UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Test Gear and Workshop Equipment

Notices

Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders.

Closed Thread
 
Thread Tools
Old 23rd Nov 2019, 9:35 am   #101
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,633
Default Re: Marconi TF144H/4S

Sounds like somethings not right,replace valve with new one? Sorry I didn't get this far so can't comment/help.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is online now  
Old 23rd Nov 2019, 3:19 pm   #102
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

Certainly, from looking at it when the crystal calibrator is selected, it doesn't look like anything more than the heater circuit is on. Visual appearance of the valve is ok. I'll have a look through my valve stash at some point and see if I have one.

Out of interest how far did you get with your set?
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.
Sinewave is offline  
Old 26th Nov 2019, 7:19 am   #103
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,633
Default Re: Marconi TF144H/4S

"Out of interest how far did you get with your set?" I checked out the PSU and got an OP, there was no OP before, but from memory I only had an OP from the direct OP but no meter deflection or OP after the attenuator.

After reading your restoration, which has been a great help, I'll go back and finish the job but it's not a priority at present.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is online now  
Old 7th Dec 2019, 8:37 pm   #104
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

Back on this, still scratching my head over this crystal oscillator. Going to take the turret out and take a good look at the valve base.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.
Sinewave is offline  
Old 7th Dec 2019, 9:53 pm   #105
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

I'm stumped over these voltages still. Maybe I should try replacing the caps which are associated with the crystal oscillator circuit, maybe something is causing something odd.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.
Sinewave is offline  
Old 8th Dec 2019, 9:27 am   #106
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,633
Default Re: Marconi TF144H/4S

"maybe something is causing something odd." Been there before, wandering round, scratching the old noggin and muttering to myself. Sorry I can't help with any suggestions, only to post the schematic of the relevant section so folk can see it without have to download a girt big file, find the relevant page etc. They're more likely to offer help.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is online now  
Old 8th Dec 2019, 2:17 pm   #107
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

Here's the schematic for the area currently of interest. I tried another valve and got nothing different.

The test for the circuit is to couple a wire around the valve to make the test mentioned earlier here. On one range I get a triangle waveform, but the frequency is of the generators output and not the crystal. The other range is a noisy signal.
Attached Thumbnails
Click image for larger version

Name:	TF144H4S CrystCal.jpg
Views:	100
Size:	55.5 KB
ID:	195043  
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.
Sinewave is offline  
Old 15th Dec 2019, 6:26 pm   #108
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

Tried another valve, made no difference.

I replaced the capacitors connected to the oscillator circuit, made no difference.

There's nothing left to replace, as apart from that it's the resistors I already did earlier.

What on earth am I missing?
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.
Sinewave is offline  
Old 15th Dec 2019, 8:34 pm   #109
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,798
Default Re: Marconi TF144H/4S

That is a rather peculiar oscillator circuit. The active device is a cathode follower with somewhat less than unity voltage gain. It relies on the resonant circuit, crystal, to do some voltage step-up. The crystal is in a shunt position, so step-up relies on those series trimmer capacitors into the parallel resonance tank of the crystal. Add in the strays of wiring to the switch and that 470pF to ground and it seems a bit iffy. It may be just that the crystals are a bit down on activity. Trying some other crystals in that frequency region might be illuminating. Dropping the value of that 470pf for a trial may also be interesting.

I don't think this circuit had much margin in the first place.

david
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is online now  
Old 15th Dec 2019, 11:21 pm   #110
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

Thanks for the comments. I tried a 270pf capacitor in place of the 470 and there was no difference.

What I can try next when I can get them are some new crystals. I still don't understand why I'm getting the high voltage readings at pin 1 and 3 of the oscillator valve.

Otherwise the set does output a good CW.

From page 2 of the manual, the 'quick operation check', I appear to hear the high pitch tone around the 1000khz mark, but it's the tests in the setting up/performance checks from page 38 that throw some doubt in the operation, plus the odd voltage readings I recorded in post #95, #97 and #99.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.
Sinewave is offline  
Old 15th Dec 2019, 11:37 pm   #111
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,798
Default Re: Marconi TF144H/4S

If there's RF about, some voltmeters go silly.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is online now  
Old 16th Dec 2019, 8:50 pm   #112
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

Yes good point. I should really use a scope for these measurements. There are completely different results with a scope compared to a high impedance digital meter or Avo 8.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.
Sinewave is offline  
Old 16th Dec 2019, 8:58 pm   #113
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

For example,

From drawing Fig 4.11 RF Box and Attenuators

V103, pin 1 voltage measurement;

112.4VDC measured with digital meter. (102.2 previously measured, depends on what I've been fiddling with)

Scope: 9.03V RMS 28.2 peak to peak.

It's likely to be the AC voltage not the DC level measured as those higher voltages would probably do more harm than just sit there. The lower voltages probably suggest what you suggested earlier with reference to the xtal voltage being the amplifier.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.

Last edited by Sinewave; 16th Dec 2019 at 9:04 pm.
Sinewave is offline  
Old 17th Dec 2019, 9:57 pm   #114
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

Ahh, on the page before the drawings, it's says voltages are DC unless marked otherwise and relative to chassis.

So I am measuring excessively high voltage then! Now to work out why....
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.
Sinewave is offline  
Old 18th Dec 2019, 9:00 pm   #115
Chris55000
Nonode
 
Join Date: Aug 2007
Location: Walsall Wood, Aldridge, Walsall, UK.
Posts: 2,853
Default Re: Marconi TF144H/4S

Hi!

Both Anode and Cathode Sections of the Master Range switch S100Af and S100Aj switch different amounts of resistance into the oscillator valve's anode and cathode circuits, when changing up from Range F/G to the higher ranges – this is done because of the differing 'Q' of the oscillator tuned circuits on Ranges G–L!

Unfortunately it looks like both sections of S100A, the anode section S100Af and the cathode section S100Aj are suffering from cruddy/open circuit contacts on ranges G–L to a greater or lesser degree, so I'm afraid you might have to be prepared to completely remove the Range switch Waters 100Af and S100Aj from the main section of S100 and give it a good scrub with contact–cleaner/degreaser, apply fresh Contact Treatment Grease, making sure you get all the making contacts on S100Af & S100Aj to better than 0.25 ohms or less!

Chris Williams
__________________
It's an enigma, that's what it is! This thing's not fixed because it doesn't want to be fixed!
Chris55000 is offline  
Old 18th Dec 2019, 10:05 pm   #116
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Marconi TF144H/4S

I wouldn't take the schematic TP voltages too seriously. The 12AU7 datasheet and the TF144H circuit diagram suggest to me that whoever marked up the DC voltages on the schematic probably made a couple of typos for V103B anode and cathode. I think they probably meant to only quote these for two specific sub ranges rather than for every sub range within (A-F) and every sub range within (G-L).

Maybe it should have shown something like this?
V103B anode (58V DC range A)
V103B cathode (22V DC range A)

V103B anode (185V DC range L) ...yes, 185V rather than just 85V here
V103B cathode (84V DC range L)


There are several reasons why the published TP voltage data is wrong for range L but the most glaring one is that the schematic suggests that V103B anode and cathode are at similar voltages at 85V DC and 84V DC respectively. This is highly suspicious and also impossible if you look at the datasheet and the rest of the TF144H schematic. If the anode is at 85V DC then there must be about 7.5mA DC passing through the 22k resistor R216 as this is fed from 250V DC. The 12AU7 datasheet Vg=0 curve suggests that there needs to be >70V difference between anode and cathode to get it to gulp current at this level. But the schematic data says there is only a 1V difference between anode and cathode here on range L. So I think it is best to ignore the test point data on the schematic and try and reverse engineer what the voltages should be for each range.

The good news is that your range L measured V103B anode (189V DC) and cathode (83V DC) voltages look reasonable to me if I look at the schematic and the datasheet and I assume that the 2MHz crystal oscillator is actually oscillating on range L. So my advice would be to stop swapping parts and to definitely leave the crystals alone for now. i.e. don't stress the crystals with a soldering iron until they are proven to be duff or a bit lossy.

There's a good chance that there is nothing significantly wrong with the voltages you are seeing so maybe the best thing to do is for us all to study the design of the oscillator and also the 12AU7 datasheet and it should be possible to come up with a revised set of test voltages that are closer to the truth than the numbers given on the schematic. It should be possible to predict/produce a set of test voltages for V103A and V103B for every range.

I can try and help here but I'm not really a valve expert. In fact I have virtually no experience of working with valves. But I do know how to do a basic oscillator analysis based on the schematic and the 12AU7 datasheet if you are interested? I can give you a couple of 'in circuit' tests to try on V103B to prove it is behaving as per the datasheet if that helps?
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 18th Dec 2019 at 10:15 pm.
G0HZU_JMR is offline  
Old 18th Dec 2019, 10:27 pm   #117
Sinewave
Octode
 
Sinewave's Avatar
 
Join Date: May 2016
Location: Oxfordshire/Bucks borders, UK.
Posts: 1,604
Default Re: Marconi TF144H/4S

Thanks Chris and Jeremy.

Chris, sounds like a good idea, I need to check those switch wafers out.

Jeremy, thanks, so it seems my measurements are along the right lines - I did wonder why I couldn't come up with any reason why/why not. All help greatly appreciated.
__________________
Avometer, vintage Fluke and Marconi collector. Also interested in vintage Yaesu and KW.
Sinewave is offline  
Old 18th Dec 2019, 11:45 pm   #118
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Marconi TF144H/4S

The best range to look at is range L because it doesn't involve V103A.

Don't worry if you aren't familiar with some of the stuff below. Besides, I'm hoping someone else with (valve) experience will have a look at it to see if I'm on the right track. This is the first time I've tried to do this with a valve so there are bound to be a few nuts in the content below...

The way I did my quick and dirty analysis of range L was to study the 12AU7 datasheet and I looked at the Vg=0V curve as highlighted in red in the first graph below. The big surprise for me (as a valve novice) is that this curve shows that the triode behaves as a non-linear resistor even when the grid bias is held at a constant Vg=0. In other words, the anode current varies significantly with changes in anode voltage right through to the upper voltage limit for the anode.

I think this curve data can be used to predict the initial startup DC conditions when V103B is powered up on range L. This is before it starts oscillating. Once it oscillates strongly, the large signal condition will cause a shift in DC levels but for now it's worth looking at the startup conditions.

I extracted a SPLINE model of this Vg=0 curve as below and this can be used as the data to model a non-linear resistor to emulate the DC behaviour of the triode at Vg=0V. The V103B circuit appears to leak the cathode bias voltage to the grid at startup so I'd expect the startup Vg to be 0V as per the Vg=0 curve in the datasheet.

1;0.0007;
10;0.0015;
20;0.002;
30;0.0035;
40;0.0045;
50;0.0055;
60;0.00675;
70;0.008;
80;0.0095;
90;0.01075;
100;0.0123;
120;0.0155;
140;0.0185;
160;0.022;
180;0.027;
200;0.029;

This is just a suite of co ordinates of voltage:current that make up the non-linear Vg=0 curve in the 12AU7 datasheet. This data can be entered into a non-linear simulator and from this, the initial DC startup conditions for V103B can be quickly predicted as in the second image below.

This predicts that at startup, V103B will be biased at just under 4mA and the anode will be at about 164V and the cathode will be at about 129V. This is a difference of just 35V and we can take a look back at the Vg=0V curve again. You can see the arrow at 35V anode-cathode voltage and this shows just under 4mA. This is a nice confidence boost that the SPLINE model is working OK as this is the correct current at 35V.

The next datasheet curve to look at is the curve for transconductance or gm. This shows that at 4mA the gm is about 1.3mmho. I'm not sure how valid this gm number is when there is only a 35V difference between anode and cathode but this gm can be used with the feedback capacitances to predict how much 'negative resistance' the valve is creating at the crystal node at this initial startup condition at 2MHz. This is a simple equation and I can show it if you are interested but the result is about -3000 ohms at 2MHz.

The absolute value of this negative resistance needs to be higher than the passive ESR of the 2MHz crystal in order for the oscillator to start up. A typical 2MHz crystal might have an ESR of 600 ohms so things look good here for a solid oscillator startup on range L.

In order to crudely predict the final DC voltages at the anode and cathode when the oscillator has full started and settled I would need to know the ESR of the 2MHz crystal. As this isn't known, I think it is fair to assume that it will be much less than 1000 ohm. When the oscillator starts up it will produce a small signal but obviously this will get bigger and bigger until something in the valve limits the oscillation level. I would expect this to happen when the valve is effectively cut off for most of the RF cycle. i.e. you can think of the crystal and feedback caps as a spinning flywheel. The amplitude of the sinewave at the grid will get bigger and bigger until a point is reached where the negative portion of the grid sinewave cuts off the valve for most of this sinewave as it will dip much lower than the cathode voltage and this will cause pinch off. So the valve becomes the equivalent of (very briefly) tapping the flywheel with a stick at a 2MHz rate and there comes a point when it reaches its maximum ability to tap up the waveform and the oscillation level will then reach a limit.

When this happens the on/off duty cycle of the anode current becomes quite low and the anode current will be like a train of narrow pulses at 2MHz. This is good for the dual role of mixer/heterodyne for V103B on range L.

The narrow pulsed nature of the anode current means that the average current taken by the valve should start to fall compared to the initial DC startup current of about 4mA. The lower the ESR of the crystal the lower this average current will get. It might drop below 3mA for example. With this in mind we can expect the anode DC voltage to creep up and the cathode DC voltage will drop as the oscillator settles.

If the final anode current falls from 4mA to 2.7mA at full oscillation levels then the voltage at V103B anode would be 250-(22000*0.0027) = 190V DC.

Conversely, the average cathode DC voltage will go down with a lower anode current. If this current is also approximately 2.7mA then the average voltage at the cathode would be something like 0.0027*33000 = 89V DC.

This seems very close to your measured results.
Attached Thumbnails
Click image for larger version

Name:	12AU7_VG0.jpg
Views:	50
Size:	171.8 KB
ID:	195592   Click image for larger version

Name:	12AU7_SPLINE.gif
Views:	52
Size:	14.2 KB
ID:	195593   Click image for larger version

Name:	12AU7_gm.jpg
Views:	44
Size:	167.9 KB
ID:	195594  
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 18th Dec 2019 at 11:57 pm.
G0HZU_JMR is offline  
Old 19th Dec 2019, 12:23 am   #119
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Marconi TF144H/4S

Probably the first test to do would be to deliberately cripple/stop the 2MHz oscillator on range L by lifting one leg of the 270pF feedback cap. Then measure the anode and cathode DC voltages of V103B on range L. I would expect there to be no oscillation and you should then be measuring the initial DC startup conditions for V103B on range L.

The datasheet based SPLINE model suggests that (with a healthy 12AU7 that mimics the datasheet curves) you would see 164V DC at the anode and 129V DC at the cathode of V103B on range L in the absence of oscillation. This assumes that R216 in the anode really is 22k ohm and the cathode resistor really is 33k ohm and the HT is 250V.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 19th Dec 2019, 3:08 pm   #120
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Marconi TF144H/4S

Probably the most relevant thing is to keep in mind is what is actually being asked of V103A and V103B in this sig gen.

From the operator point of view, they just want an accurate crystal oscillator (at either 400kHz or 2MHz) and they want to be able to hear heterodyne/beat notes in the headphones that act as audible calibration markers as per table 2.1 in the TF144H technical manual.

With this in mind, it really won't matter much if the V103A and V103B valves show some spread in DC operating point and this spread could be due to the emissivity of the valve or due to component tolerances and ageing. I suspect that the whole setup will be quite forgiving as it is deliberately designed to be non-linear and 'birdy'. As long as the heterodynes can be heard by the operator and as long as the crystal trimmer caps have been accurately set for 400kHz and 2MHz I think that is all that really matters...
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 10:49 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.