philips: LD562AB; Colette; Background for Restoration

ID: 334889
This article refers to the model: Colette LD562AB (Philips Radios - Deutschland)

philips: LD562AB; Colette; Background for Restoration 
24.Nov.13 00:08
55

Michael Watterson (IRL)
Editor
Articles: 1033
Count of Thanks: 15

Philips Colette LD562AB Restoration Part 1: Background

Also known as L5D62AB and UKW-Konzert-Koffersuper. Released in Germany 1955-1956 and in Netherlands as “Clipper” L5X62AB in 1956, though only the German model has “Colette” on the case.

Before commencing restoration the Electronics and to a lesser extent the Mechanics, case, covering, power etc must be understood.

 

Summary

  • LW, MW, SW & VHF-FM
  • Battery or Mains
  • Tone control
  • Economy Control
  • Push Pull Audio out
  • Car Aerial Socket
  • External 6V LT charging (12V on Dutch models)
  • Charge light
  • Battery status display
  • Year: 1956–1958
  • Valves / Tubes 10: DF97 DF96 DF97 DF97 DM71 DF96 DAF96 DAF96 DL96 DL96
  • IF 460/10700 kHz

 

Similar models to consider and contrast are the Philips Annette series  UKW-Koffersuper LD452AB, LD462AB(Bakelite) LD471AB, LD480AB, L4D90AB (Dutch Regenboog LX4 or L4X series)  and strangely the Vidor Vanguard. There are also related Scandinavian Dux AB57 and  Dux AB58 which are like the Annette but add the Shortwave of Colette and use a DL94 (almost twice output power of DL96)

 

Philosophy

We have to decide if we are merely repairing or restoring. A contemporary professional service person wouldn’t generally modify a model unless parts where not available. “Bodgers”  of course might even incorrectly modify a set to get it “going” with no consideration of longlivity, safety and quality. People restoring sets may want to only use original parts, or leave a model just for display rather than destroy the original. Others may use a modern part of the same value, but destroy the old part by stuffing inside it. Or perhaps leave the old part in place but out of circuit with the new part behind it. Or you may feel it’s acceptable on a 1950s model to simply fit replacement capacitors and resistors. Everyone seems to have different opinions on that. 

I’m inclined to keep the original design and not add “improvements” such as DAB, USB, Earphone sockets, digital frequency counters,  modern regulators, transistor replacements of valves, changes or enhancements in circuits UNLESS such a change was an official manufacturer’s modification in production or insisted on as field service change to solve a problem. 

Similarly I prefer to try and restore the cabinet as it was, so no Danish Oil, Polyurethane or French Polish on Wood. Damaged fabric is a problem, but unless it’s completely unusable/scrap I put up with a foxed or scuffed appearance, perhaps touching up a bit.

 

Power Supply

In the USA and UK, usually Battery / Mains “tube”/”valve” models use a series filament chain. This simplifies Mains supply and ensures that if one filament fails the supply is removed  from the rest. With Parallel filaments an unregulated supply will rise to a dangerous level for the remaining tubes, especially if the output tube fails. But in 1940 in Germany the Telefunken Y8 base tubes were optimised for NiCd (sold under DEAC brand) as each tube was different filament current so could not be operated in series. The advantage of the DEAC was two fold, that it acts as a parallel voltage regulator for Mains supply and then the LT cells which are consumed at perhaps x20 current of HT are an optional purchase. 

No surprise that this didn’t catch on in the UK or USA where Setmakers either sold batteries or were originally only battery makers! Also the major UK market was for people that only bought one set, mains OR battery, depending on if they actually had mains electricity or not.

 

LT

So like many German and Philips Battery / Mains models the Mains PSU LT is regulated by a series dropper resistor and most important is the DEAC, an NiCd cell. With age these deteriorate and the voltage may rise to 2V or more. An actual NiCd  of similar or a little higher capacity in a replica box is the best solution. A modern NiMH is however more readily available. The “C” size is about 3500mAH to 4000mA and “D” size (if not a “fake” with AA or C inside) about 7000mAH to 10000mAH. This must be very reliably connected to the terminals of the replica box as it limits the voltage. Many NiMH should not however be left indefinitely on charge. The other disadvantage compared to NiCd is that the average voltage is about 250mV higher. On charge with no load the cell will rise to 1.5V and on load fully charged is about 1.37V. If the optional D Cells are used these should not be left in the set as the NiMH will discharge them to about 50% if they are new. Alkaline are about x2 life than Zinc Carbon or Zinc Chloride in this application and less likely to leak. Interestingly  the Alkaline will recharge somewhat in a Colette or Annette if no more than 2/3rds discharged.  Any disposable cell must be removed before mains charging if exhausted or the NiMH will charge very slowly as  most of the energy is wasted in the flat cells.  Actually both the Annette series and this Colette put a very small current into the HT pack also. This would not really have charged the original “layercell” pack, but would have dramatically extended shelf life. It’s possible to fit 60 x AAA cells in  an original size replica case, so these ideally should be Alkaline. The common practice of using 10 x PP3 (9V rectangular packs) is poor value, the Alkaline are not so much better than Zinc, because the Zinc layer cells fill the can but Alkaline uses 6 off AAAA cells, wasting space.  The 60 off AAA pen cells will not mind the small HT current on mains and if less than 1/3rd  discharged may actually recharge.  

 

Grid Bias

Most Filament Tube battery sets after the use of grid batteries ceased use a resistor of 200 to 1000 ohms between LT- and HT- to provide grid bias of about -4V to -6V for the output tubes. This of course wastes HT voltage and power. It’s also less ideal for class AB or Class B outputs than Class A as the voltage is more variable and needs to be a higher drop.  The solution on the Colette is to derive bias from the local Oscillator.  At first sight this looks odd. The immediate thought would be a loosely coupled winding or sample anode and rectify it (there is an unused DAF96 diode). But actually the “solution” is simpler! With an oscillator biased by “grid leak” the grid will become negative. This voltage is simply filtered by a two stage RC filter 330K and 100pF followed by 330K and 2uF.  Note that C51 2uF +ve pin thus goes to 0V.

On VHF-FM  this oscillator is at a fixed frequency. On the AM bands the bias appears to vary somewhat between -5V and -8V  depending on band and across the band. About -5V is really class A and about -8V is close to class B. If the voltage is only about -4.5V replace the DF97 assuming there are no leaky capacitors. If C51 2uF (Electrolytic) filter is leaky or if the audio feed capacitors C91, C92 (paper, so likely WILL be leaky) 4.7nF are leaky then the bias will be too poor and DL96 or Transformer damage can occur. See the next section. A compromise design of perhaps 150 Ohms in HT- for bias to limit current in event of B4 (DF97) failure might have been a better idea!  Tthe DL96 grids are connected to 0V via effectively nearly  2.3 M Ohms grid leak (The 56K & 2 x 330K are shared) if the DF97 fails.

 

Architecture

Most Battery sets of this era use a DK96 (or possibly a DK92 for better Shortwave performance as it’s more similar to the 1L6 used in the Zenith Transoceanic). But here there isn’t one. This is why we mention the UK Vidor Vanguard, one of only two known UK tube portable with VHF-FM (the other is the Ever Ready Sky Emperor / BEREC Commander). Like the Colette, it has no Heptode/Pentagrid. Both sets use a Triodised Pentode as Local Oscillator and a Pentode as AM bands Mixer / RF amp. In both models the AM mixer Pentode is an ordinary FM IF stage for VHF. The Vanguard uses all DF97 and the triodised  VHF Self mixing/ Oscillator is AM L.O.  I have verified that the Annette LD480 and Vidor Vanguard do indeed work with a 50mA filament DC90 instead of 25mA filament Triodised DF97, one earth connection may need removed on socket, but this doesn’t seem to affect VHF operation. The  LD452AB actually does use a DC90. Presumably it will work with a DF97.  The Vanguard uses 3 x DF97 for FM IF and one of these is AM mixer. The IF and Mixer on it do also work with DF96, though it’s not quite the same pin connections or performance.  It’s unclear why the 1st FM IF on the Colette, which is an RF preamp on AM bands is a DF96 rather than DF97.

 

 

Type

AM

FM

Note

B1

DF97

OFF

Mixer/Osc

Triodized

B2

DF96

RF preamp

1st FM IF

 

B3

DF97

Mixer

2nd FM IF

Via g3, so DF96 won’t work!

B4

DF97

Local Osc.

fixed Osc.

Triodized, generate grid bias

B5

DF96

IF Amp

3rd FM IF

 

B6

DAF96

Det. AF Preamp

AF Preamp

Detector unused in FM mode

B7

DAF96

AF Driver/Tone

AF Driver/Tone

Triodised. (Diode Unused)

B8

DM71

Phase Inverter

Phase Inverter

Also Battery condition

B9

DL96

AF Output

AF Output

Class AB push pull

B10

DL96

AF Output

AF Output

Class AB push pull

 

The DM71 isn’t a tuning indicator here (it usually is) but used as phase inverter to drive B10, the 2nd DL96. One can only suppose that either there wasn’t enough gain with B6 DAF96 on its own (and some passive tone control scheme) to use the 2nd DAF96 as a phase inverter as is done on several Ever Ready Table Models (Sky Monarch AM, Sky Monarch AM/FM, Spacemaster/BEREC Skyscraper Mk. II). Actually some reports (German) of  DF64 used as phase inverter (sub-miniature hearing aid tube that can be wired under chassis).  A DM160 couldn’t be used as it probably didn’t exist! But it seems to be not purely an indicator and phase inverter. As the bias is just a 390K resistor to OV, it does appear to vary length with LT condition.  Just for comparison, the Ever Ready Sky Emperor (BEREC Commander) “solves”  the running out of places to put tubes to drive the Push Pull output by reflexing the 1st FM IF (used just as Audio stage on AM) as gain -1 / Phase inverter to drive the 2nd DL96! 

On FM the X3 and X4 germanium diodes are the FM discriminator.  So C78 5uF +V pin goes to 0V.

On “Charge” (Laden), the charging current is via an under run 12V 3W lamp as a series dropper, this would be 48 Ohms with 12V on it, but there is rather less, so it may be about 40 Ohms. Current is also via 8.2 Ohms in circuit during mains operation too.  In normal mains operation the 12V lamp is swapped for a 17 ohms resistor.  LT would thus rise to over 10V without a load if there was no DEAC or NiMH cell across the LT supply. The schematic shows what looks like 9.5V on the bridge (presumably on normal operation as that is how the switches are drawn).

Next parts we look at the restoration and typical faults. Case Restoration.

To give an idea of scale here is the Colette beside its sister Annette (LD480AB) and also the only UK portable tube model with LW, MW and VHF-FM, Vidor Vanguard (The Ever Ready/BEREC model has no LW, but SW instead). The small set is one of the last Vidor valve / tube models, with MW & LW only, the Vagabond. Both Vidor models are 1957.

Colette Annette

Colette Vanguard Vagabond

To thank the Author because you find the post helpful or well done.

 2
Bias and Local Oscillator. Colette LD562AB 
24.Nov.13 21:58
55 from 5269

Michael Watterson (IRL)
Editor
Articles: 1033
Count of Thanks: 12

Extracted here is purely the AM L.O., AM Mixer (FM 2nd IF) and the Push pull output stage without driver and phase inverter.

C47 is fixed capacitor for VHF-FM mode.  R18 330K  & C51 2uF (now miniature ceramic) are actually on a tag strip beside R53 & R54 just below the DL96 sockets. The junction of the input capacitor, 1M Ohm bias and 100K grid stopper is on the unused pin 4 as a tag.

Here is graphs of negative bias vs local oscillator for all bands. The VHF is a straight line as the AM L.O. is at a fixed  frequency of about 1.6MHz.

Bias vs tuning all bands. Voltage negative

With the original DF97 the bias was -4.5V maximum. The current DF97 is out of a box, but I didn't test it. I will test some DF97 and try one that is within spec for emission.

Vf is about 1.38V (charged NiMH) and HT 90V. Little difference on Mains or Battery.

To thank the Author because you find the post helpful or well done.

 3
More details on Design 
26.Nov.13 23:35
152 from 5269

Michael Watterson (IRL)
Editor
Articles: 1033
Count of Thanks: 15

PSU

The DEAC (or NiMH) replacement is actually charged via the 12V 3W lamp. So if it doesn't light, the DEAC may be faulty, power gone or bulb failed. If there is no red glow when "Laden" is pressed it's not charging!

If the set is on mains, and the power is unplugged or turned off, but the "Aus" (Off) hasn't been pressed, then the NiMH/DEAC will discharge into the filaments. The HT pack is connected by 4.7K Ohms, so typically the set will not operate. If there are a pair of D cells fitted they will be exhausted and leak even if Alkaline. This is why severe battery acid damage is common on the Annette & Colette tube models. The optional D cells should always be removed unless the set is being mostly used without mains.

On Economy the set uses 250mA on VHF and in normal mode 300mA. With a 10,000mAH NiMH that is 33 hours or 20 hours originally with a D6 (A 7AH Tagged GP D cell is about 23 hours). Some modern DAB radios are only 6 to 8 hours on battery!  Adding 2 x Alkaline D cells increase Portable run time to 100 Hours. An Alkaline  PP3 based HT pack will be about  30 hours and using AAA cells about 60 hours. Depending on if  FM or AM, how loud and how well the "bias" system is working.

HT is measured about 2mA more on VHF-FM mode which seems reasonable. Total HT ought to be about 14mA on AM and 16mA on VHF-FM (about twice an economical 4 tube AM only model, but only 2mA or 3mA more than a "louder" model with a DL94 tube).

Aerials

Originally most VHF-FM was horizontally polarised as it's simpler for roof mounted H aerials. Possibly sometimes slightly better coverage? So the most early portables had two telescopic aerials to act as a dipole.

The loop of wire at the telescopic aerial socket might serve three purposes:

  1. Perhaps reduce Local Oscillator Radiation (an issue with Mixer/Osc and no RF stage, on mains Radio and TV a major function of the common grid triode preamp is actually simply to reduce local oscillator radiation, the gain is poor).  
  2. It may act as an impedance match.
  3. It can act as a basic loop aerial in strong signal areas when rods are closed.

On AM and FM  the car aerial socket disconnects one telescopic rod. LW & MW use a large ferrite rod as main aerial and the FM 1st IF DF96 (B2) as a tuned RF pre-amp. The drive to the mixer then is band pass with an IF trap at 460KHz. On SW the RF in is the common of a pair of VHF chokes, so either telescopic rod has the same effect. The AM RF preamp isn't used, instead B3 DF97 mixer connects direct to the tuned circuit which has an aerial tap. Presumably the IF Trap isn't needed on SW and perhaps the idea is that an external aerial is used, which would need the output of the RF preamp to be tuned if it was kept in circuit. However the filament isn't turned off. On AM bands the B1 filament (VHF Mixer/Oscillator using triodised DF97) is disconnected.

See the Electrical/Electronics work on my set here.

To thank the Author because you find the post helpful or well done.

 4
Did Philips make a mistake? 
03.Dec.13 01:18
246 from 5269

Michael Watterson (IRL)
Editor
Articles: 1033
Count of Thanks: 13

There are models that do fine with 1 x DAF96 for detector Preamp, 1 x DAF96 for phase inverter (gain -1) and driving 2 x DL96 in Push Pull (e.g. Ferranti 945 is class AB, about -7V bias or Ia = 3.2mA per DL96). So why does the Colette LD562AB (L5X62AB) use the 2nd DAF96 as a triode amplifier and then use a DM71 as phase inverter? It's low gain so has no feedback, just using tap on previous stage. 

The Tone control can work fine on previous stage or even to ground rather than feedback on 2nd DAF96. Actually except on SW (with no external aerial) there is far too much gain. This leads to more hum on mains and prone to faint whistle around 8kHz to 10kHz  even at zero volume. 

So as an experiment I moved anode end of tone control to 0V, wired the 2nd DAF96 as gain of -1 using series resistor on grid and anode grid feedback. The DL96 driven from DM71 swapped to output of 1st DAF96. The 180pF to match response on other DL96 not needed. Audio drive to DM71 also removed.

Result still more than enough gain, but now no audible hum or whistles with volume at zero. Less distortion and clearer audio. A "bonus" is that the DM71 can work as a tuning indicator by R29 can connect to DM71 grid instead of chassis and the 390k grid leak replaced with a 3.3M Ohm from grid to negative end of C76. About 47nF on DM71 grid to 0V. This works without diodes as FM voltage is very large compared to AM strong station voltage.

There are reports too of some models using a DF64 (wire ended hearing aid type) as phase inverter. Was someone in Philips obsessed with dedicating a stage to the tone control?  Or was there a problem with gain on the Gram / PU input? Or was the idea to have overall negative feedback (there is none anywhere) but it was never added?

Looking at the volume control / gram etc we see the "earthy" end of the pot goes to the f+  of the 1st DAF96 rather than chassis / 0V for hum reduction. Or at least that's the theory. Not something I have seen on any UK model. Also would it not have been a safety compromise in the design to connect HT- via 180 to 270 ohms to limit Ia on the two DL96 if DF97 LO fails and also then make it easier to have closer to pure Class B via added negative bias from the Oscillator?

I do beleive in restoring sets to original condition, but in this case it was hard to resist testing the theory. Now will I put it back as it was or leave it in the improved condition (which actually uses the same number of parts and no tube changes or extra tubes)? 

To thank the Author because you find the post helpful or well done.