Adding a selectable line input to valve radiogram

[emeritus]

Re the potentially live metal pot shafts, where the knobs were secured using grub screws, it was often the practice to fill the grub screw hole with hard wax or the like to prevent possible shock. To remove the knob, this wax had to be dug out to access the grub screw, and was not always replaced on re-assembly.

[Rich400]

Originally Posted by dave walsh

Given that you have a professional background and a strong sense of responsibility Richard, I can see why you are concerned but this almost belt and braces" approach could be hazardous in itself. I can't give you a technical analysis but members often do start trying to add an earth and it's usually not a good idea. The really interesting aspect of live chassis design era is that there don't seem to be any examples of fatality or series injury that can be located [except in Repair Workshop's that is]. This applied to the more basic designs, where the valve heaters would be "daisy chained" at mains potential'. You seem have an Auto transformer included here. Historically, there were two reasons for the adoption of AC/DC techniques [a] Cheaper production costs [the mains transformer was a costly item] and [b] post war flexibility when many areas still had DC mains. Both types of power systems could be simply covered without a transformer in place.

It wouldn't look so good now but it was safe and it worked. People were careful and got the benefit of cheap consumer goods, using highly insulated casings and electrically separated links. Even the, perhaps higher risk, metal record deck or a valve mains/battery portable could be isolated, via a suitable capacitors, as already mentioned. Only Tape Recorders were a no go for obvious reasons perhaps. If your RG was returned to it's original standard I'm sure you would ensure that the negative lead went to chassis and not vice versa for example and be in sole charge. Often the two core lead had the same colours. That's easily updated without making a significant change I'd say but even a "live chassis is only a problem if it is live and can be accessed.

Of course I'm not making any recommendations here. It's up to you in the end as stated. I hope you ultimately get to safely enjoy your carefully refurbished Radiogram.

Dave W

Many Thanks Dave nicely put.
I do have concerns with the safety of the metal deck of the garrard 121, and the connection downstream of this directly to the chassis being made by a two core bulgin connector. This has writing on one side but no identification as to which is the fused side.
I really don’t want to remove all the internals and swap for a more modern amp with modern safety standards, but I need to satisfy myself this is safe. It’s a real shame Richards suggestion wouldn’t work. But I think there is a real justification for fitting an on board RCD and earthing this deck via a 3 core mains cable. If it’s not possible to remove the live aspect of the chassis and i can’t locate a suitable isolation transformer, then I feel I need to improve the insulation of the five points that could be touched externally. These are the four pots on the front and the record lamp blister.
I could exchange the metal pots for ABS versions and provide a more rigid cover inside the flimsy blister of the lamp unit.
I just have visions of a future grandchild visiting and pulling the knobs off (easily done) and grabbing the live pot. Or leaning on the lamp blister and doing the same to this lamp mount.
This only needs the issue of introducing the proposed line input to the stereo input to the amp resolved!

[Rich400]

Originally Posted by emeritus

Re the potentially live metal pot shafts, where the knobs were secured using grub screws, it was often the practice to fill the grub screw hole with hard wax or the like to prevent possible shock. To remove the knob, this wax had to be dug out to access the grub screw, and was not always replaced on re-assembly.

I will check this but sadly it appears that this type of knob does not have grub screws just a sprung steel flat gripping the ‘D’ flat on the shaft. Very easily pulled off.
I purchased 5 ‘new old stock’ spares I managed to find as when I purchased this a week or so ago it had one knob missing and a live pot exposed!

[Richardgr]

Hi Richard, I've been trying to read some of the fine print on those schemas you have posted but I just can't get the resolution. I was interested in the voltages from the rectifier tubes.

That capacitor C90, across the AC supply, is the one the guitar world call the 'death cap', unless I am mistaken. Nowadays that would have to be X2 or Y class - I think X2 in this case since the failure would trip your mains.

The rest of the amplifier looks to be a very nice circuit, with seperate feedback windings on the OPT and separate cathode bias for the output tubes.

Just my thoughts ...

If you can use a small isolation transformer close to the existing transformer, use a 3 wire lead and earth the chassis, then the device will be galvanically isolated, and the fact that the potentiometer spindles are metal is not such a great issue.
60VA isolating

The supply from the isolating transformer should be fused (100ma slow blow?), and the power supply switch and noise suppression capacitor should be before the isolating transformer. Then I think it is as good as any modern tube amp.

You can always buy knobs that screw on to the posts to improve safety there.

[Rich400]

Hi Richard
Great work there and really appreciated.
Good and bad news here, I’ve attached another photo hopefully this will show what you are looking for. The resolution seems to be vastly reduced when it is uploaded to this site, please let me know.
The good news is to relocate the C90 cap made me remove the back and take a look and I’m glad I did! Check out the photos for the bad news...
I think I’ll sit down later and have a think about your proposals as to what will then be live and not etc.
My immediate thoughts on the new secondary (100mA)slow blow fuse is it sounds a little low?

[Richardgr]

Originally Posted by Rich400

The good news is to relocate the C90 cap made me remove the back and take a look and I’m glad I did! Check out the photos for the bad news...

As we say over here, 'ojdå!'. (= whoops!).

All those time-limited passives will have to be hunted down and replaced, anyway. That was a good picture to illustrate the perils of those wax capacitors!

[snowman_al]

Originally Posted by Richardgr

... If you can use a small isolation transformer close to the existing transformer, use a 3 wire lead and earth the chassis, then the device will be galvanically isolated, and the fact that the potentiometer spindles are metal is not such a great issue.
60VA isolating...

That is Not powerful enough. If you work out the total current consumption, the 6.3 volt heaters and bulbs come to approximately 44 watts. Then add the HT current, guess 180mA at 240 volts that's 43 watts so you need an isolation transformer of at least 90 VA so a 120 VA transformer would be a good start, bigger is always better here.

Are you still wanting to add a line into this device?
If not then repair and renovate and leave every thing as it is, do not add an extra earth to the unit.

If you do still want another input, maybe explore a Bluetooth solution so there is no physical connection.
Or thinking out loud, if you can isolate the neutral / chassis connection on the transformer primary perhaps supply the 240 AC to the rectifiers from a separate transformer or back to back transformers. That isolates the mains from the chassis. Then you could earth the chassis.

[Rich400]

Hi Snowman thanks for your advice, it really is appreciated. Any ideas where i can find this isolation transformer?
I still intend to add a line in, specifically a Yamaha musicast input. This connects via standard RCA plugs. Unfortunately the bluetooth option (although a great suggestion for isolation) won't get me where I wish to be.

Edit: just seen your attached drawings, very much appreciated, this site has been extremely helpful and that really is brilliant.
I notice you have deleted the smoothing capacitor, is this now not required?
Would this suffice? https://uk.rs-online.com/web/p/din-r...B&gclsrc=aw.ds

[Radio Wrangler]

If your final system involves a link between neutral and earth, you can still get enough current imbalance between line and neutral to trip RCDs due to neutral voltage with respect to earth. Especially when loads are being switched on and off, you are likely to get nuisance tripping.

David

[snowman_al]

Originally Posted by Rich400

I still intend to add a line in, specifically a Yamaha musicast input. This connects via standard RCA plugs. Unfortunately the bluetooth option (although a great suggestion for isolation) won't get me where I wish to be.

Edit: just seen your attached drawings, very much appreciated, this site has been extremely helpful and that really is brilliant.
I notice you have deleted the smoothing capacitor, is this now not required?

Can you not connect a Bluetooth transmitter to the Yammy outputs? And a Bluetooth receiver in the Gram?

The cap C90 suppresses the arc you may get at switch off and protects the switch contacts. It will work fine without it, though you may get a click or pop occasionally as you switch off.
Replace it with an X rated item if it is a concern, (X = across the mains.)

[Rich400]

Originally Posted by Radio Wrangler

If your final system involves a link between neutral and earth, you can still get enough current imbalance between line and neutral to trip RCDs due to neutral voltage with respect to earth. Especially when loads are being switched on and off, you are likely to get nuisance tripping.

David

Hi thanks for the comment David
Do your concerns apply even considering the neutral to earth connection is on the secondary side of the transformer(s)?
I am thinking the primary side being monitored by the earth trip would still be balanced.

[Rich400]

Originally Posted by snowman_al

Can you not connect a Bluetooth transmitter to the Yammy outputs? And a Bluetooth receiver in the Gram?

The cap C90 suppresses the arc you may get at switch off and protects the switch contacts. It will work fine without it, though you may get a click or pop occasionally as you switch off.
Replace it with an X rated item if it is a concern, (X = across the mains.)

Hi Snowman
I suppose it i possible but concerns are bluetooth to bluetooth sync on start up, and as I am not familiar with this set up as I have only seen smart devices connected to bluetooth modules and not two dumb BT modules syncing. Thats not to say they do not exist though. As all the audio grounds will be earthed (and hence protected from a fault condition) it is just the two line in's that need to be electrically isolated. I was thinking Opto isolators but not sure of they would be fast enough. There has been a suggestion of isolation capacitors but I am not familiar with these.

[Radio Wrangler]

Originally Posted by Rich400

Hi thanks for the comment David
Do your concerns apply even considering the neutral to earth connection is on the secondary side of the transformer(s)?
I am thinking the primary side being monitored by the earth trip would still be balanced.

'fraid so. An RCD passes the neutral wire and the phase wire through the same toroidal transformer. A secondary wire into a load resistance therefore carries a scaled (by the turns ratio) replica of the difference in current between neutral and phase connections. If any current goes missing, or any extra current appears from somewhere and the discrepancy is above the RCD rating, then the RCD trips.

So if your chassis is connected to neutral, then you need to be sure that nothing that gets connected to the chassis can connect back to earth by any route. Or even to the house neutral by any other route than that one wire through that transformer.

You must also consider that neutral is not officially considered safe to touch as various faults can see it with phase voltage on it... and therefore anything with a connection which traces to your chassis gould become dangerous.

The whole business of trying to keep things safe under all circumstances is messy. Not as messy as the aftermath of an accident. Different people have different opinions on what is and isn't safe. I ask myself what would be a coroner's opinion, because that would be the absolute worst case. If you've got that one covered, you're OK, I reckon. If you haven't got that one covered, you might wish for a time machine to nip back and do things differently, so I take the view that if the worst happened, how would I wish I'd done it?

I've just spent 14 years designing radio stuff to link aircraft into automatic anti-collision systems and air traffic control, so I tend to play safe.

I'd either fit an isolating transformer and a 3-core mains lead, or I'd replace the transformer in the gram with a bigger, fully isolating one to power heaters and HT. With a full wave rectifier, the HT will be smoother and less hummy.

David

[Rich400]

Hi David
thanks for taking the time with that reply.

This is my take on this, if I have it wrong I would be pleased to be corrected.

When I was referring to the RCD not tripping as it was balanced I was remarking on an RCD fitted to monitor the primary side of the transformers.

The secondary side of the proposed isolation transformer would be effectively floating until it was earthed, and as such has no direct connection to the 'mains' neutral and is independant of it.

Considering the secondary side, in the event of a fault to earth the RCD monitoring the primary side would see more load, but it would also see it all returned, and hence remain balanced and would not trip (more's the shame). Of course excessive load in the event of a serious earth fault on the secondary would trip (eventually) the current trip of the RCBO, but this is not a protection for people it is for the equipment, the RCD takes care of that, at least on the primary side.

When you add the earth to the chassis and common it with the unfused side from the secondary side of the isolation transformer, this is only neutral in name only as it has no direct connection to the primary RCD monitored circuit.

The protection against fault to earth in the secondary side is made by the inclusion of the fuse positioned at the 'phase' connection of this secondary winding.

[Radio Wrangler]

Yes, if you have the existing heater transformer doing safety isolation to the heaters, and a new transformer doing safety isolation for the HT supply, then that should be fine... there being no connection from live or neutral to anything except for the two transformers' primaries and the motor.

Worth thinking of a full wave rectifier for less hum once you have an HT transformer.

David

[Michael Maurice]

I've serviced a few of these and never really taken much notice of the mains/chassis relationship.


I've just checked the circuit and the pick up cartridge is connected directly to chassis. So if the chassis is live, so is the cartridge.

The diagrams do not show the connections for grounding the deck.

I supposed what fooled me was the use of the two EZ80's which are normally associated with full wave rectification.

Its very odd that Decca would design it in this way.

[Rich400]

Originally Posted by Radio Wrangler

Yes, if you have the existing heater transformer doing safety isolation to the heaters, and a new transformer doing safety isolation for the HT supply, then that should be fine... there being no connection from live or neutral to anything except for the two transformers' primaries and the motor.

Worth thinking of a full wave rectifier for less hum once you have an HT transformer.

David

Many thanks David again it’s appreciated you taking the time to consider this.
I would be interested in the full wave rectification you suggest as although this unit can still operate (via a DIY earth trip) it has a very nice and hum free sound. I would hate to introduce some sound degradation with any safety mod. It is desperately in need of a service (see previous cap photos) so nothing is being done until all the mods are Theoretically finalised.
It has just occurred to me as I am writing however that the amplifier we have discussed is not the only system mounted on this chassis, of course there is the tuner! I will revisit the service manual and see if there are further issues to consider, but almost certainly there will be more load so the new T(m) transformer may have to be updated from its suggested 50 vA. I will post some photos of the tuner drawing shortly.
Thanks again

[Richardgr]

Originally Posted by Rich400

. . . I will revisit the service manual and see if there are further issues to consider, but almost certainly there will be more load so the new T(m) transformer may have to be updated from its suggested 50 vA. I will post some photos of the tuner drawing shortly.
Thanks again

Snowman_al suggested 120VA, which is better than 50VA. Transformer prices don’t vary so much at that rating, and size should not be an issue. Your favourite auction site can be an option.

[Rich400]

Originally Posted by Michael Maurice

I've serviced a few of these and never really taken much notice of the mains/chassis relationship.


I've just checked the circuit and the pick up cartridge is connected directly to chassis. So if the chassis is live, so is the cartridge.

The diagrams do not show the connections for grounding the deck.

I supposed what fooled me was the use of the two EZ80's which are normally associated with full wave rectification.

Its very odd that Decca would design it in this way.

Hi again Michael
Thanks for your comments, I will check the drawing for the cartridge connections, but if what you say is correct more reason to remove this chassis from live duty. It seems like you say though that it would be unlikely that the cartridge would be designed live.
With regards to the deck earth, this is what initially had me worried about this machine. Despite there being no isolation transformer on the mains in, the 240ac cable in is two core and goes straight to the deck motor connection then paralelled off to the two core (reversible) bulgin Plug to the chassis.
Of course there is no earth tag or similar to the metal garrard 121 deck, as there is no earth cable, and the fused live in to the chassis is reversible without indication what way it should be plugged in.
You can see why I need to improve things here.

[Edward Huggins]

This Thead seems to have taken on a complete life of its own and, to me, seems to have created it's own complexities.......
The OP appears to have good understanding of what's going on here, so why not just add a rotary 2 way switch to the gram input socket, use 2 inline "Y" series 500V DC isolation caps, fit the switch to the rear card cover panel, ensure the mains neutral lead is connected to chassis and get going?!
Perhaps: 1: This is all too simple 2: I'm missing something here 3: I've lost the plot!