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 4th Dec 2012, 11:13 pm   #41
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Feeling brave, no worries, this is a pussycat compared to the wife!

Swapped the chips over. Exactly the same, warming Q12 starts the counter displaying 10MHz, but only zeros until it's warmed. But TP goes from about 4.4V to 0.1V

I am going to swap them back and repeat the test...

Which I have just done and both chips are now changing TP4's state from 4.4V to 0.099 / 0.1V ...... but still only when Q12 is warmed
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 5th Dec 2012, 12:21 am   #42
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Quote:
Which I have just done and both chips are now changing TP4's state from 4.4V to 0.099 / 0.1V ...... but still only when Q12 is warmed
Your earlier tests only showed TP4 at about 5V all the time so has something changed?

At first the above behaviour seems really odd because TP4 drives Q12 and not the other way around but I think by warming Q12 you force the mixer into bypass mode and the signal maybe gets through and trips a detector flag to the processor and then maybe the processor progresses to another stage in its program where it correctly sets TP4 to 0V.

What I can't fathom is why it isn't setting TP4 to 0V right at the start of the search routine.

If it can set it to 0V after you heat Q12 then I guess the hardware is healthy in terms of the processor's ability to address and control the PIA port pins.

So I am a bit confused...
G0HZU_JMR is offline  
Old 5th Dec 2012, 12:31 am   #43
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Can I safely ground TP4 and see if it changes and reads the 10MHz signal?
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 5th Dec 2012, 12:46 am   #44
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Probably best to try and fathom what the counter is doing in terms of how it responds to the detector flags (RF and IF)

If you force the mixer MX1 to be bypassed (using the hot iron) then you will force the 10MHz into the IF stage and this will presumably drive the IF detector voltage up.

So this would trip the comparator U1.

But the other requirement for successful signal navigation to the o/p at port J5 is that the switch ICs U4 and U5 need to be controlled correctly for each mode.

Does the counter behaviour change if you swap identical U4 and U5 chips (CA3049) across? Only do this if they are in easy to swap sockets. Don't do it if it means desoldering chips as you could add extra faults if a chip got damaged during the desoldering.

Last edited by G0HZU_JMR; 5th Dec 2012 at 12:54 am.
G0HZU_JMR is offline  
Old 5th Dec 2012, 12:50 am   #45
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Quote:
Originally Posted by Chris Wilson View Post
Can I safely ground TP4 and see if it changes and reads the 10MHz signal?
I'd have to check the PIA datasheet to see if it could withstand this but I don't think it is worth forcing this pin to 0V. Usually PIA ports are very robust unless you do this to several of their port pins at once.

But don't do it because it isn't really a cure and it isn't worth the risk of PIA damage.
G0HZU_JMR is offline  
Old 5th Dec 2012, 1:16 am   #46
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Unfortunately U4 and U5 are soldered directly to the board. Would you say the issues are confined to A109 board?

As an aside U7 looks a lot newer, and a different style (although the same number) to U2, so maybe there's been trouble with one of these PIA's before.

Just seem the edit re R57, it's a 4.7K that's installed.

Does this bit from the service manual help? It looks like it could have three states to me. If it's unclear in the attachment I have it at:

http://www.gatesgarth.com/manual.jpg

as a full size page
Attached Thumbnails
Click image for larger version

Name:	manual.jpg
Views:	79
Size:	63.3 KB
ID:	73565  
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 6th Dec 2012, 3:41 pm   #47
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

As these posts were getting messy with a mistake in a component number,
I thought I'd tidy things up and put the salient stuff together. I have
ordered two new devices for U4 / U5 so I can change them. I found
someone with surplus NOS and they were very cheap, so what the heck!

Saga starts with counter as received, showing just dashes, not able to
do anything more than a display test, and won't perform basic 200 MHz
test function:

I have had a good look at board A108 the U6 chip is missing from, and
apart from a mysterious pair of multi turn pots the schematic parts
list seems to say are "not used" I can see no anomalies save for U6
being missing altogether.... I'll post the results of fitting this
device.

The replacement for the missing U6 chip arrived this morning and it
now completes the 200 MHz self test. It will read up to the limit of
my frequency generator on Band 3 which is only 1040 MHz, but looks
promising. Band 1 also works fine.

Band 2 sort of works, but seemingly has a serious anomaly. I cannot
get Band 2 to read below 185.000 MHz. To even get over 185 MHz needs a
lot of drive power, too. 2 volts or so. Even then this sometimes
doesn't start a display of other than zeros. If I input 184.900 it
won't read it and displays zeros. 185 and up is OK... Weird. IF
frequency? It should read from 10 MHz up to 1 GHz. It'll go to 1 GHz
plus AOK.



I believe I may have isolated the issue to the A109 board. Until the
unit warms up Band 2 is deaf and needs plenty of millivolts input to
trigger, sometimes it won't trigger at all. Once warm it sometimes
settles down to near the makers spec on how much input at what
frequencies it should need to trigger. (<NOTE -- EDIT > Seems not to
trigger at all of late, save with warming Q12) I used freeze spray to
isolate a small area, and the fine tip of a de soldering iron
connected backwards to the pump, so it blows a fine jet of hot air.
The problem is in the area marked on the schematic of board A109,
device Q12. As I have no riser board access is terribly limited.


It's a 2N4126 component number Q12 If I just touch it with a piece of
stiff, heavy copper wire wound round the tip of a small 15W iron, it
changes state almost immediately and the Band 2 is very sensitive
again, and immediately displays a frequency, down to 10 MHz, just
fine. As soon as it cools it reverts to the insensitive state. But
bear in mind above 185 MHz it works in a fashion hot or cold, with a
lot of drive... If the counter is left to warm up naturally and I cool
this device with some spray, it reads zeros again.

I am not 100% sure what it's switching function is, but I have removed
it, and isolated on my Peak semiconductor tester it varies gain form
170 cold to 190 warm, and it will suddenly go to a gain of just 4 if
it gets a bit warmer still.

R10 is also playing up. It should be a 43K 2% but measures 32K cold
and warmed a touch changes to 4.8K ! It should have a 1 PPM/C
temperature coefficient, too, so something has happened to that as
well.


I am not sure if any previous abuse would have damaged diode CR1,
shown as a ND4991 ?



New Q12 and new R10 fitted, problem just the same. Warming the new Q12
brings the display back.

Voltage tests done, results below all with a 30mV emf input to the
Band 2 socket:

TP2 No input signal 3.4 / 3.5V Displays zeros
TP2 10 MHz 3.4 / 3.5V Displays zeros
TP2 50 MHz 3.4 / 3.5 V Displays zeros
TP2 100 MHz 3.4 / 3.5V Displays zeros
TP2 400 MHz 3.4 / 3.5V Displays zeros
TP2 500 MHz 0.06V Displays frequency
TP2 900 MHz 4.9 / 5.0V Displays frequency

For TP2 the change over seems circa 450MHz

----------------------------------------

*** BUT *** Input 650 MHZ and TP2 shows 4.2 / 4.3V and displays zeros
Warming Q12 makes no difference

TP 3 No input signal 1.5V
TP3 10 MHz 1.5V Displays zeros
TP3 50 MHz 1.5V Displays zeros
TP3 100 MHz 1.5V Displays zeros
TP3 400 MHz 1.5V Displays zeros
TP3 500 MHz 4.1V Displays frequency
TP3 900 MHz 0.0V Displays frequency


-----------------------------------------

TP4 No input signal 4.5V Displays zeros
TP4 10 MHz 4.5V Displays zeros
TP4 50 MHz 4.5V Displays zeros
TP4 100 MHz 4.5V Displays zeros
TP4 400 MHz 4.5V Displays zeros
TP4 500 MHz 4.9V Displays frequency
TP4 900 MHz 4.9V Displays frequency


-----------------------------------------


TP11 No input signal 0.11V

Same voltage across all frequencies Displays zeros until it gets a 500 MHz
input, with which, and up at 900 MHz the counter displays the frequency.


-----------------------------------------

Another test:

TP2 30 mV input signal @ 10 MHz 3.4 / 3.5V displays zeros

Warm replacement Q12 with tip of barely hot iron, momentarily.

4.9V and displays frequency

------------------------------------------

On Band 3 with no input signals to the counter TP2 shows 4.9V


------------------------------------------

On Band 2 with no input signals to the counter TP2 shows 3.4 / 3.5 V
(wavers about between the two).


Q12 voltages

E 12.07V no input signal : 12.08V with 980 MHz @ 50mV frequency is shown

B 11.66V no input signal : 11.68V with 980 MHz @ 50 mV frequency is shown

C 0.22V no input signal : 0.0V with 980 MHz @ 50mV frequency is shown

C 0.22 V no input signal : 0.22V with 10 MHz @ 50mV no frequency shown (zeros)

Warm Q12 and collector voltage starts rising instantly and around
0.52V the counter shows 10 MHz.

Keep warming a bit more and collector voltage rises as high as 11.9V

As Q12 cools voltage smoothly falls and counter stops showing 10 MHz
frequency at about 0.52V and shows zeros.

A minute or so after warming collector voltage is back down to about
0.22V with the 10 MHz input signal @ 50mV and counter still shows no
frequency (zeros).


Swapped the U2 and U7 chips over. Exactly the same, warming Q12 starts
the counter displaying 10MHz, but only zeros until it's warmed. But TP
goes from about 4.4V to 0.1V

I am going to swap them back and repeat the test...

Which I have just done and both chips are now changing TP4's state
from 4.4V to 0.099 / 0.1V ...... but still only when Q12 is warmed.

I guess this means U2 is functionally OK.


__________________________________________________ ____________

This is where I am now, I hope the above makes some sense if
viewed with the service manual and the schematics. Thanks.


Someone has kindly sent me a different manual, which correctly shows the two multi turn pots on A108 board. I have put this at:

http://www.gatesgarth.com/EIP_1980.pdf
__________________
Best Regards,
CW. 2E0ILY

Last edited by Chris Wilson; 6th Dec 2012 at 4:03 pm.
Chris Wilson is offline  
Old 6th Dec 2012, 6:44 pm   #48
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

When you do your tests, are you following the flow chart fig 4.3 in the manual?

Also look at fig 4.2 for the block diagram for range 2.

At the start of the routine it surely must bias the mixer MX1 to bypass mode (using Q12 to deliver about 11V) It actually states it sets the multiplier to DC bias but I'm not sure if this is as relevant (or as correct) as the importance of setting the MX1 mixer to DC bias mode. I think MX1 must be DC biased during the early routine to allow the signals to bypass across the mixer with low loss. This is because it starts off looking to see if the RF signal is in the IF range (<200MHz)

Then it looks at your input RF signal level and compares it with the IF level with the comparator U1.

Because there is a 200MHz LPF and an amplifier before the IF level detector then the counter assumes the signal is <200MHz if the IF level is bigger than your input RF signal (because anything much above 200MHz will not get through the 200MHz LPF at a large level)

If the IF level is declared BIGGER than the input level (determined by comparator U1) then the counter has a go at measuring the frequency of the signal at the IF port. To do this it has to 'signal select' or 'route' the signal to the J5 connector using U4 and U5.

If it counts it as BELOW about 190MHz then it concludes the signal is a genuine signal in the range 2A (10-185MHz)

It will then declare that the input RF frequency is effectively the IF frequency (because it didn't need to do any downmixing) so it declares lock and displays the frequency of the IF frequency as a valid RF frequency.



It looks like your counter can't even respond correctly to a 10MHz signal in this first bit of the flowchart in fig 4.3.

So either the level arriving at the IF detector is too small wrt the RF signal or there's something wrong about the U4,U5 switching when it tries to measure the frequency.

There's lots of ways these faults could happen and the best way to faultfind the counter is to see what it does when it tries to follow the routine above.

I have no idea how quickly it is able to whizz through the flowchart in fig 4.3 but is it possible you are missing some fast action at TP4 and only reporting what happens when it gets stuck? Are you sure you are injecting at least -20dBm into it?

It probably only spends a tiny fraction of a second in any one part of that flow chart...

Last edited by G0HZU_JMR; 6th Dec 2012 at 6:51 pm.
G0HZU_JMR is offline  
Old 6th Dec 2012, 6:57 pm   #49
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

I'll revisit this later, and let you know if I can see anything.

I haven't used my scope to try and follow the signal, mainly because without riser boards access to anything save TP pins is all but impossible. Would any scope shots help if I can get it on some of the test points? I have a good Tek 7854 and a poor USB scope that's supposedly 100MHz, but is probably a lot less...

I have also collated the info and schematic for the mysterious 2 pots that I have on A108 board and my downloaded manuals bypass by saying "not used". The early manual I now have shows some info on these, and it may or may not have relevance. The info is at http://www.gatesgarth.com/latest.zip and are in the form of screenshots taken from the manual as a whole, which as i mentioned before is at http://www.gatesgarth.com/EIP_1980.pdf
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 6th Dec 2012, 7:13 pm   #50
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Because it is getting stuck so soon in the flow chart you could try the following fautfinding process.

1/ Force MX1 into bypass mode all the time by feeding in 12V to the Q12 end of R57 with Q12 removed. i.e. take out Q12 and put a 1k ohm link resistor across the vacant collector and emitter holes. So now current will flow from 12V into the MX1 port via your 1k resistor in series with R57.

Probably best also to disconnect the VCO drive signal from J1 to prevent it from contaminating the tests below.

Then stick in 10MHz at -10dBm to the counter.

You can then make measurements of the signal path from J4 right through to the IF amp at Q8 using a scope or an RF probe using short tacked on access wires as this should be OK at 10MHz. So you don't need the extender board. Just have the access wires poking out of the card safely.

The level at R6 should be 1/3rd the voltage at J4 if the first amp Q1 is healthy.

The RF level at L5 after the 200MHz LPF should be a bit lower on the scope (due to the mixer MX1 bypass loss)

The level at CR3 anode should be quite big wrt the level measured way back at J4 because of the gain of the Q8, Q9 amplifier.

You can also measure the DC voltages around U1 pins 2, 3 and 6 using a DVM or scope.

Then remove the 1k DC bias link resistor to bias off the MX1 mixer and measure all the above again and the levels after the mixer should all fall a lot because the mixer should then act as an attenuator if it has no LO signal or DC bias.

This should be enough to measure tha basic health of this signal path.

Last edited by G0HZU_JMR; 6th Dec 2012 at 7:33 pm.
G0HZU_JMR is offline  
Old 6th Dec 2012, 10:04 pm   #51
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Quote:
I have also collated the info and schematic for the mysterious 2 pots ...
I don't think they are relevant to range 2 as they appear to be calibration pots for the YIG filter. I've got a bit of works experience of playing with YIG filters although this was about 20yrs ago....

Basically, the DAC controlled YIG current sets the frequency of the YIG filter via the PIA but in our works receivers we found we had to use a custom lookup table to get it to track accurately across its range. I believe the EIP YIG will have a lookup table too. We set up our YIGs using ATE gear. Then because the YIG is also affected by temperature (even if it has an internal oven) we added correction for temperature changes. But on top of that the YIG also ages and ideally requires recalibration after a few years or so.

So my guess is that EIP maybe got fed up of rewriting PROM based calibration 'renewal' chips every year or so if the YIG aged and instead added the trim pots to allow a much more convenient and faster correction system for YIG ageing.

i.e. maybe they concluded or hoped that the trim pots were enough to keep the YIG in calibration each time it was returned for a recalibration. Just adjust for calibration slope and offset as in a manual 'y= mx +c' based trimpot control of the original (straightish?) YIG calibration lookup curve held in PROM?

Maybe they initially tried to use the same master lookup table for YIG control for some production units and used the trimpots to align it for each production model? Maybe this didn't work out?

So maybe they decided they could only do it accurately via a lookup table process so got rid of the trimpots? Or maybe the trim pots caused YIG drift?

But all this is just for input range 3 and yours doesn't work on input range 2 yet...

Last edited by G0HZU_JMR; 6th Dec 2012 at 10:16 pm.
G0HZU_JMR is offline  
Old 6th Dec 2012, 10:21 pm   #52
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Unbelievable! I started on the test by soldering a short insulated lead to the junction of R1 and R3, and left J1 disconnected. Set up the analogue scope and connected the probe to the short lead. Turned the counter on and to Band 2. Perfect readout.... What the....? Thought what had changed. Re-connected J1, still a perfect readout. Dropped input level to 30mV (-23.5dBm) and still a perfect readout from 10 MHz up.

R2 or R3 iffy?
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 6th Dec 2012, 10:48 pm   #53
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Some scope shots.

http://www.gatesgarth.com/scope/from-gen.jpg


http://www.gatesgarth.com/scope/R2-R3.jpg


First is probing X1 at the BNC connector before plugging into the Band 2 socket on the counter

Second is at the junction of R2 and R£ via a 3 inch piece of insulated wire, again probe X1

20Mhz input at your figure of -10dBm
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 6th Dec 2012, 10:51 pm   #54
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

I still think it's mischieviously messing with us both...


What isn't clear to me is if it sits there scrolling through the fig 4.3 flowchart even if there is no signal applied yet. If so it may well be at a random part of the flowchart when you connect up your 10MHz signal.

When you say it now works from 10MHz up are you testing it with inputs from 10MHz to 1000MHz on range 2 or just the 10-185MHz part of range 2?

Edit just seen your scope plots. I'll look at these shortly.

Last edited by G0HZU_JMR; 6th Dec 2012 at 11:02 pm.
G0HZU_JMR is offline  
Old 6th Dec 2012, 11:02 pm   #55
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Nooooo.... don't look at the junction of R2,R3 as this is the emitter of the common base amp. If it is running a 10mA emitter current the input impedance here will be very low. Just a few Ohms. i.e. impedance = 26/Ie = 26/10 = about 3 Ohms. So you will see a tiny signal here on a healthy counter.

I suppose the fact you see a tiny signal here at the emitter is a good indicator that the BFR90 is biased up and healthy though...

Look at the circuit closely... the input impedance of range 2 at J4 is meant to be 50 ohms. It gets this by having the 75R R2 in parallel with the 150R R1.

If you allow for the not quite zero input impedance of the BFR90 then it's actually (75R + 3 ohms emitter impedance) in parallel with 150R to get close to 50 ohm input impedance. i.e. 78R in parallel with 150R is 'about' 50 ohms.


The Q1 BFR90 is a common base amp. It transfers the input current to the output so any AC current passed into the emitter via the 75R resistor gets transferred to the RF impedance at the collector. This RF impedance is 820R in parallel with 51R in parallel with the mixer input impedance (50 Ohm?) = 25R.

So you push that 20MHz RF current into 25R via the 75R. 25/75 = 1/3. So you must get 1/3rd the RF voltage at the collector of Q1 if you push that same current into 25R. So Q1 acts like a 10dB voltage attenuator if you compare the RF voltage at R1 with the RF voltage at R6.

Last edited by G0HZU_JMR; 6th Dec 2012 at 11:14 pm.
G0HZU_JMR is offline  
Old 6th Dec 2012, 11:03 pm   #56
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Gone again now, will remove Q12 and get some measurements. How odd.
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 6th Dec 2012, 11:08 pm   #57
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Quote:
Originally Posted by G0HZU_JMR View Post
I still think it's mischieviously messing with us both...


What isn't clear to me is if it sits there scrolling through the fig 4.3 flowchart even if there is no signal applied yet. If so it may well be at a random part of the flowchart when you connect up your 10MHz signal.

When you say it now works from 10MHz up are you testing it with inputs from 10MHz to 1000MHz on range 2 or just the 10-185MHz part of range 2?

Edit just seen your scope plots. I'll look at these shortly.

All the way through from 10Mhz to 1 GHz Have removed Q12 and added a 1k resistor. Will get more scope patterns. I can tell you that it works fine on Band 2 with the Q12 removed and the resistor added. Worked straight away. Give me some actual test points you'd like to see patterns from, what about pin 1 on MX1, then before C29?
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 6th Dec 2012, 11:26 pm   #58
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Ah.... I see. You still have the 1K resistor fitted. Yes, this is much more likely to make it work reliably on 10-185MHz inputs.

But it isn't meant to be a cure

If it works solidly with the 1k fitted at Q12 then there has to be a control issue at the PIA wrt Q12 once you refit Q12.

The PIA also drives Q20 via the TP4 output of the PIA.

Anyway... if it is working really well for 10-185MHz with the 1k resistor fitted in place of Q12 then we can assume that BFR90 Q1 is OK. Also the 200MHz LPF is OK. Also the 10-200MHz IF amp Q8 and Q9 is OK and possibly the RF and IF detector diodes are OK.

Also U4 and U5 are probably OK.
What is not proven is the PIA control of TP4. You swapped PIA devices so the PIA is OK.

So either the program flow is getting confused (unlikely?) or there is 'something' screwing with the voltage control at TP4.

Time for a break and a close look at Q20 on the circuit maybe...
G0HZU_JMR is offline  
Old 6th Dec 2012, 11:29 pm   #59
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Junction R2 and R3 as before, but with Q12 out and 1k resistor in, as reference:

http://www.gatesgarth.com/R2-R3-Q12out.jpg

Probing pin 4 on MX1:

http://www.gatesgarth.com/mx1-pin4.jpg
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 6th Dec 2012, 11:34 pm   #60
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: EIP 545A 18GHz counter questions. Signature analysers?

Quote:
Originally Posted by G0HZU_JMR View Post
Ah.... I see. You still have the 1K resistor fitted. Yes, this is much more likely to make it work reliably on 10-185MHz inputs.

But it isn't meant to be a cure


If it works solidly with the 1k fitted at Q12 then there has to be a control issue at the PIA wrt Q12 once you refit Q12.

The PIA also drives Q20 via the TP4 output of the PIA.

Anyway... if it is working really well for 10-185MHz with the 1k resistor fitted in place of Q12 then we can assume that BFR90 Q1 is OK. Also the 200MHz LPF is OK. Also the 10-200MHz IF amp Q8 and Q9 is OK and possibly the RF and IF detector diodes are OK.

Also U4 and U5 are probably OK.
What is not proven is the PIA control of TP4. You swapped PIA devices so the PIA is OK.

So either the program flow is getting confused (unlikely?) or there is 'something' screwing with the voltage control at TP4.

Time for a break and a close look at Q20 on the circuit maybe...
No, I checked the signal at R2, R3 BEFORE removing Q12 and adding the 1k, and for some reason it started to work fine. But it's done this occasionally before... I was hoping disturbing R2 / R3 had effected a cure.

I have now posted scope shots with Q12 out and the resistor fitted, again at R2 / R3 and at pin 4 of MX1, but from the above I can also say it works fine, and seemingly solidly, turned on and off a few times, from 10 MHz to 1 GHz and down to 30mV.

So I need not get more shots further up the line with Q12 out?

Thanks Jeremy
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 2:25 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.