and that soon becomes evident if the battery consumption is excessive. One clue is when a radio is switched on works for two or three seconds than goes silent. If checked on a metered power supply or a multi-meter on the DC current range in series with the battery, it becomes immediately obvious that excess current is being drawn due to leaky electrolytic capacitors.

Quite a lot of the resistors in these TR82Cs had drifted high in value - some by 30%, though as the voltages around all of the transistors on the three radios were close to spec, and given that the radios are all performing well, changing them would be a pointless exercise. (As was changing the el caps, come to that).

In all of the radios the tuning caps were wobbly, as is the norm on TR82Cs due to the three supporting grommets being perished so I replaced those. Some time ago I bought some grommets of the correct size into which the brass spacers fit, which when compressed, give the necessary support to the tuning capacitor.

This third set has some cosmetic issues to sort, which I'm on with and will mention later.

Pic 1: 1kHz 200mV p-p input to audio stage, undistorted 2V p-p output.
Pic 2: Front of chassis after restoration.
Pic 3: Rear of chassis after restoration.
Pic 4: Caps replaced in second and third TR82Cs.
Pic 5: Sketch of the dimensions of required turning capacitor support grommets.

Hope these rather verbose ramblings might be of interest.

[paulsherwin]

I haven't found the need to change caps in TR82s - my own TR82C is still on its originals. Any problems are usually down to transistors which are particularly easy to change, and the wavechange switch which seems prone to gunking up. The circuit design is very robust and tolerant of out-of-spec components.

[David G4EBT]

Thanks for reading the thread Paul, and for your comments.

Yes, both the wave-change 'rocker switch' and the actual switch contacts on all these sets needed attention. I used Deoxit and as toothbrush on the contacts, and some PTFE loaded lubricant on the mechanism.

The MK1 '82C with all Mullard OCxxx seem very reliable.

Shame about the AF117s, but easily sorted and revisited many times on the forum.

One of the nicest sounding transistor sets around, and quite commonplace.

Clearly very popular in its day and deservedly so, with a long production run.

[paulsherwin]

It was quite an expensive set in its day - nearly twice as expensive as a typical 4 valve battery set. It was much cheaper to run of course.

[AD360 Rob]

As always David, a thoroughly enjoyable read and informative too. I have done many a TR82, even built a couple up from parts and with very few exceptions I have found the electrolytics to be fine. Personally I much prefer the sound from the TR130 but as for sensitivity, a TR82 beats most transistor sets of the period (and later) into a cocked hat and looks good in the process.

[kellys_eye]

Can I ask where you sourced your grommets from?

[David G4EBT]

Originally Posted by kellys_eye

Can I ask where you sourced your grommets from?

Sorry, I can't recall as it's some years ago.

However, there are countless suppliers of grommets of every shape and size in rubber and PVC. (The ones I used are PVC).

For example, the ones at the link below would I think be suitable, and are only 16.8p each, but on the face of things, the minimum order quantity is 100, so that puts them out of bounds. However, it might be worth giving them a call to see if they will accept small orders, and they have a chat line at the website. I guess they're a suppler rather than a manufacturer, so a google search might turn up some of the same or similar size.

I've attached a diagram showing the dimensions of the ones I'm referring to, which are termed: 'SKU# PV9'

https://www.reevite.co.uk/catalog/pr...8aApgdEALw_wcB

Email: sales@reevite.co.uk

Tel: 01869 252520

Reevite,
16 Murdock Road,
Bicester,
Oxfordshire,
OX26 4PP.

The critical dimensions are that the brass spacer inside the grommet in the TR82 is 6.2mm long. That ensures that when the securing screw is tightened, it doesn't foul the tuning capacitor vanes. The outside diameter of the spacer is 5.5mm so the hole in the centre of the grommet needs to be no more than 6mm. When tightened, the gap between the mounting bracket of the tuning cap and the tuning cap frame is 2.75mm, so the thickness of the rims of the grommet need to be at least 3mm so that then the mounting screw is tightened, the tuning cap won't wobble about, which is the problem we're seeking to resolve.

Most grommets are designed to simply fit a hole to prevent cables from being cut into by sharp metal edges on electrical boxes. In this application they're being use for a different purpose so the 'bulbous' ends need to be thick enough to stop the capacitor wobbling about. This all sounds a bit obscure without reference to the sketch and with the chassis in front of you.

I hope that might help a bit.

I believe that others on the forum have used different solutions to overcome the problem.

[paulsherwin]

You can buy a large selection of neoprene grommets from eBay sellers for a pound or two.

[Graham G3ZVT]

I found that actual grommets were unnecessary, synthetic rubber washers or perhaps O-rings can be used. But as I didn't have any of the above in a suitable size I made a pair rubber washers, each using a stack of five pieces of self-amalgamating tape and an office hole-punch.



SA tape is very useful stuff if you haven't used it before.