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 18th Feb 2016, 11:27 pm   #61
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

OK, so what I've found now is...

Fault Reminder.

The output of the genny appeared to vary considerably when changing the frequency within each of the 4 highest frequency ranges, I, J, K & L. The output of the ALC circuit remained constant around 1.4v. Until the generator output went very high when the ALC circuitry output strangely fell.

I had previously injected from 1 to 35 MHz CW into the ALC circuit at 4 volt p to p, with the output on R107 / R171 remaining constant at around 1V DC. I was satisfied here that the ALC was working as expected, with help from members.

With the turret on & unit powered I monitored the output at the Direct output socket with a 50 ohm BNC lead to scope. I then found the output varying from almost zero to 10V p to p as I changed frequency within range K which was the worst range.

After a big think! I then put a X10 probe on the input to the ALC circuit, under the turret, a bit awkward, this is wired straight to the direct output socket. So I though should give the same results. Here I found a constant amplitude throughout all ranges and frequencies. Going back to monitoring the direct socket this time with a T piece and a 50 ohm load the amplitude regulation was now much better. So I think reflections/standing waves and poor matching, (please tell me if I am wrong) have been confusing me for quite a while!

I did find early on, in the ALC circuit that all the four .1 uF caps were leaky and one of the diodes were extremely leaky.

This has been an extremely interesting exercise for me, and it's not put back together yet and calibrated, so it's not over yet. I am grateful for all the posts so far. I have learnt a lot. Colin
vintage_8bit is offline  
Old 25th Feb 2016, 11:49 pm   #62
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

Update.
As my generator now appears to be working after repairs to the turret and ALC, plus my confusion over the loading of the output...

I have now turned my attention to the lack of a working thermocouple. I tried the thermocouple from Rob and after removing about 50 ohm from the resistors feeding the heater it looked promising, but unfortunatly it ran out of steam over 4mhz.
I also got hold of what appeared to be a genuine replacement thermocouple but although it showed around 24 ohms for the heater and continuity on the thermocouple output its output was very low. It hardley moved the 8mV FSD meter. (Meter checks out fine and also worked with the thermocouple from Rob below 4mhz)
I am now awaiting delivery of a true power amp IC, an AD8361, to try and make a replacement circuit for the thermocouple, as suggested in a previous post (7) by Richard. Perhaps start a new thread now?
vintage_8bit is offline  
Old 26th Feb 2016, 10:19 pm   #63
trh01uk
Octode
 
trh01uk's Avatar
 
Join Date: Nov 2010
Location: Ross-on-Wye, Herefordshire, UK.
Posts: 1,648
Default Re: Marconi TF144H CT452A RF Generator

Colin,

those final results are quite puzzling. I am trying to think what could cause a thermocouple to go frequency selective, and stop working above 4MHz. I suppose its possible a feed resistor to the thermocouple is inductive, and thus increasing the impedance as the frequency increases.

There are lot of bits between the RF direct output and the thermocouple, namely RV104, R100, RV103, and R186 in parallel with R198. Not to mention the C195 limiter circuit. If any of these parts have been replaced by a modern "equivalent" it may well be that the modern part is far more inductive than a carbon composition original. Even so, 4MHz is quite low for such effects to take effect. For instance, to generate 100 ohms of inductive reactance at 4MHz you need about 4uH. That's a lot of inductance, so this explanation looks unlikely.

As for the other thermocouple that has no output... maybe the heater element has parted company from the thermoelectric junction? A few thou gap would probably be enough to prevent normal heat transfer.

I shall be very interested to hear how you get on with the AD8361!

Richard
trh01uk is offline  
Old 27th Feb 2016, 1:53 pm   #64
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

Yes Richard that roll off is strange. Im not used to working with RF as you can probably tell from my previous error measuring the output of my genny. I haven't given up on that particular thermocouple yet though, but I did measure a constant RF amplitude across the heater with change in frequancy. I had a look at a spec sheet for a list of thermocouple shunts. The style fitted origionally was designed specifically for RF. Where as the donated thermocouple style, RF is not mentioned.

I have got all the bits now for the mod now so watch this space!

Colin.
vintage_8bit is offline  
Old 20th Mar 2016, 12:21 am   #65
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

Hello again. As its been a while I thought I would just say I am still looking at a mod with a AD8361 & a AD8051 to drive the meter and give some gain. It's working but not as expected, hopefully give a better update soon.
vintage_8bit is offline  
Old 22nd Mar 2016, 10:26 pm   #66
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

Regarding the modulation controls when I set the modulation control to 50% (after setting the mod cal point to set) should the modulation trough go down to 50% of carrier peek or is it 1/3 of peek.

Using the formula I have seen mentioned in a few places, peek to peek amplitude of one side of the modulation envelope รท unmodulated peek to peek x 100 for 50% gives the 1/3 of peek I mentioned above.

I have looked at a Marconi sig gen type 2019. The 50% mod setting here gives the trough at 50% of carrier peek.
vintage_8bit is offline  
Old 23rd Mar 2016, 3:17 pm   #67
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

Ok I think Ive got it now. The formula was correct and I wasn't.
vintage_8bit is offline  
Old 12th May 2016, 11:17 pm   #68
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

Perhaps the last update on this?
With the circuit Ive built using the Ad AD8361 and a AD8051 all seemed ok setting the carrier to 2V rms. However when the modulation is applied the circuit is not responsive enough. Setting the meter to the modulation cal mark causes more than 100% modulation. So after spending many hours on this I gave up and looked again at a thermocouple shunt kindly given to me by a forum member. This thermocouple shunts heater resistance was higher at around 30 ohms so I changed the feed resisters between the 2 volt output socket and shunt to compensate. This now seems to work as well as the original. This time I also kept the leads to the shunt very short which maybe why it didnt work the 1st time I tried it.

Unfortunately this dose not solve the problem of the unabtainable shunt. So maybe someone else will have a go?
vintage_8bit is offline  
Old 13th May 2016, 9:31 am   #69
trh01uk
Octode
 
trh01uk's Avatar
 
Join Date: Nov 2010
Location: Ross-on-Wye, Herefordshire, UK.
Posts: 1,648
Default Re: Marconi TF144H CT452A RF Generator

Can you post a schematic of the AD8361 and AD8051 circuit that you used?

I don't understand your comment that the "circuit is not responsive enough". You meanit is too slow? The AD8361 shows it responding to an RF pulse in well under 50 microseconds. The AD8051 is equally fast.


Richard
trh01uk is offline  
Old 13th May 2016, 10:54 pm   #70
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

OK I'll draw out circuit and post soon.

What I meant was....

I set the carrier o/p to the "set carrier" point on the meter (2V RMS @ the direct o/p socket) and then set the modulation to the "set mod" point. The modulation is then over 100%. I can't remember what it should be just now but I think it's around 80%.

So what I am trying to say is that my circuit's o/p between no mod and 80%? mod is not giving a DC level which is in step with the difference between the "set carrier" & "set mod" witness marks on the meter.

I did actually slow the circuits response but only so it didn't bang the meter about to much during warm up and range changes. Hope this makes sense.
Colin.
vintage_8bit is offline  
Old 14th May 2016, 10:27 am   #71
trh01uk
Octode
 
trh01uk's Avatar
 
Join Date: Nov 2010
Location: Ross-on-Wye, Herefordshire, UK.
Posts: 1,648
Default Re: Marconi TF144H CT452A RF Generator

Colin,

the original circuit using the thermocouple will give a true RMS value for unmodulated or modulated carriers. According to the TF144 manual this corresponds to 2.0V DC (under test conditions) at the direct output for unmodulated, and 2.3V DC for modulated. I couldn't find a figure for mod depth for the mod set point in the manual, but we can work it out.

These test voltages represent the RMS value of modulated and unmodulated carrier. The formula for relating RMS voltages to the mod depth m is given by:

(rms modulated carrier volts/rms unmodulated carrier volts)^2 = (0.5 x m^2) + 1

When you work that through with the 2.3V and 2.0V, you get m = 80% as you stated.

As an aside here, the manual does warn that you can't get 80% if the modulating frequency is too high a percentage of the carrier frequency. This really only applies with odd situations, like trying to modulate a 10kHz carrier with 1kHz - the limit then is 30%. And the manual doesn't say how the mod set point then relates to this. But I assume that you used a more normal carrier frequency in your tests.

I think the problem you may be seeing is that the response of the AD8361 to modulation is far too fast. In other words, its DC output (at pin 7) will follow the modulation envelope with no problem - if you are using the standard circuits in the datasheet.

By contrast a thermocouple is really sluggish - it probably can't follow anything much faster than 1 - 2 Hz I would guess. Maybe even less than that.

Unless you have reduced the bandwidth of the rms measuring circuit of the AD8361 to about the same, its output at pin 7 will be following the modulation. And the meter will then give a false result. You could verify this quite easily by checking the waveform at pin 7 of the AD8361 with a scope - it should be DC, in both unmodulated and modulated situations.

I expect you took all this into account when designing your AD8361 circuit. And you will have read all the stuff in the datasheet about the FLTR pin (6), and the need to add extra capacitance.

I note the situation is even more complex than at first meets the eye, because the internal series resistance of the chip (which forms part of the filter) varies in value between high value signals (500 ohms) and low value signals (2kohms). To get to around 2Hz bandwidth with the 500 ohms resistance the external capacitor would need to be around 160uF! What value did you actually use in your test circuit?


Richardd
trh01uk is offline  
Old 19th May 2016, 10:23 am   #72
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

Sorry for the rough drawing, which should be attached? Just noticed I drew in the thermocouple on the output. It should of course be the meter coil. (Less haste, more speed)

The input resistors were chosen to keep the resistance the same as the loading of the thermocouple heater.

I had noticed the carrier waveform sitting on top of the dc output of the 8361, as you predicted. I also initially saw it looked clipped. I played around with the input resistors reducing the input into the 8361 until the carrier sitting on the dc level wasn't looking clipped. I then added the filter (22uf) to remove this. This also had a desired effect on damping the meter response to somewhere near the original response.

I left the input cap at the 0.1 value. Maybe I should have used a higher value as you suggested? I perhaps wrongly thought that the impedance would be low enough at all frequencies to not be significant.

I did try some different value caps on the filter pin without any improvement.

Unfortunately the generator is now giving an intermittent low output on all ranges, so unable to experiment just now until fixed.

I got fed up before trying to fault find with the turret in the way. This time I have salvaged a coil pack from an old turret and intend to tack this in place, once I have removed the original turret. Should give me lots of room to probe around.

Colin
Attached Thumbnails
Click image for larger version

Name:	RF power meter.jpg
Views:	117
Size:	33.3 KB
ID:	124677  

Last edited by vintage_8bit; 19th May 2016 at 10:28 am. Reason: Mistake in drawing.
vintage_8bit is offline  
Old 20th May 2016, 8:46 am   #73
trh01uk
Octode
 
trh01uk's Avatar
 
Join Date: Nov 2010
Location: Ross-on-Wye, Herefordshire, UK.
Posts: 1,648
Default Re: Marconi TF144H CT452A RF Generator

Colin,

yes, the input resistors look like a sensible mod. I take it the thermocouple heater element (a working one) measures around 23 ohms then? I checked the level you would see at the AD8361 and your values put the input nicely inside its linear range when you have 2V rms RF signal at the hi-level output of the TF144.

I didn't suggest you should change the value of the input capacitor - I was talking about the capacitor between the FILTR pin and the VPOS pin. Its that latter capacitor that you have missing from your schematic. I think you need to add one, and the value I suggested could be as high as 160uF. I would try a 100uF electrolytic for starters.

The point of this capacitor is explained in the datasheet. See "Filtering" on page 11, and "Selecting the filter capacitor" on page 14. I don't think this capacitor is equivalent to the 11k and 22uF filter you have added after the AD8361 output. That certainly has a suitably low frequency rolloff (0.6Hz), but the FILTR capacitor is conected within the squaring system of the chip, and will fulfil a somewhat different role.

I can't quite get my head around how the chip works, so I suggest a bit of experimentation is required here.

Incidentally you could test your AD8361 board on any signal generator that can produced an 80% modulated AM signal. You just need enough carrier to get the meter up to the "Set carrier" mark without the mod, and then see if the needle goes to the modulated mark when the AM is applied.

Richard
trh01uk is offline  
Old 24th May 2016, 11:05 pm   #74
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

Ok, generator fixed. I tried the capacitor across the filter pin & positive rail. I tried several values up to 220uf. They didn't make any difference to the output, with the 80% modulation applied.
Any ideas. Colin.
vintage_8bit is offline  
Old 24th May 2016, 11:47 pm   #75
trh01uk
Octode
 
trh01uk's Avatar
 
Join Date: Nov 2010
Location: Ross-on-Wye, Herefordshire, UK.
Posts: 1,648
Default Re: Marconi TF144H CT452A RF Generator

Colin,

OK - are you able to put a scope on the VRMS pin and see exactly what effect your 220uF capacitor is having? Is the modulation being removed or not?

Richard
trh01uk is offline  
Old 25th May 2016, 10:39 pm   #76
vintage_8bit
Heptode
 
Join Date: Feb 2013
Location: Rainham, Kent, UK.
Posts: 523
Default Re: Marconi TF144H CT452A RF Generator

With 1V 30khz p to p on the I/p there is 1v dc on the vrms o/p. This has a 1/2 wave rectified waveform at carrier frequancy + modulation sitting on top.This is with no cap fitted to the filter pin. With a 150uf cap fitted to the filter pin this wavefourm disappears. Leaving just the 1v dc. Removing the cap has no effect on output to the meter.
Colin
vintage_8bit is offline  
Old 26th May 2016, 4:52 am   #77
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,803
Default Re: Marconi TF144H CT452A RF Generator

Taking a sideways look at this...

An awful lot of very fine signal generators have been made in the years since this without a thermal converter, thermocouple power meter or true RMS converter in sight.

The purpose is to make a nice sinewave at RF, so the RMS to peak ratio ought to be reasonably predictable and a peak detector will be fine.

A modulated waveform is more complicated but a simple voltage average from a detector will give the carrier voltage, and comparison with a peak-held voltage from the same detector will give the AM percentage.

Marconi "went thermal" because they needed a detector that was reasonably flat across the frequency range, and then they indicated AM percentage by the increase in RMS power it showed.

Wanting to keep the original metering and operating procedure is what forces sticking with this scheme, but a carrier power meter with a switch for an AM% scale would be easier to operate.

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 26th May 2016, 8:22 am   #78
trh01uk
Octode
 
trh01uk's Avatar
 
Join Date: Nov 2010
Location: Ross-on-Wye, Herefordshire, UK.
Posts: 1,648
Default Re: Marconi TF144H CT452A RF Generator

David,

yes, all agreed. I think its a reasonable objective to see if we can just replace the thermocouple and leave everything else alone. It avoids a lot of mucking about - adding switches presumably means drilling holes in the front panel - and that's beyond what most people want to do.

The original scheme does have the beauty of extreme simplicity - and I think that's what makes it worth trying to emulate with more modern components.

Quote:
Originally Posted by vintage_8bit View Post
With 1V 30khz p to p on the I/p there is 1v dc on the vrms o/p. This has a 1/2 wave rectified waveform at carrier frequancy + modulation sitting on top.This is with no cap fitted to the filter pin. With a 150uf cap fitted to the filter pin this wavefourm disappears. Leaving just the 1v dc. Removing the cap has no effect on output to the meter.
Colin,

right so the cap does work as far as cleaning up the output of the AD8361, but even when this cap is applied, the DC levels that the meter sees is still wrong.

By the way, why are you using such a low carrier frequency? There are limits on how much mod depth you can get when the carrier is that low. You need to be at a minimum of 100kHz carrier, to get 80% mod depth with a 1KHz modulation frequency. I wonder if something odd like this is happening here? Can you retest with the carrier say at 1MHz?

Richard
trh01uk is offline  
Old 26th May 2016, 8:34 am   #79
trh01uk
Octode
 
trh01uk's Avatar
 
Join Date: Nov 2010
Location: Ross-on-Wye, Herefordshire, UK.
Posts: 1,648
Default Re: Marconi TF144H CT452A RF Generator

David, Colin,

I am also intrigued by what Colin is reporting. That a modern RMS detector is not giving the same result as the older system. Why not? I suspect there is something about the AD8361 internal workings that I have not properly understood......


Richard
trh01uk is offline  
Old 26th May 2016, 9:37 am   #80
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,803
Default Re: Marconi TF144H CT452A RF Generator

With the Gilbert Cell RMS converter, you need to be sure that it isn't even momentarily hitting saturation (been there!)

The other thing I'd be looking into is the 'law' of the two devices. Tid the thermocoupe give a voltage out proportional to the power input, or proportional to the RMS of the input voltage? There is a square-law function difference between these.

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  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 4:56 pm.


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.