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 8th Apr 2017, 1:13 am   #21
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

I've actually still got my old Marconi 2015A here and it still works. I've had this since the 1980s but it was effectively retired to storage when I bought a couple of Marconi 2019 generators in the 1990s.

I could have a look inside mine but I think it may be difficult to get to A16? It seems quite buried inside the generator according to the manual?

The last time I had it apart was during the World Cup Final and Brazil were playing. I think it was 2002 and I spent the whole of the final trying to replace the (long time broken) tuning cord whilst listening to the game. This was without any instruction manual. I managed it in the end using some special black lacing cord and it has worked ever since. I have a large drum of this lacing cord and it did a brilliant job. Just the right amount of grip without slipping.

My 2015A still works fine and I assume the first ALC loop will be healthy in it but I can't be certain. Are there any checks I can do with it? I don't want to take it apart 'too much' in case it involves going near the dreaded tuning cord again
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 10th Apr 2017, 10:41 pm   #22
paulfifties
Triode
 
Join Date: Apr 2017
Location: East Flanders, Belgium.
Posts: 17
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

Jeremy,
The measurements in attached spreadsheet were executed with a 12 MHz RF input on the desoldered mini-coax going to A16, while the negative lead of my Brymen DVM was all the time attached to the junction of D17 D2 cathode and D1 anode.
The positive lead was each time swapped between anode of A17 D2 and cathode of D1.
Remark:
the measurements were not exactly executed from low to high values, as listed in the table.
Actually I started with the 50 mV RMS value, increasing up to 90 mV, and then returning to 40, 30 and 20.

Observations:
1) All the measured values were slowly drifting down, also with the positive lead of the Brymen not connected.
2) Even for the values measured on the collector of TR3, this phenomenon was observed.
3) Not sticking to heavily to the exact values measured, we can conclude that there is definitely a tendency of decreasing voltage over A17 D1 and D2, as the RF input level on A16 was increasing ! The voltage over the diodes is in forward direction.
4) The same goes the other way around : decreasing RF input level causes the voltage drops over A17 D1 and D2 to increase.
5) The voltage drop over A17 D1 is in my case systematically smaller than the voltage drop over A17 D2 (around 5% - 6% roughly)
6) As soon as the RF input level is increasing above 55...60 mV, a distortion is observed in the trailing slope (around the zero-crossing)
7) With the RF input level as little as 10 mV, the output amplitude is increasing again, but looks like it is being compensated/stabilized.

I wonder if you find the same behaviour on your instrument. The only thing to unsolder for these measurements is the central conductor of the mini-coax cable at A12 connection nr 14. The 12 MHz RF signal is fed into this mini-coax going to A16.

Kind regards,
Paul
Attached Files
File Type: xlsx TF2015_D1_D2_TR3 measurements.xlsx (4.9 KB, 49 views)
paulfifties is offline  
Old 11th Apr 2017, 11:34 pm   #23
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

I took the base cover off and got as far as being able to see inside A17.

I did a few quick measurements using a bench DMM (Keithley 2015) and a H/H DMM for the differential measurements.

To turn the RF off I connected TR3 base to ground rather than disconnect anything.

Here's my results. They are very similar to the simulator! There are slightly higher voltage drops across the diodes but the rest looks very similar indeed.

The select on test resistors R6 and R7 were:
R6 430R
R7 560R
R5 trimpot was set very close to mid position.

No RF:

R1/R2 12.7767V DC (K2015)
R6/R7 12.7817V DC (K2015)
differential: R1/R2 (black) to R6/R7 (red) +4.8mV (H/H DMM)
differential: TR1 base (black) to TR2 base (red) -1.8mV (H/H DMM)
differential: A17 D1 cathode (black) to A17 D2 anode (red) +165mV (H/H DMM)
differential: R1/R2 (black) to TR2 base (red) +160mV (H/H DMM)



With RF and ALC settled:


RF level at A12 output coax 56mV rms at 12MHz (this feeds to A16 input?)

R1/R2 12.777V DC (K2015)
R6/R7 12.782V DC (K2015)
differential: R1/R2 (black) to R6/R7 (red) +4.7mV (H/H DMM)
differential: TR1 base (black) to TR2 base (red) -27.6mV (H/H DMM)
differential: A17 D1 cathode (black) to A17 D2 anode (red) +133mV (H/H DMM)
differential: R1/R2 (black) to TR2 base (red) +114mV (H/H DMM)

So when it all settles there must only be a 46mV drop in the A16 detector output compared to the 'no RF' condition.
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 11th Apr 2017 at 11:40 pm.
G0HZU_JMR is offline  
Old 12th Apr 2017, 12:42 am   #24
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

I've got an old Tek RSA3408A real time spectrum analyser here and I had a quick go at capturing the ALC attack characteristic once I enable the RF by releasing TR3 base. This allows the A12 oscillator to start up and the ALC then kicks in.

You can see that the oscillator envelope rises up quite quickly but then the ALC overshoots before finally settling after some ringing. The time scale is 200ms/div. The RF is taken from the generator output at about 0dBm rather than the 50mVrms at the A12 output.

I also had a quick look at the close in phase noise when the generator was set to 12MHz 0dBm output. It is surprisingly noisy and only manages -103dBc/Hz at a 5kHz offset. Not that great for an oscillator running at 12MHz.
Attached Thumbnails
Click image for larger version

Name:	2015A_alc.gif
Views:	69
Size:	23.6 KB
ID:	140812   Click image for larger version

Name:	2015_PN.gif
Views:	58
Size:	20.0 KB
ID:	140813  
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 12th Apr 2017 at 12:54 am.
G0HZU_JMR is offline  
Old 13th Apr 2017, 8:37 pm   #25
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

I overlooked the fact that the above ALC plot includes the response characteristic of the second ALC loop and this will be quite dominant.

So I just did a second capture taken at the output of the A12 module. This should be about 50mVrms and you can see that the ALC attack is very damped and slow, caused by the large capacitors in the ALC circuit. The ALC levels the signal at about 80mVpk which is about 55mVrms. I used a 1GHz Marconi x1 active RF probe to take the measurement here. This won't load the circuit.

I also added an image below of the phase noise of a fairly modern synthesised sig gen at 12MHz for comparison. You can see how much cleaner it is compared to my 2015A at 12MHz in the previous post. It is about 30dB cleaner than the 2015A at a 5kHz offset. This analyser has a snazzy looking phase noise plotting function under its Signal Source Analyser menu but I haven't bothered to use it here.
Attached Thumbnails
Click image for larger version

Name:	ALC2015SC.gif
Views:	57
Size:	20.7 KB
ID:	140921   Click image for larger version

Name:	lowPNgen.gif
Views:	70
Size:	19.9 KB
ID:	140922  
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 13th Apr 2017 at 8:42 pm.
G0HZU_JMR is offline  
Old 18th Apr 2017, 12:16 pm   #26
paulfifties
Triode
 
Join Date: Apr 2017
Location: East Flanders, Belgium.
Posts: 17
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

Going back to post #23, I will list the values I measured (no RF input!) on my TF2015 behind, marked in red.
Actually I measured with the RF-input wire to A16 desoldered at the A12 side (as I had it desoldered to allow me to connect an external 12 MHz sine wave generator to it)
Hopefully this is equivalent to shorting out the A17 R10 base resistor for TR3 ...

The select on test resistors R6 and R7 were:
R6 430R R6=470R
R7 560R R7=510R

No RF:

R1/R2 12.7767V DC (K2015) 12,738-->12,740 VDC (stable)
R6/R7 12.7817V DC (K2015) 12,669-->12,670 VDC (stable)
differential: R1/R2 (black) to R6/R7 (red) +4.8mV (H/H DMM) -70,65 mV (after about 2 sec)
differential: TR1 base (black) to TR2 base (red) -1.8mV (H/H DMM) -60,60...60,10...60,00...59,3... drifting downwards in absolute value.
differential: A17 D1 cathode (black) to A17 D2 anode (red) +165mV (H/H DMM) +103,30 mV initially, afterwards slowly decreasing.
differential: R1/R2 (black) to TR2 base (red) +160mV (H/H DMM) +12,29 mV...12,22...12,16...

I wonder what could be the reason for these strongly deviating measurements, especially on the differential values. Would I have to install different values for R6, R7 and readjust setting of pot R5 ?
Or would it be caused by the difference of the original diodes on A16 versus the two replaced BAT83S Schottky diodes for D1 and D2 on A17 ?
It puzzles me that the value measured at R1/R2 (with respect to GND) is, on my instrument, *higher* than the value at the junction of R6/R7. So I guess this is causing a wrong polarisation of the A17 D1 and D2 voltages ?

I still need to compare the forward voltages occurring over the original and replaced diode types when they are put in series, running the same forward current...
(Hope to be able to do that when I return home this evening).
paulfifties is offline  
Old 18th Apr 2017, 10:30 pm   #27
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

I'm not sure what to suggest that would be best. But if you want to try a few basic tests you could try checking the health of the diffamp and TR3 with no RF.

One dynamic way to do this would be to temporarily replace each pair of series diodes with a single 220k resistor. Then temporarily replace the R6 and R5 pair with a 1k trim pot and see if you can play with the 1k trimpot to flip the diffamp to either direction to see if you can control how it shares out the 250uA. Also see if you can control the voltage at TR3 collector with the 1 k trimpot as it sets the bias current for the A12 transistor oscillator. This would check if TR3 is OK.

Otherwise you could try the same with the diodes still in place (but with no RF) after you make sure they each have a similar Vdrop with no RF. But the above test would only take a few minutes and would give good confidence in the overall diffamp and TR3. Then it would be a case of getting the right diodes in place. I guess this part will be tougher because even Marconi needed to select on test R6 and R7 to get the thing to work inside a narrow window of operation.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 20th Apr 2017, 10:24 pm   #28
paulfifties
Triode
 
Join Date: Apr 2017
Location: East Flanders, Belgium.
Posts: 17
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

Replacing the 2 pairs of diodes as per last post of Jeremy is an excellent idea, but the A16 D1 and D2 diodes are quite hard to reach (lower side of A16 PCB).
1) So I choose to substitute on A17 the R5 potmeter and R6 resistor(=470R) by a 1K potmeter.
However, with RF-in connected (A17 R10 not shorted out), I could not get the circuit to stabilise, and the RF-out collapsed after a short time (time constant of C11 probably).
For the differential voltages TR1 base (black) and TR2 base (red), which I will refer to as "diff base voltage" in the remainder of this post, the following was observed:
At first a voltage of -76 mV was noticed, decreasing slowly to -22mV, around which value the RF-out signal suddenly collapses. Then the diff base voltage was reaching some -20,95mV. At this point, the voltage was very very slowly decreasing in absolute value (more like drifting).

With R10 shorted out (no RF-in): the diff base voltage goes immediately to -21,30 mV, and from their the same drifting (=very very slowly decreasing in absolute value) is observed.


2) Then I decided to put a fixed resistor of 1K in series with the 1K potmeter. Now the RF-out did no longer collapse !
With RF-in (A17 R10 not shorted), the diff base voltage was varying between +0,89mV...+1,06mV
When the 1K potmeter is turned downwards, at first the value of diff base voltage is jumping down to minus several mV, but then stabilising around +0,9...+1,1 mV

I could decrease the value of the 1K potmeter further down, and initially the value of the diff base voltage is jumping down to around -3 mV, but stabilises at around -0,20mV.
Turning the 1K potmeter always further down, the diff base voltage is stabilizing at negative values: -1,80mV... -2,60mV ... -4,7mV...-5,2mV...-6,0mV which seems to be about the value when the 1K potmeter reaches zero.

However, with only the 1K resistor (1K potmeter at zero) the value is no longer stabilising !
Switching the TF2015 off and on, and the RF-output is collapsing again...

3) Now the 1K potmeter is turned up again, and the diff base voltage measured when the Carrier Range switch is switched through the various ranges.
Measured 1K fix + 1K potmeter total value during below measurements : 1358 ohms


Carrier Range**************Diff base voltage (mV)
10-14,3**************-3,0
14,3-20,5**************-2,8...-3,0
20,5-29,4**************-2,5...-2,7
29,4-42**************-2,4...-2,3
42-60**************-1,6...-1,7
60-86**************-2,4...-2,6
86-124**************-2,6
124-177**************-1,4...-1,5
177-254**************-0,6...-0,7
254-363**************-0,8
363-520**************+9,0...+10,0


At every range switching, a jump respons of the diff base voltage is noticed of negative tens of mVolts (up to -25mV), then quickly decreasing to the above values.

4) Unfortunately, an important distortion is noted on the RFout signal for ranges below ca 125 MHz. At first a dip is showing on the falling edge, past the top crest.
However, while switching to lower carrier ranges, this dip in the RFout signal becomes even more pronounced, and evolves from a small 'u' shape to a large 'V' dip, and one can no longer speak of a sine wave output at all !

5) The RF-in amplitude is increasing each time a lower carrier range is selected !
At the lowest 2 ranges of the RF-in signal is showing some rounding of the lowest tops of the sinus-wave.

Conclusion: if the RF-in amplitude is to be kept rather constant over the entire Carrier Range, then something is dreadfully wrong with the differential amplifier or the diodes.
(Still need to compare these diodes though)
paulfifties is offline  
Old 22nd Apr 2017, 9:07 pm   #29
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

Quote:
but the A16 D1 and D2 diodes are quite hard to reach (lower side of A16 PCB).
I've put my 2015 back together now but I think you can unsolder the wires from terminal posts 2 and 4 on the A17 board and then solder the 220k resistor across those terminal posts 2 and 4 on A17 to effectively replace the A16 diodes with a resistor? Hope I've got that right.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 24th Apr 2017, 9:35 pm   #30
paulfifties
Triode
 
Join Date: Apr 2017
Location: East Flanders, Belgium.
Posts: 17
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

Jeremy,
The original Instruction Manual did not contain any PCB layouts.
Would you happen to have an image of the PCB layout for A16 ?
Otherwise, I will have to be desoldering the A16 D1 and D2 totally blindfold, which I find quite risky...
Unfortunately the A16 board is mounted Cu-side up, and I don't see how to remove it, without impacting the very short-wired connections adversely.
paulfifties is offline  
Old 24th Apr 2017, 11:30 pm   #31
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

I don't think you need to access the A16 board and you don't need to remove the A16 diodes. By removing the wires from the terminal pins 2 and 4 on A17 you effectively isolate the A16 board diodes.

So once the green and orange wires are unsoldered from terminals 2 and 4 on A17 you just need to add the 220k resistor across the two pins 2 and 4?
See below for an image of my A17 board.
Attached Thumbnails
Click image for larger version

Name:	2015A_A17.jpg
Views:	77
Size:	127.3 KB
ID:	141483  
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 5th May 2017, 12:33 pm   #32
paulfifties
Triode
 
Join Date: Apr 2017
Location: East Flanders, Belgium.
Posts: 17
Default Re: Automatic Level Control (ALC) drive circuit in Marconi TF2015 generator

At last, yesterday evening I have been able to continue my testing...
I carried out the proposed replacement of respectively A16 D1 and D2 and A17 D1 and D2 by a 220K resistor, unsoldering the green A17pin#2 wire.

I hooked up a DVM between the TR3 collector and ground (range DC volts), and a second one across the bases of TR1 (black) and TR2 (red)(range DC mV).
I had the generator carrier range switch at 56 MHz.
Also R5 and R6 were replaced by a 1k potmeter.

A sine wave output signal is appearing at RFout from the moment the collector voltage of TR3 is approaching 9,3V. Which corresponds to a TR2-TR1 base differential voltage of -1,0mV...-0,9mV.
Actually, it is only a very very narrow window, for which a non-distorted sine wave is produced at the RF output.
In addition, this zone is also very instable: after adjusting (using the potmeter) the TR3 collector voltage at 9,3V it takes only a short while, until the collector voltage is drifting away to 9,4 and 9,5V.
While reaching 9,5V the sine wave signal on the scope starts jittering, and shortly after, the signal is collapsing !

In short:
at TR3 collector voltage > 9,3V --> RFout collapses
at TR3 collector voltage < 9,3V --> RFout starts being distorted. At first a dip in the top crest of the sine wave, and then getting deepened and sharper as the collector voltage is drifting away downwards from 9,3V (and TR3 is slowly moving (or adjusted via the potmeter) into the direction of saturation).

I suppose my main problem is that the equilibrium state of the differential amplifier TR1-TR2 is so unstable, that the RFoutput is nearly directly driven in distortion with TR3 going into saturation.
But what would be causing this?

Another item: the voltage drop comparison of HP2835 (in circuit) or HP5082-2811 (according to schematics and parts list) versus BAT83s diode.
I used a 100µA analog meter and a 500µA analog meter for the above 100µA measurements, in series with a 10K resistor and the two diodes. Each of the two diodes had a DVM clamped onto them in the DC mV range.
See attached spreadsheet for the details.
On a second line, I noted the comparison of the BAT83s previously used, with a random different BAT83s.
Attached Files
File Type: xlsx Schottky-Current-Voltage-comparison-table.xlsx (5.4 KB, 49 views)
paulfifties is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 9:40 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.